Original Research

Conventionally, primary care physicians decided when to refer and to whom a patient should be referred.^1,2 Specialists’ assistance was sought for diagnostic or therapeutic dilemmas,^3,4 management of conditions that presented too infrequently to maintain clinical competence,⁵ and specialized procedures that fell outside a physician’s scope of practice.^3,4 In some cases, physicians referred because patients requested to see a specialist.^1,4

The reorganization of health care over the past few decades has dramatically altered the interactions between primary care physicians and specialists. The growth in multispecialty group practice arrangements⁶ has led to formal, organizationally defined linkages between practitioners. Managed health plans and medical groups⁷ encourage primary care physicians’ control over the referral process through such mechanisms as specialty referral authorizations, financial disincentives for making a referral, performance assessment of referral patterns, and referral guidelines. These changes have transformed a once-informal process into one rife with administrative restrictions on referral decision making.

The Ambulatory Sentinel Practice Network (ASPN) Referral Study was designed to describe and analyze primary care physicians’ referral decisions and their outcomes in the context of a changing health care system in the United States. The study occurred in the ASPN and other regional practice-based research networks. This report examines primary care physicians’ referral decisions. We conceptualized the referral decision as occurring in 2 phases: whether to refer followed by to whom to refer.²

Methods

Physician sample

Physicians were recruited from March 1997 to May 1998. Recruitment activities were directed to all physician members of ASPN, physicians affiliated with the Medical Group Management Association, local and regional networks (Minnesota Academy of Family Physicians Research Network, the Wisconsin Research Network, the Dartmouth Primary Cooperative Research Network (COOP), and the larger community of primary care physicians. The study was publicized via direct mailings to physicians, articles and notices in practice-based research network newsletters and journals, and presentations at conferences. Contact with physicians expressing interest was made by telephone.

Physicians were included in the study if they practiced in the United States and were not in residency or fellowship training. Of all physicians contacted, 342 expressed interest in the study and 182 completed some aspect of data collection. A total of 141 family physicians, 12 internists, and 1 pediatrician completed all phases of data collection. In this study, the 141 family physicians (41% members of ASPN) formed the physician sample. They delivered health care in 87 practices located in 31 states.

Procedures

Study protocols and materials, based on a similar practice-based research study conducted with pediatricians, were reviewed and approved by the Committee on Human Research of the Johns Hopkins School of Public Health and the Colorado Multi-Institutional Review Board.^4,8,9 We conducted a pilot test in 5 practices; this test led to further refinements of methods and questionnaires.

Data collection occurred from September 1997 to February 1999, with 94% of physicians collecting data in 1998 only. Before beginning data collection, physicians completed a questionnaire concerning their practices and personal characteristics. Each practice selected a coordinator who communicated with research staff, learned study protocols, trained office staff and physicians, and monitored data quality. Coordinators completed a questionnaire about the organizational and financial components of their practice. They kept a log of all visits made during 15 consecutive business days and occurring during regularly scheduled office hours. A business day was defined as a half or full work day, provided that the physician held routine office hours. Each patient’s date of birth (5% missing), sex (2% missing), and principal diagnosis (5% missing) were recorded.

The coordinator kept another log of all referrals made by physicians, nurses, and other office staff. Referrals made during telephone conversations with patients were included. A referral was defined as a recommendation that a patient have a face-to-face encounter with another practitioner. We excluded referrals made to laboratories, radiologic facilities, emergency departments, hospitals for inpatient admission, and “curbside consultations” (ie, when the referring physician obtains advice from a specialist but does not send the patient for a visit).

A medical record abstractor assigned ICD-9-CM codes to diagnoses provided by office staff. We matched ICD codes to an expanded set of diagnosis clusters (EDCs). EDCs group ICD codes into clinically homogeneous categories using the methods developed by Schneeweiss.¹⁰ (For more information on EDCs, see http://acg.jhsph.edu.)

When physicians made a referral, they completed a questionnaire (response rate 93.9%) with items concerning the referral decision. Reasons for referral were based on our previously developed taxonomy used in a pediatric referral study⁴ and focus groups of family physicians convened during an annual ASPN convocation.

At the study’s conclusion, physicians received a report that compared their referral practice patterns with those of the entire sample. To defray office expenses associated with data collection, each practice was given a $100 stipend in addition to $5 for each physician referral.

Generalizability analysis

We compared referral rates of the study sample with the National Ambulatory Medical Care Survey (NAMCS), a nationally representative sample of office visits made to family physicians.^11,12 We pooled surveys from 1989 to 1994, inclusive, when the majority of the items in the survey instruments remained unchanged.¹³ (The 1995–1999 surveys did not contain information on whether the visit led to referral.)

We selected visits made by patients enrolled in non-HMO health plans (NAMCS) and health plans that had neither capitated primary care physician payment nor gatekeeping arrangements (study sample). This was done because of the known effect of managed care in general, and of gatekeeping specifically, on increasing referral rates^8,13 and the unequal distribution of managed care plans between the 2 samples. Unweighted visits yielded a sample size of 37,145; of these, 11,676 met the selection criteria.

The proportions of office visits referred were compared overall and by age, sex, and health condition. The 10 most frequently referred conditions in the study sample were used for the condition-specific referral rate assessments. Statistical significance was assessed by the chi-square statistic.

Results

Descriptive information on the 141 family physician sample is presented in Table 1. Physicians spent an average of 51.3 hours per week in their jobs. About 68% of their time was devoted to direct patient care. In most practices, a staff member coordinated administrative aspects of specialty referrals; 20% permitted patients to request a referral by leaving a voice mail message.

TABLE 1
FAMILY PHYSICIAN STUDY SAMPLE

Frequency of referral

The 141 family physicians had 34,519 office visits and made 2165 referrals during 1771 practice-days; 5.1% of office visits were referred. Physicians saw an average of 19.7 patients per day (range 7.0 to 48.4) and made 1.23 referrals per full practice-day (range 0 to 3.90). Referrals made during telephone conversations with patients accounted for 18.9% of all referrals made by physicians (range 0% to 100% per physician).

An additional 369 referrals (a rate of 0.21 referrals per practice-day) were made by staff. Overall, 68% of all referrals were made by physicians during visits with patients, 18% by physicians during telephone conversations with patients, 11% by staff with physician input, and 3% by staff without physician input. In 43.6% of referrals made during telephone conversations with patients, the telephone encounter was the first presentation to medical care for the health problem.

We compared percentages of office visits in which a referral was made in the study sample with percentages of such referrals by family physicians from the NAMCS surveys (1989–1994). The overall percentages did not significantly differ between the 2 groups (4.0% vs 3.7%, P > .05). Although physicians in the study sample were statistically less likely than NAMCS counterparts to refer children (1.6% vs 2.5%, P = .030), more likely to refer the elderly (4.8% vs 4.1%, P = .045), and more likely to refer females (4.1% vs 3.9%, P = .009), these differences were small. There were no differences between the groups in condition-specific referral rates. In sum, these results show that patients in the ASPN sample were equally likely to be referred as those in the NAMCS sample.

Reasons for referral

Table 2 shows the distribution of physicians’ reasons for making the referral. Physicians endorsed a mean of 1.8 different reasons for making the typical referral. Although patients requested to see a specialist for 13.6% of referrals, physicians recorded patient request as the only reason for referral just 1.1%.

We compared referrals made for uncommon conditions (lowest tertile of practice-prevalence) with common conditions (highest practice-prevalence tertile). The calculation of practice-prevalence was based on prior research: the numerator was visits made for the index condition, and the denominator was all visits in the sample.⁵ Uncommon conditions were more likely to be referred for medical management (38.5% vs 25.4%, P < .001), patient request (19.8% vs 12.3%, P = .005), and specialist request (4.9% vs 2.1%, P = .021). Common conditions were more likely to be sent to specialists because of failed current therapy (13.6% vs 3.8%, P < .001) and endoscopy (4.3% vs 0.5%, P = .013). There were no significant differences between the 2 groups in the chances of referral for advice on either diagnosis or treatment.

TABLE 2
REASONS FOR REFERRAL

Conditions referred

Table 3 presents condition-specific referral rates and the 2 most common types of specialists referred to for the top 25 referred health problems. (A complete listing of these data for all conditions reported by study physicians can be found in Table W1.) The 50 most commonly referred health problems accounted for 76% of all referrals made during office visits. Signs or symptoms accounted for 22.4% of all referrals. Condition-specific referral rates varied from a low of 1.9% for patients with otitis media to a high of 45.7% of visits referred for patients with cholelithiasis or cholecystitis. This range in referral rates translates into 24-fold variation in the chances of referral during an office visit based solely on the presenting problem.

TABLE 3
NUMBER OF OFFICE VISITS, REFERRAL RATES, AND SPECIALISTS REFERRED TO FOR TOP 15 REFERRED CONDITIONS*

Specialist selection

Referrals were made most often to surgical subspecialists (45.4%), followed by medical subspecialists (31.0%), nonphysician clinicians (12.1%), obstetriciangynecologists (ob/gyns) (4.6%), mental health professionals (4.2%), other physicians (2.0%), and generalists (0.8%). The 5 most common specialists to whom patients were referred were orthopedic surgeons (12.1%), general surgeons (9.1%), otolaryngologists (6.9%), gastroenterologists (6.6%), and dermatologists (6.0%). Among male patients, referral to urologists was the second most common type; among female patients, referral to ob/gyns was the third most common type.

Mental health referrals were made predominantly to psychologists (2.1% of all referrals), followed by psychiatrists (1.3%) and social workers (0.4%). The most common types of nonphysician clinicians referred to were physical therapists (4.5%), podiatrists (3.0%), nutritionists (1.5%), and audiologists (1.2%).

Referring physicians recommended a specific specialist to the patient for 86.2% of referrals. In descending rank order according to the mean importance rating (range 1 to 3), the reasons for selecting a particular specialist were personal knowledge of the specialist (2.6), quality of prior feedback (2.5), technical capacity (2.3), appointment availability (2.0), patient’s request (1.6), requirements of patient’s health plan (1.6), and proximity of the specialist to the patient’s home (1.6).

Table 4 shows the 3 most common health problems referred to 10 types of specialists. (An expanded version of this table that includes 29 specialists can be found in Table W2.) The majority of referrals for each type of specialist were for 1 to 3 health problems. Family physicians made 17.1% of all referrals to practitioners within their practices. Intrapractice referrals were significantly higher than the overall average for audiologists (40.0%, P = .031), nutritionists (45.2%, P = .004), and psychologists (46.3%, P < .001) and were lower for gastroenterologists (9.3%, P = .022) and rheumatologists (4.0%, P = .005).

TABLE 4
THREE MOST COMMON CONDITIONS REFERRED TO SELECTED SPECIALISTS*

Discussion

This study shows that family physicians manage 95% of office visits without specialty referral. About one third of referrals made from primary care practices occur during encounters other than office visits. Referrals made by staff or during telephone conversations may be part of an integrated sequence of contacts between patients and physicians. Nonetheless, assisting patients in selecting a specialist, transferring relevant patient information, and scheduling specialty appointments (referral coordination activities) are more difficult to perform when patients are not seen in the office,¹⁴ because time is limited and integrating care is poorly reimbursed, if at all. When such referral decisions are made appropriately, they provide an efficient mechanism for decreasing workload in a busy primary care practice. Inappropriately made, they can lead to increased expense, unnecessary time spent with specialists, and poorly coordinated care.

We found that the rates of referral were substantially different among the most commonly referred conditions. Prior work has shown that the frequency with which conditions present to primary care physicians explains about 75% of the variation in condition-specific referral rates.⁵ The mix and severity of comorbidities are important determinants of annual patient referral rates^15,16 and the chances of referral during a visit.⁵ Thus, the epidemiology of morbidity among a patient population is a critical factor that defines the boundaries between primary care physicians and specialists. The appreciation of these clinical determinants is crucial for any valid assessment of primary care physicians’ referral patterns.

Limitations

The study’s focus was on new referral decisions made by physicians to other practitioners. No information is provided about ongoing, long-term referrals in which the patient was already under the care of a specialist. The low rates of referral for conditions such as diabetes may be a consequence of this limitation. Patients with diabetes may already have been under the care of a specialist, thereby generating few new referrals. It is also important to note that even in health plans with gatekeeping arrangements, patients self-refer to specialty care¹³; this study did not include any information on self-referral. Patient self-referral appears to be most likely among sick patients, those with established relationships with a specialist, and patients who do not have a good relationship with a primary care physician.¹⁷

We did not obtain information on the number of telephone calls fielded by physicians each day. Without these data, we were unable to determine whether our methods had failed to capture some telephone referrals or to calculate telephone referral rates. In this study, family physicians made 18.9% of all referrals during telephone conversations, in contrast with pediatricians in another study⁴ who made 27.5% of all referrals by telephone. The difference in these proportions is not large and is probably explained by pediatricians’ greater use of the telephone for patient care.

It could be argued that the volunteer physicians in this study systematically differ from the typical family physician. The average number of visits per day among study physicians (19.7) is similar to a national estimate of 19.9 visits/day for family physicians in single specialty group practices.¹⁸ Furthermore, we found similar probabilities of referral overall and for the 10 most commonly referred conditions between study physicians and a national sample, suggesting that referral propensities between the 2 groups were similar.

Why family physicians refer

No value judgments can be made about the appropriateness of physicians’ reasons for referral. Physicians most commonly referred because they were uncertain about diagnosis or treatment and sought advice from another practitioner. For about 1 in 5 referrals, physicians recorded only a sign or symptom as the diagnosis, suggesting a reasonably high level of diagnostic uncertainty. Physicians’ tolerance of uncertainty varies markedly,¹⁹ making it difficult to judge questions about appropriateness of referrals that are made to reduce this uncertainty.

Another important reason for referral was that physicians deemed the management of the health problem to be outside their scope of practice. Physicians were more likely to refer a patient with a common problem after trying out a course of treatment than was the case for uncommon problems that were more likely to be referred for medical management.

Patients may raise the topic of possible referral. When physicians agree that referral is indicated, they almost always find other reasons for making the referral. Alternatively, physicians might make a decision to refer and justify it in part as being a result of patient request. Discussions on whether a referral is needed are common in primary care. Among referrals made in an Israeli family practice network, patients raised the topic of possible referral in 27% of cases.²⁰ In a study of 856 internal medicine visits, 45% of patients indicated some desire to discuss the need for referral with their physician; however, physicians recognized these desires only about half the time.²¹

Selecting a specialist

Our results show that primary care physicians prefer to send their patients to specialists with whom they have developed a relationship. Physicians in this study maintained a high level of involvement in specialist selection, providing patients with the name of a specific practitioner for 86.2% of referrals. The most important factor in selecting a specialist in our study was the same as that found nearly 20 years ago by Ludke¹: personal knowledge of the specialist. Physicians’ dissatisfaction with the specialty referral process in managed care settings^22,23 could be a result of their reduced choice of specialists with whom they have forged personal relationships.

Slightly more than 1 in 6 referrals were made to specialists in the referring physician’s practice, consistent with movement of primary care physicians into multispecialty groups. Whether intrapractice referral holds any advantage over referrals outside the practice, such as better coordination and appointment adherence, awaits future study.

Our results show that physicians must not only select a specific practitioner but also choose among different types of practitioners. Some patients were sent to nonphysician clinicians and physicians (eg, podiatrists and orthopedic surgeons for acquired foot deformity), whereas others were sent to medical or surgical subspecialists (eg, nephrologists and urologists for urinary tract symptoms). These patterns are likely to reflect the need for multidisciplinary specialty care for some conditions. For instance, patients with diabetes may see an ophthalmologist for retinopathy evaluations and an endocrinologist for medical management consultation. For some conditions, there appears to be considerable uncertainty regarding the boundaries between specialists.²⁴ Should a patient with a skin mass be sent to a general surgeon, a dermatologist, or a plastic surgeon? When should a patient with allergic rhinitis be sent to an allergist and when to an otolaryngologist? These referral patterns may reflect local care practices and specialist availability. They may also be a consequence of a surplus of specialists in this country and competition for patients.

In a survey of family physicians that was performed in the late 1980s, respondents reported that they were more likely to refer to internal medicine subspecialists than internists for adults, but preferred general pediatricians over pediatric subspecialists.²⁵ Our findings suggest that the trend for adult patients remains, but there has been a shift away from general pediatricians toward subspecialists for pediatric referrals. These new patterns may be a consequence of greater availability of pediatric subspecialists, greater exposure of family physicians to pediatric consultants, and a larger share of family physicians who have completed residency training.

Implications for physician training

Fifty conditions accounted for 76% of all specialty referrals made during office visits in this study. Interactions with most types of specialists are generally limited to a few conditions; 3 health problems accounted for more than half of referrals to most specialties. Educators should ensure that these commonly referred conditions are emphasized in curricula that provide family physicians with the skills necessary to expand their scope of practice, when appropriate; determine when a patient should be referred; and identify the type of practitioner to whom the patient should be sent.

Physicians in training should be taught the skills required to recognize the boundaries of their clinical uncertainty and scopes of practice. A challenge for educators is to assist trainees in determining when to tolerate clinical uncertainty while employing a watchful waiting approach and when to initiate a more aggressive evaluation, including when to obtain specialty referral. Modes of implementing these approaches are likely to differ across conditions. Thus, it makes sense in physician training to place the greatest emphasis on conditions for which family physicians commonly refer.

Under certain circumstances, patient request for a specialty consultation may be a sufficient and legitimate reason for referral. For example, as we found in this study, patients with uncommon health problems may seek reassurance from specialists skilled in the management of their specific condition. Managing access to specialists, particularly when the physician is acting as an administrative gatekeeper to referrals, can be challenging.²¹ When doctors and patients disagree on the need for referral, patients may become dissatisfied with their health care²⁶ and decide to self-refer to specialty care.¹⁷ In consideration of the increasing complexity of medical care, developing skills that help physicians discuss and negotiate access to specialized services with both patients and specialists has never been more timely.

Acknowledgments

This study was funded by grant no. R01 HS09377 from the Agency for Healthcare Research and Quality. James Werner and Laurie Vorel provided technical assistance with data collection and project implementation. Many physicians collected data for this study. Their time and devotion were invaluable to the success of this study. These physicians are listed by the states in which they practice. Arizona: Scott Ekdahl, DO; Arkansas: John Scott, MD; California: Andrew Ness, MD; Colorado: Howard Corren, MD; Nell Davis, MD; Timothy Dudley, MD; Audrey Farley, MD; Tillman Farley, MD; Charles Kay, MD; Joan FAMILY PHYSICIANS’ REFERRAL DECISIONS MacEachen, MD; George Maxted, MD; John Miller, MD; Kathy Miller, MD; Steven Milligan, MD; Frank Reed, MD; Louise Schottsteadt, MD; Lynne Spicer, MD; Laura Stein, MD; Lynn Strange, MD; Dan Sullivan, MD; Georgia: Linda Casteel, MD; Randy Cronic, MD; Bruno Denis, MD; Keith Ellis, MD; Kelly Erola, MD; Craig Fabel, MD; Russell Leubbert, DO; Richard Liotta, DO; Mark Majoch, MD; David Najjar, MD; James Snow, DO; Roslyn Taylor, MD; Illinois: Steven Lidvall, MD; Anna Meenan, MD; Eduardo Scholcoff, MD; Loyd Wollstadt, MD; Indiana: Paul Daluga, MD; Steven Phillipson, MD; Iowa: Ken Miller, MD; Janet Ryan, MD; Kansas: Wendell Ellis, DO; John R. Eplee, MD; Robert Moser, MD; Daniel Sontheimer, MD; Louisiana: Linda Stewart, MD; Michigan: Linda French, MD; John Hickner, MD; Minnesota: Ravi Balasubraman, MD; Dave Bucher, MD; William Davis, MD; Richard Gebhart, MD; Katie Guthrie, MD; Anthony Jaspers, MD; Timothy Komoto, MD; Glenn McCarty, DO; Stephen Mitrione, MD; Thomas Retzinger, MD; Paul Spilseth, MD; Ashlesha Tamboli, MD; Montana: Curt Kurtz, MD; Nevada: Coleen Lyons, MD; New Hampshire: Richard Douglass, MD; Paul Friedrichs, MD; Peter Hope, MD; Jonathan Mishcon, MD; New Jersey: John Orzano, MD; Winifred Waldron, MD; New York: Carmella Abraham, MD; R. Eugene Bailey, MD; Lorne Becker, MD; John DeSimone, MD; Miguel Diaz, MD; Rebecca Elliott, MD; John Glennon, MD; James Greenwald, MD; Glenn Griffin, MD; Eileen Hoffman, MD; L. Thomas Wolff, MD; North Carolina: Ed Bujold, MD; Thomas Detesco, MD; Dave Rogers, MD; Phil Sherrod, MD; Oklahoma: Laura Miller, DO; Mike Pontious, MD; Oregon: Douglas Eliason, DO; L.J. Fagnan, MD; Jerry Flaming, DO; Tom Flaming, DO; Jeffrey Humphrey, DO; Michael Kelber, MD; John Sattenspiel, MD; Pennsylvania: John Farmer, DO; Penitha Williams, MD; South Dakota: Fred Thanel, MD; Tennessee: Dan Brewer, MD; Michael Hartsell, MD; R. Louis Murphy, MD; John Parham, MD; Texas: Michael Averitt, DO; Sharon Barber, MD; Kim Patrick Bolton, MD; Robert Cortes, MD; Paul Gerdes, MD; Robert Henry, DO; Michael Kirkpatrick, MD; John Manning, MD; Shane Maxwell, MD; Luis Moreno, MD; Larry G. Padget, MD; Peter Sullivan, MD; Utah: Scott Endsley, MD; David Flinders, MD; Jim Giovino, MD; Eric Hogenson, MD; Dwayne Roberts, MD; Virginia: Duane Lawrence, MD; James Ledwith, MD; June Tunstall, MD; George Wortley, MD; Washington: John Anderson, MD; Elizabeth Wise, MD; West Virginia: Dan Doyle, MD; J. Michael Herr, DO; Wisconsin: Richard Anstett, MD, PhD; Walter Boisvert, MD; Lea Cornell, MD; Anne Eglash, MD; Rod Erickson, MD; Tom Frisby, MD; Terry Hankey, MD; Kevin Jessen, MD; Dan Landdeck, MD; Dave Lonsdorf, MD; Michael Pace, MD; Michael Saunders, MD; Catherine Soderqueist, MD; Jon Temte, MD; Vince Winklerprins, MD; Brian Woody, MD.

Conventionally, primary care physicians decided when to refer and to whom a patient should be referred.^1,2 Specialists’ assistance was sought for diagnostic or therapeutic dilemmas,^3,4 management of conditions that presented too infrequently to maintain clinical competence,⁵ and specialized procedures that fell outside a physician’s scope of practice.^3,4 In some cases, physicians referred because patients requested to see a specialist.^1,4

The reorganization of health care over the past few decades has dramatically altered the interactions between primary care physicians and specialists. The growth in multispecialty group practice arrangements⁶ has led to formal, organizationally defined linkages between practitioners. Managed health plans and medical groups⁷ encourage primary care physicians’ control over the referral process through such mechanisms as specialty referral authorizations, financial disincentives for making a referral, performance assessment of referral patterns, and referral guidelines. These changes have transformed a once-informal process into one rife with administrative restrictions on referral decision making.

The Ambulatory Sentinel Practice Network (ASPN) Referral Study was designed to describe and analyze primary care physicians’ referral decisions and their outcomes in the context of a changing health care system in the United States. The study occurred in the ASPN and other regional practice-based research networks. This report examines primary care physicians’ referral decisions. We conceptualized the referral decision as occurring in 2 phases: whether to refer followed by to whom to refer.²

Methods

Physician sample

Physicians were recruited from March 1997 to May 1998. Recruitment activities were directed to all physician members of ASPN, physicians affiliated with the Medical Group Management Association, local and regional networks (Minnesota Academy of Family Physicians Research Network, the Wisconsin Research Network, the Dartmouth Primary Cooperative Research Network (COOP), and the larger community of primary care physicians. The study was publicized via direct mailings to physicians, articles and notices in practice-based research network newsletters and journals, and presentations at conferences. Contact with physicians expressing interest was made by telephone.

Physicians were included in the study if they practiced in the United States and were not in residency or fellowship training. Of all physicians contacted, 342 expressed interest in the study and 182 completed some aspect of data collection. A total of 141 family physicians, 12 internists, and 1 pediatrician completed all phases of data collection. In this study, the 141 family physicians (41% members of ASPN) formed the physician sample. They delivered health care in 87 practices located in 31 states.

Procedures

Study protocols and materials, based on a similar practice-based research study conducted with pediatricians, were reviewed and approved by the Committee on Human Research of the Johns Hopkins School of Public Health and the Colorado Multi-Institutional Review Board.^4,8,9 We conducted a pilot test in 5 practices; this test led to further refinements of methods and questionnaires.

Data collection occurred from September 1997 to February 1999, with 94% of physicians collecting data in 1998 only. Before beginning data collection, physicians completed a questionnaire concerning their practices and personal characteristics. Each practice selected a coordinator who communicated with research staff, learned study protocols, trained office staff and physicians, and monitored data quality. Coordinators completed a questionnaire about the organizational and financial components of their practice. They kept a log of all visits made during 15 consecutive business days and occurring during regularly scheduled office hours. A business day was defined as a half or full work day, provided that the physician held routine office hours. Each patient’s date of birth (5% missing), sex (2% missing), and principal diagnosis (5% missing) were recorded.

The coordinator kept another log of all referrals made by physicians, nurses, and other office staff. Referrals made during telephone conversations with patients were included. A referral was defined as a recommendation that a patient have a face-to-face encounter with another practitioner. We excluded referrals made to laboratories, radiologic facilities, emergency departments, hospitals for inpatient admission, and “curbside consultations” (ie, when the referring physician obtains advice from a specialist but does not send the patient for a visit).

