Clinical Inquiries

Evidence summary

The CURB-65 criteria: Having ≥2 factors increases mortality

In a split-sample (derivation and validation) analysis¹ of 3 prospective studies involving 1068 patients presenting to the hospital with the diagnosis of pneumonia, various clinical features were analyzed for their association with 30-day mortality. The 5 parameters that were most strongly associated with mortality were:

• acute confusion (odds ratio [OR]=8.1; 95% confidence interval [CI], 4.8–13.7)
• BUN >19.6 mg/dL (OR= 5.6; 95% CI, 3.1–10)
• respiratory rate ≥30 (OR=1.7; 95% CI, 1.07–2.8)
• low blood pressure (SBP <90 or DBP ≤60) (OR=2.4; 95% CI, 1.4–3.8)
• age ≥65 years (OR=5.5; 95% CI, 2.8–10.9).

The 30-day mortality estimation using these 5 criteria is called CURB-65 (Confusion, Urea, Respiratory rate, low Blood pressure, and age ≥65).

In patients with 2 or more of these factors, the associated rate of mortality increased significantly compared with patients who had none or only 1 of the factors (TABLE 1). Although albumin <3.0 g/dL was also significantly associated with an increased mortality rate (OR=4.7; 95% CI, 2.5–8.7), it was not included in CURB-65 because it is not a routine lab ordered for patients with pneumonia.

A variation of the CURB-65 score, CRB-65, uses only the clinical parameters without laboratory data (confusion, respiratory rate, blood pressure, and age). Patients with a score of 0 had a 0.9% 30-day mortality rate. However, the rate increased to 8.15% when patients had 1 or 2 of the 4 clinical criteria.

TABLE 1
CURB-65 criteria

Pneumonia Severity Index has similar sensitivity to CURB-65

A more detailed assessment using 20 parameters called the Pneumonia Severity Index (PSI) was derived and validated in separate cohorts (TABLE 2).² When compared with CURB-65 and CRB-65, the PSI has similar sensitivity and specificity in predicting 30-day mortality.³ All 3 predictive rules had high negative predictive values for mortality but a low positive predictive value at all cutoff points.

Larger proportions of patients were identified as low-risk by PSI (47.2%) and CURB-65 (43.3%) than by CRB-65 (12.6%). Therefore PSI and CURB-65 are much more helpful in identifying patients who could be treated in the outpatient setting.

TABLE 2
Pneumonia Severity Index

Recommendations from others

The 2007 Infectious Diseases Society of America (IDSA) and American Thoracic Society (ATS), in their Consensus Guidelines on the Management of Community-Acquired Pneumonia,⁴ concluded that severity-of-illness scores, such as the CURB-65 or PSI, can be used to identify patients with CAP who may be candidates for outpatient treatment (evidence level I by IDSA/ATS rating).

The guidelines recommend that objective criteria or scores always be supplemented with physician determination of subjective factors, including the ability to safely and reliably take oral medication and the availability of outpatient support resources. Also, CURB-65 is more suitable than PSI for use in the emergency department because of its simplicity of application and ability to identify low-risk patients (evidence level II).⁴

Evidence summary

The CURB-65 criteria: Having ≥2 factors increases mortality

In a split-sample (derivation and validation) analysis¹ of 3 prospective studies involving 1068 patients presenting to the hospital with the diagnosis of pneumonia, various clinical features were analyzed for their association with 30-day mortality. The 5 parameters that were most strongly associated with mortality were:

• acute confusion (odds ratio [OR]=8.1; 95% confidence interval [CI], 4.8–13.7)
• BUN >19.6 mg/dL (OR= 5.6; 95% CI, 3.1–10)
• respiratory rate ≥30 (OR=1.7; 95% CI, 1.07–2.8)
• low blood pressure (SBP <90 or DBP ≤60) (OR=2.4; 95% CI, 1.4–3.8)
• age ≥65 years (OR=5.5; 95% CI, 2.8–10.9).

The 30-day mortality estimation using these 5 criteria is called CURB-65 (Confusion, Urea, Respiratory rate, low Blood pressure, and age ≥65).

In patients with 2 or more of these factors, the associated rate of mortality increased significantly compared with patients who had none or only 1 of the factors (TABLE 1). Although albumin <3.0 g/dL was also significantly associated with an increased mortality rate (OR=4.7; 95% CI, 2.5–8.7), it was not included in CURB-65 because it is not a routine lab ordered for patients with pneumonia.

A variation of the CURB-65 score, CRB-65, uses only the clinical parameters without laboratory data (confusion, respiratory rate, blood pressure, and age). Patients with a score of 0 had a 0.9% 30-day mortality rate. However, the rate increased to 8.15% when patients had 1 or 2 of the 4 clinical criteria.

TABLE 1
CURB-65 criteria

Pneumonia Severity Index has similar sensitivity to CURB-65

A more detailed assessment using 20 parameters called the Pneumonia Severity Index (PSI) was derived and validated in separate cohorts (TABLE 2).² When compared with CURB-65 and CRB-65, the PSI has similar sensitivity and specificity in predicting 30-day mortality.³ All 3 predictive rules had high negative predictive values for mortality but a low positive predictive value at all cutoff points.

Larger proportions of patients were identified as low-risk by PSI (47.2%) and CURB-65 (43.3%) than by CRB-65 (12.6%). Therefore PSI and CURB-65 are much more helpful in identifying patients who could be treated in the outpatient setting.

TABLE 2
Pneumonia Severity Index

Recommendations from others

The 2007 Infectious Diseases Society of America (IDSA) and American Thoracic Society (ATS), in their Consensus Guidelines on the Management of Community-Acquired Pneumonia,⁴ concluded that severity-of-illness scores, such as the CURB-65 or PSI, can be used to identify patients with CAP who may be candidates for outpatient treatment (evidence level I by IDSA/ATS rating).

The guidelines recommend that objective criteria or scores always be supplemented with physician determination of subjective factors, including the ability to safely and reliably take oral medication and the availability of outpatient support resources. Also, CURB-65 is more suitable than PSI for use in the emergency department because of its simplicity of application and ability to identify low-risk patients (evidence level II).⁴

Evidence summary

The CURB-65 criteria: Having ≥2 factors increases mortality

In a split-sample (derivation and validation) analysis¹ of 3 prospective studies involving 1068 patients presenting to the hospital with the diagnosis of pneumonia, various clinical features were analyzed for their association with 30-day mortality. The 5 parameters that were most strongly associated with mortality were:

• acute confusion (odds ratio [OR]=8.1; 95% confidence interval [CI], 4.8–13.7)
• BUN >19.6 mg/dL (OR= 5.6; 95% CI, 3.1–10)
• respiratory rate ≥30 (OR=1.7; 95% CI, 1.07–2.8)
• low blood pressure (SBP <90 or DBP ≤60) (OR=2.4; 95% CI, 1.4–3.8)
• age ≥65 years (OR=5.5; 95% CI, 2.8–10.9).

The 30-day mortality estimation using these 5 criteria is called CURB-65 (Confusion, Urea, Respiratory rate, low Blood pressure, and age ≥65).

In patients with 2 or more of these factors, the associated rate of mortality increased significantly compared with patients who had none or only 1 of the factors (TABLE 1). Although albumin <3.0 g/dL was also significantly associated with an increased mortality rate (OR=4.7; 95% CI, 2.5–8.7), it was not included in CURB-65 because it is not a routine lab ordered for patients with pneumonia.

A variation of the CURB-65 score, CRB-65, uses only the clinical parameters without laboratory data (confusion, respiratory rate, blood pressure, and age). Patients with a score of 0 had a 0.9% 30-day mortality rate. However, the rate increased to 8.15% when patients had 1 or 2 of the 4 clinical criteria.

TABLE 1
CURB-65 criteria

Pneumonia Severity Index has similar sensitivity to CURB-65

A more detailed assessment using 20 parameters called the Pneumonia Severity Index (PSI) was derived and validated in separate cohorts (TABLE 2).² When compared with CURB-65 and CRB-65, the PSI has similar sensitivity and specificity in predicting 30-day mortality.³ All 3 predictive rules had high negative predictive values for mortality but a low positive predictive value at all cutoff points.

Larger proportions of patients were identified as low-risk by PSI (47.2%) and CURB-65 (43.3%) than by CRB-65 (12.6%). Therefore PSI and CURB-65 are much more helpful in identifying patients who could be treated in the outpatient setting.

TABLE 2
Pneumonia Severity Index

Recommendations from others

The 2007 Infectious Diseases Society of America (IDSA) and American Thoracic Society (ATS), in their Consensus Guidelines on the Management of Community-Acquired Pneumonia,⁴ concluded that severity-of-illness scores, such as the CURB-65 or PSI, can be used to identify patients with CAP who may be candidates for outpatient treatment (evidence level I by IDSA/ATS rating).

The guidelines recommend that objective criteria or scores always be supplemented with physician determination of subjective factors, including the ability to safely and reliably take oral medication and the availability of outpatient support resources. Also, CURB-65 is more suitable than PSI for use in the emergency department because of its simplicity of application and ability to identify low-risk patients (evidence level II).⁴

Evidence summary

Premature ejaculation is the most common male sexual dysfunction, but there is no universally accepted definition or validated screening instrument. The pathophysiology and etiology remain incompletely understood.¹ Based on surveys, prevalence rates for premature ejaculation are approximately 20% to 30%.¹

Studies in male rats have demonstrated that serotonin with various 5-HT receptor subtypes are involved in the ejaculatory process.² Based on these studies, it’s been suggested that lifelong premature ejaculation is a neurobiological phenomenon related to decreased central serotonergic neurotransmission, 5-HT_2c receptor hyposensitivity, or 5-HT_1a receptor hypersensitivity.³

Antidepressants delay ejaculation

The introduction of selective serotonin reuptake inhibitors (SSRIs) revolutionized the treatment of premature ejaculation.⁴ In 1994, the first study of SSRIs in men with premature ejaculation demonstrated a delaying effect with paroxetine (Paxil).⁵ Since that time, SSRIs have been repeatedly investigated for their propensity to delay ejaculation. Certain SSRIs and the tricyclic antidepressant clomipramine (Anafranil) have become the agents of choice for the treatment of premature ejaculation.⁶

A meta-analysis⁶ of 35 treatment studies with serotonergic antidepressants from 1943 to 2003 shows that, despite major differences in design and drug dosing, clomipramine, fluoxetine (Prozac), paroxetine, and sertraline (Zoloft) significantly delay ejaculation compared with placebo. The percentage increase in IELT was the primary outcome measured. The rank order of efficacy was:

1. paroxetine (1492% IELT increase; 95% confidence interval [CI], 918–2425)
2. sertraline (790% IELT increase; 95% CI, 532–1173)
3. clomipramine (512% IELT increase; 95% CI, 234–1122)
4. fluoxetine (295% IELT increase; 95% CI, 172–506).⁶

Of the 35 studies used in the previous meta-analysis, 8 studies (N=263) were prospective, double-blind, real-time stopwatch studies that were separately analyzed in a subsequent meta-analysis. These 8 studies evaluated clomipramine, fluoxetine, paroxetine, sertraline, citalopram (Celexa), fluvoxamine (Luvox), mirtazapine (Remeron), and nefazodone (Serzone) against placebo. Paroxetine (783% IELT increase, 95% CI, 499–1228), clomipramine (360% IELT increase, 95% CI, 200–435), sertraline (313%, 95% CI, 161–608), and fluoxetine (295%, 95% CI, 200–435) exerted a significant delay in the IELT compared with placebo.⁶

EMLA cream: “Improvement” and “cure” seen

EMLA cream, a topical anesthetic, has been evaluated as a treatment option for premature ejaculation. One double-blinded RCT⁷ (N=29) showed significant improvement in the IELT (measured by stopwatch by the subject’s partner) from baseline compared with placebo (8.45 min vs 1.95 min; P<.001) at 2 months.

Another RCT⁸ (N=84) compared EMLA cream applied 15 minutes prior to intercourse, sildenafil 50 mg orally 45 minutes prior to intercourse, EMLA cream plus sildenafil, and placebo. In the sildenafil-plus-EMLA group, 32% of the patients reported “improvement” and 54% reported “cure,” which was defined as ejaculation delayed until the patient wished it. In the EMLA-only group, 27% of the patients reported “improvement” and 50% reported “cure.” This was a statistically significant difference when compared with the placebo and sildenafil-only groups (number needed to treat [NNT]=3). There was no significant difference in reports of “improvement” or “cure” between the placebo and sildenafil-only groups.

One small RCT⁹ (N=24) compared placebo with the application of EMLA cream 20, 30, and 45 minutes prior to sexual intercourse. Improvement was seen in IELT in the 20- and 30-minute group, but penile numbness and erection loss increased in the 30- and 45-minute group.

PDE5 inhibitors: No convincing evidence

A review¹⁰ of 14 clinical trials concluded that there is no convincing evidence for PDE5 inhibitors in the treatment of men with lifelong premature ejaculation and normal erectile function. One RCT¹¹ found no increase in IELT from baseline in men taking sildenafil when compared with placebo, although patients reported overall sexual satisfaction and confidence based on a questionnaire.

