Clinical Inquiries

Evidence summary

Two prospective cohort studies have looked at the usefulness of clinical characteristics to help differentiate MRSA from MSSA infections. The studies—a 2002 observational study of 144 children and a 2007 study of 180 consecutively enrolled adults—found no clear distinguishing features for MRSA.^1,2 They did note some commonly associated risk factors, however (TABLE).^2,3

Abscess formation was the most common presentation of CA-MRSA, followed by purulent cellulitis.^3,4 The prevalence and incidence of nonpurulent CA-MRSA is not well defined.

TABLE
Is it MRSA? A look at the odds

Best treatment bet: Incision and drainage

Incision and drainage remains the mainstay of abscess treatment.^3,5 A 2007 RCT of 166 indigent, inner-city patients with confirmed MRSA investigated combining incision and drainage with 7 days of therapy using either cephalexin or placebo. The primary outcome was clinical cure or failure 7 days after incision and drainage. The trial found no advantage to adding antibiotics; MRSA would likely be resistant to cephalexin in any case.⁶

A 2006 summary from Clinical Evidence found no RCT support for any outpatient antibiotic.⁷ No evidence exists that intranasal mupirocin or antiseptic body washes reduce the recurrence rate.⁷ We found no studies evaluating the optimal treatment of purulent skin and soft tissue infections without abscesses.

Avoid fluoroquinolones

MRSA isolates demonstrate a high resistance to fluoroquinolones, so this class of drugs isn’t recommended.³

Recommendations

The Centers for Disease Control and Prevention (CDC) recommends the following treatment for CA-MRSA:

• drain all abscesses; incision and drainage alone suffices for immunocompetent patients
• for other patients, consider adjunct treatment with clindamycin, trim-ethoprim and sulfamethoxazole, tetracyclines, or linezolid.

The CDC also recommends consulting an infectious disease specialist before using linezolid and avoiding fluoroquinolone and macrolide antibiotics because resistance develops rapidly.⁸ Rifampin can be used in combination with other standard treatments.⁸

The CDC doesn’t recommend treating nonpurulent skin infections with CA-MRSA-specific antibiotics. These infections are generally caused by Streptococcus pyogenes and remain sensitive to β-lactam antibiotics. When the community prevalence of CA-MRSA is low, a β-lactam antibiotic can be used with close follow-up.⁸

The Infectious Diseases Society of America recommends incision and drainage for abscesses and treatment with CA-MRSA-specific antibiotics for purulent skin infections.⁹

Evidence summary

Two prospective cohort studies have looked at the usefulness of clinical characteristics to help differentiate MRSA from MSSA infections. The studies—a 2002 observational study of 144 children and a 2007 study of 180 consecutively enrolled adults—found no clear distinguishing features for MRSA.^1,2 They did note some commonly associated risk factors, however (TABLE).^2,3

Abscess formation was the most common presentation of CA-MRSA, followed by purulent cellulitis.^3,4 The prevalence and incidence of nonpurulent CA-MRSA is not well defined.

TABLE
Is it MRSA? A look at the odds

Best treatment bet: Incision and drainage

Incision and drainage remains the mainstay of abscess treatment.^3,5 A 2007 RCT of 166 indigent, inner-city patients with confirmed MRSA investigated combining incision and drainage with 7 days of therapy using either cephalexin or placebo. The primary outcome was clinical cure or failure 7 days after incision and drainage. The trial found no advantage to adding antibiotics; MRSA would likely be resistant to cephalexin in any case.⁶

A 2006 summary from Clinical Evidence found no RCT support for any outpatient antibiotic.⁷ No evidence exists that intranasal mupirocin or antiseptic body washes reduce the recurrence rate.⁷ We found no studies evaluating the optimal treatment of purulent skin and soft tissue infections without abscesses.

Avoid fluoroquinolones

MRSA isolates demonstrate a high resistance to fluoroquinolones, so this class of drugs isn’t recommended.³

Recommendations

The Centers for Disease Control and Prevention (CDC) recommends the following treatment for CA-MRSA:

• drain all abscesses; incision and drainage alone suffices for immunocompetent patients
• for other patients, consider adjunct treatment with clindamycin, trim-ethoprim and sulfamethoxazole, tetracyclines, or linezolid.