A medical record abstractor assigned ICD-9-CM codes to diagnoses provided by office staff. We matched ICD codes to an expanded set of diagnosis clusters (EDCs). EDCs group ICD codes into clinically homogeneous categories using the methods developed by Schneeweiss.¹⁰ (For more information on EDCs, see http://acg.jhsph.edu.)

When physicians made a referral, they completed a questionnaire (response rate 93.9%) with items concerning the referral decision. Reasons for referral were based on our previously developed taxonomy used in a pediatric referral study⁴ and focus groups of family physicians convened during an annual ASPN convocation.

At the study’s conclusion, physicians received a report that compared their referral practice patterns with those of the entire sample. To defray office expenses associated with data collection, each practice was given a $100 stipend in addition to $5 for each physician referral.

Generalizability analysis

We compared referral rates of the study sample with the National Ambulatory Medical Care Survey (NAMCS), a nationally representative sample of office visits made to family physicians.^11,12 We pooled surveys from 1989 to 1994, inclusive, when the majority of the items in the survey instruments remained unchanged.¹³ (The 1995–1999 surveys did not contain information on whether the visit led to referral.)

We selected visits made by patients enrolled in non-HMO health plans (NAMCS) and health plans that had neither capitated primary care physician payment nor gatekeeping arrangements (study sample). This was done because of the known effect of managed care in general, and of gatekeeping specifically, on increasing referral rates^8,13 and the unequal distribution of managed care plans between the 2 samples. Unweighted visits yielded a sample size of 37,145; of these, 11,676 met the selection criteria.

The proportions of office visits referred were compared overall and by age, sex, and health condition. The 10 most frequently referred conditions in the study sample were used for the condition-specific referral rate assessments. Statistical significance was assessed by the chi-square statistic.

Results

Descriptive information on the 141 family physician sample is presented in Table 1. Physicians spent an average of 51.3 hours per week in their jobs. About 68% of their time was devoted to direct patient care. In most practices, a staff member coordinated administrative aspects of specialty referrals; 20% permitted patients to request a referral by leaving a voice mail message.

TABLE 1
FAMILY PHYSICIAN STUDY SAMPLE

Frequency of referral

The 141 family physicians had 34,519 office visits and made 2165 referrals during 1771 practice-days; 5.1% of office visits were referred. Physicians saw an average of 19.7 patients per day (range 7.0 to 48.4) and made 1.23 referrals per full practice-day (range 0 to 3.90). Referrals made during telephone conversations with patients accounted for 18.9% of all referrals made by physicians (range 0% to 100% per physician).

An additional 369 referrals (a rate of 0.21 referrals per practice-day) were made by staff. Overall, 68% of all referrals were made by physicians during visits with patients, 18% by physicians during telephone conversations with patients, 11% by staff with physician input, and 3% by staff without physician input. In 43.6% of referrals made during telephone conversations with patients, the telephone encounter was the first presentation to medical care for the health problem.

We compared percentages of office visits in which a referral was made in the study sample with percentages of such referrals by family physicians from the NAMCS surveys (1989–1994). The overall percentages did not significantly differ between the 2 groups (4.0% vs 3.7%, P > .05). Although physicians in the study sample were statistically less likely than NAMCS counterparts to refer children (1.6% vs 2.5%, P = .030), more likely to refer the elderly (4.8% vs 4.1%, P = .045), and more likely to refer females (4.1% vs 3.9%, P = .009), these differences were small. There were no differences between the groups in condition-specific referral rates. In sum, these results show that patients in the ASPN sample were equally likely to be referred as those in the NAMCS sample.

Reasons for referral

Table 2 shows the distribution of physicians’ reasons for making the referral. Physicians endorsed a mean of 1.8 different reasons for making the typical referral. Although patients requested to see a specialist for 13.6% of referrals, physicians recorded patient request as the only reason for referral just 1.1%.

We compared referrals made for uncommon conditions (lowest tertile of practice-prevalence) with common conditions (highest practice-prevalence tertile). The calculation of practice-prevalence was based on prior research: the numerator was visits made for the index condition, and the denominator was all visits in the sample.⁵ Uncommon conditions were more likely to be referred for medical management (38.5% vs 25.4%, P < .001), patient request (19.8% vs 12.3%, P = .005), and specialist request (4.9% vs 2.1%, P = .021). Common conditions were more likely to be sent to specialists because of failed current therapy (13.6% vs 3.8%, P < .001) and endoscopy (4.3% vs 0.5%, P = .013). There were no significant differences between the 2 groups in the chances of referral for advice on either diagnosis or treatment.

TABLE 2
REASONS FOR REFERRAL

Conditions referred

Table 3 presents condition-specific referral rates and the 2 most common types of specialists referred to for the top 25 referred health problems. (A complete listing of these data for all conditions reported by study physicians can be found in Table W1.) The 50 most commonly referred health problems accounted for 76% of all referrals made during office visits. Signs or symptoms accounted for 22.4% of all referrals. Condition-specific referral rates varied from a low of 1.9% for patients with otitis media to a high of 45.7% of visits referred for patients with cholelithiasis or cholecystitis. This range in referral rates translates into 24-fold variation in the chances of referral during an office visit based solely on the presenting problem.

TABLE 3
NUMBER OF OFFICE VISITS, REFERRAL RATES, AND SPECIALISTS REFERRED TO FOR TOP 15 REFERRED CONDITIONS*

Specialist selection

Referrals were made most often to surgical subspecialists (45.4%), followed by medical subspecialists (31.0%), nonphysician clinicians (12.1%), obstetriciangynecologists (ob/gyns) (4.6%), mental health professionals (4.2%), other physicians (2.0%), and generalists (0.8%). The 5 most common specialists to whom patients were referred were orthopedic surgeons (12.1%), general surgeons (9.1%), otolaryngologists (6.9%), gastroenterologists (6.6%), and dermatologists (6.0%). Among male patients, referral to urologists was the second most common type; among female patients, referral to ob/gyns was the third most common type.

Mental health referrals were made predominantly to psychologists (2.1% of all referrals), followed by psychiatrists (1.3%) and social workers (0.4%). The most common types of nonphysician clinicians referred to were physical therapists (4.5%), podiatrists (3.0%), nutritionists (1.5%), and audiologists (1.2%).

Referring physicians recommended a specific specialist to the patient for 86.2% of referrals. In descending rank order according to the mean importance rating (range 1 to 3), the reasons for selecting a particular specialist were personal knowledge of the specialist (2.6), quality of prior feedback (2.5), technical capacity (2.3), appointment availability (2.0), patient’s request (1.6), requirements of patient’s health plan (1.6), and proximity of the specialist to the patient’s home (1.6).

Table 4 shows the 3 most common health problems referred to 10 types of specialists. (An expanded version of this table that includes 29 specialists can be found in Table W2.) The majority of referrals for each type of specialist were for 1 to 3 health problems. Family physicians made 17.1% of all referrals to practitioners within their practices. Intrapractice referrals were significantly higher than the overall average for audiologists (40.0%, P = .031), nutritionists (45.2%, P = .004), and psychologists (46.3%, P < .001) and were lower for gastroenterologists (9.3%, P = .022) and rheumatologists (4.0%, P = .005).

TABLE 4
THREE MOST COMMON CONDITIONS REFERRED TO SELECTED SPECIALISTS*

Discussion

This study shows that family physicians manage 95% of office visits without specialty referral. About one third of referrals made from primary care practices occur during encounters other than office visits. Referrals made by staff or during telephone conversations may be part of an integrated sequence of contacts between patients and physicians. Nonetheless, assisting patients in selecting a specialist, transferring relevant patient information, and scheduling specialty appointments (referral coordination activities) are more difficult to perform when patients are not seen in the office,¹⁴ because time is limited and integrating care is poorly reimbursed, if at all. When such referral decisions are made appropriately, they provide an efficient mechanism for decreasing workload in a busy primary care practice. Inappropriately made, they can lead to increased expense, unnecessary time spent with specialists, and poorly coordinated care.

We found that the rates of referral were substantially different among the most commonly referred conditions. Prior work has shown that the frequency with which conditions present to primary care physicians explains about 75% of the variation in condition-specific referral rates.⁵ The mix and severity of comorbidities are important determinants of annual patient referral rates^15,16 and the chances of referral during a visit.⁵ Thus, the epidemiology of morbidity among a patient population is a critical factor that defines the boundaries between primary care physicians and specialists. The appreciation of these clinical determinants is crucial for any valid assessment of primary care physicians’ referral patterns.

Limitations

The study’s focus was on new referral decisions made by physicians to other practitioners. No information is provided about ongoing, long-term referrals in which the patient was already under the care of a specialist. The low rates of referral for conditions such as diabetes may be a consequence of this limitation. Patients with diabetes may already have been under the care of a specialist, thereby generating few new referrals. It is also important to note that even in health plans with gatekeeping arrangements, patients self-refer to specialty care¹³; this study did not include any information on self-referral. Patient self-referral appears to be most likely among sick patients, those with established relationships with a specialist, and patients who do not have a good relationship with a primary care physician.¹⁷

We did not obtain information on the number of telephone calls fielded by physicians each day. Without these data, we were unable to determine whether our methods had failed to capture some telephone referrals or to calculate telephone referral rates. In this study, family physicians made 18.9% of all referrals during telephone conversations, in contrast with pediatricians in another study⁴ who made 27.5% of all referrals by telephone. The difference in these proportions is not large and is probably explained by pediatricians’ greater use of the telephone for patient care.

It could be argued that the volunteer physicians in this study systematically differ from the typical family physician. The average number of visits per day among study physicians (19.7) is similar to a national estimate of 19.9 visits/day for family physicians in single specialty group practices.¹⁸ Furthermore, we found similar probabilities of referral overall and for the 10 most commonly referred conditions between study physicians and a national sample, suggesting that referral propensities between the 2 groups were similar.

Why family physicians refer

No value judgments can be made about the appropriateness of physicians’ reasons for referral. Physicians most commonly referred because they were uncertain about diagnosis or treatment and sought advice from another practitioner. For about 1 in 5 referrals, physicians recorded only a sign or symptom as the diagnosis, suggesting a reasonably high level of diagnostic uncertainty. Physicians’ tolerance of uncertainty varies markedly,¹⁹ making it difficult to judge questions about appropriateness of referrals that are made to reduce this uncertainty.

Another important reason for referral was that physicians deemed the management of the health problem to be outside their scope of practice. Physicians were more likely to refer a patient with a common problem after trying out a course of treatment than was the case for uncommon problems that were more likely to be referred for medical management.

Patients may raise the topic of possible referral. When physicians agree that referral is indicated, they almost always find other reasons for making the referral. Alternatively, physicians might make a decision to refer and justify it in part as being a result of patient request. Discussions on whether a referral is needed are common in primary care. Among referrals made in an Israeli family practice network, patients raised the topic of possible referral in 27% of cases.²⁰ In a study of 856 internal medicine visits, 45% of patients indicated some desire to discuss the need for referral with their physician; however, physicians recognized these desires only about half the time.²¹

Selecting a specialist

Our results show that primary care physicians prefer to send their patients to specialists with whom they have developed a relationship. Physicians in this study maintained a high level of involvement in specialist selection, providing patients with the name of a specific practitioner for 86.2% of referrals. The most important factor in selecting a specialist in our study was the same as that found nearly 20 years ago by Ludke¹: personal knowledge of the specialist. Physicians’ dissatisfaction with the specialty referral process in managed care settings^22,23 could be a result of their reduced choice of specialists with whom they have forged personal relationships.

Slightly more than 1 in 6 referrals were made to specialists in the referring physician’s practice, consistent with movement of primary care physicians into multispecialty groups. Whether intrapractice referral holds any advantage over referrals outside the practice, such as better coordination and appointment adherence, awaits future study.

Our results show that physicians must not only select a specific practitioner but also choose among different types of practitioners. Some patients were sent to nonphysician clinicians and physicians (eg, podiatrists and orthopedic surgeons for acquired foot deformity), whereas others were sent to medical or surgical subspecialists (eg, nephrologists and urologists for urinary tract symptoms). These patterns are likely to reflect the need for multidisciplinary specialty care for some conditions. For instance, patients with diabetes may see an ophthalmologist for retinopathy evaluations and an endocrinologist for medical management consultation. For some conditions, there appears to be considerable uncertainty regarding the boundaries between specialists.²⁴ Should a patient with a skin mass be sent to a general surgeon, a dermatologist, or a plastic surgeon? When should a patient with allergic rhinitis be sent to an allergist and when to an otolaryngologist? These referral patterns may reflect local care practices and specialist availability. They may also be a consequence of a surplus of specialists in this country and competition for patients.

In a survey of family physicians that was performed in the late 1980s, respondents reported that they were more likely to refer to internal medicine subspecialists than internists for adults, but preferred general pediatricians over pediatric subspecialists.²⁵ Our findings suggest that the trend for adult patients remains, but there has been a shift away from general pediatricians toward subspecialists for pediatric referrals. These new patterns may be a consequence of greater availability of pediatric subspecialists, greater exposure of family physicians to pediatric consultants, and a larger share of family physicians who have completed residency training.

Implications for physician training

Fifty conditions accounted for 76% of all specialty referrals made during office visits in this study. Interactions with most types of specialists are generally limited to a few conditions; 3 health problems accounted for more than half of referrals to most specialties. Educators should ensure that these commonly referred conditions are emphasized in curricula that provide family physicians with the skills necessary to expand their scope of practice, when appropriate; determine when a patient should be referred; and identify the type of practitioner to whom the patient should be sent.

Physicians in training should be taught the skills required to recognize the boundaries of their clinical uncertainty and scopes of practice. A challenge for educators is to assist trainees in determining when to tolerate clinical uncertainty while employing a watchful waiting approach and when to initiate a more aggressive evaluation, including when to obtain specialty referral. Modes of implementing these approaches are likely to differ across conditions. Thus, it makes sense in physician training to place the greatest emphasis on conditions for which family physicians commonly refer.

Under certain circumstances, patient request for a specialty consultation may be a sufficient and legitimate reason for referral. For example, as we found in this study, patients with uncommon health problems may seek reassurance from specialists skilled in the management of their specific condition. Managing access to specialists, particularly when the physician is acting as an administrative gatekeeper to referrals, can be challenging.²¹ When doctors and patients disagree on the need for referral, patients may become dissatisfied with their health care²⁶ and decide to self-refer to specialty care.¹⁷ In consideration of the increasing complexity of medical care, developing skills that help physicians discuss and negotiate access to specialized services with both patients and specialists has never been more timely.

Acknowledgments

This study was funded by grant no. R01 HS09377 from the Agency for Healthcare Research and Quality. James Werner and Laurie Vorel provided technical assistance with data collection and project implementation. Many physicians collected data for this study. Their time and devotion were invaluable to the success of this study. These physicians are listed by the states in which they practice. Arizona: Scott Ekdahl, DO; Arkansas: John Scott, MD; California: Andrew Ness, MD; Colorado: Howard Corren, MD; Nell Davis, MD; Timothy Dudley, MD; Audrey Farley, MD; Tillman Farley, MD; Charles Kay, MD; Joan FAMILY PHYSICIANS’ REFERRAL DECISIONS MacEachen, MD; George Maxted, MD; John Miller, MD; Kathy Miller, MD; Steven Milligan, MD; Frank Reed, MD; Louise Schottsteadt, MD; Lynne Spicer, MD; Laura Stein, MD; Lynn Strange, MD; Dan Sullivan, MD; Georgia: Linda Casteel, MD; Randy Cronic, MD; Bruno Denis, MD; Keith Ellis, MD; Kelly Erola, MD; Craig Fabel, MD; Russell Leubbert, DO; Richard Liotta, DO; Mark Majoch, MD; David Najjar, MD; James Snow, DO; Roslyn Taylor, MD; Illinois: Steven Lidvall, MD; Anna Meenan, MD; Eduardo Scholcoff, MD; Loyd Wollstadt, MD; Indiana: Paul Daluga, MD; Steven Phillipson, MD; Iowa: Ken Miller, MD; Janet Ryan, MD; Kansas: Wendell Ellis, DO; John R. Eplee, MD; Robert Moser, MD; Daniel Sontheimer, MD; Louisiana: Linda Stewart, MD; Michigan: Linda French, MD; John Hickner, MD; Minnesota: Ravi Balasubraman, MD; Dave Bucher, MD; William Davis, MD; Richard Gebhart, MD; Katie Guthrie, MD; Anthony Jaspers, MD; Timothy Komoto, MD; Glenn McCarty, DO; Stephen Mitrione, MD; Thomas Retzinger, MD; Paul Spilseth, MD; Ashlesha Tamboli, MD; Montana: Curt Kurtz, MD; Nevada: Coleen Lyons, MD; New Hampshire: Richard Douglass, MD; Paul Friedrichs, MD; Peter Hope, MD; Jonathan Mishcon, MD; New Jersey: John Orzano, MD; Winifred Waldron, MD; New York: Carmella Abraham, MD; R. Eugene Bailey, MD; Lorne Becker, MD; John DeSimone, MD; Miguel Diaz, MD; Rebecca Elliott, MD; John Glennon, MD; James Greenwald, MD; Glenn Griffin, MD; Eileen Hoffman, MD; L. Thomas Wolff, MD; North Carolina: Ed Bujold, MD; Thomas Detesco, MD; Dave Rogers, MD; Phil Sherrod, MD; Oklahoma: Laura Miller, DO; Mike Pontious, MD; Oregon: Douglas Eliason, DO; L.J. Fagnan, MD; Jerry Flaming, DO; Tom Flaming, DO; Jeffrey Humphrey, DO; Michael Kelber, MD; John Sattenspiel, MD; Pennsylvania: John Farmer, DO; Penitha Williams, MD; South Dakota: Fred Thanel, MD; Tennessee: Dan Brewer, MD; Michael Hartsell, MD; R. Louis Murphy, MD; John Parham, MD; Texas: Michael Averitt, DO; Sharon Barber, MD; Kim Patrick Bolton, MD; Robert Cortes, MD; Paul Gerdes, MD; Robert Henry, DO; Michael Kirkpatrick, MD; John Manning, MD; Shane Maxwell, MD; Luis Moreno, MD; Larry G. Padget, MD; Peter Sullivan, MD; Utah: Scott Endsley, MD; David Flinders, MD; Jim Giovino, MD; Eric Hogenson, MD; Dwayne Roberts, MD; Virginia: Duane Lawrence, MD; James Ledwith, MD; June Tunstall, MD; George Wortley, MD; Washington: John Anderson, MD; Elizabeth Wise, MD; West Virginia: Dan Doyle, MD; J. Michael Herr, DO; Wisconsin: Richard Anstett, MD, PhD; Walter Boisvert, MD; Lea Cornell, MD; Anne Eglash, MD; Rod Erickson, MD; Tom Frisby, MD; Terry Hankey, MD; Kevin Jessen, MD; Dan Landdeck, MD; Dave Lonsdorf, MD; Michael Pace, MD; Michael Saunders, MD; Catherine Soderqueist, MD; Jon Temte, MD; Vince Winklerprins, MD; Brian Woody, MD.

Conventionally, primary care physicians decided when to refer and to whom a patient should be referred.^1,2 Specialists’ assistance was sought for diagnostic or therapeutic dilemmas,^3,4 management of conditions that presented too infrequently to maintain clinical competence,⁵ and specialized procedures that fell outside a physician’s scope of practice.^3,4 In some cases, physicians referred because patients requested to see a specialist.^1,4

The reorganization of health care over the past few decades has dramatically altered the interactions between primary care physicians and specialists. The growth in multispecialty group practice arrangements⁶ has led to formal, organizationally defined linkages between practitioners. Managed health plans and medical groups⁷ encourage primary care physicians’ control over the referral process through such mechanisms as specialty referral authorizations, financial disincentives for making a referral, performance assessment of referral patterns, and referral guidelines. These changes have transformed a once-informal process into one rife with administrative restrictions on referral decision making.

The Ambulatory Sentinel Practice Network (ASPN) Referral Study was designed to describe and analyze primary care physicians’ referral decisions and their outcomes in the context of a changing health care system in the United States. The study occurred in the ASPN and other regional practice-based research networks. This report examines primary care physicians’ referral decisions. We conceptualized the referral decision as occurring in 2 phases: whether to refer followed by to whom to refer.²

Methods

Physician sample

Physicians were recruited from March 1997 to May 1998. Recruitment activities were directed to all physician members of ASPN, physicians affiliated with the Medical Group Management Association, local and regional networks (Minnesota Academy of Family Physicians Research Network, the Wisconsin Research Network, the Dartmouth Primary Cooperative Research Network (COOP), and the larger community of primary care physicians. The study was publicized via direct mailings to physicians, articles and notices in practice-based research network newsletters and journals, and presentations at conferences. Contact with physicians expressing interest was made by telephone.

Physicians were included in the study if they practiced in the United States and were not in residency or fellowship training. Of all physicians contacted, 342 expressed interest in the study and 182 completed some aspect of data collection. A total of 141 family physicians, 12 internists, and 1 pediatrician completed all phases of data collection. In this study, the 141 family physicians (41% members of ASPN) formed the physician sample. They delivered health care in 87 practices located in 31 states.

Procedures

Study protocols and materials, based on a similar practice-based research study conducted with pediatricians, were reviewed and approved by the Committee on Human Research of the Johns Hopkins School of Public Health and the Colorado Multi-Institutional Review Board.^4,8,9 We conducted a pilot test in 5 practices; this test led to further refinements of methods and questionnaires.

Data collection occurred from September 1997 to February 1999, with 94% of physicians collecting data in 1998 only. Before beginning data collection, physicians completed a questionnaire concerning their practices and personal characteristics. Each practice selected a coordinator who communicated with research staff, learned study protocols, trained office staff and physicians, and monitored data quality. Coordinators completed a questionnaire about the organizational and financial components of their practice. They kept a log of all visits made during 15 consecutive business days and occurring during regularly scheduled office hours. A business day was defined as a half or full work day, provided that the physician held routine office hours. Each patient’s date of birth (5% missing), sex (2% missing), and principal diagnosis (5% missing) were recorded.

The coordinator kept another log of all referrals made by physicians, nurses, and other office staff. Referrals made during telephone conversations with patients were included. A referral was defined as a recommendation that a patient have a face-to-face encounter with another practitioner. We excluded referrals made to laboratories, radiologic facilities, emergency departments, hospitals for inpatient admission, and “curbside consultations” (ie, when the referring physician obtains advice from a specialist but does not send the patient for a visit).

A medical record abstractor assigned ICD-9-CM codes to diagnoses provided by office staff. We matched ICD codes to an expanded set of diagnosis clusters (EDCs). EDCs group ICD codes into clinically homogeneous categories using the methods developed by Schneeweiss.¹⁰ (For more information on EDCs, see http://acg.jhsph.edu.)

When physicians made a referral, they completed a questionnaire (response rate 93.9%) with items concerning the referral decision. Reasons for referral were based on our previously developed taxonomy used in a pediatric referral study⁴ and focus groups of family physicians convened during an annual ASPN convocation.

At the study’s conclusion, physicians received a report that compared their referral practice patterns with those of the entire sample. To defray office expenses associated with data collection, each practice was given a $100 stipend in addition to $5 for each physician referral.

Generalizability analysis

We compared referral rates of the study sample with the National Ambulatory Medical Care Survey (NAMCS), a nationally representative sample of office visits made to family physicians.^11,12 We pooled surveys from 1989 to 1994, inclusive, when the majority of the items in the survey instruments remained unchanged.¹³ (The 1995–1999 surveys did not contain information on whether the visit led to referral.)

We selected visits made by patients enrolled in non-HMO health plans (NAMCS) and health plans that had neither capitated primary care physician payment nor gatekeeping arrangements (study sample). This was done because of the known effect of managed care in general, and of gatekeeping specifically, on increasing referral rates^8,13 and the unequal distribution of managed care plans between the 2 samples. Unweighted visits yielded a sample size of 37,145; of these, 11,676 met the selection criteria.

The proportions of office visits referred were compared overall and by age, sex, and health condition. The 10 most frequently referred conditions in the study sample were used for the condition-specific referral rate assessments. Statistical significance was assessed by the chi-square statistic.

Results

Descriptive information on the 141 family physician sample is presented in Table 1. Physicians spent an average of 51.3 hours per week in their jobs. About 68% of their time was devoted to direct patient care. In most practices, a staff member coordinated administrative aspects of specialty referrals; 20% permitted patients to request a referral by leaving a voice mail message.

TABLE 1
FAMILY PHYSICIAN STUDY SAMPLE

Frequency of referral

The 141 family physicians had 34,519 office visits and made 2165 referrals during 1771 practice-days; 5.1% of office visits were referred. Physicians saw an average of 19.7 patients per day (range 7.0 to 48.4) and made 1.23 referrals per full practice-day (range 0 to 3.90). Referrals made during telephone conversations with patients accounted for 18.9% of all referrals made by physicians (range 0% to 100% per physician).

An additional 369 referrals (a rate of 0.21 referrals per practice-day) were made by staff. Overall, 68% of all referrals were made by physicians during visits with patients, 18% by physicians during telephone conversations with patients, 11% by staff with physician input, and 3% by staff without physician input. In 43.6% of referrals made during telephone conversations with patients, the telephone encounter was the first presentation to medical care for the health problem.

We compared percentages of office visits in which a referral was made in the study sample with percentages of such referrals by family physicians from the NAMCS surveys (1989–1994). The overall percentages did not significantly differ between the 2 groups (4.0% vs 3.7%, P > .05). Although physicians in the study sample were statistically less likely than NAMCS counterparts to refer children (1.6% vs 2.5%, P = .030), more likely to refer the elderly (4.8% vs 4.1%, P = .045), and more likely to refer females (4.1% vs 3.9%, P = .009), these differences were small. There were no differences between the groups in condition-specific referral rates. In sum, these results show that patients in the ASPN sample were equally likely to be referred as those in the NAMCS sample.