However, a study by Li et al¹² treated 45 men with premature ejaculation and comorbid erectile dysfunction with sildenafil. Eighty-nine percent reported improved erectile function, and 60% reported decreased severity of premature ejaculation.

Recommendations from others

The American Urological Association¹³ recommends antidepressants as first-line systemic therapy for premature ejaculation, specifically the SSRIs fluoxetine, paroxetine, sertraline, and the tricyclic clomipramine. Topical EMLA cream is also recommended, but the reduction of penile sensation may limit the acceptability of this treatment option.

The British Association for Sexual Health and HIV Special Interest Group for Sexual Dysfunction¹⁴ also recommends SSRIs and clomipramine as they have the strongest evidence for their efficacy. The group emphasizes the importance of combining behavioral and pharmacologic therapies as the management approach should be tailored to the individual patient.

Acknowledgments

The views expressed in this article are those of the authors and do not necessarily reflect the official policy or position of the Department of the Navy, Department of Defense, nor the US Government.

Evidence summary

Premature ejaculation is the most common male sexual dysfunction, but there is no universally accepted definition or validated screening instrument. The pathophysiology and etiology remain incompletely understood.¹ Based on surveys, prevalence rates for premature ejaculation are approximately 20% to 30%.¹

Studies in male rats have demonstrated that serotonin with various 5-HT receptor subtypes are involved in the ejaculatory process.² Based on these studies, it’s been suggested that lifelong premature ejaculation is a neurobiological phenomenon related to decreased central serotonergic neurotransmission, 5-HT_2c receptor hyposensitivity, or 5-HT_1a receptor hypersensitivity.³

Antidepressants delay ejaculation

The introduction of selective serotonin reuptake inhibitors (SSRIs) revolutionized the treatment of premature ejaculation.⁴ In 1994, the first study of SSRIs in men with premature ejaculation demonstrated a delaying effect with paroxetine (Paxil).⁵ Since that time, SSRIs have been repeatedly investigated for their propensity to delay ejaculation. Certain SSRIs and the tricyclic antidepressant clomipramine (Anafranil) have become the agents of choice for the treatment of premature ejaculation.⁶

A meta-analysis⁶ of 35 treatment studies with serotonergic antidepressants from 1943 to 2003 shows that, despite major differences in design and drug dosing, clomipramine, fluoxetine (Prozac), paroxetine, and sertraline (Zoloft) significantly delay ejaculation compared with placebo. The percentage increase in IELT was the primary outcome measured. The rank order of efficacy was:

1. paroxetine (1492% IELT increase; 95% confidence interval [CI], 918–2425)
2. sertraline (790% IELT increase; 95% CI, 532–1173)
3. clomipramine (512% IELT increase; 95% CI, 234–1122)
4. fluoxetine (295% IELT increase; 95% CI, 172–506).⁶

Of the 35 studies used in the previous meta-analysis, 8 studies (N=263) were prospective, double-blind, real-time stopwatch studies that were separately analyzed in a subsequent meta-analysis. These 8 studies evaluated clomipramine, fluoxetine, paroxetine, sertraline, citalopram (Celexa), fluvoxamine (Luvox), mirtazapine (Remeron), and nefazodone (Serzone) against placebo. Paroxetine (783% IELT increase, 95% CI, 499–1228), clomipramine (360% IELT increase, 95% CI, 200–435), sertraline (313%, 95% CI, 161–608), and fluoxetine (295%, 95% CI, 200–435) exerted a significant delay in the IELT compared with placebo.⁶

EMLA cream: “Improvement” and “cure” seen

EMLA cream, a topical anesthetic, has been evaluated as a treatment option for premature ejaculation. One double-blinded RCT⁷ (N=29) showed significant improvement in the IELT (measured by stopwatch by the subject’s partner) from baseline compared with placebo (8.45 min vs 1.95 min; P<.001) at 2 months.

Another RCT⁸ (N=84) compared EMLA cream applied 15 minutes prior to intercourse, sildenafil 50 mg orally 45 minutes prior to intercourse, EMLA cream plus sildenafil, and placebo. In the sildenafil-plus-EMLA group, 32% of the patients reported “improvement” and 54% reported “cure,” which was defined as ejaculation delayed until the patient wished it. In the EMLA-only group, 27% of the patients reported “improvement” and 50% reported “cure.” This was a statistically significant difference when compared with the placebo and sildenafil-only groups (number needed to treat [NNT]=3). There was no significant difference in reports of “improvement” or “cure” between the placebo and sildenafil-only groups.

One small RCT⁹ (N=24) compared placebo with the application of EMLA cream 20, 30, and 45 minutes prior to sexual intercourse. Improvement was seen in IELT in the 20- and 30-minute group, but penile numbness and erection loss increased in the 30- and 45-minute group.

PDE5 inhibitors: No convincing evidence

A review¹⁰ of 14 clinical trials concluded that there is no convincing evidence for PDE5 inhibitors in the treatment of men with lifelong premature ejaculation and normal erectile function. One RCT¹¹ found no increase in IELT from baseline in men taking sildenafil when compared with placebo, although patients reported overall sexual satisfaction and confidence based on a questionnaire.

However, a study by Li et al¹² treated 45 men with premature ejaculation and comorbid erectile dysfunction with sildenafil. Eighty-nine percent reported improved erectile function, and 60% reported decreased severity of premature ejaculation.

Recommendations from others

The American Urological Association¹³ recommends antidepressants as first-line systemic therapy for premature ejaculation, specifically the SSRIs fluoxetine, paroxetine, sertraline, and the tricyclic clomipramine. Topical EMLA cream is also recommended, but the reduction of penile sensation may limit the acceptability of this treatment option.

The British Association for Sexual Health and HIV Special Interest Group for Sexual Dysfunction¹⁴ also recommends SSRIs and clomipramine as they have the strongest evidence for their efficacy. The group emphasizes the importance of combining behavioral and pharmacologic therapies as the management approach should be tailored to the individual patient.

Acknowledgments

The views expressed in this article are those of the authors and do not necessarily reflect the official policy or position of the Department of the Navy, Department of Defense, nor the US Government.

Evidence summary

Premature ejaculation is the most common male sexual dysfunction, but there is no universally accepted definition or validated screening instrument. The pathophysiology and etiology remain incompletely understood.¹ Based on surveys, prevalence rates for premature ejaculation are approximately 20% to 30%.¹

Studies in male rats have demonstrated that serotonin with various 5-HT receptor subtypes are involved in the ejaculatory process.² Based on these studies, it’s been suggested that lifelong premature ejaculation is a neurobiological phenomenon related to decreased central serotonergic neurotransmission, 5-HT_2c receptor hyposensitivity, or 5-HT_1a receptor hypersensitivity.³

Antidepressants delay ejaculation

The introduction of selective serotonin reuptake inhibitors (SSRIs) revolutionized the treatment of premature ejaculation.⁴ In 1994, the first study of SSRIs in men with premature ejaculation demonstrated a delaying effect with paroxetine (Paxil).⁵ Since that time, SSRIs have been repeatedly investigated for their propensity to delay ejaculation. Certain SSRIs and the tricyclic antidepressant clomipramine (Anafranil) have become the agents of choice for the treatment of premature ejaculation.⁶

A meta-analysis⁶ of 35 treatment studies with serotonergic antidepressants from 1943 to 2003 shows that, despite major differences in design and drug dosing, clomipramine, fluoxetine (Prozac), paroxetine, and sertraline (Zoloft) significantly delay ejaculation compared with placebo. The percentage increase in IELT was the primary outcome measured. The rank order of efficacy was:

1. paroxetine (1492% IELT increase; 95% confidence interval [CI], 918–2425)
2. sertraline (790% IELT increase; 95% CI, 532–1173)
3. clomipramine (512% IELT increase; 95% CI, 234–1122)
4. fluoxetine (295% IELT increase; 95% CI, 172–506).⁶

Of the 35 studies used in the previous meta-analysis, 8 studies (N=263) were prospective, double-blind, real-time stopwatch studies that were separately analyzed in a subsequent meta-analysis. These 8 studies evaluated clomipramine, fluoxetine, paroxetine, sertraline, citalopram (Celexa), fluvoxamine (Luvox), mirtazapine (Remeron), and nefazodone (Serzone) against placebo. Paroxetine (783% IELT increase, 95% CI, 499–1228), clomipramine (360% IELT increase, 95% CI, 200–435), sertraline (313%, 95% CI, 161–608), and fluoxetine (295%, 95% CI, 200–435) exerted a significant delay in the IELT compared with placebo.⁶

EMLA cream: “Improvement” and “cure” seen

EMLA cream, a topical anesthetic, has been evaluated as a treatment option for premature ejaculation. One double-blinded RCT⁷ (N=29) showed significant improvement in the IELT (measured by stopwatch by the subject’s partner) from baseline compared with placebo (8.45 min vs 1.95 min; P<.001) at 2 months.

Another RCT⁸ (N=84) compared EMLA cream applied 15 minutes prior to intercourse, sildenafil 50 mg orally 45 minutes prior to intercourse, EMLA cream plus sildenafil, and placebo. In the sildenafil-plus-EMLA group, 32% of the patients reported “improvement” and 54% reported “cure,” which was defined as ejaculation delayed until the patient wished it. In the EMLA-only group, 27% of the patients reported “improvement” and 50% reported “cure.” This was a statistically significant difference when compared with the placebo and sildenafil-only groups (number needed to treat [NNT]=3). There was no significant difference in reports of “improvement” or “cure” between the placebo and sildenafil-only groups.

One small RCT⁹ (N=24) compared placebo with the application of EMLA cream 20, 30, and 45 minutes prior to sexual intercourse. Improvement was seen in IELT in the 20- and 30-minute group, but penile numbness and erection loss increased in the 30- and 45-minute group.

PDE5 inhibitors: No convincing evidence

A review¹⁰ of 14 clinical trials concluded that there is no convincing evidence for PDE5 inhibitors in the treatment of men with lifelong premature ejaculation and normal erectile function. One RCT¹¹ found no increase in IELT from baseline in men taking sildenafil when compared with placebo, although patients reported overall sexual satisfaction and confidence based on a questionnaire.

However, a study by Li et al¹² treated 45 men with premature ejaculation and comorbid erectile dysfunction with sildenafil. Eighty-nine percent reported improved erectile function, and 60% reported decreased severity of premature ejaculation.

Recommendations from others

The American Urological Association¹³ recommends antidepressants as first-line systemic therapy for premature ejaculation, specifically the SSRIs fluoxetine, paroxetine, sertraline, and the tricyclic clomipramine. Topical EMLA cream is also recommended, but the reduction of penile sensation may limit the acceptability of this treatment option.

The British Association for Sexual Health and HIV Special Interest Group for Sexual Dysfunction¹⁴ also recommends SSRIs and clomipramine as they have the strongest evidence for their efficacy. The group emphasizes the importance of combining behavioral and pharmacologic therapies as the management approach should be tailored to the individual patient.

Acknowledgments

The views expressed in this article are those of the authors and do not necessarily reflect the official policy or position of the Department of the Navy, Department of Defense, nor the US Government.

Evidence summary

For cervicogenic headaches: Spinal manipulative therapy reduces pain

Three studies¹ evaluated SMT for treatment of recurrent cervicogenic headaches). A multicenter trial² randomized 200 patients with cervicogenic headaches to either SMT (8–12 sessions over 6 weeks) or placebo. The SMT group had significantly reduced pain (at 1 week, effect size [ES]=0.7; 95% confidence interval [CI], 0.3–1.2; and at 12 months, ES=0.4; 95% CI, 0.0–0.8) and fewer headaches (ES=0.7; 95% CI, 0.3–1.1 at both time points) than placebo.

Another RCT³ with 105 patients compared SMT (3 times a week for 3 weeks) with placebo. The SMT group reported significantly less pain after 3 weeks (ES=2.2; 95% CI, 1.7–2.7).

A third trial⁴ randomized 30 patients to either SMT, mobilization (small oscillatory movements to a joint within its normal range), or wait-list placement. At the end of treatment, there was a nonsignificant trend toward greater pain reduction in patients receiving SMT than either those receiving mobilization (ES=0.4; 95% CI, –0.5 to 1.4) or those on the wait list (ES=0.6; 95% CI, –0.4 to 1.5).

For tension-type headaches: Results are mixed

Two trials⁵ investigated the efficacy of SMT on tension-type headaches. The first, an RCT with 150 patients with recurrent headaches, compared SMT (2 sessions per week) with amitriptyline (10 mg daily week 1, 20 mg daily week 2, then 30 mg daily) for 6 weeks. At the end of 6 weeks, the SMT group reported a nonsignificant trend toward more headache pain (ES for SMT vs amitriptyline= –0.4; 95% CI, –0.8 to 0.0), but fewer side effects. They had similar headache frequency and medication use.

Another study⁶ of 22 patients compared SMT with 2 different controls (palpation and rest) for acute tension-type headache. The SMT group was significantly more likely to experience immediate improvement (ES=1.8; 95% CI, 0.4–3.2).

For migraine: Spinal manipulative therapy is similar to amitriptyline

In 1 trial⁷ of migraine prophylaxis, 218 patients were randomized to either 14 sessions of SMT for 2 months or oral amitriptyline (titrated up weekly during the first month and continued at 100 mg daily over the second month). The headache index (a measure of daily pain intensity) was equivalent in both groups in the last 4 weeks of treatment (ES for SMT vs amitriptyline= –0.1; 95% CI, –0.5 to 0.3).