The CDC also recommends consulting an infectious disease specialist before using linezolid and avoiding fluoroquinolone and macrolide antibiotics because resistance develops rapidly.⁸ Rifampin can be used in combination with other standard treatments.⁸

The CDC doesn’t recommend treating nonpurulent skin infections with CA-MRSA-specific antibiotics. These infections are generally caused by Streptococcus pyogenes and remain sensitive to β-lactam antibiotics. When the community prevalence of CA-MRSA is low, a β-lactam antibiotic can be used with close follow-up.⁸

The Infectious Diseases Society of America recommends incision and drainage for abscesses and treatment with CA-MRSA-specific antibiotics for purulent skin infections.⁹

Evidence summary

Two prospective cohort studies have looked at the usefulness of clinical characteristics to help differentiate MRSA from MSSA infections. The studies—a 2002 observational study of 144 children and a 2007 study of 180 consecutively enrolled adults—found no clear distinguishing features for MRSA.^1,2 They did note some commonly associated risk factors, however (TABLE).^2,3

Abscess formation was the most common presentation of CA-MRSA, followed by purulent cellulitis.^3,4 The prevalence and incidence of nonpurulent CA-MRSA is not well defined.

TABLE
Is it MRSA? A look at the odds

Best treatment bet: Incision and drainage

Incision and drainage remains the mainstay of abscess treatment.^3,5 A 2007 RCT of 166 indigent, inner-city patients with confirmed MRSA investigated combining incision and drainage with 7 days of therapy using either cephalexin or placebo. The primary outcome was clinical cure or failure 7 days after incision and drainage. The trial found no advantage to adding antibiotics; MRSA would likely be resistant to cephalexin in any case.⁶

A 2006 summary from Clinical Evidence found no RCT support for any outpatient antibiotic.⁷ No evidence exists that intranasal mupirocin or antiseptic body washes reduce the recurrence rate.⁷ We found no studies evaluating the optimal treatment of purulent skin and soft tissue infections without abscesses.

Avoid fluoroquinolones

MRSA isolates demonstrate a high resistance to fluoroquinolones, so this class of drugs isn’t recommended.³

Recommendations

The Centers for Disease Control and Prevention (CDC) recommends the following treatment for CA-MRSA:

• drain all abscesses; incision and drainage alone suffices for immunocompetent patients
• for other patients, consider adjunct treatment with clindamycin, trim-ethoprim and sulfamethoxazole, tetracyclines, or linezolid.

The CDC also recommends consulting an infectious disease specialist before using linezolid and avoiding fluoroquinolone and macrolide antibiotics because resistance develops rapidly.⁸ Rifampin can be used in combination with other standard treatments.⁸

The CDC doesn’t recommend treating nonpurulent skin infections with CA-MRSA-specific antibiotics. These infections are generally caused by Streptococcus pyogenes and remain sensitive to β-lactam antibiotics. When the community prevalence of CA-MRSA is low, a β-lactam antibiotic can be used with close follow-up.⁸

The Infectious Diseases Society of America recommends incision and drainage for abscesses and treatment with CA-MRSA-specific antibiotics for purulent skin infections.⁹

Evidence summary

Prokinetic agents promote transit of intestinal contents by increasing the frequency or strength of small intestine contractions. Available prokinetics include erythromycin and metoclopramide. Metoclopramide has been tested only for upper gastrointestinal mobility. The only randomized controlled trials (RCTs) of erythromycin for constipation have been conducted in children. Cisapride and tegaserod have been withdrawn from general use because of adverse side effects. The TABLE summarizes the available data.