Reasons for referral

Table 2 shows the distribution of physicians’ reasons for making the referral. Physicians endorsed a mean of 1.8 different reasons for making the typical referral. Although patients requested to see a specialist for 13.6% of referrals, physicians recorded patient request as the only reason for referral just 1.1%.

We compared referrals made for uncommon conditions (lowest tertile of practice-prevalence) with common conditions (highest practice-prevalence tertile). The calculation of practice-prevalence was based on prior research: the numerator was visits made for the index condition, and the denominator was all visits in the sample.⁵ Uncommon conditions were more likely to be referred for medical management (38.5% vs 25.4%, P < .001), patient request (19.8% vs 12.3%, P = .005), and specialist request (4.9% vs 2.1%, P = .021). Common conditions were more likely to be sent to specialists because of failed current therapy (13.6% vs 3.8%, P < .001) and endoscopy (4.3% vs 0.5%, P = .013). There were no significant differences between the 2 groups in the chances of referral for advice on either diagnosis or treatment.

TABLE 2
REASONS FOR REFERRAL

Conditions referred

Table 3 presents condition-specific referral rates and the 2 most common types of specialists referred to for the top 25 referred health problems. (A complete listing of these data for all conditions reported by study physicians can be found in Table W1.) The 50 most commonly referred health problems accounted for 76% of all referrals made during office visits. Signs or symptoms accounted for 22.4% of all referrals. Condition-specific referral rates varied from a low of 1.9% for patients with otitis media to a high of 45.7% of visits referred for patients with cholelithiasis or cholecystitis. This range in referral rates translates into 24-fold variation in the chances of referral during an office visit based solely on the presenting problem.

TABLE 3
NUMBER OF OFFICE VISITS, REFERRAL RATES, AND SPECIALISTS REFERRED TO FOR TOP 15 REFERRED CONDITIONS*

Specialist selection

Referrals were made most often to surgical subspecialists (45.4%), followed by medical subspecialists (31.0%), nonphysician clinicians (12.1%), obstetriciangynecologists (ob/gyns) (4.6%), mental health professionals (4.2%), other physicians (2.0%), and generalists (0.8%). The 5 most common specialists to whom patients were referred were orthopedic surgeons (12.1%), general surgeons (9.1%), otolaryngologists (6.9%), gastroenterologists (6.6%), and dermatologists (6.0%). Among male patients, referral to urologists was the second most common type; among female patients, referral to ob/gyns was the third most common type.

Mental health referrals were made predominantly to psychologists (2.1% of all referrals), followed by psychiatrists (1.3%) and social workers (0.4%). The most common types of nonphysician clinicians referred to were physical therapists (4.5%), podiatrists (3.0%), nutritionists (1.5%), and audiologists (1.2%).

Referring physicians recommended a specific specialist to the patient for 86.2% of referrals. In descending rank order according to the mean importance rating (range 1 to 3), the reasons for selecting a particular specialist were personal knowledge of the specialist (2.6), quality of prior feedback (2.5), technical capacity (2.3), appointment availability (2.0), patient’s request (1.6), requirements of patient’s health plan (1.6), and proximity of the specialist to the patient’s home (1.6).

Table 4 shows the 3 most common health problems referred to 10 types of specialists. (An expanded version of this table that includes 29 specialists can be found in Table W2.) The majority of referrals for each type of specialist were for 1 to 3 health problems. Family physicians made 17.1% of all referrals to practitioners within their practices. Intrapractice referrals were significantly higher than the overall average for audiologists (40.0%, P = .031), nutritionists (45.2%, P = .004), and psychologists (46.3%, P < .001) and were lower for gastroenterologists (9.3%, P = .022) and rheumatologists (4.0%, P = .005).

TABLE 4
THREE MOST COMMON CONDITIONS REFERRED TO SELECTED SPECIALISTS*

Discussion

This study shows that family physicians manage 95% of office visits without specialty referral. About one third of referrals made from primary care practices occur during encounters other than office visits. Referrals made by staff or during telephone conversations may be part of an integrated sequence of contacts between patients and physicians. Nonetheless, assisting patients in selecting a specialist, transferring relevant patient information, and scheduling specialty appointments (referral coordination activities) are more difficult to perform when patients are not seen in the office,¹⁴ because time is limited and integrating care is poorly reimbursed, if at all. When such referral decisions are made appropriately, they provide an efficient mechanism for decreasing workload in a busy primary care practice. Inappropriately made, they can lead to increased expense, unnecessary time spent with specialists, and poorly coordinated care.

We found that the rates of referral were substantially different among the most commonly referred conditions. Prior work has shown that the frequency with which conditions present to primary care physicians explains about 75% of the variation in condition-specific referral rates.⁵ The mix and severity of comorbidities are important determinants of annual patient referral rates^15,16 and the chances of referral during a visit.⁵ Thus, the epidemiology of morbidity among a patient population is a critical factor that defines the boundaries between primary care physicians and specialists. The appreciation of these clinical determinants is crucial for any valid assessment of primary care physicians’ referral patterns.

Limitations

The study’s focus was on new referral decisions made by physicians to other practitioners. No information is provided about ongoing, long-term referrals in which the patient was already under the care of a specialist. The low rates of referral for conditions such as diabetes may be a consequence of this limitation. Patients with diabetes may already have been under the care of a specialist, thereby generating few new referrals. It is also important to note that even in health plans with gatekeeping arrangements, patients self-refer to specialty care¹³; this study did not include any information on self-referral. Patient self-referral appears to be most likely among sick patients, those with established relationships with a specialist, and patients who do not have a good relationship with a primary care physician.¹⁷

We did not obtain information on the number of telephone calls fielded by physicians each day. Without these data, we were unable to determine whether our methods had failed to capture some telephone referrals or to calculate telephone referral rates. In this study, family physicians made 18.9% of all referrals during telephone conversations, in contrast with pediatricians in another study⁴ who made 27.5% of all referrals by telephone. The difference in these proportions is not large and is probably explained by pediatricians’ greater use of the telephone for patient care.

It could be argued that the volunteer physicians in this study systematically differ from the typical family physician. The average number of visits per day among study physicians (19.7) is similar to a national estimate of 19.9 visits/day for family physicians in single specialty group practices.¹⁸ Furthermore, we found similar probabilities of referral overall and for the 10 most commonly referred conditions between study physicians and a national sample, suggesting that referral propensities between the 2 groups were similar.

Why family physicians refer

No value judgments can be made about the appropriateness of physicians’ reasons for referral. Physicians most commonly referred because they were uncertain about diagnosis or treatment and sought advice from another practitioner. For about 1 in 5 referrals, physicians recorded only a sign or symptom as the diagnosis, suggesting a reasonably high level of diagnostic uncertainty. Physicians’ tolerance of uncertainty varies markedly,¹⁹ making it difficult to judge questions about appropriateness of referrals that are made to reduce this uncertainty.

Another important reason for referral was that physicians deemed the management of the health problem to be outside their scope of practice. Physicians were more likely to refer a patient with a common problem after trying out a course of treatment than was the case for uncommon problems that were more likely to be referred for medical management.

Patients may raise the topic of possible referral. When physicians agree that referral is indicated, they almost always find other reasons for making the referral. Alternatively, physicians might make a decision to refer and justify it in part as being a result of patient request. Discussions on whether a referral is needed are common in primary care. Among referrals made in an Israeli family practice network, patients raised the topic of possible referral in 27% of cases.²⁰ In a study of 856 internal medicine visits, 45% of patients indicated some desire to discuss the need for referral with their physician; however, physicians recognized these desires only about half the time.²¹

Selecting a specialist

Our results show that primary care physicians prefer to send their patients to specialists with whom they have developed a relationship. Physicians in this study maintained a high level of involvement in specialist selection, providing patients with the name of a specific practitioner for 86.2% of referrals. The most important factor in selecting a specialist in our study was the same as that found nearly 20 years ago by Ludke¹: personal knowledge of the specialist. Physicians’ dissatisfaction with the specialty referral process in managed care settings^22,23 could be a result of their reduced choice of specialists with whom they have forged personal relationships.

Slightly more than 1 in 6 referrals were made to specialists in the referring physician’s practice, consistent with movement of primary care physicians into multispecialty groups. Whether intrapractice referral holds any advantage over referrals outside the practice, such as better coordination and appointment adherence, awaits future study.

Our results show that physicians must not only select a specific practitioner but also choose among different types of practitioners. Some patients were sent to nonphysician clinicians and physicians (eg, podiatrists and orthopedic surgeons for acquired foot deformity), whereas others were sent to medical or surgical subspecialists (eg, nephrologists and urologists for urinary tract symptoms). These patterns are likely to reflect the need for multidisciplinary specialty care for some conditions. For instance, patients with diabetes may see an ophthalmologist for retinopathy evaluations and an endocrinologist for medical management consultation. For some conditions, there appears to be considerable uncertainty regarding the boundaries between specialists.²⁴ Should a patient with a skin mass be sent to a general surgeon, a dermatologist, or a plastic surgeon? When should a patient with allergic rhinitis be sent to an allergist and when to an otolaryngologist? These referral patterns may reflect local care practices and specialist availability. They may also be a consequence of a surplus of specialists in this country and competition for patients.

In a survey of family physicians that was performed in the late 1980s, respondents reported that they were more likely to refer to internal medicine subspecialists than internists for adults, but preferred general pediatricians over pediatric subspecialists.²⁵ Our findings suggest that the trend for adult patients remains, but there has been a shift away from general pediatricians toward subspecialists for pediatric referrals. These new patterns may be a consequence of greater availability of pediatric subspecialists, greater exposure of family physicians to pediatric consultants, and a larger share of family physicians who have completed residency training.

Implications for physician training

Fifty conditions accounted for 76% of all specialty referrals made during office visits in this study. Interactions with most types of specialists are generally limited to a few conditions; 3 health problems accounted for more than half of referrals to most specialties. Educators should ensure that these commonly referred conditions are emphasized in curricula that provide family physicians with the skills necessary to expand their scope of practice, when appropriate; determine when a patient should be referred; and identify the type of practitioner to whom the patient should be sent.

Physicians in training should be taught the skills required to recognize the boundaries of their clinical uncertainty and scopes of practice. A challenge for educators is to assist trainees in determining when to tolerate clinical uncertainty while employing a watchful waiting approach and when to initiate a more aggressive evaluation, including when to obtain specialty referral. Modes of implementing these approaches are likely to differ across conditions. Thus, it makes sense in physician training to place the greatest emphasis on conditions for which family physicians commonly refer.

Under certain circumstances, patient request for a specialty consultation may be a sufficient and legitimate reason for referral. For example, as we found in this study, patients with uncommon health problems may seek reassurance from specialists skilled in the management of their specific condition. Managing access to specialists, particularly when the physician is acting as an administrative gatekeeper to referrals, can be challenging.²¹ When doctors and patients disagree on the need for referral, patients may become dissatisfied with their health care²⁶ and decide to self-refer to specialty care.¹⁷ In consideration of the increasing complexity of medical care, developing skills that help physicians discuss and negotiate access to specialized services with both patients and specialists has never been more timely.

Acknowledgments

This study was funded by grant no. R01 HS09377 from the Agency for Healthcare Research and Quality. James Werner and Laurie Vorel provided technical assistance with data collection and project implementation. Many physicians collected data for this study. Their time and devotion were invaluable to the success of this study. These physicians are listed by the states in which they practice. Arizona: Scott Ekdahl, DO; Arkansas: John Scott, MD; California: Andrew Ness, MD; Colorado: Howard Corren, MD; Nell Davis, MD; Timothy Dudley, MD; Audrey Farley, MD; Tillman Farley, MD; Charles Kay, MD; Joan FAMILY PHYSICIANS’ REFERRAL DECISIONS MacEachen, MD; George Maxted, MD; John Miller, MD; Kathy Miller, MD; Steven Milligan, MD; Frank Reed, MD; Louise Schottsteadt, MD; Lynne Spicer, MD; Laura Stein, MD; Lynn Strange, MD; Dan Sullivan, MD; Georgia: Linda Casteel, MD; Randy Cronic, MD; Bruno Denis, MD; Keith Ellis, MD; Kelly Erola, MD; Craig Fabel, MD; Russell Leubbert, DO; Richard Liotta, DO; Mark Majoch, MD; David Najjar, MD; James Snow, DO; Roslyn Taylor, MD; Illinois: Steven Lidvall, MD; Anna Meenan, MD; Eduardo Scholcoff, MD; Loyd Wollstadt, MD; Indiana: Paul Daluga, MD; Steven Phillipson, MD; Iowa: Ken Miller, MD; Janet Ryan, MD; Kansas: Wendell Ellis, DO; John R. Eplee, MD; Robert Moser, MD; Daniel Sontheimer, MD; Louisiana: Linda Stewart, MD; Michigan: Linda French, MD; John Hickner, MD; Minnesota: Ravi Balasubraman, MD; Dave Bucher, MD; William Davis, MD; Richard Gebhart, MD; Katie Guthrie, MD; Anthony Jaspers, MD; Timothy Komoto, MD; Glenn McCarty, DO; Stephen Mitrione, MD; Thomas Retzinger, MD; Paul Spilseth, MD; Ashlesha Tamboli, MD; Montana: Curt Kurtz, MD; Nevada: Coleen Lyons, MD; New Hampshire: Richard Douglass, MD; Paul Friedrichs, MD; Peter Hope, MD; Jonathan Mishcon, MD; New Jersey: John Orzano, MD; Winifred Waldron, MD; New York: Carmella Abraham, MD; R. Eugene Bailey, MD; Lorne Becker, MD; John DeSimone, MD; Miguel Diaz, MD; Rebecca Elliott, MD; John Glennon, MD; James Greenwald, MD; Glenn Griffin, MD; Eileen Hoffman, MD; L. Thomas Wolff, MD; North Carolina: Ed Bujold, MD; Thomas Detesco, MD; Dave Rogers, MD; Phil Sherrod, MD; Oklahoma: Laura Miller, DO; Mike Pontious, MD; Oregon: Douglas Eliason, DO; L.J. Fagnan, MD; Jerry Flaming, DO; Tom Flaming, DO; Jeffrey Humphrey, DO; Michael Kelber, MD; John Sattenspiel, MD; Pennsylvania: John Farmer, DO; Penitha Williams, MD; South Dakota: Fred Thanel, MD; Tennessee: Dan Brewer, MD; Michael Hartsell, MD; R. Louis Murphy, MD; John Parham, MD; Texas: Michael Averitt, DO; Sharon Barber, MD; Kim Patrick Bolton, MD; Robert Cortes, MD; Paul Gerdes, MD; Robert Henry, DO; Michael Kirkpatrick, MD; John Manning, MD; Shane Maxwell, MD; Luis Moreno, MD; Larry G. Padget, MD; Peter Sullivan, MD; Utah: Scott Endsley, MD; David Flinders, MD; Jim Giovino, MD; Eric Hogenson, MD; Dwayne Roberts, MD; Virginia: Duane Lawrence, MD; James Ledwith, MD; June Tunstall, MD; George Wortley, MD; Washington: John Anderson, MD; Elizabeth Wise, MD; West Virginia: Dan Doyle, MD; J. Michael Herr, DO; Wisconsin: Richard Anstett, MD, PhD; Walter Boisvert, MD; Lea Cornell, MD; Anne Eglash, MD; Rod Erickson, MD; Tom Frisby, MD; Terry Hankey, MD; Kevin Jessen, MD; Dan Landdeck, MD; Dave Lonsdorf, MD; Michael Pace, MD; Michael Saunders, MD; Catherine Soderqueist, MD; Jon Temte, MD; Vince Winklerprins, MD; Brian Woody, MD.

Data from the medical literature^1-3 and from 2 ad hoc international consensus conferences^4,5 suggest that antibiotic therapy aimed at eradication of Helicobacter pylori causes persistent healing of peptic ulcer and should therefore be the treatment of choice for patients with peptic ulcer disease (PUD). While administering eradication therapy to H pylori–positive patients with nonulcer dyspepsia (NUD) remains under debate,^6,7 such therapy is generally not recommended.

How these findings are used in clinical practice is largely unknown. Two surveys have reported rates of eradication therapy given by primary care physicians (PCPs) of close to 90% in PUD patients and 50% in NUD patients.^8,9 This information is strongly biased, however, in fact that it was derived from answers to specific questions asked by mail. Responses indicate treatment under ideal conditions rather than real ones. To our knowledge, no reliable analysis on this subject is available.

Our work evaluated the frequency with which eradication therapy was administered in Italy to dyspeptic patients with and without PUD from September 1998 to September 2000. We assessed whether the performance of a complete diagnostic workup affected the rate at which eradication therapy was prescribed. We also evaluated the combination therapies that physicians used.

Methods

The study population included 7336 patients with a PCP’s diagnosis of PUD or NUD from September 1998 to September 2000. A total of 470 (6.4%) of these patients were referred to a gastroenterologist and excluded, leaving a final study population of 6866 patients.

Selection of physicians and data collection

In 1994, software designed by a team of epidemiologists and computer experts from Tor Vergata University, Rome, Italy, was given to 19,000 Italian PCPs. The software was designed to help physicians collect data from their patients during each visit.¹⁰ Data on the number of visits to PCPs during 1993 were obtained. One year later, 2000 physicians agreed to compare the percentage of patients included in their database with the total number of visits and to return their accumulated databases for quality control.

Among the 371 physicians with a quality database of good quality (defined as including at least 95% of their patients in the database and declaring a similar number of patients as in the previous year, 1993), 120 agreed to participate in our study. New software was designed to gather data concerning the performance of eradication therapy for H pylori in the past, the requests and results of upper gastrointestinal endoscopy and abdominal sonography, and the prescribed treatment. The diagnosis was required in the database (ie, it was not possible to have access to subsequent fields in the absence of these data). Furthermore, the software was able to recognize the pharmacologic class of each drug from its generic name. Regarding antibiotic prescriptions for patients in whom a diagnosis of PUD or NUD had been made, the physician was asked if the treatment was intended for the diagnosed disease or for unrelated conditions.

From the start of the study, physicians were asked to include in the new database all patients coming to the office for an initial visit to evaluate dyspepsia of at least 3 months’ duration. If new data concerning diagnostic procedures or treatment emerged during subsequent visits, these were added to the database. It was possible for the PCP to change the diagnosis on the basis of new findings. The diagnosis made during the last visit was considered the final diagnosis.

Although participating PCPs knew they were involved in a study concerning their behavior in treating dyspepsia, they did not know the study’s goal: to determine the rate at which they had prescribed eradication therapy to dyspeptic patients. Physicians were given personal computers as an incentive to participate in the study.

All computers were linked to a central server, located in the epidemiology laboratory of Tor Vergata University of Rome, to which all data were transferred weekly. Each patient was identified by a code number assigned by the attending physician.

The physicians who agreed to participate in the study were stratified according to the following geographic criteria. Approximately one half were in the north of Italy and the other half were in the south. Within each of these areas, approximately one half of participating PCPs worked in cities with 100,000 inhabitants or fewer and the other half in towns with more than 100,000 inhabitants. Forty PCPs were randomly excluded from the study to avoid overrepresentation of certain areas of the country, particularly large cities. Therefore, the data in our study refer to a total of 80 PCPs. The age range of participating physicians was 32 years to 63 years; 61 were men; and all had been practicing PCPs for at least 5 years (range: 5 years to 36 years). Eight PCPs were specialists: 6 in internal medicine, 1 in gynecology, and 1 in rheumatology.

A definitive diagnosis of peptic ulcer was based on findings obtained by the reference standard examination (upper gastrointestinal endoscopy for the definitive diagnosis of PUD) or by a combination of findings (a definitive diagnosis of NUD required normal findings at upper gastrointestinal endoscopy and abdominal sonography). In the other cases, the diagnosis was considered presumptive.

Therapy

We considered the combination therapies most frequently evaluated in clinical trials: bismuth-based triple therapy (bismuth plus metronidazole and tetracycline; bismuth plus clarithromycin and tetracycline; bismuth plus clarithromycin and amoxicillin; bismuth plus metronidazole and amoxicillin^11,12); proton pump inhibitor (PPI)–based triple therapy (PPI plus 2 of the following: amoxicillin, clarithromycin, or metronidazole¹³); PPI-based dual therapy (PPI plus amoxicillin or clarithromycin or metronidazole¹⁴); and others (none of the former).

Statistical analysis

The chi-squared test was used to compare the frequency of discrete variables. A P value of less than .05 was required for statistical significance. Statistical Package for the Social Sciences software was used for the evaluation of significance.

Results

PUD was diagnosed in 2162 patients (1412 men, average age = 45 ± 15.8 years). NUD was diagnosed in 4704 patients (1328 men, average age = 42 ± 13.2 years). Among the 2162 patients with PUD, eradication therapy was prescribed for 481. However, since 115 of the 2162 patients had received such therapy before entering the study, the total number of patients who received antibiotic therapy was 596 (27.6%). Other treatments (mostly H₂-receptor antagonists or PPIs) were prescribed to the remaining 1566 patients with PUD.

Eradication therapy was given to 94 (2%) of 4704 patients diagnosed with NUD (17 patients had received treatment before 1998). Table 1 shows that eradication therapy was prescribed more frequently for patients with PUD than for those with NUD (27.6% vs 2%; P = .0001).

Among patients with PUD, eradication therapy was prescribed more frequently for those with a definitive diagnosis than for those with a presumptive diagnosis (37.7% vs 20.3%, respectively; P < .0001), but the reverse was observed in NUD patients (0.9% vs 2.2%; P = .025) (Table 2). The latter difference is of uncertain clinical significance.

Of 904 patients with a definitive diagnosis of PUD, 223 had a newly diagnosed peptic ulcer; 97 (43.5%) of these were treated with eradication therapy. We observed no change in the percentage of patients with PUD receiving eradication therapy during the study period: 220 of 1005 (22%) during the first year and 261 out of 1157 (22.6%) during the second year.

Of the 80 PCPs, 72 prescribed some kind of eradication therapy. Seven of the 8 physicians who had never prescribed eradication therapy were living in small towns in the south of Italy. Other characteristics of the nonprescribers, such as age and sex, were similar to those of the remaining physicians.

Of 690 patients who received eradication therapy, the type of combination was known for 558. The combination regimen used for the 132 patients treated before the study began was not available. Of 558 patients, 301 (54%) were given PPI-based dual therapy and 225 (40.3%) received PPI-based triple therapy. Other treatments were prescribed to 32 (5.7%) patients (Table 3).

Among patients in whom dual therapy was prescribed, PPI plus clarithromycin was used in 242 patients (43.4% of the total population of treated patients; 80.4% of the subgroup receiving dual therapy). The combination of PPI, clarithromycin, and metronidazole was the most widely used treatment in patients who received triple therapy: it was prescribed to 192 patients (34.4% of all treated patients; 85.3% of the subgroup given PPI-based triple therapy).

Most patients for whom other therapies were prescribed received bismuth-containing combinations. A combination of bismuth and PPI was prescribed to 10 patients (1.8% of the total population). The drug was added to PPI-based triple therapy in 10 patients (1.8% of the total population) and to PPI-based dual therapy in 2 patients (0.4% of the total population). The remaining 10 patients were treated as follows: a combination of 2 antibiotics without PPI (2 patients), H₂-based triple therapy (6 patients), or antibiotic monotherapy (2 patients). None of these combinations is known to eradicate H pylori effectively.

TABLE 1
FREQUENCY OF ERADICATION THERAPY OF HELICOBACTER PYLORI

TABLE 2
FREQUENCY OF ATTEMPTED ERADICATION BY DEFINITIVE OR PRESUMPTIVE DIAGNOSIS

TABLE 3
ERADICATION REGIMENS USED BY ITALIAN PRIMARY CARE PHYSICIANS

Discussion

The data from our study indicate that from 1998 to 2000, the majority of patients with PUD seen by the PCPs participating in the study were not treated with antibiotic therapy aimed at the eradication of H pylori.

In our series, only approximately one third of patients with a definitive diagnosis of PUD were treated with antibiotic therapy, a figure much lower than the 90% reported in nationwide surveys during 1995 and 1996 in the United States and Germany.^8,9 We believe that the most important factor underlying this difference may be the study design. Our study was based on the actual treatment given by the physicians to their patients; previous studies, however, were based on responses to a mailed questionnaire. While the previous studies may reflect ways in which PCPs would ideally treat their patients, some discrepancy is unavoidable when passing from theory to practice.

Other studies based on real-world prescription data had results similar to ours, despite having a small sample size^15,16 or studying underserved populations.^17,18 In light of these data, we suspect that the underuse of antibiotic therapy for PUD disease is common in many areas of the Western world.

Since PCPs were not required to include information on H pylori testing in the database, we did not have reliable data on the frequency of testing or on the relative frequency of positive and negative results. Therefore, it is possible that the low number of prescriptions of eradication therapy for patients with PUD was caused by a high rate of H pylori–negative peptic ulcer. This seems improbable, however, since a high rate of H pylori–positive peptic ulcers has been reported in Italian patients.¹⁹

One likely reason for the low prescription rate of eradication therapy by PCPs was concern about patient compliance and the side effects of antibiotics. Although a recent study reported discontinuation of therapy because of adverse events or noncompliance in less than 10% of patients,⁷ it is well known that data on compliance coming from research studies are not automatically transferable to clinical practice.²⁰

Since eradication therapy was prescribed by the majority of PCPs involved in our study, our findings suggest that an awareness of new information does not necessarily effect changes in physicians’ prescribing patterns.^21-23 Both knowledge-oriented strategies (ie, purely educational interventions) and behavior-oriented interventions (ie, strategies intended to alter behavior, usually by incentives and penalties) are necessary to change physicians’ prescribing patterns regarding PUD. Furthermore, change strategies should be matched to the type of clinician. Our data suggests that most PCPs involved in our study are pragmatists.²⁴ These physicians will not change their behavior in a way that would increase their workload or conflict with patient expectations. Therefore, to increase the rate of prescriptions of eradication therapy for PUD, it is crucial to remove obstacles (eg, facilitate the performance of H pylori testing and endoscopy) and to focus educational interventions on practical issues (eg, place emphasis on the fact that prescribing eradication therapy to these patients may lead to a reduction of visits in the future).