A month after both therapies were stopped, there was a nonsignificant trend toward a lower headache index in the group that had received SMT than the group that had received amitriptyline (ES=0.4; 95% CI, 0.0–0.8). Ten percent of the medication group withdrew from this study due to side effects; no side effects were reported from SMT.⁷

Another RCT⁸ of migraine prophylaxis with 88 patients compared SMT twice weekly for 8 weeks with mobilization techniques. At 8 weeks post-treatment, there was a nonsignificant trend favoring SMT over mobilization in decreasing pain (ES=0.4; 95% CI, –0.2 to 1.0).

Recommendations from others

The National Headache Foundation⁹ states that “the value and cost-effectiveness of chiropractic, osteopathic medicine, and physical therapy in migraine have not been proven in clinical trials. Conflicting results and poor clinical trial design limit the ability to judge the effectiveness of manipulative treatments. Physical therapy, although limited in its study, has proven more effective than manipulative treatment in selective cases.”

Evidence summary

For cervicogenic headaches: Spinal manipulative therapy reduces pain

Three studies¹ evaluated SMT for treatment of recurrent cervicogenic headaches). A multicenter trial² randomized 200 patients with cervicogenic headaches to either SMT (8–12 sessions over 6 weeks) or placebo. The SMT group had significantly reduced pain (at 1 week, effect size [ES]=0.7; 95% confidence interval [CI], 0.3–1.2; and at 12 months, ES=0.4; 95% CI, 0.0–0.8) and fewer headaches (ES=0.7; 95% CI, 0.3–1.1 at both time points) than placebo.

Another RCT³ with 105 patients compared SMT (3 times a week for 3 weeks) with placebo. The SMT group reported significantly less pain after 3 weeks (ES=2.2; 95% CI, 1.7–2.7).

A third trial⁴ randomized 30 patients to either SMT, mobilization (small oscillatory movements to a joint within its normal range), or wait-list placement. At the end of treatment, there was a nonsignificant trend toward greater pain reduction in patients receiving SMT than either those receiving mobilization (ES=0.4; 95% CI, –0.5 to 1.4) or those on the wait list (ES=0.6; 95% CI, –0.4 to 1.5).

For tension-type headaches: Results are mixed

Two trials⁵ investigated the efficacy of SMT on tension-type headaches. The first, an RCT with 150 patients with recurrent headaches, compared SMT (2 sessions per week) with amitriptyline (10 mg daily week 1, 20 mg daily week 2, then 30 mg daily) for 6 weeks. At the end of 6 weeks, the SMT group reported a nonsignificant trend toward more headache pain (ES for SMT vs amitriptyline= –0.4; 95% CI, –0.8 to 0.0), but fewer side effects. They had similar headache frequency and medication use.

Another study⁶ of 22 patients compared SMT with 2 different controls (palpation and rest) for acute tension-type headache. The SMT group was significantly more likely to experience immediate improvement (ES=1.8; 95% CI, 0.4–3.2).

For migraine: Spinal manipulative therapy is similar to amitriptyline

In 1 trial⁷ of migraine prophylaxis, 218 patients were randomized to either 14 sessions of SMT for 2 months or oral amitriptyline (titrated up weekly during the first month and continued at 100 mg daily over the second month). The headache index (a measure of daily pain intensity) was equivalent in both groups in the last 4 weeks of treatment (ES for SMT vs amitriptyline= –0.1; 95% CI, –0.5 to 0.3).

A month after both therapies were stopped, there was a nonsignificant trend toward a lower headache index in the group that had received SMT than the group that had received amitriptyline (ES=0.4; 95% CI, 0.0–0.8). Ten percent of the medication group withdrew from this study due to side effects; no side effects were reported from SMT.⁷

Another RCT⁸ of migraine prophylaxis with 88 patients compared SMT twice weekly for 8 weeks with mobilization techniques. At 8 weeks post-treatment, there was a nonsignificant trend favoring SMT over mobilization in decreasing pain (ES=0.4; 95% CI, –0.2 to 1.0).

Recommendations from others

The National Headache Foundation⁹ states that “the value and cost-effectiveness of chiropractic, osteopathic medicine, and physical therapy in migraine have not been proven in clinical trials. Conflicting results and poor clinical trial design limit the ability to judge the effectiveness of manipulative treatments. Physical therapy, although limited in its study, has proven more effective than manipulative treatment in selective cases.”

Evidence summary

For cervicogenic headaches: Spinal manipulative therapy reduces pain

Three studies¹ evaluated SMT for treatment of recurrent cervicogenic headaches). A multicenter trial² randomized 200 patients with cervicogenic headaches to either SMT (8–12 sessions over 6 weeks) or placebo. The SMT group had significantly reduced pain (at 1 week, effect size [ES]=0.7; 95% confidence interval [CI], 0.3–1.2; and at 12 months, ES=0.4; 95% CI, 0.0–0.8) and fewer headaches (ES=0.7; 95% CI, 0.3–1.1 at both time points) than placebo.

Another RCT³ with 105 patients compared SMT (3 times a week for 3 weeks) with placebo. The SMT group reported significantly less pain after 3 weeks (ES=2.2; 95% CI, 1.7–2.7).

A third trial⁴ randomized 30 patients to either SMT, mobilization (small oscillatory movements to a joint within its normal range), or wait-list placement. At the end of treatment, there was a nonsignificant trend toward greater pain reduction in patients receiving SMT than either those receiving mobilization (ES=0.4; 95% CI, –0.5 to 1.4) or those on the wait list (ES=0.6; 95% CI, –0.4 to 1.5).

For tension-type headaches: Results are mixed

Two trials⁵ investigated the efficacy of SMT on tension-type headaches. The first, an RCT with 150 patients with recurrent headaches, compared SMT (2 sessions per week) with amitriptyline (10 mg daily week 1, 20 mg daily week 2, then 30 mg daily) for 6 weeks. At the end of 6 weeks, the SMT group reported a nonsignificant trend toward more headache pain (ES for SMT vs amitriptyline= –0.4; 95% CI, –0.8 to 0.0), but fewer side effects. They had similar headache frequency and medication use.

Another study⁶ of 22 patients compared SMT with 2 different controls (palpation and rest) for acute tension-type headache. The SMT group was significantly more likely to experience immediate improvement (ES=1.8; 95% CI, 0.4–3.2).

For migraine: Spinal manipulative therapy is similar to amitriptyline

In 1 trial⁷ of migraine prophylaxis, 218 patients were randomized to either 14 sessions of SMT for 2 months or oral amitriptyline (titrated up weekly during the first month and continued at 100 mg daily over the second month). The headache index (a measure of daily pain intensity) was equivalent in both groups in the last 4 weeks of treatment (ES for SMT vs amitriptyline= –0.1; 95% CI, –0.5 to 0.3).

A month after both therapies were stopped, there was a nonsignificant trend toward a lower headache index in the group that had received SMT than the group that had received amitriptyline (ES=0.4; 95% CI, 0.0–0.8). Ten percent of the medication group withdrew from this study due to side effects; no side effects were reported from SMT.⁷

Another RCT⁸ of migraine prophylaxis with 88 patients compared SMT twice weekly for 8 weeks with mobilization techniques. At 8 weeks post-treatment, there was a nonsignificant trend favoring SMT over mobilization in decreasing pain (ES=0.4; 95% CI, –0.2 to 1.0).

Recommendations from others

The National Headache Foundation⁹ states that “the value and cost-effectiveness of chiropractic, osteopathic medicine, and physical therapy in migraine have not been proven in clinical trials. Conflicting results and poor clinical trial design limit the ability to judge the effectiveness of manipulative treatments. Physical therapy, although limited in its study, has proven more effective than manipulative treatment in selective cases.”

Evidence summary

Chronic leg edema is defined as palpable swelling caused by an increase in interstitial fluid volume lasting at least 72 hours.¹

We were able to find only 1 moderate-quality study regarding the diagnosis of bilateral lower extremity edema that included a thorough cardiovascular evaluation.

Single study in bilateral leg edema: What the FPs thought…

A nonconsecutive cohort study² evaluated the causes of bilateral leg edema among 58 ambulatory adults (between 29 and 83 years of age) enrolled from an inner-city family medicine clinic in Cleveland. Edema was present for >3 months in 78% of patients, and 84% were obese. Patients were excluded if the edema was known to be due to nifedipine, intra-abdominal malignancy, hypothyroidism, or idiopathic cyclic edema.

Family physicians obtained a history and performed a physical exam on all patients and recorded a clinical diagnosis for the edema. Initial clinical impressions included: venous insufficiency (71%), congestive heart failure (18%), nephrotic syndrome (13%), uncertain (7%), and other causes (2% each). Other causes included lymphedema, pulmonary hypertension, cor pulmonale, hypoalbuminemia, NSAID or corticosteroid use, sleep apnea, and obesity. (Total percentages exceed 100% because some patients had multiple conditions.)

All patients were then evaluated with a serum albumin, a 24-hour urine protein collection, echocardiogram, and lower extremity duplex venous ultrasound (13 of 58 patients did not complete the echocardiogram and venous duplex ultrasound). Investigators developed a final diagnosis using the results of this evaluation in conjunction with the clinical information obtained by the physician.

…and what the testing revealed

Final diagnoses included pulmonary hypertension/borderline pulmonary hyper-tension (>30 mm Hg) (42%), congestive heart failure (29%), idiopathic edema (27%), venous insufficiency (22%), medication use (15%, primarily corticosteroids and NSAIDs), proteinuria >1 g/day (15%), and other causes (2% each). These other causes included transient renal disease, hypoalbuminemia, lymphedema, and stenosis of the inferior vena cava. (Total percentages again exceed 100% because some patients had multiple diagnoses.)

All 15 patients with cardiac conditions and 17 of 19 patients with pulmonary hypertension were over 45 years of age. Of the 19 patients with pulmonary hypertension, 6 had CHF, 4 had chronic obstructive pulmonary disease, 4 had sleep apnea diagnosed in subsequent testing, 1 had an atrial septal defect, and 4 appeared to have primary pulmonary hypertension.

Investigators recommend 3 steps

The investigators recommended 3 steps to evaluate chronic leg edema.

1. Stop any potentially causative medicines, such as NSAIDs or calcium-channel blockers.
2. Obtain an echocardiogram if the patient is 45 years of age or older.
3. Obtain a sleep study if the echo-cardiogram reveals pulmonary hypertension without an apparent cause.

Recommendations from others

The authors of a recent narrative systematic review¹ stated that “most patients with chronic leg edema can be assumed to have venous insufficiency, CHF, or cyclic edema, unless another cause is suspected after a history and physical examination.” They also stated that: “pulmonary hypertension and early CHF can both cause leg edema before they become clinically obvious” and reiterate that patients over 45 years of age with edema of unclear cause should have an echocardiogram to rule out pulmonary hypertension.

Evidence summary

Chronic leg edema is defined as palpable swelling caused by an increase in interstitial fluid volume lasting at least 72 hours.¹

We were able to find only 1 moderate-quality study regarding the diagnosis of bilateral lower extremity edema that included a thorough cardiovascular evaluation.

Single study in bilateral leg edema: What the FPs thought…

A nonconsecutive cohort study² evaluated the causes of bilateral leg edema among 58 ambulatory adults (between 29 and 83 years of age) enrolled from an inner-city family medicine clinic in Cleveland. Edema was present for >3 months in 78% of patients, and 84% were obese. Patients were excluded if the edema was known to be due to nifedipine, intra-abdominal malignancy, hypothyroidism, or idiopathic cyclic edema.

Family physicians obtained a history and performed a physical exam on all patients and recorded a clinical diagnosis for the edema. Initial clinical impressions included: venous insufficiency (71%), congestive heart failure (18%), nephrotic syndrome (13%), uncertain (7%), and other causes (2% each). Other causes included lymphedema, pulmonary hypertension, cor pulmonale, hypoalbuminemia, NSAID or corticosteroid use, sleep apnea, and obesity. (Total percentages exceed 100% because some patients had multiple conditions.)

All patients were then evaluated with a serum albumin, a 24-hour urine protein collection, echocardiogram, and lower extremity duplex venous ultrasound (13 of 58 patients did not complete the echocardiogram and venous duplex ultrasound). Investigators developed a final diagnosis using the results of this evaluation in conjunction with the clinical information obtained by the physician.

…and what the testing revealed

Final diagnoses included pulmonary hypertension/borderline pulmonary hyper-tension (>30 mm Hg) (42%), congestive heart failure (29%), idiopathic edema (27%), venous insufficiency (22%), medication use (15%, primarily corticosteroids and NSAIDs), proteinuria >1 g/day (15%), and other causes (2% each). These other causes included transient renal disease, hypoalbuminemia, lymphedema, and stenosis of the inferior vena cava. (Total percentages again exceed 100% because some patients had multiple diagnoses.)

All 15 patients with cardiac conditions and 17 of 19 patients with pulmonary hypertension were over 45 years of age. Of the 19 patients with pulmonary hypertension, 6 had CHF, 4 had chronic obstructive pulmonary disease, 4 had sleep apnea diagnosed in subsequent testing, 1 had an atrial septal defect, and 4 appeared to have primary pulmonary hypertension.