TABLE
Prokinetics for constipation: What the research tells us

Pediatric constipation: Erythromycin helps; watch dosage

A small RCT of 14 children between 4 and 13 years of age showed that erythromycin improved symptoms of constipation and decreased laxative use (number needed to treat [NNT]=10).¹ Two RCTs in neonates demonstrated that erythromycin shortened intestinal transit time and improved feeding tolerance.^2,3

The erythromycin dose used in these studies was lower than the dosage for antibiotic purposes; no adverse effects were reported. However, cardiac arrhythmias and death have occurred when erythromycin is given to adults and children at the usual antibiotic doses.⁴

Adult constipation: The options are limited

One RCT of cisapride for constipation showed that it improved symptoms,⁵ whereas another demonstrated no significant difference between cisapride and placebo in constipation-predominant irritable bowel syndrome.⁶ Reports of fatal arrhythmias have prompted restrictions on the use of the drug.

In 2 RCTs of tegaserod for constipation, patients exhibited improved abdominal symptoms and increased spontaneous bowel movements (NNT=6 for 2 mg and 5 for 6 mg in the first study; NNT=11 for 2 mg and 7 for 6 mg in the second study).^7,8 A pooled analysis of RCTs of tegaserod revealed an increase in cardiovascular events, prompting withdrawal of the drug from the market (number needed to harm=1000).⁹ Tegaserod is available only for emergency and investigational use.

Renzapride, a newer prokinetic similar to cisapride, is under investigation. It is one tenth the strength of cisapride and carries a lower potential risk of cardiac complications.¹⁰ Two small placebo-controlled trials demonstrated improved abdominal pain and stool consistency, but only 1 showed statistically significant results compared with placebo.^11,12

Recommendations

The North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition states that the benefits of cisapride do not outweigh the risks.¹³ The American College of Gastroenterology’s Chronic Constipation Task Force states that tegaserod effectively treats chronic constipation.¹⁴ Neither guideline includes recommendations regarding other prokinetic agents.

Evidence summary

Prokinetic agents promote transit of intestinal contents by increasing the frequency or strength of small intestine contractions. Available prokinetics include erythromycin and metoclopramide. Metoclopramide has been tested only for upper gastrointestinal mobility. The only randomized controlled trials (RCTs) of erythromycin for constipation have been conducted in children. Cisapride and tegaserod have been withdrawn from general use because of adverse side effects. The TABLE summarizes the available data.

TABLE
Prokinetics for constipation: What the research tells us

Pediatric constipation: Erythromycin helps; watch dosage

A small RCT of 14 children between 4 and 13 years of age showed that erythromycin improved symptoms of constipation and decreased laxative use (number needed to treat [NNT]=10).¹ Two RCTs in neonates demonstrated that erythromycin shortened intestinal transit time and improved feeding tolerance.^2,3

The erythromycin dose used in these studies was lower than the dosage for antibiotic purposes; no adverse effects were reported. However, cardiac arrhythmias and death have occurred when erythromycin is given to adults and children at the usual antibiotic doses.⁴

Adult constipation: The options are limited

One RCT of cisapride for constipation showed that it improved symptoms,⁵ whereas another demonstrated no significant difference between cisapride and placebo in constipation-predominant irritable bowel syndrome.⁶ Reports of fatal arrhythmias have prompted restrictions on the use of the drug.

In 2 RCTs of tegaserod for constipation, patients exhibited improved abdominal symptoms and increased spontaneous bowel movements (NNT=6 for 2 mg and 5 for 6 mg in the first study; NNT=11 for 2 mg and 7 for 6 mg in the second study).^7,8 A pooled analysis of RCTs of tegaserod revealed an increase in cardiovascular events, prompting withdrawal of the drug from the market (number needed to harm=1000).⁹ Tegaserod is available only for emergency and investigational use.

Renzapride, a newer prokinetic similar to cisapride, is under investigation. It is one tenth the strength of cisapride and carries a lower potential risk of cardiac complications.¹⁰ Two small placebo-controlled trials demonstrated improved abdominal pain and stool consistency, but only 1 showed statistically significant results compared with placebo.^11,12

Recommendations

The North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition states that the benefits of cisapride do not outweigh the risks.¹³ The American College of Gastroenterology’s Chronic Constipation Task Force states that tegaserod effectively treats chronic constipation.¹⁴ Neither guideline includes recommendations regarding other prokinetic agents.