The very low rate of eradication therapy (less than 3%) for patients with NUD in our study seems at odds with the high prevalence (more than 50%) of eradication therapy prescribed by US physicians for patients with NUD.⁹ The same factors explaining the different rates of eradication therapy for patients with PUD apply to differing rates in patients with NUD.

Interestingly, antibiotic therapy was prescribed at a significantly lower rate for NUD patients with a definitive diagnosis. This fact suggests that physicians did not expect important benefits from H pylori eradication in patients who did not have gastroduodenal lesions. Another important finding was that most treated patients received less than optimal treatment. The majority of patients receiving eradication therapy were given a regimen consisting of no more than 2 drugs although this regimen is less effective²⁵ and less convenient from a cost–benefit perspective than is a 3-drug combination.^26,27 Our findings are strikingly similar to those of a small study performed in Scotland²⁸ that showed that more than 55% of patients receiving eradication therapy were treated with PPI-based dual therapy. These data suggest that PCPs are choosing their prescribing options in relation to short-term cost minimization rather than long-term cost effectiveness.

Generalizing our data to the entire Italian health care system may not be valid. Although much attention was paid to the reliability of collected data and to creating a wide geographic distribution of physicians involved in the study, stringent criteria were used for inclusion: ownership of a personal computer, capability of using fairly complex software, and willingness to participate. The majority of Italian PCPs do not share these characteristics. This hypothesis is demonstrated by the fact that only 80 physicians were selected from the initial pool of 19,000. There is no reason, however, to suggest that the above-mentioned characteristics interfere with changing clinicians’ practice patterns.

Conclusions

Our study shows that recommendations for eradication therapy for PUD did not translate into clinical practice in Italy until at least 2000. This means that Italian PCPs failed to reap the clinical and financial benefits resulting from this treatment. This finding, in conjunction with the administration of suboptimal eradication therapy to treated patients, indicates the need for both educational efforts and behavior-oriented interventions aimed at causing the prescribing patterns of eradication therapy of Italian PCPs to conform to the standard reported in the literature.

Acknowledgments

The authors are indebted to Paola Piccolo, MD, for her assistance in the preparation of this manuscript and for English language consultation.

Data from the medical literature^1-3 and from 2 ad hoc international consensus conferences^4,5 suggest that antibiotic therapy aimed at eradication of Helicobacter pylori causes persistent healing of peptic ulcer and should therefore be the treatment of choice for patients with peptic ulcer disease (PUD). While administering eradication therapy to H pylori–positive patients with nonulcer dyspepsia (NUD) remains under debate,^6,7 such therapy is generally not recommended.

How these findings are used in clinical practice is largely unknown. Two surveys have reported rates of eradication therapy given by primary care physicians (PCPs) of close to 90% in PUD patients and 50% in NUD patients.^8,9 This information is strongly biased, however, in fact that it was derived from answers to specific questions asked by mail. Responses indicate treatment under ideal conditions rather than real ones. To our knowledge, no reliable analysis on this subject is available.

Our work evaluated the frequency with which eradication therapy was administered in Italy to dyspeptic patients with and without PUD from September 1998 to September 2000. We assessed whether the performance of a complete diagnostic workup affected the rate at which eradication therapy was prescribed. We also evaluated the combination therapies that physicians used.

Methods

The study population included 7336 patients with a PCP’s diagnosis of PUD or NUD from September 1998 to September 2000. A total of 470 (6.4%) of these patients were referred to a gastroenterologist and excluded, leaving a final study population of 6866 patients.

Selection of physicians and data collection

In 1994, software designed by a team of epidemiologists and computer experts from Tor Vergata University, Rome, Italy, was given to 19,000 Italian PCPs. The software was designed to help physicians collect data from their patients during each visit.¹⁰ Data on the number of visits to PCPs during 1993 were obtained. One year later, 2000 physicians agreed to compare the percentage of patients included in their database with the total number of visits and to return their accumulated databases for quality control.

Among the 371 physicians with a quality database of good quality (defined as including at least 95% of their patients in the database and declaring a similar number of patients as in the previous year, 1993), 120 agreed to participate in our study. New software was designed to gather data concerning the performance of eradication therapy for H pylori in the past, the requests and results of upper gastrointestinal endoscopy and abdominal sonography, and the prescribed treatment. The diagnosis was required in the database (ie, it was not possible to have access to subsequent fields in the absence of these data). Furthermore, the software was able to recognize the pharmacologic class of each drug from its generic name. Regarding antibiotic prescriptions for patients in whom a diagnosis of PUD or NUD had been made, the physician was asked if the treatment was intended for the diagnosed disease or for unrelated conditions.

From the start of the study, physicians were asked to include in the new database all patients coming to the office for an initial visit to evaluate dyspepsia of at least 3 months’ duration. If new data concerning diagnostic procedures or treatment emerged during subsequent visits, these were added to the database. It was possible for the PCP to change the diagnosis on the basis of new findings. The diagnosis made during the last visit was considered the final diagnosis.

Although participating PCPs knew they were involved in a study concerning their behavior in treating dyspepsia, they did not know the study’s goal: to determine the rate at which they had prescribed eradication therapy to dyspeptic patients. Physicians were given personal computers as an incentive to participate in the study.

All computers were linked to a central server, located in the epidemiology laboratory of Tor Vergata University of Rome, to which all data were transferred weekly. Each patient was identified by a code number assigned by the attending physician.

The physicians who agreed to participate in the study were stratified according to the following geographic criteria. Approximately one half were in the north of Italy and the other half were in the south. Within each of these areas, approximately one half of participating PCPs worked in cities with 100,000 inhabitants or fewer and the other half in towns with more than 100,000 inhabitants. Forty PCPs were randomly excluded from the study to avoid overrepresentation of certain areas of the country, particularly large cities. Therefore, the data in our study refer to a total of 80 PCPs. The age range of participating physicians was 32 years to 63 years; 61 were men; and all had been practicing PCPs for at least 5 years (range: 5 years to 36 years). Eight PCPs were specialists: 6 in internal medicine, 1 in gynecology, and 1 in rheumatology.

A definitive diagnosis of peptic ulcer was based on findings obtained by the reference standard examination (upper gastrointestinal endoscopy for the definitive diagnosis of PUD) or by a combination of findings (a definitive diagnosis of NUD required normal findings at upper gastrointestinal endoscopy and abdominal sonography). In the other cases, the diagnosis was considered presumptive.

Therapy

We considered the combination therapies most frequently evaluated in clinical trials: bismuth-based triple therapy (bismuth plus metronidazole and tetracycline; bismuth plus clarithromycin and tetracycline; bismuth plus clarithromycin and amoxicillin; bismuth plus metronidazole and amoxicillin^11,12); proton pump inhibitor (PPI)–based triple therapy (PPI plus 2 of the following: amoxicillin, clarithromycin, or metronidazole¹³); PPI-based dual therapy (PPI plus amoxicillin or clarithromycin or metronidazole¹⁴); and others (none of the former).

Statistical analysis

The chi-squared test was used to compare the frequency of discrete variables. A P value of less than .05 was required for statistical significance. Statistical Package for the Social Sciences software was used for the evaluation of significance.

Results

PUD was diagnosed in 2162 patients (1412 men, average age = 45 ± 15.8 years). NUD was diagnosed in 4704 patients (1328 men, average age = 42 ± 13.2 years). Among the 2162 patients with PUD, eradication therapy was prescribed for 481. However, since 115 of the 2162 patients had received such therapy before entering the study, the total number of patients who received antibiotic therapy was 596 (27.6%). Other treatments (mostly H₂-receptor antagonists or PPIs) were prescribed to the remaining 1566 patients with PUD.

Eradication therapy was given to 94 (2%) of 4704 patients diagnosed with NUD (17 patients had received treatment before 1998). Table 1 shows that eradication therapy was prescribed more frequently for patients with PUD than for those with NUD (27.6% vs 2%; P = .0001).

Among patients with PUD, eradication therapy was prescribed more frequently for those with a definitive diagnosis than for those with a presumptive diagnosis (37.7% vs 20.3%, respectively; P < .0001), but the reverse was observed in NUD patients (0.9% vs 2.2%; P = .025) (Table 2). The latter difference is of uncertain clinical significance.

Of 904 patients with a definitive diagnosis of PUD, 223 had a newly diagnosed peptic ulcer; 97 (43.5%) of these were treated with eradication therapy. We observed no change in the percentage of patients with PUD receiving eradication therapy during the study period: 220 of 1005 (22%) during the first year and 261 out of 1157 (22.6%) during the second year.

Of the 80 PCPs, 72 prescribed some kind of eradication therapy. Seven of the 8 physicians who had never prescribed eradication therapy were living in small towns in the south of Italy. Other characteristics of the nonprescribers, such as age and sex, were similar to those of the remaining physicians.

Of 690 patients who received eradication therapy, the type of combination was known for 558. The combination regimen used for the 132 patients treated before the study began was not available. Of 558 patients, 301 (54%) were given PPI-based dual therapy and 225 (40.3%) received PPI-based triple therapy. Other treatments were prescribed to 32 (5.7%) patients (Table 3).

Among patients in whom dual therapy was prescribed, PPI plus clarithromycin was used in 242 patients (43.4% of the total population of treated patients; 80.4% of the subgroup receiving dual therapy). The combination of PPI, clarithromycin, and metronidazole was the most widely used treatment in patients who received triple therapy: it was prescribed to 192 patients (34.4% of all treated patients; 85.3% of the subgroup given PPI-based triple therapy).

Most patients for whom other therapies were prescribed received bismuth-containing combinations. A combination of bismuth and PPI was prescribed to 10 patients (1.8% of the total population). The drug was added to PPI-based triple therapy in 10 patients (1.8% of the total population) and to PPI-based dual therapy in 2 patients (0.4% of the total population). The remaining 10 patients were treated as follows: a combination of 2 antibiotics without PPI (2 patients), H₂-based triple therapy (6 patients), or antibiotic monotherapy (2 patients). None of these combinations is known to eradicate H pylori effectively.

TABLE 1
FREQUENCY OF ERADICATION THERAPY OF HELICOBACTER PYLORI

TABLE 2
FREQUENCY OF ATTEMPTED ERADICATION BY DEFINITIVE OR PRESUMPTIVE DIAGNOSIS

TABLE 3
ERADICATION REGIMENS USED BY ITALIAN PRIMARY CARE PHYSICIANS

Discussion

The data from our study indicate that from 1998 to 2000, the majority of patients with PUD seen by the PCPs participating in the study were not treated with antibiotic therapy aimed at the eradication of H pylori.

In our series, only approximately one third of patients with a definitive diagnosis of PUD were treated with antibiotic therapy, a figure much lower than the 90% reported in nationwide surveys during 1995 and 1996 in the United States and Germany.^8,9 We believe that the most important factor underlying this difference may be the study design. Our study was based on the actual treatment given by the physicians to their patients; previous studies, however, were based on responses to a mailed questionnaire. While the previous studies may reflect ways in which PCPs would ideally treat their patients, some discrepancy is unavoidable when passing from theory to practice.

Other studies based on real-world prescription data had results similar to ours, despite having a small sample size^15,16 or studying underserved populations.^17,18 In light of these data, we suspect that the underuse of antibiotic therapy for PUD disease is common in many areas of the Western world.

Since PCPs were not required to include information on H pylori testing in the database, we did not have reliable data on the frequency of testing or on the relative frequency of positive and negative results. Therefore, it is possible that the low number of prescriptions of eradication therapy for patients with PUD was caused by a high rate of H pylori–negative peptic ulcer. This seems improbable, however, since a high rate of H pylori–positive peptic ulcers has been reported in Italian patients.¹⁹

One likely reason for the low prescription rate of eradication therapy by PCPs was concern about patient compliance and the side effects of antibiotics. Although a recent study reported discontinuation of therapy because of adverse events or noncompliance in less than 10% of patients,⁷ it is well known that data on compliance coming from research studies are not automatically transferable to clinical practice.²⁰

Since eradication therapy was prescribed by the majority of PCPs involved in our study, our findings suggest that an awareness of new information does not necessarily effect changes in physicians’ prescribing patterns.^21-23 Both knowledge-oriented strategies (ie, purely educational interventions) and behavior-oriented interventions (ie, strategies intended to alter behavior, usually by incentives and penalties) are necessary to change physicians’ prescribing patterns regarding PUD. Furthermore, change strategies should be matched to the type of clinician. Our data suggests that most PCPs involved in our study are pragmatists.²⁴ These physicians will not change their behavior in a way that would increase their workload or conflict with patient expectations. Therefore, to increase the rate of prescriptions of eradication therapy for PUD, it is crucial to remove obstacles (eg, facilitate the performance of H pylori testing and endoscopy) and to focus educational interventions on practical issues (eg, place emphasis on the fact that prescribing eradication therapy to these patients may lead to a reduction of visits in the future).

The very low rate of eradication therapy (less than 3%) for patients with NUD in our study seems at odds with the high prevalence (more than 50%) of eradication therapy prescribed by US physicians for patients with NUD.⁹ The same factors explaining the different rates of eradication therapy for patients with PUD apply to differing rates in patients with NUD.

Interestingly, antibiotic therapy was prescribed at a significantly lower rate for NUD patients with a definitive diagnosis. This fact suggests that physicians did not expect important benefits from H pylori eradication in patients who did not have gastroduodenal lesions. Another important finding was that most treated patients received less than optimal treatment. The majority of patients receiving eradication therapy were given a regimen consisting of no more than 2 drugs although this regimen is less effective²⁵ and less convenient from a cost–benefit perspective than is a 3-drug combination.^26,27 Our findings are strikingly similar to those of a small study performed in Scotland²⁸ that showed that more than 55% of patients receiving eradication therapy were treated with PPI-based dual therapy. These data suggest that PCPs are choosing their prescribing options in relation to short-term cost minimization rather than long-term cost effectiveness.

Generalizing our data to the entire Italian health care system may not be valid. Although much attention was paid to the reliability of collected data and to creating a wide geographic distribution of physicians involved in the study, stringent criteria were used for inclusion: ownership of a personal computer, capability of using fairly complex software, and willingness to participate. The majority of Italian PCPs do not share these characteristics. This hypothesis is demonstrated by the fact that only 80 physicians were selected from the initial pool of 19,000. There is no reason, however, to suggest that the above-mentioned characteristics interfere with changing clinicians’ practice patterns.

Conclusions

Our study shows that recommendations for eradication therapy for PUD did not translate into clinical practice in Italy until at least 2000. This means that Italian PCPs failed to reap the clinical and financial benefits resulting from this treatment. This finding, in conjunction with the administration of suboptimal eradication therapy to treated patients, indicates the need for both educational efforts and behavior-oriented interventions aimed at causing the prescribing patterns of eradication therapy of Italian PCPs to conform to the standard reported in the literature.

Acknowledgments

The authors are indebted to Paola Piccolo, MD, for her assistance in the preparation of this manuscript and for English language consultation.

Data from the medical literature^1-3 and from 2 ad hoc international consensus conferences^4,5 suggest that antibiotic therapy aimed at eradication of Helicobacter pylori causes persistent healing of peptic ulcer and should therefore be the treatment of choice for patients with peptic ulcer disease (PUD). While administering eradication therapy to H pylori–positive patients with nonulcer dyspepsia (NUD) remains under debate,^6,7 such therapy is generally not recommended.

How these findings are used in clinical practice is largely unknown. Two surveys have reported rates of eradication therapy given by primary care physicians (PCPs) of close to 90% in PUD patients and 50% in NUD patients.^8,9 This information is strongly biased, however, in fact that it was derived from answers to specific questions asked by mail. Responses indicate treatment under ideal conditions rather than real ones. To our knowledge, no reliable analysis on this subject is available.

Our work evaluated the frequency with which eradication therapy was administered in Italy to dyspeptic patients with and without PUD from September 1998 to September 2000. We assessed whether the performance of a complete diagnostic workup affected the rate at which eradication therapy was prescribed. We also evaluated the combination therapies that physicians used.

Methods

The study population included 7336 patients with a PCP’s diagnosis of PUD or NUD from September 1998 to September 2000. A total of 470 (6.4%) of these patients were referred to a gastroenterologist and excluded, leaving a final study population of 6866 patients.

Selection of physicians and data collection

In 1994, software designed by a team of epidemiologists and computer experts from Tor Vergata University, Rome, Italy, was given to 19,000 Italian PCPs. The software was designed to help physicians collect data from their patients during each visit.¹⁰ Data on the number of visits to PCPs during 1993 were obtained. One year later, 2000 physicians agreed to compare the percentage of patients included in their database with the total number of visits and to return their accumulated databases for quality control.

Among the 371 physicians with a quality database of good quality (defined as including at least 95% of their patients in the database and declaring a similar number of patients as in the previous year, 1993), 120 agreed to participate in our study. New software was designed to gather data concerning the performance of eradication therapy for H pylori in the past, the requests and results of upper gastrointestinal endoscopy and abdominal sonography, and the prescribed treatment. The diagnosis was required in the database (ie, it was not possible to have access to subsequent fields in the absence of these data). Furthermore, the software was able to recognize the pharmacologic class of each drug from its generic name. Regarding antibiotic prescriptions for patients in whom a diagnosis of PUD or NUD had been made, the physician was asked if the treatment was intended for the diagnosed disease or for unrelated conditions.

From the start of the study, physicians were asked to include in the new database all patients coming to the office for an initial visit to evaluate dyspepsia of at least 3 months’ duration. If new data concerning diagnostic procedures or treatment emerged during subsequent visits, these were added to the database. It was possible for the PCP to change the diagnosis on the basis of new findings. The diagnosis made during the last visit was considered the final diagnosis.

Although participating PCPs knew they were involved in a study concerning their behavior in treating dyspepsia, they did not know the study’s goal: to determine the rate at which they had prescribed eradication therapy to dyspeptic patients. Physicians were given personal computers as an incentive to participate in the study.

All computers were linked to a central server, located in the epidemiology laboratory of Tor Vergata University of Rome, to which all data were transferred weekly. Each patient was identified by a code number assigned by the attending physician.

The physicians who agreed to participate in the study were stratified according to the following geographic criteria. Approximately one half were in the north of Italy and the other half were in the south. Within each of these areas, approximately one half of participating PCPs worked in cities with 100,000 inhabitants or fewer and the other half in towns with more than 100,000 inhabitants. Forty PCPs were randomly excluded from the study to avoid overrepresentation of certain areas of the country, particularly large cities. Therefore, the data in our study refer to a total of 80 PCPs. The age range of participating physicians was 32 years to 63 years; 61 were men; and all had been practicing PCPs for at least 5 years (range: 5 years to 36 years). Eight PCPs were specialists: 6 in internal medicine, 1 in gynecology, and 1 in rheumatology.

A definitive diagnosis of peptic ulcer was based on findings obtained by the reference standard examination (upper gastrointestinal endoscopy for the definitive diagnosis of PUD) or by a combination of findings (a definitive diagnosis of NUD required normal findings at upper gastrointestinal endoscopy and abdominal sonography). In the other cases, the diagnosis was considered presumptive.

Therapy

We considered the combination therapies most frequently evaluated in clinical trials: bismuth-based triple therapy (bismuth plus metronidazole and tetracycline; bismuth plus clarithromycin and tetracycline; bismuth plus clarithromycin and amoxicillin; bismuth plus metronidazole and amoxicillin^11,12); proton pump inhibitor (PPI)–based triple therapy (PPI plus 2 of the following: amoxicillin, clarithromycin, or metronidazole¹³); PPI-based dual therapy (PPI plus amoxicillin or clarithromycin or metronidazole¹⁴); and others (none of the former).

Statistical analysis

The chi-squared test was used to compare the frequency of discrete variables. A P value of less than .05 was required for statistical significance. Statistical Package for the Social Sciences software was used for the evaluation of significance.

Results

PUD was diagnosed in 2162 patients (1412 men, average age = 45 ± 15.8 years). NUD was diagnosed in 4704 patients (1328 men, average age = 42 ± 13.2 years). Among the 2162 patients with PUD, eradication therapy was prescribed for 481. However, since 115 of the 2162 patients had received such therapy before entering the study, the total number of patients who received antibiotic therapy was 596 (27.6%). Other treatments (mostly H₂-receptor antagonists or PPIs) were prescribed to the remaining 1566 patients with PUD.

Eradication therapy was given to 94 (2%) of 4704 patients diagnosed with NUD (17 patients had received treatment before 1998). Table 1 shows that eradication therapy was prescribed more frequently for patients with PUD than for those with NUD (27.6% vs 2%; P = .0001).

Among patients with PUD, eradication therapy was prescribed more frequently for those with a definitive diagnosis than for those with a presumptive diagnosis (37.7% vs 20.3%, respectively; P < .0001), but the reverse was observed in NUD patients (0.9% vs 2.2%; P = .025) (Table 2). The latter difference is of uncertain clinical significance.

Of 904 patients with a definitive diagnosis of PUD, 223 had a newly diagnosed peptic ulcer; 97 (43.5%) of these were treated with eradication therapy. We observed no change in the percentage of patients with PUD receiving eradication therapy during the study period: 220 of 1005 (22%) during the first year and 261 out of 1157 (22.6%) during the second year.

Of the 80 PCPs, 72 prescribed some kind of eradication therapy. Seven of the 8 physicians who had never prescribed eradication therapy were living in small towns in the south of Italy. Other characteristics of the nonprescribers, such as age and sex, were similar to those of the remaining physicians.

Of 690 patients who received eradication therapy, the type of combination was known for 558. The combination regimen used for the 132 patients treated before the study began was not available. Of 558 patients, 301 (54%) were given PPI-based dual therapy and 225 (40.3%) received PPI-based triple therapy. Other treatments were prescribed to 32 (5.7%) patients (Table 3).

Among patients in whom dual therapy was prescribed, PPI plus clarithromycin was used in 242 patients (43.4% of the total population of treated patients; 80.4% of the subgroup receiving dual therapy). The combination of PPI, clarithromycin, and metronidazole was the most widely used treatment in patients who received triple therapy: it was prescribed to 192 patients (34.4% of all treated patients; 85.3% of the subgroup given PPI-based triple therapy).

Most patients for whom other therapies were prescribed received bismuth-containing combinations. A combination of bismuth and PPI was prescribed to 10 patients (1.8% of the total population). The drug was added to PPI-based triple therapy in 10 patients (1.8% of the total population) and to PPI-based dual therapy in 2 patients (0.4% of the total population). The remaining 10 patients were treated as follows: a combination of 2 antibiotics without PPI (2 patients), H₂-based triple therapy (6 patients), or antibiotic monotherapy (2 patients). None of these combinations is known to eradicate H pylori effectively.

TABLE 1
FREQUENCY OF ERADICATION THERAPY OF HELICOBACTER PYLORI

TABLE 2
FREQUENCY OF ATTEMPTED ERADICATION BY DEFINITIVE OR PRESUMPTIVE DIAGNOSIS

TABLE 3
ERADICATION REGIMENS USED BY ITALIAN PRIMARY CARE PHYSICIANS

Discussion

The data from our study indicate that from 1998 to 2000, the majority of patients with PUD seen by the PCPs participating in the study were not treated with antibiotic therapy aimed at the eradication of H pylori.

In our series, only approximately one third of patients with a definitive diagnosis of PUD were treated with antibiotic therapy, a figure much lower than the 90% reported in nationwide surveys during 1995 and 1996 in the United States and Germany.^8,9 We believe that the most important factor underlying this difference may be the study design. Our study was based on the actual treatment given by the physicians to their patients; previous studies, however, were based on responses to a mailed questionnaire. While the previous studies may reflect ways in which PCPs would ideally treat their patients, some discrepancy is unavoidable when passing from theory to practice.

Other studies based on real-world prescription data had results similar to ours, despite having a small sample size^15,16 or studying underserved populations.^17,18 In light of these data, we suspect that the underuse of antibiotic therapy for PUD disease is common in many areas of the Western world.

Since PCPs were not required to include information on H pylori testing in the database, we did not have reliable data on the frequency of testing or on the relative frequency of positive and negative results. Therefore, it is possible that the low number of prescriptions of eradication therapy for patients with PUD was caused by a high rate of H pylori–negative peptic ulcer. This seems improbable, however, since a high rate of H pylori–positive peptic ulcers has been reported in Italian patients.¹⁹

One likely reason for the low prescription rate of eradication therapy by PCPs was concern about patient compliance and the side effects of antibiotics. Although a recent study reported discontinuation of therapy because of adverse events or noncompliance in less than 10% of patients,⁷ it is well known that data on compliance coming from research studies are not automatically transferable to clinical practice.²⁰

Since eradication therapy was prescribed by the majority of PCPs involved in our study, our findings suggest that an awareness of new information does not necessarily effect changes in physicians’ prescribing patterns.^21-23 Both knowledge-oriented strategies (ie, purely educational interventions) and behavior-oriented interventions (ie, strategies intended to alter behavior, usually by incentives and penalties) are necessary to change physicians’ prescribing patterns regarding PUD. Furthermore, change strategies should be matched to the type of clinician. Our data suggests that most PCPs involved in our study are pragmatists.²⁴ These physicians will not change their behavior in a way that would increase their workload or conflict with patient expectations. Therefore, to increase the rate of prescriptions of eradication therapy for PUD, it is crucial to remove obstacles (eg, facilitate the performance of H pylori testing and endoscopy) and to focus educational interventions on practical issues (eg, place emphasis on the fact that prescribing eradication therapy to these patients may lead to a reduction of visits in the future).

The very low rate of eradication therapy (less than 3%) for patients with NUD in our study seems at odds with the high prevalence (more than 50%) of eradication therapy prescribed by US physicians for patients with NUD.⁹ The same factors explaining the different rates of eradication therapy for patients with PUD apply to differing rates in patients with NUD.