Investigators recommend 3 steps

The investigators recommended 3 steps to evaluate chronic leg edema.

1. Stop any potentially causative medicines, such as NSAIDs or calcium-channel blockers.
2. Obtain an echocardiogram if the patient is 45 years of age or older.
3. Obtain a sleep study if the echo-cardiogram reveals pulmonary hypertension without an apparent cause.

Recommendations from others

The authors of a recent narrative systematic review¹ stated that “most patients with chronic leg edema can be assumed to have venous insufficiency, CHF, or cyclic edema, unless another cause is suspected after a history and physical examination.” They also stated that: “pulmonary hypertension and early CHF can both cause leg edema before they become clinically obvious” and reiterate that patients over 45 years of age with edema of unclear cause should have an echocardiogram to rule out pulmonary hypertension.

Evidence summary

Chronic leg edema is defined as palpable swelling caused by an increase in interstitial fluid volume lasting at least 72 hours.¹

We were able to find only 1 moderate-quality study regarding the diagnosis of bilateral lower extremity edema that included a thorough cardiovascular evaluation.

Single study in bilateral leg edema: What the FPs thought…

A nonconsecutive cohort study² evaluated the causes of bilateral leg edema among 58 ambulatory adults (between 29 and 83 years of age) enrolled from an inner-city family medicine clinic in Cleveland. Edema was present for >3 months in 78% of patients, and 84% were obese. Patients were excluded if the edema was known to be due to nifedipine, intra-abdominal malignancy, hypothyroidism, or idiopathic cyclic edema.

Family physicians obtained a history and performed a physical exam on all patients and recorded a clinical diagnosis for the edema. Initial clinical impressions included: venous insufficiency (71%), congestive heart failure (18%), nephrotic syndrome (13%), uncertain (7%), and other causes (2% each). Other causes included lymphedema, pulmonary hypertension, cor pulmonale, hypoalbuminemia, NSAID or corticosteroid use, sleep apnea, and obesity. (Total percentages exceed 100% because some patients had multiple conditions.)

All patients were then evaluated with a serum albumin, a 24-hour urine protein collection, echocardiogram, and lower extremity duplex venous ultrasound (13 of 58 patients did not complete the echocardiogram and venous duplex ultrasound). Investigators developed a final diagnosis using the results of this evaluation in conjunction with the clinical information obtained by the physician.

…and what the testing revealed

Final diagnoses included pulmonary hypertension/borderline pulmonary hyper-tension (>30 mm Hg) (42%), congestive heart failure (29%), idiopathic edema (27%), venous insufficiency (22%), medication use (15%, primarily corticosteroids and NSAIDs), proteinuria >1 g/day (15%), and other causes (2% each). These other causes included transient renal disease, hypoalbuminemia, lymphedema, and stenosis of the inferior vena cava. (Total percentages again exceed 100% because some patients had multiple diagnoses.)

All 15 patients with cardiac conditions and 17 of 19 patients with pulmonary hypertension were over 45 years of age. Of the 19 patients with pulmonary hypertension, 6 had CHF, 4 had chronic obstructive pulmonary disease, 4 had sleep apnea diagnosed in subsequent testing, 1 had an atrial septal defect, and 4 appeared to have primary pulmonary hypertension.

Investigators recommend 3 steps

The investigators recommended 3 steps to evaluate chronic leg edema.

1. Stop any potentially causative medicines, such as NSAIDs or calcium-channel blockers.
2. Obtain an echocardiogram if the patient is 45 years of age or older.
3. Obtain a sleep study if the echo-cardiogram reveals pulmonary hypertension without an apparent cause.

Recommendations from others

The authors of a recent narrative systematic review¹ stated that “most patients with chronic leg edema can be assumed to have venous insufficiency, CHF, or cyclic edema, unless another cause is suspected after a history and physical examination.” They also stated that: “pulmonary hypertension and early CHF can both cause leg edema before they become clinically obvious” and reiterate that patients over 45 years of age with edema of unclear cause should have an echocardiogram to rule out pulmonary hypertension.

Evidence summary

Feeding difficulties, inadequate weight gain, and unusual physical exam findings may lead providers to assess thyroid function in newborns. Additionally, while measurement of free T4 is more common, some states use TSH as the required newborn screening assay to evaluate for congenital hypothyroidism. Whether ordered as a screening test or in response to symptoms, an elevated TSH in a newborn requires further investigation.

Congenital hypothyroidism is the most serious cause of an elevated TSH in a newborn. If left untreated, congenital hypothyroidism leads to developmental delay and mental retardation; however, with early treatment, intellectual outcomes are greatly improved. Newborns with an elevated TSH should be evaluated with repeat TSH and free T4 measurements in order to assess for congenital hypothyroidism.

TSH >50 mU/L increases chances of congenital hypothyroidism

Not surprisingly, higher levels of TSH increase the likelihood of congenital hypothyroidism. A study from the Netherlands examined diagnoses of infants with an increased TSH on newborn screening heelstick. Of 112 newborns with a TSH >50 mU/L, 110 (98%) had congenital hypothyroidism on further examination. However, only 34 of 594 (5.7%) newborns with a TSH between 9 and 20 mU/L were diagnosed with congenital hypothyroidism. Nineteen of 46 newborns (41%) with levels between 20 and 50 mU/L had congenital hypothyroidism.¹

Other common causes of hypothyroidism

While TSH is generally an accurate measurement of thyroid function, other factors can also lead to an elevated level. The same Dutch study mentioned above explored the presumed causes of elevated TSH among children who were diagnosed with transient hypothyroidism (initially elevated TSH level found to be normal on follow-up testing). The most common causes were thyroid-binding globulin deficiency (200/548 or 36% of newborns with transient hypothyroidism), severe illness (36%), prematurity (8%), and errors in screening procedures (4%).

Another study confirmed that TSH levels were higher in infants born preterm; babies with the earliest gestational ages had the highest TSH levels. The same study also found that TSH levels increased with increasing degrees of illness. Very preterm babies, those with cerebral pathology, low Apgar scores, respiratory distress syndrome, persistent ductus arteriosus requiring treatment, and necrotizing enterocolitis were at highest risk for having abnormally elevated TSH levels in this study.² If a sample is drawn from a newborn exhibiting symptoms (such as poor feeding or hypotonia), the TSH level may be elevated in spite of normal thyroid function.

Maternal thyroid disease can also cause a suppression of thyroid function in the newborn. One study¹ found that of 34 children with transient hypothyroidism, 10 had mothers with undertreated or unrecognized Graves disease.

Finally, either iodine excess or iodine deficiency can cause transient hypothyroidism. Case reports have directly demonstrated the effects of topical iodine exposure on newborn TSH levels.^3-5 Deficiency of dietary iodine is a common cause of both congenital and transient hypothyroidism in newborns worldwide, although it is rare in the United States. The World Health Organization lists 54 countries with inadequate iodine intake; consider children from these countries at high risk for hypothyroidism due to iodine deficiency.⁶

Draw TSH on the second or third day of life

In term, healthy newborns, TSH levels normally increase to levels of 60 mU/L within 30 minutes of delivery. This is followed by a rapid decline in TSH levels over the first 5 days of life to <10 mU/L. An Australian study found more elevated TSH levels for samples drawn on day 2 of life compared with day 3 of life, likely reflecting normal postnatal physiology.³ Age-specific reference ranges are necessary for interpretation of TSH levels during the first 5 days of life. The second or third day of life remains the optimal time for screening when appropriate reference ranges are used.

Recommendations from others

The United States Preventive Services Task Force (USPSTF),⁷ the American Academy of Family Physicians,⁸ the American Academy of Pediatrics,⁹ and the American Thyroid Association (ATA)¹⁰ all recommend routine screening of asymptomatic newborns for congenital hypothyroidism. The USPSTF recommends that clinicians evaluate abnormal thyroid screening results with a supplemental lab test, using TSH as the primary test and T4 as the supplemental test.⁷ Additionally, the ATA endorses a second thyroid screening at 7 to 14 days of life to increase specificity of congenital hypothyroidism screening.¹⁰

Evidence summary

Feeding difficulties, inadequate weight gain, and unusual physical exam findings may lead providers to assess thyroid function in newborns. Additionally, while measurement of free T4 is more common, some states use TSH as the required newborn screening assay to evaluate for congenital hypothyroidism. Whether ordered as a screening test or in response to symptoms, an elevated TSH in a newborn requires further investigation.

Congenital hypothyroidism is the most serious cause of an elevated TSH in a newborn. If left untreated, congenital hypothyroidism leads to developmental delay and mental retardation; however, with early treatment, intellectual outcomes are greatly improved. Newborns with an elevated TSH should be evaluated with repeat TSH and free T4 measurements in order to assess for congenital hypothyroidism.

TSH >50 mU/L increases chances of congenital hypothyroidism

Not surprisingly, higher levels of TSH increase the likelihood of congenital hypothyroidism. A study from the Netherlands examined diagnoses of infants with an increased TSH on newborn screening heelstick. Of 112 newborns with a TSH >50 mU/L, 110 (98%) had congenital hypothyroidism on further examination. However, only 34 of 594 (5.7%) newborns with a TSH between 9 and 20 mU/L were diagnosed with congenital hypothyroidism. Nineteen of 46 newborns (41%) with levels between 20 and 50 mU/L had congenital hypothyroidism.¹

Other common causes of hypothyroidism

While TSH is generally an accurate measurement of thyroid function, other factors can also lead to an elevated level. The same Dutch study mentioned above explored the presumed causes of elevated TSH among children who were diagnosed with transient hypothyroidism (initially elevated TSH level found to be normal on follow-up testing). The most common causes were thyroid-binding globulin deficiency (200/548 or 36% of newborns with transient hypothyroidism), severe illness (36%), prematurity (8%), and errors in screening procedures (4%).

Another study confirmed that TSH levels were higher in infants born preterm; babies with the earliest gestational ages had the highest TSH levels. The same study also found that TSH levels increased with increasing degrees of illness. Very preterm babies, those with cerebral pathology, low Apgar scores, respiratory distress syndrome, persistent ductus arteriosus requiring treatment, and necrotizing enterocolitis were at highest risk for having abnormally elevated TSH levels in this study.² If a sample is drawn from a newborn exhibiting symptoms (such as poor feeding or hypotonia), the TSH level may be elevated in spite of normal thyroid function.

Maternal thyroid disease can also cause a suppression of thyroid function in the newborn. One study¹ found that of 34 children with transient hypothyroidism, 10 had mothers with undertreated or unrecognized Graves disease.

Finally, either iodine excess or iodine deficiency can cause transient hypothyroidism. Case reports have directly demonstrated the effects of topical iodine exposure on newborn TSH levels.^3-5 Deficiency of dietary iodine is a common cause of both congenital and transient hypothyroidism in newborns worldwide, although it is rare in the United States. The World Health Organization lists 54 countries with inadequate iodine intake; consider children from these countries at high risk for hypothyroidism due to iodine deficiency.⁶

Draw TSH on the second or third day of life

In term, healthy newborns, TSH levels normally increase to levels of 60 mU/L within 30 minutes of delivery. This is followed by a rapid decline in TSH levels over the first 5 days of life to <10 mU/L. An Australian study found more elevated TSH levels for samples drawn on day 2 of life compared with day 3 of life, likely reflecting normal postnatal physiology.³ Age-specific reference ranges are necessary for interpretation of TSH levels during the first 5 days of life. The second or third day of life remains the optimal time for screening when appropriate reference ranges are used.

Recommendations from others

The United States Preventive Services Task Force (USPSTF),⁷ the American Academy of Family Physicians,⁸ the American Academy of Pediatrics,⁹ and the American Thyroid Association (ATA)¹⁰ all recommend routine screening of asymptomatic newborns for congenital hypothyroidism. The USPSTF recommends that clinicians evaluate abnormal thyroid screening results with a supplemental lab test, using TSH as the primary test and T4 as the supplemental test.⁷ Additionally, the ATA endorses a second thyroid screening at 7 to 14 days of life to increase specificity of congenital hypothyroidism screening.¹⁰

Evidence summary

Feeding difficulties, inadequate weight gain, and unusual physical exam findings may lead providers to assess thyroid function in newborns. Additionally, while measurement of free T4 is more common, some states use TSH as the required newborn screening assay to evaluate for congenital hypothyroidism. Whether ordered as a screening test or in response to symptoms, an elevated TSH in a newborn requires further investigation.

Congenital hypothyroidism is the most serious cause of an elevated TSH in a newborn. If left untreated, congenital hypothyroidism leads to developmental delay and mental retardation; however, with early treatment, intellectual outcomes are greatly improved. Newborns with an elevated TSH should be evaluated with repeat TSH and free T4 measurements in order to assess for congenital hypothyroidism.