Evidence summary

Prokinetic agents promote transit of intestinal contents by increasing the frequency or strength of small intestine contractions. Available prokinetics include erythromycin and metoclopramide. Metoclopramide has been tested only for upper gastrointestinal mobility. The only randomized controlled trials (RCTs) of erythromycin for constipation have been conducted in children. Cisapride and tegaserod have been withdrawn from general use because of adverse side effects. The TABLE summarizes the available data.

TABLE
Prokinetics for constipation: What the research tells us

Pediatric constipation: Erythromycin helps; watch dosage

A small RCT of 14 children between 4 and 13 years of age showed that erythromycin improved symptoms of constipation and decreased laxative use (number needed to treat [NNT]=10).¹ Two RCTs in neonates demonstrated that erythromycin shortened intestinal transit time and improved feeding tolerance.^2,3

The erythromycin dose used in these studies was lower than the dosage for antibiotic purposes; no adverse effects were reported. However, cardiac arrhythmias and death have occurred when erythromycin is given to adults and children at the usual antibiotic doses.⁴

Adult constipation: The options are limited

One RCT of cisapride for constipation showed that it improved symptoms,⁵ whereas another demonstrated no significant difference between cisapride and placebo in constipation-predominant irritable bowel syndrome.⁶ Reports of fatal arrhythmias have prompted restrictions on the use of the drug.

In 2 RCTs of tegaserod for constipation, patients exhibited improved abdominal symptoms and increased spontaneous bowel movements (NNT=6 for 2 mg and 5 for 6 mg in the first study; NNT=11 for 2 mg and 7 for 6 mg in the second study).^7,8 A pooled analysis of RCTs of tegaserod revealed an increase in cardiovascular events, prompting withdrawal of the drug from the market (number needed to harm=1000).⁹ Tegaserod is available only for emergency and investigational use.

Renzapride, a newer prokinetic similar to cisapride, is under investigation. It is one tenth the strength of cisapride and carries a lower potential risk of cardiac complications.¹⁰ Two small placebo-controlled trials demonstrated improved abdominal pain and stool consistency, but only 1 showed statistically significant results compared with placebo.^11,12

Recommendations

The North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition states that the benefits of cisapride do not outweigh the risks.¹³ The American College of Gastroenterology’s Chronic Constipation Task Force states that tegaserod effectively treats chronic constipation.¹⁴ Neither guideline includes recommendations regarding other prokinetic agents.

Evidence summary

Nearly 50% of late-pregnancy IUFDs have no associated risk factors. Fetal demise, however, may be heralded by decreased fetal movement followed by cessation of movement at least 12 hours before death.¹ Maternal monitoring of fetal movement by kick counts has been proposed as a method to verify fetal well-being and decrease the rate of IUFD in the general obstetric population.

Counting doesn’t reduce antenatal death, large study shows

A well-done RCT randomized 68,654 women to either usual care or structured, daily monitoring of fetal movement using the count-to-10 method—daily maternal documentation of the amount of time it takes to perceive 10 fetal movements. Usual care was comprised of a query about fetal movement at antenatal visits and instruction to perform fetal movement monitoring at the provider’s discretion. Mothers were told to visit their health-care provider for evaluation if they felt no movement in 24 hours or fewer than 10 movements in 10 hours during a 48-hour period. The trial showed no benefit from monitoring in reducing the rate of antenatal death from all causes.

The rate of all fetal deaths in the counting group was 2.9 per 1000 normally formed, live, singleton births; the rate in the control group was 2.67 (absolute risk reduction=0.24; 95% confidence interval [CI], –0.5 to 0.98). Women in the counting group spent an average of 160 hours counting during pregnancy and had a statistically significant increase in fetal non-stress-test (NST) monitoring (odds ratio [OR]=1.39; 95% CI, 1.31-1.49; number needed to harm [NNH]=50 to cause 1 additional NST). A statistically insignificant trend toward increased antepartum admissions was also noted in the counting group.²

Maternal perception of less movement not linked to fetal outcome

A retrospective cohort study of 6793 patients compared pregnancy outcomes of 463 women who presented for evaluation of decreased fetal movement with outcomes among the general obstetric population. The study excluded women who reported complete cessation of fetal movement.