Interestingly, antibiotic therapy was prescribed at a significantly lower rate for NUD patients with a definitive diagnosis. This fact suggests that physicians did not expect important benefits from H pylori eradication in patients who did not have gastroduodenal lesions. Another important finding was that most treated patients received less than optimal treatment. The majority of patients receiving eradication therapy were given a regimen consisting of no more than 2 drugs although this regimen is less effective²⁵ and less convenient from a cost–benefit perspective than is a 3-drug combination.^26,27 Our findings are strikingly similar to those of a small study performed in Scotland²⁸ that showed that more than 55% of patients receiving eradication therapy were treated with PPI-based dual therapy. These data suggest that PCPs are choosing their prescribing options in relation to short-term cost minimization rather than long-term cost effectiveness.

Generalizing our data to the entire Italian health care system may not be valid. Although much attention was paid to the reliability of collected data and to creating a wide geographic distribution of physicians involved in the study, stringent criteria were used for inclusion: ownership of a personal computer, capability of using fairly complex software, and willingness to participate. The majority of Italian PCPs do not share these characteristics. This hypothesis is demonstrated by the fact that only 80 physicians were selected from the initial pool of 19,000. There is no reason, however, to suggest that the above-mentioned characteristics interfere with changing clinicians’ practice patterns.

Conclusions

Our study shows that recommendations for eradication therapy for PUD did not translate into clinical practice in Italy until at least 2000. This means that Italian PCPs failed to reap the clinical and financial benefits resulting from this treatment. This finding, in conjunction with the administration of suboptimal eradication therapy to treated patients, indicates the need for both educational efforts and behavior-oriented interventions aimed at causing the prescribing patterns of eradication therapy of Italian PCPs to conform to the standard reported in the literature.

Acknowledgments

The authors are indebted to Paola Piccolo, MD, for her assistance in the preparation of this manuscript and for English language consultation.

An emerging body of research supports the inclusion of spiritual issues in healthcare. Studies have correlated religious commitment with health.^1-3 Many patients affirm the importance of spiritual factors in their lives.^4-5 Recent studies demonstrate that many patients wish to have spirituality considered in their health care, especially during grave illness or emotional crisis.^4-6 How to accomplish this objective is less clear. Although physicians possess spiritual assessment tools,^7-11 broader issues such as physician attitudes, roles, and varied ways of dealing with spirituality have not been widely studied. Understanding this context is crucial if physicians are to include spiritual assessment in patient care.

Two studies of Midwestern family physicians found strong support for addressing patients’ spiritual concerns. In one survey, family physicians in Illinois (n = 210) believed that strong religious convictions positively affect older patients’ mental health (68%) and physical health (42%).¹² These doctors supported physicians’ pursuing spiritual issues at patients’ request (88%) and when patients faced bereavement or impending death (66%). Similarly, Missouri family physicians (n = 231) affirmed that spiritual well-being is an important health component (96%) and that hospitalized patients with spiritual concerns should be referred to chaplains (86%).¹³ A far smaller percentage of these physicians, however, felt they should personally address patients’ spiritual questions (58%).

Despite acknowledging the importance of spiritual issues, the Missouri physicians seldom engaged patients in conversations about death and dying, meditation or quiet reflection, prayer, forgiveness, giving and receiving love, the role of a deity in illness, and the meaning or purpose of illness. They reported such barriers to spiritual discussions as lack of time (71%), inadequate training for taking spiritual histories (59%), and difficulty in identifying patients who want to discuss spiritual issues (56%). The gulf between physicians’ attitudes and practice of spiritual assessment suggests an incomplete understanding of their role in spiritual health.

A study by Craigie and Hobbs¹⁴ of 12 family physicians who are themselves deeply spiritual represents early progress toward understanding this role. These physicians perceived that their spirituality enabled them to experience sacredness in patient encounters, to view medicine as a mission, to maintain centeredness, and to serve as instruments of healing. They described themselves as facilitators and encouragers of patients’ spiritual values and resources. We reasoned that unlike the deeply spiritual respondents in the Craigie and Hobbs study, family physicians in general are likely to have a broad range of attitudes and practices regarding spiritual assessment. We sought to better understand the spectrum of views about the physician’s role in spiritual encounters, to describe family physicians’ approaches to addressing spiritual issues, and to further explore barriers to spiritual discussions and facilitators of these discussions.

Methods

We conducted semistructured interviews¹⁵ with 13 family physicians. Participants assessed their frequency of addressing patients’ spiritual issues and provided demographic information and practice characteristics. Interview topics included spirituality in the doctor-patient relationship, the practice of addressing spiritual issues, and perceived facilitators and barriers to discussing spiritual issues. Interviews were conducted by one of the authors (either A.D.B. or D.H.) or by a research assistant trained in qualitative investigation techniques. Interviews averaged 45 minutes in duration.

To guard against bias in advocating a particular stance toward spiritual assessment, we stressed to respondents that we wanted their honest observations and confirmed their statements throughout the interview. Before analyzing the data, we noted our preconceptions toward spiritual assessment. We consciously sought to avoid these biases while reviewing the data.¹⁶ To further reduce the likelihood of bias, we selected a research team whose members represented multiple academic disciplines and religious backgrounds.

Qualitative research aims to uncover new information and perspectives rather than to draw definitive conclusions from a representative study sample.¹⁷ Study participants were deliberately selected¹⁸ to represent a range of demographic factors (sex, age, religious background), practice types (academic or community practice; urban and rural), and practice regarding physicians’ role in addressing patients’ spiritual issues.

All study participants were board-certified family physicians in Missouri. Three participants were white women; 10 were white men. They ranged in age from 37 to 63 years. Three were in full-time community practice; all others were medical school or residency faculty. All but 1 faculty member reported previous community practice experience. Two participants practice in rural locations; 2 in community health centers; 1 in a metropolitan community practice; 4 in metropolitan community-based residency clinics; and 4 in a metropolitan university-based residency clinic. Subjects’ religious affiliations were Jewish (1), Christian (6), “Unitarian Universalist with Muslim leanings” (1), “Unitarian Universalist with Buddhist leanings” (1), “Unitarian” (2), “none” (1), and agnostic (1).

Interviews took place in participants’ offices. We informed them of the use of audiotapes during the telephone recruitment and obtained verbal consent before audiotaping. An Institutional Review Board approved our study.

Study staff transcribed the interviews verbatim. Investigators verified interview content through comparison with interviewers’ notes and entered the text into Ethnograph,¹⁹ a computer database program designed to organize textual material. Investigators used an iterative process to make an initial template for organizing and coding data.²⁰ Our multiple readings of interviews led to further code revisions until consensus was reached regarding salient issues or themes.^21,22 We solicited respondents’ views of the validity of the final codes and themes and of the accuracy of illustrative quotations.

Results

Six respondents reported regularly addressing spiritual issues with patients. One respondent reported an intermediate level of involvement; 6 reported that they do not regularly address spiritual issues. One physician was opposed to physicians’ addressing spiritual issues with patients.

The themes that emerged from the coded interviews were associated with 5 issues: (1) the appropriate role for physicians in addressing spiritual concerns; (2) situations in which physicians focus on spiritual issues (the nature and setting of these discussions); (3) how physicians address spiritual issues; (4) barriers to addressing spiritual issues; and (5) facilitators of spiritual assessment.

Physician’s role

Physicians who regularly discussed spirituality believed that the scientific evidence linking spirituality and positive health outcomes justified their actions. One study participant stated, “Every physician ought to be dealing with [patients’] spiritual issues. [For example,] how can you justify not talking about spirituality to a patient with depression when you can prove scientifically that strengthening faith commitment helps them? It really comes down to a quality of care issue.”

Some respondents believed that the primacy of spirituality in life provided a justification for addressing spiritual issues with patients. As one stated, “These values . . . get at the core of who you are. I would hope that I would be respectful and supportive” [whether or not I was a physician].

The respondents universally viewed themselves as supportive resources for patients through listening, validating spiritual beliefs, and remaining with patients during times of need. One expressed that healing occurs as physicians and patients connect as people, stating, “I don’t have to be a spiritual master. I can be a human being, trying to connect with another human being. That is a healing experience.”

Although several participants seldom addressed spiritual matters, only one strongly opposed the initiation of such discussions, out of concern about role definition and invasion of patients’ privacy. This participant felt that spiritual matters were “no more in the physician’s domain than questions regarding patients’ finances or their most evil thoughts.”

Nature and setting of discussions

Respondents reported specific patient illnesses and stressors that are likely to prompt spiritual discussions. These included terminal illness; chronic illness; specific conditions, such as heart disease, cancer, or miscarriage; depression, anxiety, or other psychiatric illness; pregnancy; and life stressors, including traumatic illness in the family. Other patient situations associated with spiritual discussions included the presence of symptoms without an explanation (eg, pain, insomnia, anorexia), loss of a bodily function, role change within the family, or an illness that erodes one’s self-concept.

Physician respondents also reported factors that often prompt them to ask spiritual questions. These included intensive care unit admission, new diagnosis of terminal illness, treatment failures, patients’ dissatisfaction with progress of treatment, and discussion of advanced care directives. The respondents who regularly address spiritual issues use screening questions that they tend to ask in response to a patient’s cues or crisis (Table 1).

Some respondents asserted that patients’ spiritual questions arise from their unique reactions to life stress and illness. One physician stated that patients’ questions have “more to do with their view of their illness than what the illness really is.” Spiritual questions commonly asked by patients covered a wide range of spiritual themes (Table 1).

TABLE 1
SELECTED SPIRITUAL QUESTIONS OF PATIENTS AND PHYSICIANS

Manner of addressing spiritual health issues

The physicians in our study believed that in most circumstances, patients should initiate spiritual discussions. One said, “It’s one of those areas where you need a small amount of the patient’s permission to get started and a lot more of the patient’s permission to finish.”

Those who regularly address spiritual issues reported using a variety of techniques and approaches (Table 2). These physicians allow for an inclusive definition of spirituality; they try to normalize spiritual discussions and to integrate spiritual discussions into the ongoing doctor–patient relationship. One said that “bringing [spirituality] to the table” along with other potentially sensitive issues helps patients know “what you’re interested in and gives them the option of deciding to pursue it or not.”

The physicians who address spiritual issues follow principles of spiritual assessment (Table 2). All respondents affirmed that spiritual discussions should be approached with sensitivity and integrity to avoid imposing their own belief systems on their patients. One said, “I can’t even describe how negative it [would be] for me to impose my spiritual beliefs on [my] patients.” Another respondent agreed, but also described a tension between faith-based and profession-based thoughts: “[Discussing one’s faith with a patient risks being] an abuse of power; yet if a patient dies tonight and I haven’t shared the Good News that I have . . . I’m neglecting something that’s very important. . . . How do we do this . . . with both gentleness toward the patient and reverence toward God?”

Respondents expressed divergent viewpoints concerning routine spiritual history taking. Although some considered this to be an essential skill, those who seldom addressed spiritual issues found it less pressing and more time consuming than medical concerns. None reported the routine use of currently available spiritual assessment tools. A respondent opposed to initiating spiritual discussions noted a Judeo-Christian bias in these tools, calling their use “cultural imperialism.”

TABLE 2
PHYSICIANS’ APPROACHES TO ADDRESSING SPIRITUAL ISSUES

Barriers to spiritual assessment

Our respondents noted significant barriers, including physician barriers, mutual physician–patient barriers, physician-perceived patient barriers, and situational barriers (Table 3). An example of a physician–patient barrier is the mutual feeling that neither wants to raise issues of spirituality for fear of alienating or causing discomfort in the other.

TABLE 3
SELECTED BARRIERS TO SPIRITUAL DISCUSSIONS AND FACILITATORS OF THEM

Facilitators of spiritual discussions

Respondents noted that characteristics facilitating patients’ discussions of sexuality and other sensitive issues also facilitate conversations about spirituality. These characteristics include communicating a willingness to engage in (and having the time for) such discussions and assuring patients that spiritual confidences will be received in a nonjudgmental fashion.

Physicians who are more spiritually inclined may be more likely to address spiritual issues with patients. As one respondent stated, “When I have conversations about spiritual issues, it’s [sic] usually been at my initiation . . . because I’m more concerned about religious sorts of things than many physicians.”

A final theme expressed by respondents is that physicians who model a life characterized by balance and spiritual maturity can facilitate patients’ spiritual growth. One stated, “My patients perceive something about my balance and spiritual strength that makes them believe they can do anything. It allows me to move to the next level with them . . . [by showing them how to foster] that strength in themselves with the help of family, community, and God.” Other facilitators are listed in Table 3.

Discussion

The relationship between religiosity and positive health outcomes does much to justify spiritual assessment.^1-3 Other justifications include enhanced coping in chronic illness states,²³ providing patients with hope in illness-coping and recovery;^24,25 the possibility that neglect of spiritual needs may drive patients away from medical treatment,²⁴ and evidence that some patients desire physicians to raise spiritual issues.^6,25,26

We sought to explore the context of spiritual assessment rather than to further justify such assessments. The context of spiritual assessment refers to the philosophical question of whether physicians should address spiritual questions and to practical questions of how spiritual matters arise, how physicians approach them, and what barriers and facilitators they perceive with regard to discussing spirituality. Our study adds to knowledge about this context in several important ways.

We found variance of opinion concerning the physician’s role in spiritual assessment. Respondents reporting infrequent spiritual assessment expressed the view that spiritual issues have lower priority than other medical concerns. Yet those who regularly address spiritual issues justified this with scientific evidence associating spirituality and health. They also proposed a justification not found in previous studies: that spirituality is central to life and therefore important for its own sake rather than simply as a means to a medical end. These findings support and augment previously cited justifications for physicians assisting patients with spiritual health issues.^1-3,6,24-26

Our study results add to the list of categories that prompt discussions of spiritual issues. Respondents affirmed a role for physicians in discussing end-of-life issues and advanced care directives, as in previous studies.^27-29 In addition, they observed that patients’ spiritual questions arise from their unique responses to chronic illness, terminal illness, and life stressors. They identified 2 new categories prompting spiritual discussions: unexplained symptoms and treatment failure.

All respondents affirmed a role for physicians in supporting patients who initiate spiritual discussions. As in a previous study,¹⁴ they viewed themselves as facilitators and encouragers of patients’ spiritual values and as resources rather than as spiritual counselors. The most reticent physicians believed in responding to patients’ questions rather than initiating discussions, an approach that may fail to identify spiritual issues. All respondents supported a patient-centered approach to spiritual assessment in which physicians act with integrity and take care not to abuse their position.

Many physicians saw value in spiritual history taking, though none reported routine use of spiritual assessment tools. The potential Judeo-Christian bias in assessment questions noted by one respondent highlights the need to use culturally sensitive, generic assessment tools³⁰ and to work toward further development of such tools.

We identified new barriers to spiritual assessment, including a physician’s upbringing and culture, lack of spiritual inclination or awareness, resistance to exposing personal beliefs, and belief that spiritual discussions will not influence patients’ illnesses or lives. Respondents also postulated patient barriers, including fears that their physician might judge them for their spiritual views or consider their raising spiritual questions inappropriate.

We identified facilitators of spiritual discussions, such as communicating a willingness to have these discussions. One respondent noted that physicians whose lives are characterized by spiritual maturity might serve as agents of patients’ spiritual growth, consistent with a previous study’s themes of caregiver spirituality and physician vocation and mission.¹⁴

Limitations

Because qualitative research aims to uncover new perspectives rather than to make generalizable assessments, our findings may not apply to all physicians or to all family physicians. Although our respondents did not represent all major world religions, ethnic groups, and cultures, they did offer a diversity of spiritual and religious perspectives. Finally, our study gives only physicians’ perspectives. We are currently studying patients’ perspectives of situations that elicit spiritual questions and of potential barriers to spiritual assessment. We will use themes from our patient and physician qualitative studies to frame questions for a national patient questionnaire regarding physicians’ spiritual assessment.

Conclusions

Physicians differ in their comfort and practice of addressing spiritual issues with patients, but affirm a role for themselves in responding to patients’ spiritual concerns. Perceived barriers, physicians’ role definition, familiarity with factors likely to prompt spiritual questions, and the recognition of principles guiding spiritual discussions form the context for family physicians’ discussions of spiritual issues with patients. This context is important to consider when training medical students and residents in spiritual assessment. Careful attention to this context will also enhance the practicing physician’s skill in providing patient-centered assistance with spiritual health concerns.

Acknowledgments

The authors wish to acknowledge Arej Sawani, who assisted in data collection; Sheri Price, who assisted in manuscript preparation; and Richard Ellis, MD, MPH, Daniel Vinson, MD, MSPH, Steven Zweig, MD, MSPH, and Dale Smith, who reviewed the manuscript and offered editorial suggestions.

An emerging body of research supports the inclusion of spiritual issues in healthcare. Studies have correlated religious commitment with health.^1-3 Many patients affirm the importance of spiritual factors in their lives.^4-5 Recent studies demonstrate that many patients wish to have spirituality considered in their health care, especially during grave illness or emotional crisis.^4-6 How to accomplish this objective is less clear. Although physicians possess spiritual assessment tools,^7-11 broader issues such as physician attitudes, roles, and varied ways of dealing with spirituality have not been widely studied. Understanding this context is crucial if physicians are to include spiritual assessment in patient care.

Two studies of Midwestern family physicians found strong support for addressing patients’ spiritual concerns. In one survey, family physicians in Illinois (n = 210) believed that strong religious convictions positively affect older patients’ mental health (68%) and physical health (42%).¹² These doctors supported physicians’ pursuing spiritual issues at patients’ request (88%) and when patients faced bereavement or impending death (66%). Similarly, Missouri family physicians (n = 231) affirmed that spiritual well-being is an important health component (96%) and that hospitalized patients with spiritual concerns should be referred to chaplains (86%).¹³ A far smaller percentage of these physicians, however, felt they should personally address patients’ spiritual questions (58%).

Despite acknowledging the importance of spiritual issues, the Missouri physicians seldom engaged patients in conversations about death and dying, meditation or quiet reflection, prayer, forgiveness, giving and receiving love, the role of a deity in illness, and the meaning or purpose of illness. They reported such barriers to spiritual discussions as lack of time (71%), inadequate training for taking spiritual histories (59%), and difficulty in identifying patients who want to discuss spiritual issues (56%). The gulf between physicians’ attitudes and practice of spiritual assessment suggests an incomplete understanding of their role in spiritual health.

A study by Craigie and Hobbs¹⁴ of 12 family physicians who are themselves deeply spiritual represents early progress toward understanding this role. These physicians perceived that their spirituality enabled them to experience sacredness in patient encounters, to view medicine as a mission, to maintain centeredness, and to serve as instruments of healing. They described themselves as facilitators and encouragers of patients’ spiritual values and resources. We reasoned that unlike the deeply spiritual respondents in the Craigie and Hobbs study, family physicians in general are likely to have a broad range of attitudes and practices regarding spiritual assessment. We sought to better understand the spectrum of views about the physician’s role in spiritual encounters, to describe family physicians’ approaches to addressing spiritual issues, and to further explore barriers to spiritual discussions and facilitators of these discussions.

Methods

We conducted semistructured interviews¹⁵ with 13 family physicians. Participants assessed their frequency of addressing patients’ spiritual issues and provided demographic information and practice characteristics. Interview topics included spirituality in the doctor-patient relationship, the practice of addressing spiritual issues, and perceived facilitators and barriers to discussing spiritual issues. Interviews were conducted by one of the authors (either A.D.B. or D.H.) or by a research assistant trained in qualitative investigation techniques. Interviews averaged 45 minutes in duration.

To guard against bias in advocating a particular stance toward spiritual assessment, we stressed to respondents that we wanted their honest observations and confirmed their statements throughout the interview. Before analyzing the data, we noted our preconceptions toward spiritual assessment. We consciously sought to avoid these biases while reviewing the data.¹⁶ To further reduce the likelihood of bias, we selected a research team whose members represented multiple academic disciplines and religious backgrounds.

Qualitative research aims to uncover new information and perspectives rather than to draw definitive conclusions from a representative study sample.¹⁷ Study participants were deliberately selected¹⁸ to represent a range of demographic factors (sex, age, religious background), practice types (academic or community practice; urban and rural), and practice regarding physicians’ role in addressing patients’ spiritual issues.

All study participants were board-certified family physicians in Missouri. Three participants were white women; 10 were white men. They ranged in age from 37 to 63 years. Three were in full-time community practice; all others were medical school or residency faculty. All but 1 faculty member reported previous community practice experience. Two participants practice in rural locations; 2 in community health centers; 1 in a metropolitan community practice; 4 in metropolitan community-based residency clinics; and 4 in a metropolitan university-based residency clinic. Subjects’ religious affiliations were Jewish (1), Christian (6), “Unitarian Universalist with Muslim leanings” (1), “Unitarian Universalist with Buddhist leanings” (1), “Unitarian” (2), “none” (1), and agnostic (1).

Interviews took place in participants’ offices. We informed them of the use of audiotapes during the telephone recruitment and obtained verbal consent before audiotaping. An Institutional Review Board approved our study.

Study staff transcribed the interviews verbatim. Investigators verified interview content through comparison with interviewers’ notes and entered the text into Ethnograph,¹⁹ a computer database program designed to organize textual material. Investigators used an iterative process to make an initial template for organizing and coding data.²⁰ Our multiple readings of interviews led to further code revisions until consensus was reached regarding salient issues or themes.^21,22 We solicited respondents’ views of the validity of the final codes and themes and of the accuracy of illustrative quotations.

Results

Six respondents reported regularly addressing spiritual issues with patients. One respondent reported an intermediate level of involvement; 6 reported that they do not regularly address spiritual issues. One physician was opposed to physicians’ addressing spiritual issues with patients.

The themes that emerged from the coded interviews were associated with 5 issues: (1) the appropriate role for physicians in addressing spiritual concerns; (2) situations in which physicians focus on spiritual issues (the nature and setting of these discussions); (3) how physicians address spiritual issues; (4) barriers to addressing spiritual issues; and (5) facilitators of spiritual assessment.

Physician’s role

Physicians who regularly discussed spirituality believed that the scientific evidence linking spirituality and positive health outcomes justified their actions. One study participant stated, “Every physician ought to be dealing with [patients’] spiritual issues. [For example,] how can you justify not talking about spirituality to a patient with depression when you can prove scientifically that strengthening faith commitment helps them? It really comes down to a quality of care issue.”

Some respondents believed that the primacy of spirituality in life provided a justification for addressing spiritual issues with patients. As one stated, “These values . . . get at the core of who you are. I would hope that I would be respectful and supportive” [whether or not I was a physician].

The respondents universally viewed themselves as supportive resources for patients through listening, validating spiritual beliefs, and remaining with patients during times of need. One expressed that healing occurs as physicians and patients connect as people, stating, “I don’t have to be a spiritual master. I can be a human being, trying to connect with another human being. That is a healing experience.”

Although several participants seldom addressed spiritual matters, only one strongly opposed the initiation of such discussions, out of concern about role definition and invasion of patients’ privacy. This participant felt that spiritual matters were “no more in the physician’s domain than questions regarding patients’ finances or their most evil thoughts.”

Nature and setting of discussions

Respondents reported specific patient illnesses and stressors that are likely to prompt spiritual discussions. These included terminal illness; chronic illness; specific conditions, such as heart disease, cancer, or miscarriage; depression, anxiety, or other psychiatric illness; pregnancy; and life stressors, including traumatic illness in the family. Other patient situations associated with spiritual discussions included the presence of symptoms without an explanation (eg, pain, insomnia, anorexia), loss of a bodily function, role change within the family, or an illness that erodes one’s self-concept.

Physician respondents also reported factors that often prompt them to ask spiritual questions. These included intensive care unit admission, new diagnosis of terminal illness, treatment failures, patients’ dissatisfaction with progress of treatment, and discussion of advanced care directives. The respondents who regularly address spiritual issues use screening questions that they tend to ask in response to a patient’s cues or crisis (Table 1).

Some respondents asserted that patients’ spiritual questions arise from their unique reactions to life stress and illness. One physician stated that patients’ questions have “more to do with their view of their illness than what the illness really is.” Spiritual questions commonly asked by patients covered a wide range of spiritual themes (Table 1).

TABLE 1
SELECTED SPIRITUAL QUESTIONS OF PATIENTS AND PHYSICIANS

Manner of addressing spiritual health issues

The physicians in our study believed that in most circumstances, patients should initiate spiritual discussions. One said, “It’s one of those areas where you need a small amount of the patient’s permission to get started and a lot more of the patient’s permission to finish.”

Those who regularly address spiritual issues reported using a variety of techniques and approaches (Table 2). These physicians allow for an inclusive definition of spirituality; they try to normalize spiritual discussions and to integrate spiritual discussions into the ongoing doctor–patient relationship. One said that “bringing [spirituality] to the table” along with other potentially sensitive issues helps patients know “what you’re interested in and gives them the option of deciding to pursue it or not.”

The physicians who address spiritual issues follow principles of spiritual assessment (Table 2). All respondents affirmed that spiritual discussions should be approached with sensitivity and integrity to avoid imposing their own belief systems on their patients. One said, “I can’t even describe how negative it [would be] for me to impose my spiritual beliefs on [my] patients.” Another respondent agreed, but also described a tension between faith-based and profession-based thoughts: “[Discussing one’s faith with a patient risks being] an abuse of power; yet if a patient dies tonight and I haven’t shared the Good News that I have . . . I’m neglecting something that’s very important. . . . How do we do this . . . with both gentleness toward the patient and reverence toward God?”

Respondents expressed divergent viewpoints concerning routine spiritual history taking. Although some considered this to be an essential skill, those who seldom addressed spiritual issues found it less pressing and more time consuming than medical concerns. None reported the routine use of currently available spiritual assessment tools. A respondent opposed to initiating spiritual discussions noted a Judeo-Christian bias in these tools, calling their use “cultural imperialism.”