TSH >50 mU/L increases chances of congenital hypothyroidism

Not surprisingly, higher levels of TSH increase the likelihood of congenital hypothyroidism. A study from the Netherlands examined diagnoses of infants with an increased TSH on newborn screening heelstick. Of 112 newborns with a TSH >50 mU/L, 110 (98%) had congenital hypothyroidism on further examination. However, only 34 of 594 (5.7%) newborns with a TSH between 9 and 20 mU/L were diagnosed with congenital hypothyroidism. Nineteen of 46 newborns (41%) with levels between 20 and 50 mU/L had congenital hypothyroidism.¹

Other common causes of hypothyroidism

While TSH is generally an accurate measurement of thyroid function, other factors can also lead to an elevated level. The same Dutch study mentioned above explored the presumed causes of elevated TSH among children who were diagnosed with transient hypothyroidism (initially elevated TSH level found to be normal on follow-up testing). The most common causes were thyroid-binding globulin deficiency (200/548 or 36% of newborns with transient hypothyroidism), severe illness (36%), prematurity (8%), and errors in screening procedures (4%).

Another study confirmed that TSH levels were higher in infants born preterm; babies with the earliest gestational ages had the highest TSH levels. The same study also found that TSH levels increased with increasing degrees of illness. Very preterm babies, those with cerebral pathology, low Apgar scores, respiratory distress syndrome, persistent ductus arteriosus requiring treatment, and necrotizing enterocolitis were at highest risk for having abnormally elevated TSH levels in this study.² If a sample is drawn from a newborn exhibiting symptoms (such as poor feeding or hypotonia), the TSH level may be elevated in spite of normal thyroid function.

Maternal thyroid disease can also cause a suppression of thyroid function in the newborn. One study¹ found that of 34 children with transient hypothyroidism, 10 had mothers with undertreated or unrecognized Graves disease.

Finally, either iodine excess or iodine deficiency can cause transient hypothyroidism. Case reports have directly demonstrated the effects of topical iodine exposure on newborn TSH levels.^3-5 Deficiency of dietary iodine is a common cause of both congenital and transient hypothyroidism in newborns worldwide, although it is rare in the United States. The World Health Organization lists 54 countries with inadequate iodine intake; consider children from these countries at high risk for hypothyroidism due to iodine deficiency.⁶

Draw TSH on the second or third day of life

In term, healthy newborns, TSH levels normally increase to levels of 60 mU/L within 30 minutes of delivery. This is followed by a rapid decline in TSH levels over the first 5 days of life to <10 mU/L. An Australian study found more elevated TSH levels for samples drawn on day 2 of life compared with day 3 of life, likely reflecting normal postnatal physiology.³ Age-specific reference ranges are necessary for interpretation of TSH levels during the first 5 days of life. The second or third day of life remains the optimal time for screening when appropriate reference ranges are used.

Recommendations from others

The United States Preventive Services Task Force (USPSTF),⁷ the American Academy of Family Physicians,⁸ the American Academy of Pediatrics,⁹ and the American Thyroid Association (ATA)¹⁰ all recommend routine screening of asymptomatic newborns for congenital hypothyroidism. The USPSTF recommends that clinicians evaluate abnormal thyroid screening results with a supplemental lab test, using TSH as the primary test and T4 as the supplemental test.⁷ Additionally, the ATA endorses a second thyroid screening at 7 to 14 days of life to increase specificity of congenital hypothyroidism screening.¹⁰

Evidence summary

Surveys suggest that patients prefer a detailed disclosure

Our review of the medical literature since 2000 found 238 articles regarding disclosure of medical errors. Of these, 17 contain empirical data. There have been no randomized controlled trials regarding the effects of disclosure, and evidence is limited about the best method for carrying out a disclosure or the consequences of disclosure in clinical settings.

The 17 studies we selected all presented case scenarios involving medical errors—either as surveys or in focus groups—to physicians or lay persons. Respondents were asked to imagine a given scenario and then describe the feelings or emotions that were generated by specific errors, whether disclosures should be made, and whether disclosure affects the likelihood of patients seeking legal advice.

Survey data suggest that patients prefer detailed disclosure about what happened, why it happened, the consequences, and strategies for preventing future errors.^1-4 Regardless of whether a full disclosure occurred, patients are more likely to seek legal advice if they perceive the error as having serious consequences.^2,5

Physicians and nurses describe negative emotional consequences when they realize they have made an error.^6-7 They discuss errors among themselves, but are reluctant to disclose errors to patients.⁸

Even though physicians feel that disclosure of errors is important, they may lack the skills to make a successful disclosure or feel they do not have institutional support for disclosure.⁹ Most data suggest that physician well-being is improved by discussion of errors with patients and colleagues.^1,10

Legal ramifications of disclosure are unclear

Very little evidence exists regarding the effect of disclosure on the likelihood of legal action in actual practice.⁵ Eighty-five percent of respondents to one survey indicated that financial compensation of patients affected by medical errors is appropriate. But the spectrum of repercussions can vary from waiving charges for minor incidents to considering early settlement in serious cases.²

Recommendations from others

Disclosure of medical errors is recommended by numerous medical ethicists and is a key component of patient safety initiatives. Guidelines developed in Australia are in the TABLE.¹¹

The American Medical Association’s Code of Ethics states: “Physicians must offer professional and compassionate concern toward patients who have been harmed, regardless of whether the harm was caused by a health care error. An expression of concern need not be an admission of responsibility. When patient harm has been caused by an error, physicians should offer a general explanation regarding the nature of the error and the measures being taken to prevent similar occurrences in the future. Such communication is fundamental to the trust that underlies the patient-physician relationship, and may help reduce the risk of liability.”¹⁰

TABLE
How to manage a medical error

Evidence summary

Surveys suggest that patients prefer a detailed disclosure

Our review of the medical literature since 2000 found 238 articles regarding disclosure of medical errors. Of these, 17 contain empirical data. There have been no randomized controlled trials regarding the effects of disclosure, and evidence is limited about the best method for carrying out a disclosure or the consequences of disclosure in clinical settings.

The 17 studies we selected all presented case scenarios involving medical errors—either as surveys or in focus groups—to physicians or lay persons. Respondents were asked to imagine a given scenario and then describe the feelings or emotions that were generated by specific errors, whether disclosures should be made, and whether disclosure affects the likelihood of patients seeking legal advice.

Survey data suggest that patients prefer detailed disclosure about what happened, why it happened, the consequences, and strategies for preventing future errors.^1-4 Regardless of whether a full disclosure occurred, patients are more likely to seek legal advice if they perceive the error as having serious consequences.^2,5

Physicians and nurses describe negative emotional consequences when they realize they have made an error.^6-7 They discuss errors among themselves, but are reluctant to disclose errors to patients.⁸

Even though physicians feel that disclosure of errors is important, they may lack the skills to make a successful disclosure or feel they do not have institutional support for disclosure.⁹ Most data suggest that physician well-being is improved by discussion of errors with patients and colleagues.^1,10

Legal ramifications of disclosure are unclear

Very little evidence exists regarding the effect of disclosure on the likelihood of legal action in actual practice.⁵ Eighty-five percent of respondents to one survey indicated that financial compensation of patients affected by medical errors is appropriate. But the spectrum of repercussions can vary from waiving charges for minor incidents to considering early settlement in serious cases.²

Recommendations from others

Disclosure of medical errors is recommended by numerous medical ethicists and is a key component of patient safety initiatives. Guidelines developed in Australia are in the TABLE.¹¹

The American Medical Association’s Code of Ethics states: “Physicians must offer professional and compassionate concern toward patients who have been harmed, regardless of whether the harm was caused by a health care error. An expression of concern need not be an admission of responsibility. When patient harm has been caused by an error, physicians should offer a general explanation regarding the nature of the error and the measures being taken to prevent similar occurrences in the future. Such communication is fundamental to the trust that underlies the patient-physician relationship, and may help reduce the risk of liability.”¹⁰

TABLE
How to manage a medical error

Evidence summary

Surveys suggest that patients prefer a detailed disclosure

Our review of the medical literature since 2000 found 238 articles regarding disclosure of medical errors. Of these, 17 contain empirical data. There have been no randomized controlled trials regarding the effects of disclosure, and evidence is limited about the best method for carrying out a disclosure or the consequences of disclosure in clinical settings.

The 17 studies we selected all presented case scenarios involving medical errors—either as surveys or in focus groups—to physicians or lay persons. Respondents were asked to imagine a given scenario and then describe the feelings or emotions that were generated by specific errors, whether disclosures should be made, and whether disclosure affects the likelihood of patients seeking legal advice.

Survey data suggest that patients prefer detailed disclosure about what happened, why it happened, the consequences, and strategies for preventing future errors.^1-4 Regardless of whether a full disclosure occurred, patients are more likely to seek legal advice if they perceive the error as having serious consequences.^2,5

Physicians and nurses describe negative emotional consequences when they realize they have made an error.^6-7 They discuss errors among themselves, but are reluctant to disclose errors to patients.⁸

Even though physicians feel that disclosure of errors is important, they may lack the skills to make a successful disclosure or feel they do not have institutional support for disclosure.⁹ Most data suggest that physician well-being is improved by discussion of errors with patients and colleagues.^1,10

Legal ramifications of disclosure are unclear

Very little evidence exists regarding the effect of disclosure on the likelihood of legal action in actual practice.⁵ Eighty-five percent of respondents to one survey indicated that financial compensation of patients affected by medical errors is appropriate. But the spectrum of repercussions can vary from waiving charges for minor incidents to considering early settlement in serious cases.²

Recommendations from others

Disclosure of medical errors is recommended by numerous medical ethicists and is a key component of patient safety initiatives. Guidelines developed in Australia are in the TABLE.¹¹

The American Medical Association’s Code of Ethics states: “Physicians must offer professional and compassionate concern toward patients who have been harmed, regardless of whether the harm was caused by a health care error. An expression of concern need not be an admission of responsibility. When patient harm has been caused by an error, physicians should offer a general explanation regarding the nature of the error and the measures being taken to prevent similar occurrences in the future. Such communication is fundamental to the trust that underlies the patient-physician relationship, and may help reduce the risk of liability.”¹⁰

TABLE
How to manage a medical error

Evidence summary

Studies without risk stratification find little benefit from beta-blockers

A systematic review including 25 randomized controlled trials (RCTs) evaluated perioperative beta-blocker therapy for noncardiac surgery in a total of 2722 patients who were not stratified according to cardiac risk status.¹

Perioperative beta-blockers produced no significant effect on:

• all-cause mortality (odds ratio [OR]=0.78; 95% confidence interval [CI], 0.33–1.87),
• acute myocardial infarction (OR=0.59; 95% CI, 0.25–1.39),
• atrial fibrillation/flutter and other supraventricular arrhythmias (OR=0.43; 95% CI, 0.14–1.37), or
• length of hospital stay (weighted mean difference, –5.6 days; 95% CI, –12.2, 1.04).

However, in this review perioperative beta-blocker therapy reduced perioperative myocardial ischemia (OR=0.38; 95% CI, 0.21–0.69), and increased 2 adverse outcomes: hemodynamically significant bradycardia (OR=1.98; 95% CI, 1.08– 3.66) and hypotension requiring treatment (OR=2.52; 95% CI, 1.94–3.28).¹ This study was limited by the inconsistent definition and assessment of outcomes among the individual trials.

No effect on total mortality, but a benefit in a composite outcome. An earlier systematic review with 22 RCTs and 2437 total patients, that was also not stratified according to cardiac risk status, found no effect from perioperative beta-blockers on total mortality or cardiovascular mortality alone. However, it did demonstrate a composite outcome of reduced cardiovascular mortality, reduced nonfatal myocardial infarction, and reduced nonfatal cardiac arrest (relative risk [RR]=0.44; 95% CI, 0.20–0.97) 30 days after surgery.²

Beta-blockers benefit certain high-risk patients
A retrospective cohort study evaluated the effect of perioperative beta-blocker therapy on perioperative mortality, according to preoperative RCRI assessment.³ The study population included 663,635 adults (mean age, 62 years) undergoing major noncardiac surgery at 329 US hospitals. Researchers calculated individual RCRI scores (TABLE).⁴ Half the patients had an RCRI of 0, 38% had an RCRI of 1, 10% had an RCRI of 2, and only 2% had an RCRI of 3 or greater.

Be wary of beta-blockers in low-risk patients. In the 580,665 patients with low cardiac risk, perioperative beta-blocker therapy increased the risk of in-hospital death: for all patients with an RCRI of 0: OR=1.36 (95% CI, 1.27–1.45); number needed to harm (NNH)=208; for all patients with an RCRI of 1: OR=1.09 (95% CI, 1.01–1.19); NNH=504).

A different story for high-risk cardiac patients. Perioperative beta-blocker therapy reduced the risk of in-hospital death in patients with an RCRI of 2 (OR=0.88; 95% CI, 0.80–0.98; number needed to treat [NNT]=227), an RCRI of 3 (OR=0.71; 95% CI, 0.63–0.80; NNT=62), or an RCRI of 4 or more (OR=0.58; 95% CI, 0.50–0.67; NNT=33).

TABLE
Before surgery, calculate your patient’s cardiac risk

Long-acting beta-blockers=fewer MIs

A population-based, retrospective cohort analysis with 37,151 patients over 65 years of age compared perioperative beta-blocker therapy using atenolol (a long-acting beta-blocker) with metoprolol (a shorter-acting beta-blocker) for elective surgery.⁵ Investigators excluded patients with symptomatic coronary disease.