Pregnancies evaluated for decreased fetal movement were less likely to have an Apgar score <7 at 5 minutes (relative risk [RR]=0.56; 95% CI, 0.29-0.96; P=.05) and less likely to be preterm (RR=0.68; 95% CI, 0.48-0.94; P=.02). No significant difference in cesarean section for fetal distress or admission to the neonatal intensive care unit was noted between the study and control groups. The study suggests that maternal perception of decreased fetal movement is not associated with poor fetal outcome.³

A recent rigorous systematic review yielded no significant outcome effect related to fetal kick counts.⁴ A prospective cohort study of 4383 births in California, using historical controls, found a drop in fetal mortality from 8.7 to 2.1 deaths/1000. The historical control rate was higher than statewide data from the same time period, however. The overall weaker design of the study and probable effect of regression to the mean significantly limit the interpretation of outcomes.⁵

Recommendations

The American College of Obstetrics and Gynecology (ACOG) makes no recommendation for or against assessing daily fetal movement in routine pregnancies. ACOG notes that no consistent evidence suggests that formal assessment of fetal movement decreases IUFD.⁶

The Institute for Clinical Systems Improvement recommends instructing patients on “daily identification of fetal movement at the 28-week visit.” The institute doesn’t recommend specific criteria for evaluating fetal movements or offer recommendations for follow-up of a maternal report of decreased fetal movement.¹ The National Institute for Clinical Excellence in Great Britain recommends against routine formal fetal-movement counting.⁷

Evidence summary

Nearly 50% of late-pregnancy IUFDs have no associated risk factors. Fetal demise, however, may be heralded by decreased fetal movement followed by cessation of movement at least 12 hours before death.¹ Maternal monitoring of fetal movement by kick counts has been proposed as a method to verify fetal well-being and decrease the rate of IUFD in the general obstetric population.

Counting doesn’t reduce antenatal death, large study shows

A well-done RCT randomized 68,654 women to either usual care or structured, daily monitoring of fetal movement using the count-to-10 method—daily maternal documentation of the amount of time it takes to perceive 10 fetal movements. Usual care was comprised of a query about fetal movement at antenatal visits and instruction to perform fetal movement monitoring at the provider’s discretion. Mothers were told to visit their health-care provider for evaluation if they felt no movement in 24 hours or fewer than 10 movements in 10 hours during a 48-hour period. The trial showed no benefit from monitoring in reducing the rate of antenatal death from all causes.

The rate of all fetal deaths in the counting group was 2.9 per 1000 normally formed, live, singleton births; the rate in the control group was 2.67 (absolute risk reduction=0.24; 95% confidence interval [CI], –0.5 to 0.98). Women in the counting group spent an average of 160 hours counting during pregnancy and had a statistically significant increase in fetal non-stress-test (NST) monitoring (odds ratio [OR]=1.39; 95% CI, 1.31-1.49; number needed to harm [NNH]=50 to cause 1 additional NST). A statistically insignificant trend toward increased antepartum admissions was also noted in the counting group.²

Maternal perception of less movement not linked to fetal outcome

A retrospective cohort study of 6793 patients compared pregnancy outcomes of 463 women who presented for evaluation of decreased fetal movement with outcomes among the general obstetric population. The study excluded women who reported complete cessation of fetal movement.

Pregnancies evaluated for decreased fetal movement were less likely to have an Apgar score <7 at 5 minutes (relative risk [RR]=0.56; 95% CI, 0.29-0.96; P=.05) and less likely to be preterm (RR=0.68; 95% CI, 0.48-0.94; P=.02). No significant difference in cesarean section for fetal distress or admission to the neonatal intensive care unit was noted between the study and control groups. The study suggests that maternal perception of decreased fetal movement is not associated with poor fetal outcome.³

A recent rigorous systematic review yielded no significant outcome effect related to fetal kick counts.⁴ A prospective cohort study of 4383 births in California, using historical controls, found a drop in fetal mortality from 8.7 to 2.1 deaths/1000. The historical control rate was higher than statewide data from the same time period, however. The overall weaker design of the study and probable effect of regression to the mean significantly limit the interpretation of outcomes.⁵