TABLE 2
PHYSICIANS’ APPROACHES TO ADDRESSING SPIRITUAL ISSUES

Barriers to spiritual assessment

Our respondents noted significant barriers, including physician barriers, mutual physician–patient barriers, physician-perceived patient barriers, and situational barriers (Table 3). An example of a physician–patient barrier is the mutual feeling that neither wants to raise issues of spirituality for fear of alienating or causing discomfort in the other.

TABLE 3
SELECTED BARRIERS TO SPIRITUAL DISCUSSIONS AND FACILITATORS OF THEM

Facilitators of spiritual discussions

Respondents noted that characteristics facilitating patients’ discussions of sexuality and other sensitive issues also facilitate conversations about spirituality. These characteristics include communicating a willingness to engage in (and having the time for) such discussions and assuring patients that spiritual confidences will be received in a nonjudgmental fashion.

Physicians who are more spiritually inclined may be more likely to address spiritual issues with patients. As one respondent stated, “When I have conversations about spiritual issues, it’s [sic] usually been at my initiation . . . because I’m more concerned about religious sorts of things than many physicians.”

A final theme expressed by respondents is that physicians who model a life characterized by balance and spiritual maturity can facilitate patients’ spiritual growth. One stated, “My patients perceive something about my balance and spiritual strength that makes them believe they can do anything. It allows me to move to the next level with them . . . [by showing them how to foster] that strength in themselves with the help of family, community, and God.” Other facilitators are listed in Table 3.

Discussion

The relationship between religiosity and positive health outcomes does much to justify spiritual assessment.^1-3 Other justifications include enhanced coping in chronic illness states,²³ providing patients with hope in illness-coping and recovery;^24,25 the possibility that neglect of spiritual needs may drive patients away from medical treatment,²⁴ and evidence that some patients desire physicians to raise spiritual issues.^6,25,26

We sought to explore the context of spiritual assessment rather than to further justify such assessments. The context of spiritual assessment refers to the philosophical question of whether physicians should address spiritual questions and to practical questions of how spiritual matters arise, how physicians approach them, and what barriers and facilitators they perceive with regard to discussing spirituality. Our study adds to knowledge about this context in several important ways.

We found variance of opinion concerning the physician’s role in spiritual assessment. Respondents reporting infrequent spiritual assessment expressed the view that spiritual issues have lower priority than other medical concerns. Yet those who regularly address spiritual issues justified this with scientific evidence associating spirituality and health. They also proposed a justification not found in previous studies: that spirituality is central to life and therefore important for its own sake rather than simply as a means to a medical end. These findings support and augment previously cited justifications for physicians assisting patients with spiritual health issues.^1-3,6,24-26

Our study results add to the list of categories that prompt discussions of spiritual issues. Respondents affirmed a role for physicians in discussing end-of-life issues and advanced care directives, as in previous studies.^27-29 In addition, they observed that patients’ spiritual questions arise from their unique responses to chronic illness, terminal illness, and life stressors. They identified 2 new categories prompting spiritual discussions: unexplained symptoms and treatment failure.

All respondents affirmed a role for physicians in supporting patients who initiate spiritual discussions. As in a previous study,¹⁴ they viewed themselves as facilitators and encouragers of patients’ spiritual values and as resources rather than as spiritual counselors. The most reticent physicians believed in responding to patients’ questions rather than initiating discussions, an approach that may fail to identify spiritual issues. All respondents supported a patient-centered approach to spiritual assessment in which physicians act with integrity and take care not to abuse their position.

Many physicians saw value in spiritual history taking, though none reported routine use of spiritual assessment tools. The potential Judeo-Christian bias in assessment questions noted by one respondent highlights the need to use culturally sensitive, generic assessment tools³⁰ and to work toward further development of such tools.

We identified new barriers to spiritual assessment, including a physician’s upbringing and culture, lack of spiritual inclination or awareness, resistance to exposing personal beliefs, and belief that spiritual discussions will not influence patients’ illnesses or lives. Respondents also postulated patient barriers, including fears that their physician might judge them for their spiritual views or consider their raising spiritual questions inappropriate.

We identified facilitators of spiritual discussions, such as communicating a willingness to have these discussions. One respondent noted that physicians whose lives are characterized by spiritual maturity might serve as agents of patients’ spiritual growth, consistent with a previous study’s themes of caregiver spirituality and physician vocation and mission.¹⁴

Limitations

Because qualitative research aims to uncover new perspectives rather than to make generalizable assessments, our findings may not apply to all physicians or to all family physicians. Although our respondents did not represent all major world religions, ethnic groups, and cultures, they did offer a diversity of spiritual and religious perspectives. Finally, our study gives only physicians’ perspectives. We are currently studying patients’ perspectives of situations that elicit spiritual questions and of potential barriers to spiritual assessment. We will use themes from our patient and physician qualitative studies to frame questions for a national patient questionnaire regarding physicians’ spiritual assessment.

Conclusions

Physicians differ in their comfort and practice of addressing spiritual issues with patients, but affirm a role for themselves in responding to patients’ spiritual concerns. Perceived barriers, physicians’ role definition, familiarity with factors likely to prompt spiritual questions, and the recognition of principles guiding spiritual discussions form the context for family physicians’ discussions of spiritual issues with patients. This context is important to consider when training medical students and residents in spiritual assessment. Careful attention to this context will also enhance the practicing physician’s skill in providing patient-centered assistance with spiritual health concerns.

Acknowledgments

The authors wish to acknowledge Arej Sawani, who assisted in data collection; Sheri Price, who assisted in manuscript preparation; and Richard Ellis, MD, MPH, Daniel Vinson, MD, MSPH, Steven Zweig, MD, MSPH, and Dale Smith, who reviewed the manuscript and offered editorial suggestions.

An emerging body of research supports the inclusion of spiritual issues in healthcare. Studies have correlated religious commitment with health.^1-3 Many patients affirm the importance of spiritual factors in their lives.^4-5 Recent studies demonstrate that many patients wish to have spirituality considered in their health care, especially during grave illness or emotional crisis.^4-6 How to accomplish this objective is less clear. Although physicians possess spiritual assessment tools,^7-11 broader issues such as physician attitudes, roles, and varied ways of dealing with spirituality have not been widely studied. Understanding this context is crucial if physicians are to include spiritual assessment in patient care.

Two studies of Midwestern family physicians found strong support for addressing patients’ spiritual concerns. In one survey, family physicians in Illinois (n = 210) believed that strong religious convictions positively affect older patients’ mental health (68%) and physical health (42%).¹² These doctors supported physicians’ pursuing spiritual issues at patients’ request (88%) and when patients faced bereavement or impending death (66%). Similarly, Missouri family physicians (n = 231) affirmed that spiritual well-being is an important health component (96%) and that hospitalized patients with spiritual concerns should be referred to chaplains (86%).¹³ A far smaller percentage of these physicians, however, felt they should personally address patients’ spiritual questions (58%).

Despite acknowledging the importance of spiritual issues, the Missouri physicians seldom engaged patients in conversations about death and dying, meditation or quiet reflection, prayer, forgiveness, giving and receiving love, the role of a deity in illness, and the meaning or purpose of illness. They reported such barriers to spiritual discussions as lack of time (71%), inadequate training for taking spiritual histories (59%), and difficulty in identifying patients who want to discuss spiritual issues (56%). The gulf between physicians’ attitudes and practice of spiritual assessment suggests an incomplete understanding of their role in spiritual health.

A study by Craigie and Hobbs¹⁴ of 12 family physicians who are themselves deeply spiritual represents early progress toward understanding this role. These physicians perceived that their spirituality enabled them to experience sacredness in patient encounters, to view medicine as a mission, to maintain centeredness, and to serve as instruments of healing. They described themselves as facilitators and encouragers of patients’ spiritual values and resources. We reasoned that unlike the deeply spiritual respondents in the Craigie and Hobbs study, family physicians in general are likely to have a broad range of attitudes and practices regarding spiritual assessment. We sought to better understand the spectrum of views about the physician’s role in spiritual encounters, to describe family physicians’ approaches to addressing spiritual issues, and to further explore barriers to spiritual discussions and facilitators of these discussions.

Methods

We conducted semistructured interviews¹⁵ with 13 family physicians. Participants assessed their frequency of addressing patients’ spiritual issues and provided demographic information and practice characteristics. Interview topics included spirituality in the doctor-patient relationship, the practice of addressing spiritual issues, and perceived facilitators and barriers to discussing spiritual issues. Interviews were conducted by one of the authors (either A.D.B. or D.H.) or by a research assistant trained in qualitative investigation techniques. Interviews averaged 45 minutes in duration.

To guard against bias in advocating a particular stance toward spiritual assessment, we stressed to respondents that we wanted their honest observations and confirmed their statements throughout the interview. Before analyzing the data, we noted our preconceptions toward spiritual assessment. We consciously sought to avoid these biases while reviewing the data.¹⁶ To further reduce the likelihood of bias, we selected a research team whose members represented multiple academic disciplines and religious backgrounds.

Qualitative research aims to uncover new information and perspectives rather than to draw definitive conclusions from a representative study sample.¹⁷ Study participants were deliberately selected¹⁸ to represent a range of demographic factors (sex, age, religious background), practice types (academic or community practice; urban and rural), and practice regarding physicians’ role in addressing patients’ spiritual issues.

All study participants were board-certified family physicians in Missouri. Three participants were white women; 10 were white men. They ranged in age from 37 to 63 years. Three were in full-time community practice; all others were medical school or residency faculty. All but 1 faculty member reported previous community practice experience. Two participants practice in rural locations; 2 in community health centers; 1 in a metropolitan community practice; 4 in metropolitan community-based residency clinics; and 4 in a metropolitan university-based residency clinic. Subjects’ religious affiliations were Jewish (1), Christian (6), “Unitarian Universalist with Muslim leanings” (1), “Unitarian Universalist with Buddhist leanings” (1), “Unitarian” (2), “none” (1), and agnostic (1).

Interviews took place in participants’ offices. We informed them of the use of audiotapes during the telephone recruitment and obtained verbal consent before audiotaping. An Institutional Review Board approved our study.

Study staff transcribed the interviews verbatim. Investigators verified interview content through comparison with interviewers’ notes and entered the text into Ethnograph,¹⁹ a computer database program designed to organize textual material. Investigators used an iterative process to make an initial template for organizing and coding data.²⁰ Our multiple readings of interviews led to further code revisions until consensus was reached regarding salient issues or themes.^21,22 We solicited respondents’ views of the validity of the final codes and themes and of the accuracy of illustrative quotations.

Results

Six respondents reported regularly addressing spiritual issues with patients. One respondent reported an intermediate level of involvement; 6 reported that they do not regularly address spiritual issues. One physician was opposed to physicians’ addressing spiritual issues with patients.

The themes that emerged from the coded interviews were associated with 5 issues: (1) the appropriate role for physicians in addressing spiritual concerns; (2) situations in which physicians focus on spiritual issues (the nature and setting of these discussions); (3) how physicians address spiritual issues; (4) barriers to addressing spiritual issues; and (5) facilitators of spiritual assessment.

Physician’s role

Physicians who regularly discussed spirituality believed that the scientific evidence linking spirituality and positive health outcomes justified their actions. One study participant stated, “Every physician ought to be dealing with [patients’] spiritual issues. [For example,] how can you justify not talking about spirituality to a patient with depression when you can prove scientifically that strengthening faith commitment helps them? It really comes down to a quality of care issue.”

Some respondents believed that the primacy of spirituality in life provided a justification for addressing spiritual issues with patients. As one stated, “These values . . . get at the core of who you are. I would hope that I would be respectful and supportive” [whether or not I was a physician].

The respondents universally viewed themselves as supportive resources for patients through listening, validating spiritual beliefs, and remaining with patients during times of need. One expressed that healing occurs as physicians and patients connect as people, stating, “I don’t have to be a spiritual master. I can be a human being, trying to connect with another human being. That is a healing experience.”

Although several participants seldom addressed spiritual matters, only one strongly opposed the initiation of such discussions, out of concern about role definition and invasion of patients’ privacy. This participant felt that spiritual matters were “no more in the physician’s domain than questions regarding patients’ finances or their most evil thoughts.”

Nature and setting of discussions

Respondents reported specific patient illnesses and stressors that are likely to prompt spiritual discussions. These included terminal illness; chronic illness; specific conditions, such as heart disease, cancer, or miscarriage; depression, anxiety, or other psychiatric illness; pregnancy; and life stressors, including traumatic illness in the family. Other patient situations associated with spiritual discussions included the presence of symptoms without an explanation (eg, pain, insomnia, anorexia), loss of a bodily function, role change within the family, or an illness that erodes one’s self-concept.

Physician respondents also reported factors that often prompt them to ask spiritual questions. These included intensive care unit admission, new diagnosis of terminal illness, treatment failures, patients’ dissatisfaction with progress of treatment, and discussion of advanced care directives. The respondents who regularly address spiritual issues use screening questions that they tend to ask in response to a patient’s cues or crisis (Table 1).

Some respondents asserted that patients’ spiritual questions arise from their unique reactions to life stress and illness. One physician stated that patients’ questions have “more to do with their view of their illness than what the illness really is.” Spiritual questions commonly asked by patients covered a wide range of spiritual themes (Table 1).

TABLE 1
SELECTED SPIRITUAL QUESTIONS OF PATIENTS AND PHYSICIANS

Manner of addressing spiritual health issues

The physicians in our study believed that in most circumstances, patients should initiate spiritual discussions. One said, “It’s one of those areas where you need a small amount of the patient’s permission to get started and a lot more of the patient’s permission to finish.”

Those who regularly address spiritual issues reported using a variety of techniques and approaches (Table 2). These physicians allow for an inclusive definition of spirituality; they try to normalize spiritual discussions and to integrate spiritual discussions into the ongoing doctor–patient relationship. One said that “bringing [spirituality] to the table” along with other potentially sensitive issues helps patients know “what you’re interested in and gives them the option of deciding to pursue it or not.”

The physicians who address spiritual issues follow principles of spiritual assessment (Table 2). All respondents affirmed that spiritual discussions should be approached with sensitivity and integrity to avoid imposing their own belief systems on their patients. One said, “I can’t even describe how negative it [would be] for me to impose my spiritual beliefs on [my] patients.” Another respondent agreed, but also described a tension between faith-based and profession-based thoughts: “[Discussing one’s faith with a patient risks being] an abuse of power; yet if a patient dies tonight and I haven’t shared the Good News that I have . . . I’m neglecting something that’s very important. . . . How do we do this . . . with both gentleness toward the patient and reverence toward God?”

Respondents expressed divergent viewpoints concerning routine spiritual history taking. Although some considered this to be an essential skill, those who seldom addressed spiritual issues found it less pressing and more time consuming than medical concerns. None reported the routine use of currently available spiritual assessment tools. A respondent opposed to initiating spiritual discussions noted a Judeo-Christian bias in these tools, calling their use “cultural imperialism.”

TABLE 2
PHYSICIANS’ APPROACHES TO ADDRESSING SPIRITUAL ISSUES

Barriers to spiritual assessment

Our respondents noted significant barriers, including physician barriers, mutual physician–patient barriers, physician-perceived patient barriers, and situational barriers (Table 3). An example of a physician–patient barrier is the mutual feeling that neither wants to raise issues of spirituality for fear of alienating or causing discomfort in the other.

TABLE 3
SELECTED BARRIERS TO SPIRITUAL DISCUSSIONS AND FACILITATORS OF THEM

Facilitators of spiritual discussions

Respondents noted that characteristics facilitating patients’ discussions of sexuality and other sensitive issues also facilitate conversations about spirituality. These characteristics include communicating a willingness to engage in (and having the time for) such discussions and assuring patients that spiritual confidences will be received in a nonjudgmental fashion.

Physicians who are more spiritually inclined may be more likely to address spiritual issues with patients. As one respondent stated, “When I have conversations about spiritual issues, it’s [sic] usually been at my initiation . . . because I’m more concerned about religious sorts of things than many physicians.”

A final theme expressed by respondents is that physicians who model a life characterized by balance and spiritual maturity can facilitate patients’ spiritual growth. One stated, “My patients perceive something about my balance and spiritual strength that makes them believe they can do anything. It allows me to move to the next level with them . . . [by showing them how to foster] that strength in themselves with the help of family, community, and God.” Other facilitators are listed in Table 3.

Discussion

The relationship between religiosity and positive health outcomes does much to justify spiritual assessment.^1-3 Other justifications include enhanced coping in chronic illness states,²³ providing patients with hope in illness-coping and recovery;^24,25 the possibility that neglect of spiritual needs may drive patients away from medical treatment,²⁴ and evidence that some patients desire physicians to raise spiritual issues.^6,25,26

We sought to explore the context of spiritual assessment rather than to further justify such assessments. The context of spiritual assessment refers to the philosophical question of whether physicians should address spiritual questions and to practical questions of how spiritual matters arise, how physicians approach them, and what barriers and facilitators they perceive with regard to discussing spirituality. Our study adds to knowledge about this context in several important ways.

We found variance of opinion concerning the physician’s role in spiritual assessment. Respondents reporting infrequent spiritual assessment expressed the view that spiritual issues have lower priority than other medical concerns. Yet those who regularly address spiritual issues justified this with scientific evidence associating spirituality and health. They also proposed a justification not found in previous studies: that spirituality is central to life and therefore important for its own sake rather than simply as a means to a medical end. These findings support and augment previously cited justifications for physicians assisting patients with spiritual health issues.^1-3,6,24-26

Our study results add to the list of categories that prompt discussions of spiritual issues. Respondents affirmed a role for physicians in discussing end-of-life issues and advanced care directives, as in previous studies.^27-29 In addition, they observed that patients’ spiritual questions arise from their unique responses to chronic illness, terminal illness, and life stressors. They identified 2 new categories prompting spiritual discussions: unexplained symptoms and treatment failure.

All respondents affirmed a role for physicians in supporting patients who initiate spiritual discussions. As in a previous study,¹⁴ they viewed themselves as facilitators and encouragers of patients’ spiritual values and as resources rather than as spiritual counselors. The most reticent physicians believed in responding to patients’ questions rather than initiating discussions, an approach that may fail to identify spiritual issues. All respondents supported a patient-centered approach to spiritual assessment in which physicians act with integrity and take care not to abuse their position.

Many physicians saw value in spiritual history taking, though none reported routine use of spiritual assessment tools. The potential Judeo-Christian bias in assessment questions noted by one respondent highlights the need to use culturally sensitive, generic assessment tools³⁰ and to work toward further development of such tools.

We identified new barriers to spiritual assessment, including a physician’s upbringing and culture, lack of spiritual inclination or awareness, resistance to exposing personal beliefs, and belief that spiritual discussions will not influence patients’ illnesses or lives. Respondents also postulated patient barriers, including fears that their physician might judge them for their spiritual views or consider their raising spiritual questions inappropriate.

We identified facilitators of spiritual discussions, such as communicating a willingness to have these discussions. One respondent noted that physicians whose lives are characterized by spiritual maturity might serve as agents of patients’ spiritual growth, consistent with a previous study’s themes of caregiver spirituality and physician vocation and mission.¹⁴

Limitations

Because qualitative research aims to uncover new perspectives rather than to make generalizable assessments, our findings may not apply to all physicians or to all family physicians. Although our respondents did not represent all major world religions, ethnic groups, and cultures, they did offer a diversity of spiritual and religious perspectives. Finally, our study gives only physicians’ perspectives. We are currently studying patients’ perspectives of situations that elicit spiritual questions and of potential barriers to spiritual assessment. We will use themes from our patient and physician qualitative studies to frame questions for a national patient questionnaire regarding physicians’ spiritual assessment.

Conclusions

Physicians differ in their comfort and practice of addressing spiritual issues with patients, but affirm a role for themselves in responding to patients’ spiritual concerns. Perceived barriers, physicians’ role definition, familiarity with factors likely to prompt spiritual questions, and the recognition of principles guiding spiritual discussions form the context for family physicians’ discussions of spiritual issues with patients. This context is important to consider when training medical students and residents in spiritual assessment. Careful attention to this context will also enhance the practicing physician’s skill in providing patient-centered assistance with spiritual health concerns.

Acknowledgments

The authors wish to acknowledge Arej Sawani, who assisted in data collection; Sheri Price, who assisted in manuscript preparation; and Richard Ellis, MD, MPH, Daniel Vinson, MD, MSPH, Steven Zweig, MD, MSPH, and Dale Smith, who reviewed the manuscript and offered editorial suggestions.

Insomnia, one of the most common complaints in primary care practice, affects more than 60 million Americans.¹ Inadequate sleep has been associated with reduced physical health,² subsequent decline in health status, and increased mortality.^3,4 Patients with chronic insomnia are more likely to develop affective disorders.^5,6 Insomnia may worsen somatic symptoms.⁷ Recent studies in health maintenance organization enrollees have demonstrated that insomnia is independently associated with significantly greater functional impairment, more days of disability related to health problems, and greater use of medical services.^8,9

Patients often present to primary care physicians with chronic comorbidities that may adversely affect sleep quality and that may also compromise functional status. While previous investigations have consistently shown decreased functional status in subjects with insomnia, these studies have tended to focus on general population samples,^10-12 health plan enrollees,^8,9,13 or volunteers recruited by media advertisement.¹⁴ In addition, most of the studies either lacked detailed assessment of medical or psychiatric comorbidities or did not control for the presence of these comorbidities in the analysis. Thus, our study aims to (1) determine whether insomnia is independently associated with decreased health-related quality of life (HRQOL) in patients with chronic conditions, and (2) compare the decrease in quality of life associated with insomnia with that associated with other chronic conditions.

Methods

We conducted a cross-sectional analysis of data from the Medical Outcomes Study (MOS), an observational study of health outcomes for patients with chronic medical and psychiatric conditions.^15-21

Sample and data collection

Study participants were English-speaking adults who had had an office visit with 1 of 523 clinicians trained in family practice, general internal medicine, cardiology, endocrinology, psychiatry, or clinical psychology during 9-day screening periods held from February to November 1986. Patients who were eligible for inclusion in the MOS baseline panel completed questionnaires addressing general health status (the Medical Outcomes Study Short Form Health Survey [SF-36]), alcohol use, exercise, and sleep.

Data from standardized physician-completed forms were used to identify patients with 5 index conditions: hypertension, diabetes, congestive heart failure, recent myocardial infarction, and depression.^15-20 We identified patients with depression using a short form of the Center for Epidemiologic Studies—Depression (CES-D) Scale²² and then used the National Institute of Mental Health Diagnostic Interview Schedule (DIS) to assess its severity.^18,22 The definition of current depressive disorder was based on Diagnostic and Statistical Manual, 3rd ed. (DSM-III ), criteria for lifetime major depression or dysthymia during the previous 12 months (DSM-IV was unavailable at the time of the MOS). Patients with depressive symptoms who did not satisfy this definition were considered to have subthreshold depression.²³

Measures of HRQOL

The SF-36 is a 36-item generic quality-of-life measure that assesses 8 domains: (1) physical functioning; (2) role limitation due to physical health problems (role physical); (3) bodily pain; (4) general health perceptions; (5) vitality; (6) social functioning; (7) role limitations due to emotional health problems (role emotional); and (8) mental health.^24-26 Physical function and role physical scales best distinguish between groups differing in severity of chronic medical conditions and have the purest interpretation with regard to physical health; mental health and role-emotional scales best distinguish between groups differing in severity of psychiatric disorders and have the purest interpretation with regard to mental health. Social function, vitality, and general health perception scales measure both physical and mental health status.²⁴ All health measures were scored on scales of 0 to 100, with higher scores indicating better health.

Measures of sleep

Insomnia was defined as the complaint of initiating or maintaining sleep (CIMS) using a 6-point categorical scale, with responses ranging from “all of the time” to “never.”²⁷ Mild insomnia was defined by responses to 2 of the MOS sleep items indicating difficulty in initiating or maintaining sleep over the prior 4 weeks “some” or “a good bit” of the time; severe insomnia was defined by difficulty “most” or “all” of the time. By inquiring about sleep over the prior 4 weeks, the 2-item measure of DSM-III captures key elements of the DSM-IV definition²⁸ and is similar to definitions used in other epidemiologic investigations of insomnia.^29,30

The measure provides a simple tool for the identification of chronic insomnia in busy primary care practice. In addition, the validity of this 2-item categorical measure is supported by excellent concordance between this measure and a continuous 5-item measure of insomnia (details available on request from the author). The 2-item measure, it is important to note, does not include the DSM-IV item pertaining to impairment of daytime functioning, as this item is closely related to the dependent variables of interest in this study.

Measures of potential confounders

We used an indicator of poverty by dichotomizing per capita household income (in 1985 dollars) at a cut-off point of 200% of the poverty level. Alcohol and smoking status were assessed with a 3-point scale: no history of use, past user, or current user. Frequency of exercise was assessed with the question, “How often do you exercise?” on a 6-point Likert-type scale, with responses ranging from “daily or almost daily” to “almost never or never.” Subjects were overweight if body mass index (BMI) exceeded 25 and obese if BMI exceeded 30.^31,32

We identified 16 common medical conditions comorbid to the index conditions by using data from the MOS standardized health examination. Data on medications were excluded because they do not reflect contemporary patterns of medication use and because of collinearity between medication and comorbidity variables already in the model.

Statistical analysis

We used multiple linear regression to identify the association between insomnia and HRQOL. We adjusted for sociodemographic characteristics, health habits, index conditions, severity of index conditions,¹⁹ a count of the 16 medical comorbidities,³⁴ and study location. To account for the potentially nonlinear relationship between age and HRQOL, we included 3 dummy variables for age: 40 to 55 years, 56 to 65 years, and older than 65 years (age younger than 40 years was the holdout category). Similarly, we included dummy variables for education (less than 12 years or exactly 12 years; more than 12 years was the holdout category) and exercise (at least 4 times a week or less than once a week; 1 to 3 times a week was the holdout category).