Patients taking atenolol had fewer MIs (1.6% vs 2.0%, P=.004) and fewer deaths (1.2% vs 1.6%, P=.007) when compared with metoprolol. Atenolol produced a 13% relative risk reduction over metoprolol for MI or death after adjusting for age, sex, type of surgery, and use of furosemide, calcium-channel blockers, angiotensin-converting enzyme inhibitors, and statins (comparative NNT=165).

Dose titration by heart rate

An observational cohort study with 272 patients undergoing elective major vascular surgery (mean age 67.4 years, 80% male) evaluated whether higher doses of beta-blockers and tight heart rate control reduced perioperative myocardial ischemia and troponin T release.⁶ Patients with higher beta-blocker doses, lower heart rates, and lower absolute change in heart rate during the perioperative period had significantly less perioperative myocardial ischemia and troponin T release (P<.0001).

The DECREASE-II trial, a prospective cohort study with 1476 patients undergoing elective open abdominal aortic or infrainguinal arterial reconstruction also found that patients with heart rates <65 beats per minute had a significantly lower risk of cardiac death or MI at 30 days postoperatively (1.3% vs 5.2%, OR=0.24; 95% CI, 0.09–0.66).⁷

Begin therapy 30 days before surgery

Authors of a systematic review including 5 RCTs (586 total patients) evaluating perioperative beta-blocker therapy in noncardiac surgery concluded that beta-blocker therapy should begin as long as 30 days prior to surgery to allow for titration of dose to the target heart rate and continue at least throughout hospitalization (longer if adequate medical follow-up can be arranged postoperatively).⁸

Recommendations from others

The American College of Cardiology/American Heart Association Task Force on Practice Guidelines⁹ recommends:

• continuing beta-blockers for patients already receiving them to treat angina, symptomatic arrhythmias, and hypertension before their surgical risk evaluation.
• initiating perioperative beta-blocker therapy for patients undergoing vascular surgery if they are at high cardiac risk, as evidenced by ischemia on preoperative testing.
• considering perioperative beta-blocker therapy for patients undergoing intermediate-to high-risk procedures if preoperative risk assessment identifies them as having intermediate or higher cardiac risk, and for patients undergoing vascular surgery who are at low cardiac risk.

Evidence summary

Studies without risk stratification find little benefit from beta-blockers

A systematic review including 25 randomized controlled trials (RCTs) evaluated perioperative beta-blocker therapy for noncardiac surgery in a total of 2722 patients who were not stratified according to cardiac risk status.¹

Perioperative beta-blockers produced no significant effect on:

• all-cause mortality (odds ratio [OR]=0.78; 95% confidence interval [CI], 0.33–1.87),
• acute myocardial infarction (OR=0.59; 95% CI, 0.25–1.39),
• atrial fibrillation/flutter and other supraventricular arrhythmias (OR=0.43; 95% CI, 0.14–1.37), or
• length of hospital stay (weighted mean difference, –5.6 days; 95% CI, –12.2, 1.04).

However, in this review perioperative beta-blocker therapy reduced perioperative myocardial ischemia (OR=0.38; 95% CI, 0.21–0.69), and increased 2 adverse outcomes: hemodynamically significant bradycardia (OR=1.98; 95% CI, 1.08– 3.66) and hypotension requiring treatment (OR=2.52; 95% CI, 1.94–3.28).¹ This study was limited by the inconsistent definition and assessment of outcomes among the individual trials.

No effect on total mortality, but a benefit in a composite outcome. An earlier systematic review with 22 RCTs and 2437 total patients, that was also not stratified according to cardiac risk status, found no effect from perioperative beta-blockers on total mortality or cardiovascular mortality alone. However, it did demonstrate a composite outcome of reduced cardiovascular mortality, reduced nonfatal myocardial infarction, and reduced nonfatal cardiac arrest (relative risk [RR]=0.44; 95% CI, 0.20–0.97) 30 days after surgery.²

Beta-blockers benefit certain high-risk patients
A retrospective cohort study evaluated the effect of perioperative beta-blocker therapy on perioperative mortality, according to preoperative RCRI assessment.³ The study population included 663,635 adults (mean age, 62 years) undergoing major noncardiac surgery at 329 US hospitals. Researchers calculated individual RCRI scores (TABLE).⁴ Half the patients had an RCRI of 0, 38% had an RCRI of 1, 10% had an RCRI of 2, and only 2% had an RCRI of 3 or greater.

Be wary of beta-blockers in low-risk patients. In the 580,665 patients with low cardiac risk, perioperative beta-blocker therapy increased the risk of in-hospital death: for all patients with an RCRI of 0: OR=1.36 (95% CI, 1.27–1.45); number needed to harm (NNH)=208; for all patients with an RCRI of 1: OR=1.09 (95% CI, 1.01–1.19); NNH=504).

A different story for high-risk cardiac patients. Perioperative beta-blocker therapy reduced the risk of in-hospital death in patients with an RCRI of 2 (OR=0.88; 95% CI, 0.80–0.98; number needed to treat [NNT]=227), an RCRI of 3 (OR=0.71; 95% CI, 0.63–0.80; NNT=62), or an RCRI of 4 or more (OR=0.58; 95% CI, 0.50–0.67; NNT=33).

TABLE
Before surgery, calculate your patient’s cardiac risk

Long-acting beta-blockers=fewer MIs

A population-based, retrospective cohort analysis with 37,151 patients over 65 years of age compared perioperative beta-blocker therapy using atenolol (a long-acting beta-blocker) with metoprolol (a shorter-acting beta-blocker) for elective surgery.⁵ Investigators excluded patients with symptomatic coronary disease.

Patients taking atenolol had fewer MIs (1.6% vs 2.0%, P=.004) and fewer deaths (1.2% vs 1.6%, P=.007) when compared with metoprolol. Atenolol produced a 13% relative risk reduction over metoprolol for MI or death after adjusting for age, sex, type of surgery, and use of furosemide, calcium-channel blockers, angiotensin-converting enzyme inhibitors, and statins (comparative NNT=165).

Dose titration by heart rate

An observational cohort study with 272 patients undergoing elective major vascular surgery (mean age 67.4 years, 80% male) evaluated whether higher doses of beta-blockers and tight heart rate control reduced perioperative myocardial ischemia and troponin T release.⁶ Patients with higher beta-blocker doses, lower heart rates, and lower absolute change in heart rate during the perioperative period had significantly less perioperative myocardial ischemia and troponin T release (P<.0001).

The DECREASE-II trial, a prospective cohort study with 1476 patients undergoing elective open abdominal aortic or infrainguinal arterial reconstruction also found that patients with heart rates <65 beats per minute had a significantly lower risk of cardiac death or MI at 30 days postoperatively (1.3% vs 5.2%, OR=0.24; 95% CI, 0.09–0.66).⁷

Begin therapy 30 days before surgery

Authors of a systematic review including 5 RCTs (586 total patients) evaluating perioperative beta-blocker therapy in noncardiac surgery concluded that beta-blocker therapy should begin as long as 30 days prior to surgery to allow for titration of dose to the target heart rate and continue at least throughout hospitalization (longer if adequate medical follow-up can be arranged postoperatively).⁸

Recommendations from others

The American College of Cardiology/American Heart Association Task Force on Practice Guidelines⁹ recommends:

• continuing beta-blockers for patients already receiving them to treat angina, symptomatic arrhythmias, and hypertension before their surgical risk evaluation.
• initiating perioperative beta-blocker therapy for patients undergoing vascular surgery if they are at high cardiac risk, as evidenced by ischemia on preoperative testing.
• considering perioperative beta-blocker therapy for patients undergoing intermediate-to high-risk procedures if preoperative risk assessment identifies them as having intermediate or higher cardiac risk, and for patients undergoing vascular surgery who are at low cardiac risk.

Evidence summary

Studies without risk stratification find little benefit from beta-blockers

A systematic review including 25 randomized controlled trials (RCTs) evaluated perioperative beta-blocker therapy for noncardiac surgery in a total of 2722 patients who were not stratified according to cardiac risk status.¹

Perioperative beta-blockers produced no significant effect on:

• all-cause mortality (odds ratio [OR]=0.78; 95% confidence interval [CI], 0.33–1.87),
• acute myocardial infarction (OR=0.59; 95% CI, 0.25–1.39),
• atrial fibrillation/flutter and other supraventricular arrhythmias (OR=0.43; 95% CI, 0.14–1.37), or
• length of hospital stay (weighted mean difference, –5.6 days; 95% CI, –12.2, 1.04).

However, in this review perioperative beta-blocker therapy reduced perioperative myocardial ischemia (OR=0.38; 95% CI, 0.21–0.69), and increased 2 adverse outcomes: hemodynamically significant bradycardia (OR=1.98; 95% CI, 1.08– 3.66) and hypotension requiring treatment (OR=2.52; 95% CI, 1.94–3.28).¹ This study was limited by the inconsistent definition and assessment of outcomes among the individual trials.

No effect on total mortality, but a benefit in a composite outcome. An earlier systematic review with 22 RCTs and 2437 total patients, that was also not stratified according to cardiac risk status, found no effect from perioperative beta-blockers on total mortality or cardiovascular mortality alone. However, it did demonstrate a composite outcome of reduced cardiovascular mortality, reduced nonfatal myocardial infarction, and reduced nonfatal cardiac arrest (relative risk [RR]=0.44; 95% CI, 0.20–0.97) 30 days after surgery.²

Beta-blockers benefit certain high-risk patients
A retrospective cohort study evaluated the effect of perioperative beta-blocker therapy on perioperative mortality, according to preoperative RCRI assessment.³ The study population included 663,635 adults (mean age, 62 years) undergoing major noncardiac surgery at 329 US hospitals. Researchers calculated individual RCRI scores (TABLE).⁴ Half the patients had an RCRI of 0, 38% had an RCRI of 1, 10% had an RCRI of 2, and only 2% had an RCRI of 3 or greater.

Be wary of beta-blockers in low-risk patients. In the 580,665 patients with low cardiac risk, perioperative beta-blocker therapy increased the risk of in-hospital death: for all patients with an RCRI of 0: OR=1.36 (95% CI, 1.27–1.45); number needed to harm (NNH)=208; for all patients with an RCRI of 1: OR=1.09 (95% CI, 1.01–1.19); NNH=504).

A different story for high-risk cardiac patients. Perioperative beta-blocker therapy reduced the risk of in-hospital death in patients with an RCRI of 2 (OR=0.88; 95% CI, 0.80–0.98; number needed to treat [NNT]=227), an RCRI of 3 (OR=0.71; 95% CI, 0.63–0.80; NNT=62), or an RCRI of 4 or more (OR=0.58; 95% CI, 0.50–0.67; NNT=33).

TABLE
Before surgery, calculate your patient’s cardiac risk

Long-acting beta-blockers=fewer MIs

A population-based, retrospective cohort analysis with 37,151 patients over 65 years of age compared perioperative beta-blocker therapy using atenolol (a long-acting beta-blocker) with metoprolol (a shorter-acting beta-blocker) for elective surgery.⁵ Investigators excluded patients with symptomatic coronary disease.

Patients taking atenolol had fewer MIs (1.6% vs 2.0%, P=.004) and fewer deaths (1.2% vs 1.6%, P=.007) when compared with metoprolol. Atenolol produced a 13% relative risk reduction over metoprolol for MI or death after adjusting for age, sex, type of surgery, and use of furosemide, calcium-channel blockers, angiotensin-converting enzyme inhibitors, and statins (comparative NNT=165).

Dose titration by heart rate

An observational cohort study with 272 patients undergoing elective major vascular surgery (mean age 67.4 years, 80% male) evaluated whether higher doses of beta-blockers and tight heart rate control reduced perioperative myocardial ischemia and troponin T release.⁶ Patients with higher beta-blocker doses, lower heart rates, and lower absolute change in heart rate during the perioperative period had significantly less perioperative myocardial ischemia and troponin T release (P<.0001).

The DECREASE-II trial, a prospective cohort study with 1476 patients undergoing elective open abdominal aortic or infrainguinal arterial reconstruction also found that patients with heart rates <65 beats per minute had a significantly lower risk of cardiac death or MI at 30 days postoperatively (1.3% vs 5.2%, OR=0.24; 95% CI, 0.09–0.66).⁷

Begin therapy 30 days before surgery

Authors of a systematic review including 5 RCTs (586 total patients) evaluating perioperative beta-blocker therapy in noncardiac surgery concluded that beta-blocker therapy should begin as long as 30 days prior to surgery to allow for titration of dose to the target heart rate and continue at least throughout hospitalization (longer if adequate medical follow-up can be arranged postoperatively).⁸

Recommendations from others

The American College of Cardiology/American Heart Association Task Force on Practice Guidelines⁹ recommends:

• continuing beta-blockers for patients already receiving them to treat angina, symptomatic arrhythmias, and hypertension before their surgical risk evaluation.
• initiating perioperative beta-blocker therapy for patients undergoing vascular surgery if they are at high cardiac risk, as evidenced by ischemia on preoperative testing.
• considering perioperative beta-blocker therapy for patients undergoing intermediate-to high-risk procedures if preoperative risk assessment identifies them as having intermediate or higher cardiac risk, and for patients undergoing vascular surgery who are at low cardiac risk.