Recommendations

The American College of Obstetrics and Gynecology (ACOG) makes no recommendation for or against assessing daily fetal movement in routine pregnancies. ACOG notes that no consistent evidence suggests that formal assessment of fetal movement decreases IUFD.⁶

The Institute for Clinical Systems Improvement recommends instructing patients on “daily identification of fetal movement at the 28-week visit.” The institute doesn’t recommend specific criteria for evaluating fetal movements or offer recommendations for follow-up of a maternal report of decreased fetal movement.¹ The National Institute for Clinical Excellence in Great Britain recommends against routine formal fetal-movement counting.⁷

Evidence summary

Nearly 50% of late-pregnancy IUFDs have no associated risk factors. Fetal demise, however, may be heralded by decreased fetal movement followed by cessation of movement at least 12 hours before death.¹ Maternal monitoring of fetal movement by kick counts has been proposed as a method to verify fetal well-being and decrease the rate of IUFD in the general obstetric population.

Counting doesn’t reduce antenatal death, large study shows

A well-done RCT randomized 68,654 women to either usual care or structured, daily monitoring of fetal movement using the count-to-10 method—daily maternal documentation of the amount of time it takes to perceive 10 fetal movements. Usual care was comprised of a query about fetal movement at antenatal visits and instruction to perform fetal movement monitoring at the provider’s discretion. Mothers were told to visit their health-care provider for evaluation if they felt no movement in 24 hours or fewer than 10 movements in 10 hours during a 48-hour period. The trial showed no benefit from monitoring in reducing the rate of antenatal death from all causes.

The rate of all fetal deaths in the counting group was 2.9 per 1000 normally formed, live, singleton births; the rate in the control group was 2.67 (absolute risk reduction=0.24; 95% confidence interval [CI], –0.5 to 0.98). Women in the counting group spent an average of 160 hours counting during pregnancy and had a statistically significant increase in fetal non-stress-test (NST) monitoring (odds ratio [OR]=1.39; 95% CI, 1.31-1.49; number needed to harm [NNH]=50 to cause 1 additional NST). A statistically insignificant trend toward increased antepartum admissions was also noted in the counting group.²

Maternal perception of less movement not linked to fetal outcome

A retrospective cohort study of 6793 patients compared pregnancy outcomes of 463 women who presented for evaluation of decreased fetal movement with outcomes among the general obstetric population. The study excluded women who reported complete cessation of fetal movement.

Pregnancies evaluated for decreased fetal movement were less likely to have an Apgar score <7 at 5 minutes (relative risk [RR]=0.56; 95% CI, 0.29-0.96; P=.05) and less likely to be preterm (RR=0.68; 95% CI, 0.48-0.94; P=.02). No significant difference in cesarean section for fetal distress or admission to the neonatal intensive care unit was noted between the study and control groups. The study suggests that maternal perception of decreased fetal movement is not associated with poor fetal outcome.³

A recent rigorous systematic review yielded no significant outcome effect related to fetal kick counts.⁴ A prospective cohort study of 4383 births in California, using historical controls, found a drop in fetal mortality from 8.7 to 2.1 deaths/1000. The historical control rate was higher than statewide data from the same time period, however. The overall weaker design of the study and probable effect of regression to the mean significantly limit the interpretation of outcomes.⁵

Recommendations

The American College of Obstetrics and Gynecology (ACOG) makes no recommendation for or against assessing daily fetal movement in routine pregnancies. ACOG notes that no consistent evidence suggests that formal assessment of fetal movement decreases IUFD.⁶

The Institute for Clinical Systems Improvement recommends instructing patients on “daily identification of fetal movement at the 28-week visit.” The institute doesn’t recommend specific criteria for evaluating fetal movements or offer recommendations for follow-up of a maternal report of decreased fetal movement.¹ The National Institute for Clinical Excellence in Great Britain recommends against routine formal fetal-movement counting.⁷

Evidence summary

Early feeding of solid food—defined as introduction of solids before 4 months of age—has been implicated as a potential cause of several adverse outcomes, including atopy, adiposity, gastrointestinal illness, and celiac disease. But what does the evidence tell us?