We report the average deviation in HRQOL values for mild and severe insomnia and for 2 comparison conditions: clinical depression and congestive heart failure (CHF). These conditions were selected because they are representative of conditions with predominant effects on physical functioning (CHF) and mental health (depression).^17,24,25 The average deviation in HRQOL is represented by the regression coefficients corresponding to the terms for these conditions in each HRQOL model.¹⁷ Because all patients in the current study had at least 1 of the 5 physician-identified conditions, we used the subgroup of patients with mild hypertension¹⁹ and without insomnia as the reference group.

We also performed a subset analysis of 2197 patients who had completed a screening version of the DIS for anxiety disorders (generalized anxiety disorder, phobia, or panic disorder) at the baseline health evaluation. In this patient subset, we constructed a series of regression models and examined the change in average deviation in HRQOL associated with insomnia with the addition of groups of covariates (sociodemographics, health habits, medical conditions, depression, and anxiety) to a base model including only insomnia and study location variables.

We also examined whether our results were robust by using logistic regression. Because most of the dependent variables are highly skewed (and thus may not satisfy the distributional assumptions of linear regression), we dichotomized each dependent variable as categorical (lowest tertile versus middle and upper tertiles) and determined the odds ratios for mild and severe insomnia associated with the lowest tertile of each HRQOL measure. Because we assessed the significance of both mild and severe insomnia in 8 different HRQOL domains, we used the Bonferroni correction to adjust for multiple comparisons (only P values ≥.003 were considered statistically significant).

Finally, we checked for selected 2-way interaction terms to determine whether the association between insomnia and HRQOL differed significantly by age, gender, race, education, and burden of comorbidity.^28,35,36

Results

Baesd on our definition of insomnia, 16% of study patients had severe insomnia and 34% had mild insomnia. Patients with mild and severe insomnia were more likely to be female, nonwhite, poor, unemployed, and unmarried. They were also more likely to have clinical (or subthreshold) depression, congestive heart failure, and a greater number of medical comorbidities (Table 1).

We found a significant independent association between insomnia and HRQOL, even after we had statistically controlled for sociodemographic characteristics, health habits, BMI, medical comorbidities, and the presence of clinical and subthreshold depression. All covariates were forced into each HRQOL model simultaneously. Insomnia showed the strongest associations with mental health, vitality, and social subscales (based on the Wald statistic).

Relative to patients without insomnia, increased severity of insomnia was associated with a progressively greater deviation in HRQOL (Table 2). The decreases in HRQOL observed for mild and severe insomnia were pervasive across all SF-36 domains and were similar when the sample was restricted to those patients seen by a primary care clinician (family physician or general internist, n = 2463). The magnitude of the average deviation in physical function for severe insomnia was comparable with that observed for CHF; similarly, the average deviation in the mental health domain for severe insomnia was about 60% that observed for clinical depression.

Mild and severe insomnia were associated with diminished HRQOL across all SF-36 domains using logistic regression (relative to patients without insomnia); however, the magnitude of associations for subscales differed somewhat. In these analyses, the adjusted odds ratios for mild and severe insomnia were greatest for mental health (3.5 and 10.2, respectively), vitality (2.4 and 7.4, respectively), and general health perceptions (2.0 and 5.1, respectively). Other subscale results are available on request from the author.

The incremental effect of controlling for chronic medical conditions, depression, and anxiety on the decrements in HRQOL associated with insomnia is demonstrated in Table 3 for selected domains. Addition of medical variables reduces the strength of association between insomnia and physical function by 28% and 28%, vitality by 17% and 18%, and mental health by 12% and 12% (for mild and severe insomnia, respectively). Subsequent addition of depression and anxiety variables reduces the strength of association between insomnia and mental health by 33% and 38%, vitality by 18% and 23%, and physical function by 11% and 10% (for mild and severe insomnia, respectively). Even after accounting for depression and anxiety, both mild and severe insomnia account for significantly decreased HRQOL in this subset analysis.

In tests of interactions, the association between insomnia and HRQOL was similar across age, gender, race, education, and comorbidity (≥3 versus <3 comorbid conditions). Of a total of 64 interaction terms across all SF-36 subscales, only 3 were statistically significant and no consistent pattern was observed; this is what would be expected by chance alone.

TABLE 1
PATIENT CHARACTERISTICS ACROSS INSOMNIA GROUPS

TABLE 2
AVERAGE DEVIATION IN HEALTH-RELATED QUALITY OF LIFE (HRQOL) DOMAINS ASSOCIATED WITH INSOMNIA IN THE MEDICAL OUTCOMES STUDY CROSS-SECTIONAL SAMPLE (N=3445)

TABLE 3
THE EFFECT OF ADDING SPECIFIC VARIABLE GROUPS ON THE AVERAGE DEVIATION IN HEALTH-RELATED QUALITY OF LIFE (HRQOL) DOMAINS ASSOCIATED WITH INSOMNIA AT BASELINE

Discussion

Physicians often ignore sleep disorders such as insomnia in clinical practice, in part because of increasing time constraints and because of poor preparation to deal with sleep disorders.^37,38 The National Commission on Sleep Disorders Research (NCSDR) reviewed 10,000 medical records from 9 family practice clinics, finding only 123 records that contained a note about the patient’s sleep; not a single record suggested an effective response to the patient’s sleep complaint.¹ Ignoring insomnia can have major implications for public health, however. Insomnia increases the risk of occupational injury and traffic accidents; a recent study concluded that driving while sleepy was at least as dangerous as the risk of driving under the influence of alcohol.³⁹ Similarly, a 1991 national survey showed that respondents with chronic insomnia are more likely to feel sleepy when driving, to have impaired concentration, and to have problems in accomplishing daily tasks.¹⁰

Our results extend the results of previous investigations by demonstrating that insomnia is independently associated with a significant decrease in overall quality of life for patients with chronic illness. The magnitude of this decrease for severe insomnia is comparable with that observed for chronic conditions such as CHF and clinical depression. We found that such decrements worsened with increased levels of sleep disturbance and that these decreased values persisted even after accounting for medical comorbidity, depression, and anxiety. These findings are similar to the results of a recent population-based investigation, in which insomnia was associated with global decreases in HRQOL, even after excluding subjects who met DSM-IV criteria for depression or anxiety.¹²

Another possible explanation for the negative association between insomnia and HRQOL is that patients with insomnia overreport functional impairment in a systematic fashion. This phenomenon has been described in depressed patients as “negative thinking bias,”⁴⁰ and it is possible that insomnia also leads to biased reporting. While it is difficult to rule this out, analyses of the health care behavior of patients with insomnia provides indirect evidence of the functional impairment of these patients, as measured by their increased use of medical and mental health services.^8,9,13

Limitations

The limitations of our study deserve comment. First, the MOS included a select group of physicians and their patients who agreed to participate in a comprehensive study of medical care. That the MOS sampled only patients with 5 chronic conditions who were insured and had a continuous relationship with a provider in 3 large urban areas limits the ability to generalize. Noteworthy, however, is that the prevalence of mild or severe insomnia in the MOS sample (50%) was comparable with that reported in a recent study of managed-care enrollees (46%).⁹

Second, this analysis is based on cross-sectional data. We cannot rule out the possibility that decreased quality of life leads to insomnia (ie, reverse causality). The causal relation of insomnia to diminished HRQOL is supported by longitudinal data suggesting that insomnia is an important precursor of depression.^5,41 Alternatively, insomnia may partially mediate the effect of chronic conditions such as CHF on HRQOL. Indeed, prior work has shown that worsening of chronic conditions tends to be associated with worsening of insomnia during follow-up, and vice versa.⁴²

Third, sleep problems were assessed by self-report. No attempt was made to validate our findings with polysomnographic measurements. Polysomnography has limited value in the evaluation of insomnia; many self-identified insomniacs do not show objective sleep abnormalities on polysomnography⁴³ and tend to have high night-to-night variability in the quality of their sleep.⁴⁴ Indeed, some discrepancies may exist between self-reported sleep and laboratory data.^45,46 Data obtained from MOS patients regarding other measures of sleep (including total sleep time, sleep latency, and adequacy of sleep) appear to verify the existence of sleep problems.⁴²

Fourth, the definition of depression was based on DSM-III criteria, which differ somewhat from those of DSM-IV. We note, however, that the primary emphasis in revising the DSM-III criteria for these conditions was to improve discrimination between major depression and dysthymia,⁴⁷ for both of which our analysis controlled. In addition, the diagnostic criteria for major depression and dysthymia incorporate sleep disturbance; thus, adjustment for depression would tend to reduce the estimated association between insomnia and HRQOL (as confirmed in Table 3).

Interpretation of the above findings should account for the nature of the study sample. Because all patients in the study sample had chronic conditions, average deviations in HRQOL scores were computed relative to the group with mild hypertension and without insomnia (reference group). Because even mild hypertension and its treatment may have an impact on several HRQOL domains,⁴⁸ the associations between insomnia and HRQOL in the current study would probably have been even stronger if the reference group comprised patients with no chronic conditions.

While depression and insomnia (one of the diagnostic criteria for depression) overlap considerably, pharmacologic and behavioral approaches, including light therapy^49-52 and exercise,⁵³ have potential to significantly lessen the burden of chronic insomnia and its effect on HRQOL.⁵⁴ In addition, clinicians should identify (and treat, if possible) predisposing conditions associated with insomnia.^28,36,42 Finally, future clinical trials involving treatment of insomnia should incorporate measures, such as the SF-36, that can detect clinically meaningful effects of insomnia on HRQOL in addition to standard laboratory and self-reported measures of sleep disturbance.

Conclusions

Even after accounting for the presence of chronic medical comorbidities, depression, and anxiety, insomnia is associated with significantly worsened HRQOL, particularly in the domains of mental health, vitality, and general health perceptions. By identifying and treating chronic insomnia, including any predisposing conditions that may perpetuate insomnia, clinicians can significantly improve the quality of life of patients with chronic illness.

Acknowledgments

Dr Katz’s work on this study was supported by institutional funds from the University of Wisconsin Medical School, Graduate School, and Department of Medicine. Dr McHorney’s work on this study was supported by the Department of Veterans Affairs (HSR&D HFP #96-001; RR&D C-2016; HSR&D IIR #95-033). Dr Katz is a Robert Wood Johnson Generalist Physician Faculty Scholar.

Insomnia, one of the most common complaints in primary care practice, affects more than 60 million Americans.¹ Inadequate sleep has been associated with reduced physical health,² subsequent decline in health status, and increased mortality.^3,4 Patients with chronic insomnia are more likely to develop affective disorders.^5,6 Insomnia may worsen somatic symptoms.⁷ Recent studies in health maintenance organization enrollees have demonstrated that insomnia is independently associated with significantly greater functional impairment, more days of disability related to health problems, and greater use of medical services.^8,9

Patients often present to primary care physicians with chronic comorbidities that may adversely affect sleep quality and that may also compromise functional status. While previous investigations have consistently shown decreased functional status in subjects with insomnia, these studies have tended to focus on general population samples,^10-12 health plan enrollees,^8,9,13 or volunteers recruited by media advertisement.¹⁴ In addition, most of the studies either lacked detailed assessment of medical or psychiatric comorbidities or did not control for the presence of these comorbidities in the analysis. Thus, our study aims to (1) determine whether insomnia is independently associated with decreased health-related quality of life (HRQOL) in patients with chronic conditions, and (2) compare the decrease in quality of life associated with insomnia with that associated with other chronic conditions.

Methods

We conducted a cross-sectional analysis of data from the Medical Outcomes Study (MOS), an observational study of health outcomes for patients with chronic medical and psychiatric conditions.^15-21

Sample and data collection

Study participants were English-speaking adults who had had an office visit with 1 of 523 clinicians trained in family practice, general internal medicine, cardiology, endocrinology, psychiatry, or clinical psychology during 9-day screening periods held from February to November 1986. Patients who were eligible for inclusion in the MOS baseline panel completed questionnaires addressing general health status (the Medical Outcomes Study Short Form Health Survey [SF-36]), alcohol use, exercise, and sleep.

Data from standardized physician-completed forms were used to identify patients with 5 index conditions: hypertension, diabetes, congestive heart failure, recent myocardial infarction, and depression.^15-20 We identified patients with depression using a short form of the Center for Epidemiologic Studies—Depression (CES-D) Scale²² and then used the National Institute of Mental Health Diagnostic Interview Schedule (DIS) to assess its severity.^18,22 The definition of current depressive disorder was based on Diagnostic and Statistical Manual, 3rd ed. (DSM-III ), criteria for lifetime major depression or dysthymia during the previous 12 months (DSM-IV was unavailable at the time of the MOS). Patients with depressive symptoms who did not satisfy this definition were considered to have subthreshold depression.²³

Measures of HRQOL

The SF-36 is a 36-item generic quality-of-life measure that assesses 8 domains: (1) physical functioning; (2) role limitation due to physical health problems (role physical); (3) bodily pain; (4) general health perceptions; (5) vitality; (6) social functioning; (7) role limitations due to emotional health problems (role emotional); and (8) mental health.^24-26 Physical function and role physical scales best distinguish between groups differing in severity of chronic medical conditions and have the purest interpretation with regard to physical health; mental health and role-emotional scales best distinguish between groups differing in severity of psychiatric disorders and have the purest interpretation with regard to mental health. Social function, vitality, and general health perception scales measure both physical and mental health status.²⁴ All health measures were scored on scales of 0 to 100, with higher scores indicating better health.

Measures of sleep

Insomnia was defined as the complaint of initiating or maintaining sleep (CIMS) using a 6-point categorical scale, with responses ranging from “all of the time” to “never.”²⁷ Mild insomnia was defined by responses to 2 of the MOS sleep items indicating difficulty in initiating or maintaining sleep over the prior 4 weeks “some” or “a good bit” of the time; severe insomnia was defined by difficulty “most” or “all” of the time. By inquiring about sleep over the prior 4 weeks, the 2-item measure of DSM-III captures key elements of the DSM-IV definition²⁸ and is similar to definitions used in other epidemiologic investigations of insomnia.^29,30

The measure provides a simple tool for the identification of chronic insomnia in busy primary care practice. In addition, the validity of this 2-item categorical measure is supported by excellent concordance between this measure and a continuous 5-item measure of insomnia (details available on request from the author). The 2-item measure, it is important to note, does not include the DSM-IV item pertaining to impairment of daytime functioning, as this item is closely related to the dependent variables of interest in this study.

Measures of potential confounders

We used an indicator of poverty by dichotomizing per capita household income (in 1985 dollars) at a cut-off point of 200% of the poverty level. Alcohol and smoking status were assessed with a 3-point scale: no history of use, past user, or current user. Frequency of exercise was assessed with the question, “How often do you exercise?” on a 6-point Likert-type scale, with responses ranging from “daily or almost daily” to “almost never or never.” Subjects were overweight if body mass index (BMI) exceeded 25 and obese if BMI exceeded 30.^31,32

We identified 16 common medical conditions comorbid to the index conditions by using data from the MOS standardized health examination. Data on medications were excluded because they do not reflect contemporary patterns of medication use and because of collinearity between medication and comorbidity variables already in the model.

Statistical analysis

We used multiple linear regression to identify the association between insomnia and HRQOL. We adjusted for sociodemographic characteristics, health habits, index conditions, severity of index conditions,¹⁹ a count of the 16 medical comorbidities,³⁴ and study location. To account for the potentially nonlinear relationship between age and HRQOL, we included 3 dummy variables for age: 40 to 55 years, 56 to 65 years, and older than 65 years (age younger than 40 years was the holdout category). Similarly, we included dummy variables for education (less than 12 years or exactly 12 years; more than 12 years was the holdout category) and exercise (at least 4 times a week or less than once a week; 1 to 3 times a week was the holdout category).

We report the average deviation in HRQOL values for mild and severe insomnia and for 2 comparison conditions: clinical depression and congestive heart failure (CHF). These conditions were selected because they are representative of conditions with predominant effects on physical functioning (CHF) and mental health (depression).^17,24,25 The average deviation in HRQOL is represented by the regression coefficients corresponding to the terms for these conditions in each HRQOL model.¹⁷ Because all patients in the current study had at least 1 of the 5 physician-identified conditions, we used the subgroup of patients with mild hypertension¹⁹ and without insomnia as the reference group.

We also performed a subset analysis of 2197 patients who had completed a screening version of the DIS for anxiety disorders (generalized anxiety disorder, phobia, or panic disorder) at the baseline health evaluation. In this patient subset, we constructed a series of regression models and examined the change in average deviation in HRQOL associated with insomnia with the addition of groups of covariates (sociodemographics, health habits, medical conditions, depression, and anxiety) to a base model including only insomnia and study location variables.

We also examined whether our results were robust by using logistic regression. Because most of the dependent variables are highly skewed (and thus may not satisfy the distributional assumptions of linear regression), we dichotomized each dependent variable as categorical (lowest tertile versus middle and upper tertiles) and determined the odds ratios for mild and severe insomnia associated with the lowest tertile of each HRQOL measure. Because we assessed the significance of both mild and severe insomnia in 8 different HRQOL domains, we used the Bonferroni correction to adjust for multiple comparisons (only P values ≥.003 were considered statistically significant).

Finally, we checked for selected 2-way interaction terms to determine whether the association between insomnia and HRQOL differed significantly by age, gender, race, education, and burden of comorbidity.^28,35,36

Results

Baesd on our definition of insomnia, 16% of study patients had severe insomnia and 34% had mild insomnia. Patients with mild and severe insomnia were more likely to be female, nonwhite, poor, unemployed, and unmarried. They were also more likely to have clinical (or subthreshold) depression, congestive heart failure, and a greater number of medical comorbidities (Table 1).

We found a significant independent association between insomnia and HRQOL, even after we had statistically controlled for sociodemographic characteristics, health habits, BMI, medical comorbidities, and the presence of clinical and subthreshold depression. All covariates were forced into each HRQOL model simultaneously. Insomnia showed the strongest associations with mental health, vitality, and social subscales (based on the Wald statistic).

Relative to patients without insomnia, increased severity of insomnia was associated with a progressively greater deviation in HRQOL (Table 2). The decreases in HRQOL observed for mild and severe insomnia were pervasive across all SF-36 domains and were similar when the sample was restricted to those patients seen by a primary care clinician (family physician or general internist, n = 2463). The magnitude of the average deviation in physical function for severe insomnia was comparable with that observed for CHF; similarly, the average deviation in the mental health domain for severe insomnia was about 60% that observed for clinical depression.

Mild and severe insomnia were associated with diminished HRQOL across all SF-36 domains using logistic regression (relative to patients without insomnia); however, the magnitude of associations for subscales differed somewhat. In these analyses, the adjusted odds ratios for mild and severe insomnia were greatest for mental health (3.5 and 10.2, respectively), vitality (2.4 and 7.4, respectively), and general health perceptions (2.0 and 5.1, respectively). Other subscale results are available on request from the author.

The incremental effect of controlling for chronic medical conditions, depression, and anxiety on the decrements in HRQOL associated with insomnia is demonstrated in Table 3 for selected domains. Addition of medical variables reduces the strength of association between insomnia and physical function by 28% and 28%, vitality by 17% and 18%, and mental health by 12% and 12% (for mild and severe insomnia, respectively). Subsequent addition of depression and anxiety variables reduces the strength of association between insomnia and mental health by 33% and 38%, vitality by 18% and 23%, and physical function by 11% and 10% (for mild and severe insomnia, respectively). Even after accounting for depression and anxiety, both mild and severe insomnia account for significantly decreased HRQOL in this subset analysis.

In tests of interactions, the association between insomnia and HRQOL was similar across age, gender, race, education, and comorbidity (≥3 versus <3 comorbid conditions). Of a total of 64 interaction terms across all SF-36 subscales, only 3 were statistically significant and no consistent pattern was observed; this is what would be expected by chance alone.

TABLE 1
PATIENT CHARACTERISTICS ACROSS INSOMNIA GROUPS

TABLE 2
AVERAGE DEVIATION IN HEALTH-RELATED QUALITY OF LIFE (HRQOL) DOMAINS ASSOCIATED WITH INSOMNIA IN THE MEDICAL OUTCOMES STUDY CROSS-SECTIONAL SAMPLE (N=3445)

TABLE 3
THE EFFECT OF ADDING SPECIFIC VARIABLE GROUPS ON THE AVERAGE DEVIATION IN HEALTH-RELATED QUALITY OF LIFE (HRQOL) DOMAINS ASSOCIATED WITH INSOMNIA AT BASELINE

Discussion

Physicians often ignore sleep disorders such as insomnia in clinical practice, in part because of increasing time constraints and because of poor preparation to deal with sleep disorders.^37,38 The National Commission on Sleep Disorders Research (NCSDR) reviewed 10,000 medical records from 9 family practice clinics, finding only 123 records that contained a note about the patient’s sleep; not a single record suggested an effective response to the patient’s sleep complaint.¹ Ignoring insomnia can have major implications for public health, however. Insomnia increases the risk of occupational injury and traffic accidents; a recent study concluded that driving while sleepy was at least as dangerous as the risk of driving under the influence of alcohol.³⁹ Similarly, a 1991 national survey showed that respondents with chronic insomnia are more likely to feel sleepy when driving, to have impaired concentration, and to have problems in accomplishing daily tasks.¹⁰

Our results extend the results of previous investigations by demonstrating that insomnia is independently associated with a significant decrease in overall quality of life for patients with chronic illness. The magnitude of this decrease for severe insomnia is comparable with that observed for chronic conditions such as CHF and clinical depression. We found that such decrements worsened with increased levels of sleep disturbance and that these decreased values persisted even after accounting for medical comorbidity, depression, and anxiety. These findings are similar to the results of a recent population-based investigation, in which insomnia was associated with global decreases in HRQOL, even after excluding subjects who met DSM-IV criteria for depression or anxiety.¹²

Another possible explanation for the negative association between insomnia and HRQOL is that patients with insomnia overreport functional impairment in a systematic fashion. This phenomenon has been described in depressed patients as “negative thinking bias,”⁴⁰ and it is possible that insomnia also leads to biased reporting. While it is difficult to rule this out, analyses of the health care behavior of patients with insomnia provides indirect evidence of the functional impairment of these patients, as measured by their increased use of medical and mental health services.^8,9,13

Limitations

The limitations of our study deserve comment. First, the MOS included a select group of physicians and their patients who agreed to participate in a comprehensive study of medical care. That the MOS sampled only patients with 5 chronic conditions who were insured and had a continuous relationship with a provider in 3 large urban areas limits the ability to generalize. Noteworthy, however, is that the prevalence of mild or severe insomnia in the MOS sample (50%) was comparable with that reported in a recent study of managed-care enrollees (46%).⁹

Second, this analysis is based on cross-sectional data. We cannot rule out the possibility that decreased quality of life leads to insomnia (ie, reverse causality). The causal relation of insomnia to diminished HRQOL is supported by longitudinal data suggesting that insomnia is an important precursor of depression.^5,41 Alternatively, insomnia may partially mediate the effect of chronic conditions such as CHF on HRQOL. Indeed, prior work has shown that worsening of chronic conditions tends to be associated with worsening of insomnia during follow-up, and vice versa.⁴²

Third, sleep problems were assessed by self-report. No attempt was made to validate our findings with polysomnographic measurements. Polysomnography has limited value in the evaluation of insomnia; many self-identified insomniacs do not show objective sleep abnormalities on polysomnography⁴³ and tend to have high night-to-night variability in the quality of their sleep.⁴⁴ Indeed, some discrepancies may exist between self-reported sleep and laboratory data.^45,46 Data obtained from MOS patients regarding other measures of sleep (including total sleep time, sleep latency, and adequacy of sleep) appear to verify the existence of sleep problems.⁴²

Fourth, the definition of depression was based on DSM-III criteria, which differ somewhat from those of DSM-IV. We note, however, that the primary emphasis in revising the DSM-III criteria for these conditions was to improve discrimination between major depression and dysthymia,⁴⁷ for both of which our analysis controlled. In addition, the diagnostic criteria for major depression and dysthymia incorporate sleep disturbance; thus, adjustment for depression would tend to reduce the estimated association between insomnia and HRQOL (as confirmed in Table 3).

Interpretation of the above findings should account for the nature of the study sample. Because all patients in the study sample had chronic conditions, average deviations in HRQOL scores were computed relative to the group with mild hypertension and without insomnia (reference group). Because even mild hypertension and its treatment may have an impact on several HRQOL domains,⁴⁸ the associations between insomnia and HRQOL in the current study would probably have been even stronger if the reference group comprised patients with no chronic conditions.

While depression and insomnia (one of the diagnostic criteria for depression) overlap considerably, pharmacologic and behavioral approaches, including light therapy^49-52 and exercise,⁵³ have potential to significantly lessen the burden of chronic insomnia and its effect on HRQOL.⁵⁴ In addition, clinicians should identify (and treat, if possible) predisposing conditions associated with insomnia.^28,36,42 Finally, future clinical trials involving treatment of insomnia should incorporate measures, such as the SF-36, that can detect clinically meaningful effects of insomnia on HRQOL in addition to standard laboratory and self-reported measures of sleep disturbance.

Conclusions

Even after accounting for the presence of chronic medical comorbidities, depression, and anxiety, insomnia is associated with significantly worsened HRQOL, particularly in the domains of mental health, vitality, and general health perceptions. By identifying and treating chronic insomnia, including any predisposing conditions that may perpetuate insomnia, clinicians can significantly improve the quality of life of patients with chronic illness.

Acknowledgments

Dr Katz’s work on this study was supported by institutional funds from the University of Wisconsin Medical School, Graduate School, and Department of Medicine. Dr McHorney’s work on this study was supported by the Department of Veterans Affairs (HSR&D HFP #96-001; RR&D C-2016; HSR&D IIR #95-033). Dr Katz is a Robert Wood Johnson Generalist Physician Faculty Scholar.