Evidence summary

The prevalence of temporal arteritis (also called giant cell arteritis) increases significantly with age. For those under 50 years of age, this condition is extremely rare; the prevalence increases exponentially with age.¹

Jaw claudication quadruples likelihood of temporal arteritis

A 2002 systematic review² and a 2005 decision analysis³ examined validating cohort studies to determine the likelihood ratios of symptoms, signs, and blood tests (TABLE). These cohort studies are subject to verification bias, as most cohorts represent a selected sample of patients who had a positive temporal artery biopsy. The authors of the 2005 decision analysis note that unilateral temporal artery biopsy has a mean sensitivity of 86.9% (95% confidence interval [CI], 83.1%–90.6%) when compared with a gold standard derived from bilateral artery biopsy, American College of Rheumatology criteria, or clinical diagnosis.³

A headache—even a temporal headache—has a low positive likelihood ratio. Diplopia doubles and jaw claudication quadruples the likelihood of temporal arteritis, but the presence of other symptoms (such as anorexia, weight loss, arthralgia, fatigue, fever, polymyalgia rheumatica, vertigo, and unilateral visual loss) does not significantly increase the probability of temporal arteritis. An abnormal temporal artery on physical examination doubles the likelihood of temporal arteritis.²

TABLE
Jaw claudication and thrombocytosis increase likelihood of temporal arteritis^2,3

Testing for thrombocytosis more helpful than ESR

Lab tests using ESR have been examined in various studies. A high ESR (>100 mm/hour) may only slightly increase the chance of temporal arteritis (TABLE). Five studies³ have documented that thrombocytosis (platelets >375,000/mm³) is more helpful for ruling in temporal arteritis than an elevated ESR.⁴ Conversely, normal platelets are more accurate for ruling out temporal arteritis than a normal ESR.

Recommendations from others

According to the American College of Rheumatology, a patient is said to have temporal arteritis when 3 of the following 5 criteria are met:

• age of onset ≥50 years of age
• new onset of localized headache
• temporal artery tenderness or decreased pulsation
• ESR >50 mm/hour
• abnormal artery biopsy.^5,6

Evidence summary

The prevalence of temporal arteritis (also called giant cell arteritis) increases significantly with age. For those under 50 years of age, this condition is extremely rare; the prevalence increases exponentially with age.¹

Jaw claudication quadruples likelihood of temporal arteritis

A 2002 systematic review² and a 2005 decision analysis³ examined validating cohort studies to determine the likelihood ratios of symptoms, signs, and blood tests (TABLE). These cohort studies are subject to verification bias, as most cohorts represent a selected sample of patients who had a positive temporal artery biopsy. The authors of the 2005 decision analysis note that unilateral temporal artery biopsy has a mean sensitivity of 86.9% (95% confidence interval [CI], 83.1%–90.6%) when compared with a gold standard derived from bilateral artery biopsy, American College of Rheumatology criteria, or clinical diagnosis.³

A headache—even a temporal headache—has a low positive likelihood ratio. Diplopia doubles and jaw claudication quadruples the likelihood of temporal arteritis, but the presence of other symptoms (such as anorexia, weight loss, arthralgia, fatigue, fever, polymyalgia rheumatica, vertigo, and unilateral visual loss) does not significantly increase the probability of temporal arteritis. An abnormal temporal artery on physical examination doubles the likelihood of temporal arteritis.²

TABLE
Jaw claudication and thrombocytosis increase likelihood of temporal arteritis^2,3

Testing for thrombocytosis more helpful than ESR

Lab tests using ESR have been examined in various studies. A high ESR (>100 mm/hour) may only slightly increase the chance of temporal arteritis (TABLE). Five studies³ have documented that thrombocytosis (platelets >375,000/mm³) is more helpful for ruling in temporal arteritis than an elevated ESR.⁴ Conversely, normal platelets are more accurate for ruling out temporal arteritis than a normal ESR.

Recommendations from others

According to the American College of Rheumatology, a patient is said to have temporal arteritis when 3 of the following 5 criteria are met:

• age of onset ≥50 years of age
• new onset of localized headache
• temporal artery tenderness or decreased pulsation
• ESR >50 mm/hour
• abnormal artery biopsy.^5,6

Evidence summary

The prevalence of temporal arteritis (also called giant cell arteritis) increases significantly with age. For those under 50 years of age, this condition is extremely rare; the prevalence increases exponentially with age.¹

Jaw claudication quadruples likelihood of temporal arteritis

A 2002 systematic review² and a 2005 decision analysis³ examined validating cohort studies to determine the likelihood ratios of symptoms, signs, and blood tests (TABLE). These cohort studies are subject to verification bias, as most cohorts represent a selected sample of patients who had a positive temporal artery biopsy. The authors of the 2005 decision analysis note that unilateral temporal artery biopsy has a mean sensitivity of 86.9% (95% confidence interval [CI], 83.1%–90.6%) when compared with a gold standard derived from bilateral artery biopsy, American College of Rheumatology criteria, or clinical diagnosis.³

A headache—even a temporal headache—has a low positive likelihood ratio. Diplopia doubles and jaw claudication quadruples the likelihood of temporal arteritis, but the presence of other symptoms (such as anorexia, weight loss, arthralgia, fatigue, fever, polymyalgia rheumatica, vertigo, and unilateral visual loss) does not significantly increase the probability of temporal arteritis. An abnormal temporal artery on physical examination doubles the likelihood of temporal arteritis.²

TABLE
Jaw claudication and thrombocytosis increase likelihood of temporal arteritis^2,3

Testing for thrombocytosis more helpful than ESR

Lab tests using ESR have been examined in various studies. A high ESR (>100 mm/hour) may only slightly increase the chance of temporal arteritis (TABLE). Five studies³ have documented that thrombocytosis (platelets >375,000/mm³) is more helpful for ruling in temporal arteritis than an elevated ESR.⁴ Conversely, normal platelets are more accurate for ruling out temporal arteritis than a normal ESR.

Recommendations from others

According to the American College of Rheumatology, a patient is said to have temporal arteritis when 3 of the following 5 criteria are met:

• age of onset ≥50 years of age
• new onset of localized headache
• temporal artery tenderness or decreased pulsation
• ESR >50 mm/hour
• abnormal artery biopsy.^5,6

Evidence summary

These criteria help determine who needs an x-ray

The Ottawa Knee Rules recommend knee x-rays for any patient meeting one of these criteria:¹

1. ≥55 years of age
2. Tenderness over the head of the fibula or isolated to the patella without other bony tenderness
3. Unable to flex the knee to 90°, or unable to bear weight (for at least 4 steps) both immediately and in the emergency department.

Should you order an x-ray?

Ottawa rules for knee x-ray

A knee x-ray is required only for knee injury patients with any of these findings:

• 55 years of age or older
• isolated tenderness of the patella (no bone tenderness of the knee other than the patella)
• tenderness at the head of the fibula
• inability to flex the knee to 90°
• inability to weight bear both immediately and in the ED (4 steps, unable to transfer weight twice onto each lower limb regardless of limping). (www.gp-training.net/rheum/ottawa.htm)

A downhill skier is catching the inside edge of the front of a ski in the snow, forcing external rotation of the tibia relative to the femur. The detail of the knee joint shows injury to the medial collateral ligament and anterior cruciate ligament.

In a systematic review,² the Ottawa rules had a pooled negative likelihood ratio (LR–) of 0.05 (95% confidence interval [CI], 0.02–0.23). A prospective study of the Ottawa rules among children with acute knee injury demonstrated a positive likelihood ratio (LR+) of 1.81 (95% CI, 1.47–2.21) and an LR– of 0.16 (95% CI, 0.02–1.04).³ However, limited information is available in the pediatric population.

Ultrasound and MRI both perform well

Physical examination (including the Lachman test, Drawer sign, and McMurray test) by an orthopedist or sports medicine–trained physician was 74% to 88% sensitive and 72% to 95% specific for suspected meniscal or ligamentous injuries; MRI added marginal value in referral decisions regarding these conditions.⁴

In a small nonrandomized study of adult knee trauma, sonographic diagnosis of an effusion (by an expert) had an LR+ for diagnosing an internal derangement of the knee of 2.0 (95% CI, 0.67–5.96) and an LR– of 0.33 (95% CI, 0.12–0.96), as compared with the gold standard of MRI.⁵

In a nonrandomized prospective study of adults receiving MRI of the knee prior to arthroscopy, MRI identified meniscal and ligamentous lesions of the knee satisfactorily (Table 1).⁶ In a retrospective study of adolescents (11 to 17 years of age) having knee MRIs and undergoing knee surgery, MRI compared favorably with the operative (gold standard) diagnosis (Table 2).⁷

TABLE 1
MRI can accurately identify meniscal and ligamentous knee lesions in adults

TABLE 2
MRI compares well with operative Dx in identifying knee lesions in adolescents

Recommendations from others

University of Michigan Health System⁸ guidelines indicate the following:

• Most knee pain is caused by patellofemoral syndrome and osteoarthritis.
• MRI of the knee has not been proven to be superior to the clinical exam by an experienced examiner in the evaluation of acute knee injuries.
• MRI may be useful to assess bone pathology underlying chronic knee pain.
• Differentiating between knee pain without constitutional symptoms, knee pain with constitutional symptoms, and traumatic knee pain is helpful in determining a diagnosis.
• Patients with knee pain and swelling who have non-bloody aspirates may also have serious knee pathology.

Evidence summary

These criteria help determine who needs an x-ray

The Ottawa Knee Rules recommend knee x-rays for any patient meeting one of these criteria:¹

1. ≥55 years of age
2. Tenderness over the head of the fibula or isolated to the patella without other bony tenderness
3. Unable to flex the knee to 90°, or unable to bear weight (for at least 4 steps) both immediately and in the emergency department.

Should you order an x-ray?

Ottawa rules for knee x-ray

A knee x-ray is required only for knee injury patients with any of these findings:

• 55 years of age or older
• isolated tenderness of the patella (no bone tenderness of the knee other than the patella)
• tenderness at the head of the fibula
• inability to flex the knee to 90°
• inability to weight bear both immediately and in the ED (4 steps, unable to transfer weight twice onto each lower limb regardless of limping). (www.gp-training.net/rheum/ottawa.htm)

A downhill skier is catching the inside edge of the front of a ski in the snow, forcing external rotation of the tibia relative to the femur. The detail of the knee joint shows injury to the medial collateral ligament and anterior cruciate ligament.

In a systematic review,² the Ottawa rules had a pooled negative likelihood ratio (LR–) of 0.05 (95% confidence interval [CI], 0.02–0.23). A prospective study of the Ottawa rules among children with acute knee injury demonstrated a positive likelihood ratio (LR+) of 1.81 (95% CI, 1.47–2.21) and an LR– of 0.16 (95% CI, 0.02–1.04).³ However, limited information is available in the pediatric population.

Ultrasound and MRI both perform well

Physical examination (including the Lachman test, Drawer sign, and McMurray test) by an orthopedist or sports medicine–trained physician was 74% to 88% sensitive and 72% to 95% specific for suspected meniscal or ligamentous injuries; MRI added marginal value in referral decisions regarding these conditions.⁴

In a small nonrandomized study of adult knee trauma, sonographic diagnosis of an effusion (by an expert) had an LR+ for diagnosing an internal derangement of the knee of 2.0 (95% CI, 0.67–5.96) and an LR– of 0.33 (95% CI, 0.12–0.96), as compared with the gold standard of MRI.⁵

In a nonrandomized prospective study of adults receiving MRI of the knee prior to arthroscopy, MRI identified meniscal and ligamentous lesions of the knee satisfactorily (Table 1).⁶ In a retrospective study of adolescents (11 to 17 years of age) having knee MRIs and undergoing knee surgery, MRI compared favorably with the operative (gold standard) diagnosis (Table 2).⁷

TABLE 1
MRI can accurately identify meniscal and ligamentous knee lesions in adults

TABLE 2
MRI compares well with operative Dx in identifying knee lesions in adolescents

Recommendations from others

University of Michigan Health System⁸ guidelines indicate the following:

• Most knee pain is caused by patellofemoral syndrome and osteoarthritis.
• MRI of the knee has not been proven to be superior to the clinical exam by an experienced examiner in the evaluation of acute knee injuries.
• MRI may be useful to assess bone pathology underlying chronic knee pain.
• Differentiating between knee pain without constitutional symptoms, knee pain with constitutional symptoms, and traumatic knee pain is helpful in determining a diagnosis.
• Patients with knee pain and swelling who have non-bloody aspirates may also have serious knee pathology.

Evidence summary

These criteria help determine who needs an x-ray

The Ottawa Knee Rules recommend knee x-rays for any patient meeting one of these criteria:¹

1. ≥55 years of age
2. Tenderness over the head of the fibula or isolated to the patella without other bony tenderness
3. Unable to flex the knee to 90°, or unable to bear weight (for at least 4 steps) both immediately and in the emergency department.

Should you order an x-ray?