Atopy: Conflicting evidence, little support

A 2006 meta-analysis of 13 studies examining the risk of atopic diseases associated with early introduction of solids concluded that “there is insufficient evidence to suggest that, on its own, the early introduction of solids to infants is associated with an increased risk of asthma, food allergy, allergic rhinitis, or animal allergies.”¹

The meta-analysis showed some evidence of an association between early solid introduction and eczema, the strongest being a 10-year New Zealand cohort study (N=1265).² The study concluded that introducing solids and increasing food diversity before 4 months of age was associated with eczema at 2 to 10 years.

Subsequent studies have not produced similar results.^3-6 A well-designed 6-year German birth cohort study (N=2612) found that delaying introduction of solids didn’t decrease the odds of asthma, allergic rhinitis, or food sensitization. Increased food diversity at 4 months (trying 3-8 foods vs none) was associated with some, but not all, measures of eczema.⁴

A German prospective cohort study (N=5991) found no benefit in delaying solids longer than 4 months or potentially allergenic solids longer than 6 months to prevent eczema.⁵ A Dutch prospective cohort study (N=2558) not only confirmed these findings, but found a higher risk of atopy at 2 years of age after delayed introduction of solids, even after accounting for reverse causation (delaying solids in children with early atopic symptoms).⁶

Adiposity and GI illness: No significant association

A randomized, prospective trial (N=165) compared early and late introduction of solids with fat mass at 3, 6, and 12 months of age. No significant difference in adiposity between treatment groups was noted at any end point.⁷ Several cohort studies (N=316, N=313, N=54) have found no statistically significant relationship between early solid feeding and obesity at 6 months, 5 years, and 6 years, respectively.^8-10

A cohort study (N=455) in Dundee, Scotland, reported that infants fed solids before 12 weeks were “significantly heavier” at 26 weeks but not at 52 or 104 weeks; this finding has limited validity, however, because the study didn’t control for parental weight.¹¹ The Dundee study also found no significant association between gastrointestinal illness and the timing of introduction of solids.¹¹

Celiac disease: Timing key for high-risk babies

A prospective observational study (N=1560) of infants at increased risk of celiac disease (defined as having either the HLA-DR3 or DR4 allele) determined that the optimum time for introducing gluten-containing foods is 4 to 6 months. Infants exposed in the first 3 months of life had a 5-fold increased risk of developing celiac disease, and babies exposed after the sixth month had nearly a 2-fold increased risk.¹²

Recommendations

Strong consensus exists for withholding solids until 4 to 6 months of age, including recommendations from the American Academy of Pediatrics,¹³ the European Academy of Allergy and Clinical Immunology,¹⁴ and the World Health Organization.¹⁵ The American College of Allergy, Asthma, and Immunology has adopted a more precise recommendation that solids be introduced no earlier than 6 months.¹⁶

Acknowledgements

The views expressed in this article are those of the authors and do not necessarily reflect the official position of the Department of the Navy, Air Force, Department of Defense, or the United States Government.

Evidence summary

Early feeding of solid food—defined as introduction of solids before 4 months of age—has been implicated as a potential cause of several adverse outcomes, including atopy, adiposity, gastrointestinal illness, and celiac disease. But what does the evidence tell us?

Atopy: Conflicting evidence, little support

A 2006 meta-analysis of 13 studies examining the risk of atopic diseases associated with early introduction of solids concluded that “there is insufficient evidence to suggest that, on its own, the early introduction of solids to infants is associated with an increased risk of asthma, food allergy, allergic rhinitis, or animal allergies.”¹

The meta-analysis showed some evidence of an association between early solid introduction and eczema, the strongest being a 10-year New Zealand cohort study (N=1265).² The study concluded that introducing solids and increasing food diversity before 4 months of age was associated with eczema at 2 to 10 years.