Insomnia, one of the most common complaints in primary care practice, affects more than 60 million Americans.¹ Inadequate sleep has been associated with reduced physical health,² subsequent decline in health status, and increased mortality.^3,4 Patients with chronic insomnia are more likely to develop affective disorders.^5,6 Insomnia may worsen somatic symptoms.⁷ Recent studies in health maintenance organization enrollees have demonstrated that insomnia is independently associated with significantly greater functional impairment, more days of disability related to health problems, and greater use of medical services.^8,9

Patients often present to primary care physicians with chronic comorbidities that may adversely affect sleep quality and that may also compromise functional status. While previous investigations have consistently shown decreased functional status in subjects with insomnia, these studies have tended to focus on general population samples,^10-12 health plan enrollees,^8,9,13 or volunteers recruited by media advertisement.¹⁴ In addition, most of the studies either lacked detailed assessment of medical or psychiatric comorbidities or did not control for the presence of these comorbidities in the analysis. Thus, our study aims to (1) determine whether insomnia is independently associated with decreased health-related quality of life (HRQOL) in patients with chronic conditions, and (2) compare the decrease in quality of life associated with insomnia with that associated with other chronic conditions.

Methods

We conducted a cross-sectional analysis of data from the Medical Outcomes Study (MOS), an observational study of health outcomes for patients with chronic medical and psychiatric conditions.^15-21

Sample and data collection

Study participants were English-speaking adults who had had an office visit with 1 of 523 clinicians trained in family practice, general internal medicine, cardiology, endocrinology, psychiatry, or clinical psychology during 9-day screening periods held from February to November 1986. Patients who were eligible for inclusion in the MOS baseline panel completed questionnaires addressing general health status (the Medical Outcomes Study Short Form Health Survey [SF-36]), alcohol use, exercise, and sleep.

Data from standardized physician-completed forms were used to identify patients with 5 index conditions: hypertension, diabetes, congestive heart failure, recent myocardial infarction, and depression.^15-20 We identified patients with depression using a short form of the Center for Epidemiologic Studies—Depression (CES-D) Scale²² and then used the National Institute of Mental Health Diagnostic Interview Schedule (DIS) to assess its severity.^18,22 The definition of current depressive disorder was based on Diagnostic and Statistical Manual, 3rd ed. (DSM-III ), criteria for lifetime major depression or dysthymia during the previous 12 months (DSM-IV was unavailable at the time of the MOS). Patients with depressive symptoms who did not satisfy this definition were considered to have subthreshold depression.²³

Measures of HRQOL

The SF-36 is a 36-item generic quality-of-life measure that assesses 8 domains: (1) physical functioning; (2) role limitation due to physical health problems (role physical); (3) bodily pain; (4) general health perceptions; (5) vitality; (6) social functioning; (7) role limitations due to emotional health problems (role emotional); and (8) mental health.^24-26 Physical function and role physical scales best distinguish between groups differing in severity of chronic medical conditions and have the purest interpretation with regard to physical health; mental health and role-emotional scales best distinguish between groups differing in severity of psychiatric disorders and have the purest interpretation with regard to mental health. Social function, vitality, and general health perception scales measure both physical and mental health status.²⁴ All health measures were scored on scales of 0 to 100, with higher scores indicating better health.

Measures of sleep

Insomnia was defined as the complaint of initiating or maintaining sleep (CIMS) using a 6-point categorical scale, with responses ranging from “all of the time” to “never.”²⁷ Mild insomnia was defined by responses to 2 of the MOS sleep items indicating difficulty in initiating or maintaining sleep over the prior 4 weeks “some” or “a good bit” of the time; severe insomnia was defined by difficulty “most” or “all” of the time. By inquiring about sleep over the prior 4 weeks, the 2-item measure of DSM-III captures key elements of the DSM-IV definition²⁸ and is similar to definitions used in other epidemiologic investigations of insomnia.^29,30

The measure provides a simple tool for the identification of chronic insomnia in busy primary care practice. In addition, the validity of this 2-item categorical measure is supported by excellent concordance between this measure and a continuous 5-item measure of insomnia (details available on request from the author). The 2-item measure, it is important to note, does not include the DSM-IV item pertaining to impairment of daytime functioning, as this item is closely related to the dependent variables of interest in this study.

Measures of potential confounders

We used an indicator of poverty by dichotomizing per capita household income (in 1985 dollars) at a cut-off point of 200% of the poverty level. Alcohol and smoking status were assessed with a 3-point scale: no history of use, past user, or current user. Frequency of exercise was assessed with the question, “How often do you exercise?” on a 6-point Likert-type scale, with responses ranging from “daily or almost daily” to “almost never or never.” Subjects were overweight if body mass index (BMI) exceeded 25 and obese if BMI exceeded 30.^31,32

We identified 16 common medical conditions comorbid to the index conditions by using data from the MOS standardized health examination. Data on medications were excluded because they do not reflect contemporary patterns of medication use and because of collinearity between medication and comorbidity variables already in the model.

Statistical analysis

We used multiple linear regression to identify the association between insomnia and HRQOL. We adjusted for sociodemographic characteristics, health habits, index conditions, severity of index conditions,¹⁹ a count of the 16 medical comorbidities,³⁴ and study location. To account for the potentially nonlinear relationship between age and HRQOL, we included 3 dummy variables for age: 40 to 55 years, 56 to 65 years, and older than 65 years (age younger than 40 years was the holdout category). Similarly, we included dummy variables for education (less than 12 years or exactly 12 years; more than 12 years was the holdout category) and exercise (at least 4 times a week or less than once a week; 1 to 3 times a week was the holdout category).

We report the average deviation in HRQOL values for mild and severe insomnia and for 2 comparison conditions: clinical depression and congestive heart failure (CHF). These conditions were selected because they are representative of conditions with predominant effects on physical functioning (CHF) and mental health (depression).^17,24,25 The average deviation in HRQOL is represented by the regression coefficients corresponding to the terms for these conditions in each HRQOL model.¹⁷ Because all patients in the current study had at least 1 of the 5 physician-identified conditions, we used the subgroup of patients with mild hypertension¹⁹ and without insomnia as the reference group.

We also performed a subset analysis of 2197 patients who had completed a screening version of the DIS for anxiety disorders (generalized anxiety disorder, phobia, or panic disorder) at the baseline health evaluation. In this patient subset, we constructed a series of regression models and examined the change in average deviation in HRQOL associated with insomnia with the addition of groups of covariates (sociodemographics, health habits, medical conditions, depression, and anxiety) to a base model including only insomnia and study location variables.

We also examined whether our results were robust by using logistic regression. Because most of the dependent variables are highly skewed (and thus may not satisfy the distributional assumptions of linear regression), we dichotomized each dependent variable as categorical (lowest tertile versus middle and upper tertiles) and determined the odds ratios for mild and severe insomnia associated with the lowest tertile of each HRQOL measure. Because we assessed the significance of both mild and severe insomnia in 8 different HRQOL domains, we used the Bonferroni correction to adjust for multiple comparisons (only P values ≥.003 were considered statistically significant).

Finally, we checked for selected 2-way interaction terms to determine whether the association between insomnia and HRQOL differed significantly by age, gender, race, education, and burden of comorbidity.^28,35,36

Results

Baesd on our definition of insomnia, 16% of study patients had severe insomnia and 34% had mild insomnia. Patients with mild and severe insomnia were more likely to be female, nonwhite, poor, unemployed, and unmarried. They were also more likely to have clinical (or subthreshold) depression, congestive heart failure, and a greater number of medical comorbidities (Table 1).

We found a significant independent association between insomnia and HRQOL, even after we had statistically controlled for sociodemographic characteristics, health habits, BMI, medical comorbidities, and the presence of clinical and subthreshold depression. All covariates were forced into each HRQOL model simultaneously. Insomnia showed the strongest associations with mental health, vitality, and social subscales (based on the Wald statistic).

Relative to patients without insomnia, increased severity of insomnia was associated with a progressively greater deviation in HRQOL (Table 2). The decreases in HRQOL observed for mild and severe insomnia were pervasive across all SF-36 domains and were similar when the sample was restricted to those patients seen by a primary care clinician (family physician or general internist, n = 2463). The magnitude of the average deviation in physical function for severe insomnia was comparable with that observed for CHF; similarly, the average deviation in the mental health domain for severe insomnia was about 60% that observed for clinical depression.

Mild and severe insomnia were associated with diminished HRQOL across all SF-36 domains using logistic regression (relative to patients without insomnia); however, the magnitude of associations for subscales differed somewhat. In these analyses, the adjusted odds ratios for mild and severe insomnia were greatest for mental health (3.5 and 10.2, respectively), vitality (2.4 and 7.4, respectively), and general health perceptions (2.0 and 5.1, respectively). Other subscale results are available on request from the author.

The incremental effect of controlling for chronic medical conditions, depression, and anxiety on the decrements in HRQOL associated with insomnia is demonstrated in Table 3 for selected domains. Addition of medical variables reduces the strength of association between insomnia and physical function by 28% and 28%, vitality by 17% and 18%, and mental health by 12% and 12% (for mild and severe insomnia, respectively). Subsequent addition of depression and anxiety variables reduces the strength of association between insomnia and mental health by 33% and 38%, vitality by 18% and 23%, and physical function by 11% and 10% (for mild and severe insomnia, respectively). Even after accounting for depression and anxiety, both mild and severe insomnia account for significantly decreased HRQOL in this subset analysis.

In tests of interactions, the association between insomnia and HRQOL was similar across age, gender, race, education, and comorbidity (≥3 versus <3 comorbid conditions). Of a total of 64 interaction terms across all SF-36 subscales, only 3 were statistically significant and no consistent pattern was observed; this is what would be expected by chance alone.

TABLE 1
PATIENT CHARACTERISTICS ACROSS INSOMNIA GROUPS

TABLE 2
AVERAGE DEVIATION IN HEALTH-RELATED QUALITY OF LIFE (HRQOL) DOMAINS ASSOCIATED WITH INSOMNIA IN THE MEDICAL OUTCOMES STUDY CROSS-SECTIONAL SAMPLE (N=3445)

TABLE 3
THE EFFECT OF ADDING SPECIFIC VARIABLE GROUPS ON THE AVERAGE DEVIATION IN HEALTH-RELATED QUALITY OF LIFE (HRQOL) DOMAINS ASSOCIATED WITH INSOMNIA AT BASELINE

Discussion

Physicians often ignore sleep disorders such as insomnia in clinical practice, in part because of increasing time constraints and because of poor preparation to deal with sleep disorders.^37,38 The National Commission on Sleep Disorders Research (NCSDR) reviewed 10,000 medical records from 9 family practice clinics, finding only 123 records that contained a note about the patient’s sleep; not a single record suggested an effective response to the patient’s sleep complaint.¹ Ignoring insomnia can have major implications for public health, however. Insomnia increases the risk of occupational injury and traffic accidents; a recent study concluded that driving while sleepy was at least as dangerous as the risk of driving under the influence of alcohol.³⁹ Similarly, a 1991 national survey showed that respondents with chronic insomnia are more likely to feel sleepy when driving, to have impaired concentration, and to have problems in accomplishing daily tasks.¹⁰

Our results extend the results of previous investigations by demonstrating that insomnia is independently associated with a significant decrease in overall quality of life for patients with chronic illness. The magnitude of this decrease for severe insomnia is comparable with that observed for chronic conditions such as CHF and clinical depression. We found that such decrements worsened with increased levels of sleep disturbance and that these decreased values persisted even after accounting for medical comorbidity, depression, and anxiety. These findings are similar to the results of a recent population-based investigation, in which insomnia was associated with global decreases in HRQOL, even after excluding subjects who met DSM-IV criteria for depression or anxiety.¹²

Another possible explanation for the negative association between insomnia and HRQOL is that patients with insomnia overreport functional impairment in a systematic fashion. This phenomenon has been described in depressed patients as “negative thinking bias,”⁴⁰ and it is possible that insomnia also leads to biased reporting. While it is difficult to rule this out, analyses of the health care behavior of patients with insomnia provides indirect evidence of the functional impairment of these patients, as measured by their increased use of medical and mental health services.^8,9,13

Limitations

The limitations of our study deserve comment. First, the MOS included a select group of physicians and their patients who agreed to participate in a comprehensive study of medical care. That the MOS sampled only patients with 5 chronic conditions who were insured and had a continuous relationship with a provider in 3 large urban areas limits the ability to generalize. Noteworthy, however, is that the prevalence of mild or severe insomnia in the MOS sample (50%) was comparable with that reported in a recent study of managed-care enrollees (46%).⁹

Second, this analysis is based on cross-sectional data. We cannot rule out the possibility that decreased quality of life leads to insomnia (ie, reverse causality). The causal relation of insomnia to diminished HRQOL is supported by longitudinal data suggesting that insomnia is an important precursor of depression.^5,41 Alternatively, insomnia may partially mediate the effect of chronic conditions such as CHF on HRQOL. Indeed, prior work has shown that worsening of chronic conditions tends to be associated with worsening of insomnia during follow-up, and vice versa.⁴²

Third, sleep problems were assessed by self-report. No attempt was made to validate our findings with polysomnographic measurements. Polysomnography has limited value in the evaluation of insomnia; many self-identified insomniacs do not show objective sleep abnormalities on polysomnography⁴³ and tend to have high night-to-night variability in the quality of their sleep.⁴⁴ Indeed, some discrepancies may exist between self-reported sleep and laboratory data.^45,46 Data obtained from MOS patients regarding other measures of sleep (including total sleep time, sleep latency, and adequacy of sleep) appear to verify the existence of sleep problems.⁴²

Fourth, the definition of depression was based on DSM-III criteria, which differ somewhat from those of DSM-IV. We note, however, that the primary emphasis in revising the DSM-III criteria for these conditions was to improve discrimination between major depression and dysthymia,⁴⁷ for both of which our analysis controlled. In addition, the diagnostic criteria for major depression and dysthymia incorporate sleep disturbance; thus, adjustment for depression would tend to reduce the estimated association between insomnia and HRQOL (as confirmed in Table 3).

Interpretation of the above findings should account for the nature of the study sample. Because all patients in the study sample had chronic conditions, average deviations in HRQOL scores were computed relative to the group with mild hypertension and without insomnia (reference group). Because even mild hypertension and its treatment may have an impact on several HRQOL domains,⁴⁸ the associations between insomnia and HRQOL in the current study would probably have been even stronger if the reference group comprised patients with no chronic conditions.

While depression and insomnia (one of the diagnostic criteria for depression) overlap considerably, pharmacologic and behavioral approaches, including light therapy^49-52 and exercise,⁵³ have potential to significantly lessen the burden of chronic insomnia and its effect on HRQOL.⁵⁴ In addition, clinicians should identify (and treat, if possible) predisposing conditions associated with insomnia.^28,36,42 Finally, future clinical trials involving treatment of insomnia should incorporate measures, such as the SF-36, that can detect clinically meaningful effects of insomnia on HRQOL in addition to standard laboratory and self-reported measures of sleep disturbance.

Conclusions

Even after accounting for the presence of chronic medical comorbidities, depression, and anxiety, insomnia is associated with significantly worsened HRQOL, particularly in the domains of mental health, vitality, and general health perceptions. By identifying and treating chronic insomnia, including any predisposing conditions that may perpetuate insomnia, clinicians can significantly improve the quality of life of patients with chronic illness.

Acknowledgments

Dr Katz’s work on this study was supported by institutional funds from the University of Wisconsin Medical School, Graduate School, and Department of Medicine. Dr McHorney’s work on this study was supported by the Department of Veterans Affairs (HSR&D HFP #96-001; RR&D C-2016; HSR&D IIR #95-033). Dr Katz is a Robert Wood Johnson Generalist Physician Faculty Scholar.

The common cold, usually caused by viral infection of the upper respiratory tract, is a very prevalent illness. On average, US adults suffer from 1 to 4 episodes per year.^1-3 This high incidence, along with significant symptomatic and functional impairment, combine to make this syndrome an important health problem. Hundreds of trials have attempted to demonstrate effective treatments.^4,5 Unfortunately, few efforts have been made to develop and validate instruments to measure the symptomatic and functional impact of the common cold.

The term “upper respiratory infection” (URI) is a nosologic category constructed by physicians and other health professionals to reflect an upper airway, mucus-producing, inflammatory reaction to infection, usually viral. It is a disease category. The terms rhinitis, rhinosinusitis, pharyngitis, and bronchitis are often used to indicate the anatomic area most affected. The term “common cold” is an illness term constructed and used by the general populace. This distinction between professional (disease) and popular (illness) conceptions⁶ provides the reasoning for participant-based, patient-oriented qualitative development of measurement tools. While many medical professionals may choose to measure URI disease by physical examination, viral culture, or laboratory analysis of blood or nasal discharge, we believe that most people are more interested in how they can reduce the severity and duration of their symptoms and the functional impairments that result from their illness.

George Gee Jackson and colleagues⁷ began experimental work in the 1950s, observing and recording the cold symptoms produced by challenging more than 1,000 volunteers with filtered nasal secretions obtained from cold-sufferers. Eight symptoms–sneezing, headache, malaise, chilliness, nasal discharge, nasal obstruction, sore throat, and cough–were selected for evaluation and graded as absent (0), mild (1), moderate (2), or severe (3) every day for 6 days after inoculation. A score of 14 or higher was chosen as the cutoff value that best distinguished infected from noninfected participants. Thus, the original Jackson scale was apparently designed to discriminate between those with and without demonstrable viral infection, and not as a measure of severity. The tables and graphics in Jackson’s seminal works point toward reasonable internal consistency and discriminate validity.^7-9 However, other important measurement properties, such as precision, reliability, responsiveness, and stability, were not reported. Despite these limitations, Jackson’s scale has been used for decades by most of the major common cold research groups.^10-15

Using various modifications of the Jackson scale, researchers of the cold have characterized the frequency and severity of the 8 symptoms noted above in both natural colds and experimentally induced rhinovirus infections. Variability in symptom expression remains a hallmark of URI. Although specific pathogens are associated with the severity and distribution of symptoms at the population level, symptoms are poor predictors of etiology at the individual level. Infection itself is an imperfect predictor of symptom expression, as asymptomatic infections occur frequently, and as URI-like symptoms occur in people in whom it is not possible to demonstrate infections.¹⁶ Even among people with documented experimental infections of single strains of virus, variance outweighs central tendency in all symptom measurements.^17,18

The search for objective disease measures with which to compare symptom scores has also progressed. To date, the following measures have been evaluated: detection of virus with culture or polymerase chain reaction,^16,19 cytokine measurement,^20-22 serologic markers,²³ physical examination,^9,24 radiologic imaging,^25,26 rhinomanometry,^27,28 mucus weight,²⁹ mucus velocity, and number of tissues used.³⁰ None have been shown to be superior to self-reported symptoms in terms of precision, reliability, or responsiveness or in their ability to predict functional impairment or subsequent illness. Perhaps more important, none have been shown to reflect the values of the people who experience colds. Although a number of quality-of-life instruments have been developed to assess allergic rhinitis,^31-36 we have been unable to locate any specifically developed to assess URI.

We therefore decided to develop the Wisconsin Upper Respiratory Symptom Survey (WURSS) to provide a standardized measure for evaluating the negative consequences of the common cold. We were particularly interested in developing a health-related quality of life instrument that would represent the symptomatic and functional dimensions that are important to cold-sufferers.^37-40 The instrument should be able to discriminate accurately between active intervention and placebo effects in randomized therapeutic trials and should balance brevity and ease of use with optimal precision, reliability, and responsiveness.^41-43 It should be based on self-diagnosis and self-assessment because neither accepted criteria nor adequate tests are available to diagnose “upper respiratory infection” or “acute infectious rhinosinusitis”(with or without “pharyngitis”) and because the vast majority of cold treatments will be taken without professional input after self-diagnosis.

Methods

Phase 1: Initial development during a randomized trial

The development of this study began in 1998 during the design of a randomized controlled trial (RCT) of echinacea as a cold treatment. We created our first instrument by showing successive drafts to friends and colleagues (mostly family physicians), stopping once we were satisfied that the questionnaire had reasonable face validity. This initial instrument rated global severity of illness (“How sick do you feel today?”) and 15 individual symptoms on a 9-point Likert-type scale. The 15 symptom-measuring items were complemented by 4 dichotomous (yes/no) functional outcome questions, adapted with permission from the validated Medical Outcomes Study 36-item Short-Form Health Survey (SF-36).⁴⁴

This initial instrument was used in the spring of 1999 in the echinacea RCT.⁴⁵ This experience provided a good initial test of our instrument, as the participants were recruited within 36 hours of their first symptom and monitored each day until they had answered “No” to the question, “Do you think that you are still sick today?” for 2 days in a row. Each participant was asked to fill out the questionnaire both on paper and on a computerized data-collecting facsimile (available at http://www.fammed.wisc.edu/samplecold).

Phase 2: Further instrument development using qualitative methods

After the echinacea RCT was completed, our primary concern was that we might be overlooking or under-representing important illness domains. We also suspected that wording, question order, response range, and other formatting concerns could be improved. To achieve these goals, we used qualitative instrument-development methodologies, involving the people we wanted to measure–cold-sufferers–in the development process.^40,46-50

After obtaining approval from the University of Wisconsin Medical School Human Subjects Committee, we began interviewing Madison-area adults who responded to community advertising asking for volunteers with colds. Inclusion criteria required answering “Yes” to the question, “Do you believe that you have a cold?” For an interview to be arranged, at least one cold symptom had to be present, and the research assistant had to be convinced that the caller was indeed suffering from a common cold. Prospective participants with itchy eyes, sneezing, or a history of allergy were excluded if either the participant or the interviewer felt that any current symptoms might have been caused by allergy. Interviews were held in a location of mutual convenience and with the aid of an interview guide developed by our research team. Interviewers were carefully trained in the research protocol and used interview guides for both the initial telephone screen and the in-person interviews. Interviewers included both clinicians and nonclinicians.

The semistructured interview guide used open-ended questions designed to elicit the participants’ own terminology for describing their colds (Table 1).^51-53 We aimed for an understanding of how the experience of the cold influenced the lives of the participants. Participants were first asked to list all their symptoms, then to describe how each symptom bothered them. Next, we asked which symptom(s) appeared first and which one(s) followed. We then asked which symptoms were most bothersome and why. Participants were asked to describe what they did to relieve their cold symptoms, why, and whether the therapy provided any relief. Participants were then asked about how their cold affected their lifestyle with regard to work, relationships, activities, and so forth. Additionally, we asked about symptoms and effects of previous colds. This exploratory phase of the interview lasted approximately 20 to 30 minutes.

Once the interviewer had a thorough description and understanding of the participant’s cold, the participant was asked to complete the questionnaire-in-development. After marking answers on the questionnaire (which took 3 to 5 minutes), each participant was asked to comment on its ease of use, item wording, formatting, and response range as well as whether it accurately and comprehensively measured the symptoms and functional impact they were experiencing. The instrument development phase of the interview lasted for another 20 to 30 minutes.

We used focus group methods in the final month of the study as an additional window into participants’ experiences.^54-56 The focus groups used the same inclusion criteria as the long interviews and followed the same general format, first using open-ended questions to elicit symptoms and their impact, then administrating the questionnaire and discussing item inclusion and formatting. However, we encouraged discussion rather than self-assessment, as the focus group methodology derives its strength from the interactive nature of conversation. For instance, a statement made by one participant would spark interest or recall in another, thereby generating a richer, fuller, and more representative description of symptoms and functional impact.

Individual interviews were held by 1 of 5 trained interviewers (B.B., L.L., R.M., E.S., J.S.). All 3 focus groups were run by the lead author, with at least 1 other research team member assisting. Interviews and focus groups were arranged as soon as possible after the initial telephone contact so that participants would still have cold symptoms while being interviewed. All interviews and focus groups were discussed in biweekly group meetings. Decisions on item inclusion, wording, and questionnaire format were made by research group consensus. Several versions of the questionnaire were brought back to cold-sufferers for further cognitive testing. The diversity of interviewers and respondents provided protection against personal bias in ascertaining and interpreting symptoms and impairments.

TABLE 1
QUESTIONS ASKED DURING INTERVIEW

Results

Phase 1

Of the 148 college students enrolled, 142 followed protocol and were included in the analysis. Of the 853 person-days documented, 546 (64%) were covered by both data systems; 287 (33.6%) came from paper surveys only; and 18 (2.1%) were filled out via computer only. Because only 2 (0.2%) questionnaires were missing any data, our data capture rate was 99.8%. Comparing data from the computerized and paper data sources provided evidence of consistency. Of the 546 days in which both paper and computer instruments provided data, 512 yielded identical responses (94% concordant) to the global severity of illness question. Of the 34 (6%) discrepancies, 29 were off by 1 point on the 9-point Likert-type scale and 5 discrepancies were off by 2 points. Comparing computer and paper responses with the 15 specific symptom questions also yielded high levels of concordance. Of 8190 item responses, 7777 (95%) were concordant, while 413 (5%) were classified as data discrepancies. Of these, 293 were off by 1 point on the 9-point scale; 68 were off by 2 points; 27, by 3 points; 17, by 4 points; 7, by 5 points; and 1 by 6 points.

Factor analysis of the data provided further evidence of internal validity. Structural equation modeling techniques^57,58 were used to model symptom severities over time. A 4-dimensional symptom-recovery model (df = 71; P = .000025) provided a goodness of fit index of 0.88, a root mean square residual of .095, and a chi-squared/df ratio of 139/71 = 1.95. From the pool of 15 scaled symptom scores, 14 items contributed significantly to the model. (In this data set, loss of appetite was an infrequent symptom contributing insignificantly toward the model, and was dropped.) The 14 symptoms naturally aggregated into 4 underlying symptomatic dimensions: cough, throat, nasal, and fever and aches. Table 2 provides the reliability coefficients, standardized item loading coefficients, and standard errors of these loadings for the 4 dimensions. The reliability coefficients of the symptom dimensions were calculated using a procedure proposed by Dillon and Goldstein.⁵⁹ Scale recovery curves, generated using a mixed modeling approach,^60,61 were internally predictive, responsive,^37,62 and consistent with what is known about the natural history of URI.