Ottawa rules for knee x-ray

A knee x-ray is required only for knee injury patients with any of these findings:

• 55 years of age or older
• isolated tenderness of the patella (no bone tenderness of the knee other than the patella)
• tenderness at the head of the fibula
• inability to flex the knee to 90°
• inability to weight bear both immediately and in the ED (4 steps, unable to transfer weight twice onto each lower limb regardless of limping). (www.gp-training.net/rheum/ottawa.htm)

A downhill skier is catching the inside edge of the front of a ski in the snow, forcing external rotation of the tibia relative to the femur. The detail of the knee joint shows injury to the medial collateral ligament and anterior cruciate ligament.

In a systematic review,² the Ottawa rules had a pooled negative likelihood ratio (LR–) of 0.05 (95% confidence interval [CI], 0.02–0.23). A prospective study of the Ottawa rules among children with acute knee injury demonstrated a positive likelihood ratio (LR+) of 1.81 (95% CI, 1.47–2.21) and an LR– of 0.16 (95% CI, 0.02–1.04).³ However, limited information is available in the pediatric population.

Ultrasound and MRI both perform well

Physical examination (including the Lachman test, Drawer sign, and McMurray test) by an orthopedist or sports medicine–trained physician was 74% to 88% sensitive and 72% to 95% specific for suspected meniscal or ligamentous injuries; MRI added marginal value in referral decisions regarding these conditions.⁴

In a small nonrandomized study of adult knee trauma, sonographic diagnosis of an effusion (by an expert) had an LR+ for diagnosing an internal derangement of the knee of 2.0 (95% CI, 0.67–5.96) and an LR– of 0.33 (95% CI, 0.12–0.96), as compared with the gold standard of MRI.⁵

In a nonrandomized prospective study of adults receiving MRI of the knee prior to arthroscopy, MRI identified meniscal and ligamentous lesions of the knee satisfactorily (Table 1).⁶ In a retrospective study of adolescents (11 to 17 years of age) having knee MRIs and undergoing knee surgery, MRI compared favorably with the operative (gold standard) diagnosis (Table 2).⁷

TABLE 1
MRI can accurately identify meniscal and ligamentous knee lesions in adults

TABLE 2
MRI compares well with operative Dx in identifying knee lesions in adolescents

Recommendations from others

University of Michigan Health System⁸ guidelines indicate the following:

• Most knee pain is caused by patellofemoral syndrome and osteoarthritis.
• MRI of the knee has not been proven to be superior to the clinical exam by an experienced examiner in the evaluation of acute knee injuries.
• MRI may be useful to assess bone pathology underlying chronic knee pain.
• Differentiating between knee pain without constitutional symptoms, knee pain with constitutional symptoms, and traumatic knee pain is helpful in determining a diagnosis.
• Patients with knee pain and swelling who have non-bloody aspirates may also have serious knee pathology.

Evidence summary

Herbs and probiotics may help. A Cochrane review of herbal therapies evaluated 75 randomized controlled trials (RCTs), including 7957 patients; it concluded that some herbal preparations may reduce symptoms of IBS. However, more rigorous studies are needed: There was never more than 1 trial comparing a given herbal medicine with a specific control, making it difficult to combine trials in a meaningful way.¹

A multicenter RCT compared 2 herbal formulations with placebo.² The first contained extracts of bitter candy-tuft, chamomile, peppermint, caraway, and licorice. The second, a commercial preparation called Iberogast, had the same ingredients as the first, as well as lemon balm, celandine, angelica, and milk thistle. The study demonstrated global IBS symptom reduction, with a relative risk [RR] of 1.68 (99% confidence interval [CI], 1.00–2.8) for Iberogast, and RR=1.90 (99% CI, 1.15–3.14) for the first formulation. Both formulations were well tolerated and more effective than placebo in the treatment of IBS, regardless of the dominant symptom.

One RCT compared the probiotic formula Bifidobacterium infantis 35624 with Lactobacillus and placebo, and found that IBS symptom scores and quality of life measures were better with the Bifidobacterium preparation (P<.05).³

Soluble fiber or an elimination diet can reduce symptoms. A systematic review of 17 RCTs (9 of soluble fiber, 8 of insoluble) (N=1363) found soluble fiber more effective than placebo (pooled RR=1.55; 95% CI, 1.35–1.78) for reduction in global symptoms and constipation.⁴ Insoluble fiber was no better than placebo (pooled RR=0.89; 95% CI, 0.72–1.11). Abdominal pain was not reduced with either fiber.

An elimination diet, based on the presence of IgG antibodies to various foods, was compared with a sham diet in an RCT.⁵ Those who fully adhered to the diet reported symptom reduction, with a mean symptom score 98 points lower (95% CI, 52–144; NNT=2.5). (A 50-point drop was considered clinically significant.)

Therapy, self-help book, and hypnosis provide relief. A single-blinded study comparing cognitive behavioral therapy (CBT) with patient education alone found CBT more effective.⁶ The effect size was 0.50 (defined as moderate; 95% CI, 0.2–0.8). Analysis of response rates found that 73% responded to CBT vs 41.3% in the education group (NNT=3.2).

Another RCT tested the use of an educational self-help guidebook.⁷ Outcomes were symptom score, perception of improvement, and primary care consultation rates. At 1 year, the self-help guidebook group had 1.56 visits/year less than the control group (95% CI, 1.15–1.98), a 60% decrease. The self-help guidebook group reported a higher degree of perceived improvement, with a mean effect of 0.51 (95% CI, 0.23–0.79); there were no differences in severity scores.

Hypnosis has been evaluated in several studies. A systematic review found 6 studies with a control and 8 without, for a total of 644 patients. An average of 80% of the patients reported global IBS symptom relief. Patients with typical IBS responded to hypnosis; however, males with diarrhea-predominant symptoms, and all subjects with atypical symptoms or comorbid psychopathology were less likely to respond.⁸

Recommendations from others

The American College of Gastroenterology⁹ recommends behavioral therapies (such as relaxation therapy, hypnotherapy, and psychotherapy) for the treatment of individual IBS symptoms. They also report that bulking agents such as insoluble fiber (eg, wheat bran) and soluble fiber (eg, psyllium) are no more effective than placebo at relieving global IBS symptoms.

The American Gastroenterological Association¹⁰ endorses multiple nonpharmacological treatments for IBS, including therapeutic physician-patient relationship, patient education, dietary and lifestyle modifications, symptom monitoring, and behavioral therapies.

Evidence summary

Herbs and probiotics may help. A Cochrane review of herbal therapies evaluated 75 randomized controlled trials (RCTs), including 7957 patients; it concluded that some herbal preparations may reduce symptoms of IBS. However, more rigorous studies are needed: There was never more than 1 trial comparing a given herbal medicine with a specific control, making it difficult to combine trials in a meaningful way.¹

A multicenter RCT compared 2 herbal formulations with placebo.² The first contained extracts of bitter candy-tuft, chamomile, peppermint, caraway, and licorice. The second, a commercial preparation called Iberogast, had the same ingredients as the first, as well as lemon balm, celandine, angelica, and milk thistle. The study demonstrated global IBS symptom reduction, with a relative risk [RR] of 1.68 (99% confidence interval [CI], 1.00–2.8) for Iberogast, and RR=1.90 (99% CI, 1.15–3.14) for the first formulation. Both formulations were well tolerated and more effective than placebo in the treatment of IBS, regardless of the dominant symptom.

One RCT compared the probiotic formula Bifidobacterium infantis 35624 with Lactobacillus and placebo, and found that IBS symptom scores and quality of life measures were better with the Bifidobacterium preparation (P<.05).³

Soluble fiber or an elimination diet can reduce symptoms. A systematic review of 17 RCTs (9 of soluble fiber, 8 of insoluble) (N=1363) found soluble fiber more effective than placebo (pooled RR=1.55; 95% CI, 1.35–1.78) for reduction in global symptoms and constipation.⁴ Insoluble fiber was no better than placebo (pooled RR=0.89; 95% CI, 0.72–1.11). Abdominal pain was not reduced with either fiber.

An elimination diet, based on the presence of IgG antibodies to various foods, was compared with a sham diet in an RCT.⁵ Those who fully adhered to the diet reported symptom reduction, with a mean symptom score 98 points lower (95% CI, 52–144; NNT=2.5). (A 50-point drop was considered clinically significant.)

Therapy, self-help book, and hypnosis provide relief. A single-blinded study comparing cognitive behavioral therapy (CBT) with patient education alone found CBT more effective.⁶ The effect size was 0.50 (defined as moderate; 95% CI, 0.2–0.8). Analysis of response rates found that 73% responded to CBT vs 41.3% in the education group (NNT=3.2).

Another RCT tested the use of an educational self-help guidebook.⁷ Outcomes were symptom score, perception of improvement, and primary care consultation rates. At 1 year, the self-help guidebook group had 1.56 visits/year less than the control group (95% CI, 1.15–1.98), a 60% decrease. The self-help guidebook group reported a higher degree of perceived improvement, with a mean effect of 0.51 (95% CI, 0.23–0.79); there were no differences in severity scores.

Hypnosis has been evaluated in several studies. A systematic review found 6 studies with a control and 8 without, for a total of 644 patients. An average of 80% of the patients reported global IBS symptom relief. Patients with typical IBS responded to hypnosis; however, males with diarrhea-predominant symptoms, and all subjects with atypical symptoms or comorbid psychopathology were less likely to respond.⁸

Recommendations from others

The American College of Gastroenterology⁹ recommends behavioral therapies (such as relaxation therapy, hypnotherapy, and psychotherapy) for the treatment of individual IBS symptoms. They also report that bulking agents such as insoluble fiber (eg, wheat bran) and soluble fiber (eg, psyllium) are no more effective than placebo at relieving global IBS symptoms.

The American Gastroenterological Association¹⁰ endorses multiple nonpharmacological treatments for IBS, including therapeutic physician-patient relationship, patient education, dietary and lifestyle modifications, symptom monitoring, and behavioral therapies.

Evidence summary

Herbs and probiotics may help. A Cochrane review of herbal therapies evaluated 75 randomized controlled trials (RCTs), including 7957 patients; it concluded that some herbal preparations may reduce symptoms of IBS. However, more rigorous studies are needed: There was never more than 1 trial comparing a given herbal medicine with a specific control, making it difficult to combine trials in a meaningful way.¹

A multicenter RCT compared 2 herbal formulations with placebo.² The first contained extracts of bitter candy-tuft, chamomile, peppermint, caraway, and licorice. The second, a commercial preparation called Iberogast, had the same ingredients as the first, as well as lemon balm, celandine, angelica, and milk thistle. The study demonstrated global IBS symptom reduction, with a relative risk [RR] of 1.68 (99% confidence interval [CI], 1.00–2.8) for Iberogast, and RR=1.90 (99% CI, 1.15–3.14) for the first formulation. Both formulations were well tolerated and more effective than placebo in the treatment of IBS, regardless of the dominant symptom.

One RCT compared the probiotic formula Bifidobacterium infantis 35624 with Lactobacillus and placebo, and found that IBS symptom scores and quality of life measures were better with the Bifidobacterium preparation (P<.05).³

Soluble fiber or an elimination diet can reduce symptoms. A systematic review of 17 RCTs (9 of soluble fiber, 8 of insoluble) (N=1363) found soluble fiber more effective than placebo (pooled RR=1.55; 95% CI, 1.35–1.78) for reduction in global symptoms and constipation.⁴ Insoluble fiber was no better than placebo (pooled RR=0.89; 95% CI, 0.72–1.11). Abdominal pain was not reduced with either fiber.

An elimination diet, based on the presence of IgG antibodies to various foods, was compared with a sham diet in an RCT.⁵ Those who fully adhered to the diet reported symptom reduction, with a mean symptom score 98 points lower (95% CI, 52–144; NNT=2.5). (A 50-point drop was considered clinically significant.)

Therapy, self-help book, and hypnosis provide relief. A single-blinded study comparing cognitive behavioral therapy (CBT) with patient education alone found CBT more effective.⁶ The effect size was 0.50 (defined as moderate; 95% CI, 0.2–0.8). Analysis of response rates found that 73% responded to CBT vs 41.3% in the education group (NNT=3.2).

Another RCT tested the use of an educational self-help guidebook.⁷ Outcomes were symptom score, perception of improvement, and primary care consultation rates. At 1 year, the self-help guidebook group had 1.56 visits/year less than the control group (95% CI, 1.15–1.98), a 60% decrease. The self-help guidebook group reported a higher degree of perceived improvement, with a mean effect of 0.51 (95% CI, 0.23–0.79); there were no differences in severity scores.

Hypnosis has been evaluated in several studies. A systematic review found 6 studies with a control and 8 without, for a total of 644 patients. An average of 80% of the patients reported global IBS symptom relief. Patients with typical IBS responded to hypnosis; however, males with diarrhea-predominant symptoms, and all subjects with atypical symptoms or comorbid psychopathology were less likely to respond.⁸

Recommendations from others

The American College of Gastroenterology⁹ recommends behavioral therapies (such as relaxation therapy, hypnotherapy, and psychotherapy) for the treatment of individual IBS symptoms. They also report that bulking agents such as insoluble fiber (eg, wheat bran) and soluble fiber (eg, psyllium) are no more effective than placebo at relieving global IBS symptoms.

The American Gastroenterological Association¹⁰ endorses multiple nonpharmacological treatments for IBS, including therapeutic physician-patient relationship, patient education, dietary and lifestyle modifications, symptom monitoring, and behavioral therapies.