Subsequent studies have not produced similar results.^3-6 A well-designed 6-year German birth cohort study (N=2612) found that delaying introduction of solids didn’t decrease the odds of asthma, allergic rhinitis, or food sensitization. Increased food diversity at 4 months (trying 3-8 foods vs none) was associated with some, but not all, measures of eczema.⁴

A German prospective cohort study (N=5991) found no benefit in delaying solids longer than 4 months or potentially allergenic solids longer than 6 months to prevent eczema.⁵ A Dutch prospective cohort study (N=2558) not only confirmed these findings, but found a higher risk of atopy at 2 years of age after delayed introduction of solids, even after accounting for reverse causation (delaying solids in children with early atopic symptoms).⁶

Adiposity and GI illness: No significant association

A randomized, prospective trial (N=165) compared early and late introduction of solids with fat mass at 3, 6, and 12 months of age. No significant difference in adiposity between treatment groups was noted at any end point.⁷ Several cohort studies (N=316, N=313, N=54) have found no statistically significant relationship between early solid feeding and obesity at 6 months, 5 years, and 6 years, respectively.^8-10

A cohort study (N=455) in Dundee, Scotland, reported that infants fed solids before 12 weeks were “significantly heavier” at 26 weeks but not at 52 or 104 weeks; this finding has limited validity, however, because the study didn’t control for parental weight.¹¹ The Dundee study also found no significant association between gastrointestinal illness and the timing of introduction of solids.¹¹

Celiac disease: Timing key for high-risk babies

A prospective observational study (N=1560) of infants at increased risk of celiac disease (defined as having either the HLA-DR3 or DR4 allele) determined that the optimum time for introducing gluten-containing foods is 4 to 6 months. Infants exposed in the first 3 months of life had a 5-fold increased risk of developing celiac disease, and babies exposed after the sixth month had nearly a 2-fold increased risk.¹²

Recommendations

Strong consensus exists for withholding solids until 4 to 6 months of age, including recommendations from the American Academy of Pediatrics,¹³ the European Academy of Allergy and Clinical Immunology,¹⁴ and the World Health Organization.¹⁵ The American College of Allergy, Asthma, and Immunology has adopted a more precise recommendation that solids be introduced no earlier than 6 months.¹⁶

Acknowledgements

The views expressed in this article are those of the authors and do not necessarily reflect the official position of the Department of the Navy, Air Force, Department of Defense, or the United States Government.

Evidence summary

Early feeding of solid food—defined as introduction of solids before 4 months of age—has been implicated as a potential cause of several adverse outcomes, including atopy, adiposity, gastrointestinal illness, and celiac disease. But what does the evidence tell us?

Atopy: Conflicting evidence, little support

A 2006 meta-analysis of 13 studies examining the risk of atopic diseases associated with early introduction of solids concluded that “there is insufficient evidence to suggest that, on its own, the early introduction of solids to infants is associated with an increased risk of asthma, food allergy, allergic rhinitis, or animal allergies.”¹

The meta-analysis showed some evidence of an association between early solid introduction and eczema, the strongest being a 10-year New Zealand cohort study (N=1265).² The study concluded that introducing solids and increasing food diversity before 4 months of age was associated with eczema at 2 to 10 years.

Subsequent studies have not produced similar results.^3-6 A well-designed 6-year German birth cohort study (N=2612) found that delaying introduction of solids didn’t decrease the odds of asthma, allergic rhinitis, or food sensitization. Increased food diversity at 4 months (trying 3-8 foods vs none) was associated with some, but not all, measures of eczema.⁴

A German prospective cohort study (N=5991) found no benefit in delaying solids longer than 4 months or potentially allergenic solids longer than 6 months to prevent eczema.⁵ A Dutch prospective cohort study (N=2558) not only confirmed these findings, but found a higher risk of atopy at 2 years of age after delayed introduction of solids, even after accounting for reverse causation (delaying solids in children with early atopic symptoms).⁶

Adiposity and GI illness: No significant association

A randomized, prospective trial (N=165) compared early and late introduction of solids with fat mass at 3, 6, and 12 months of age. No significant difference in adiposity between treatment groups was noted at any end point.⁷ Several cohort studies (N=316, N=313, N=54) have found no statistically significant relationship between early solid feeding and obesity at 6 months, 5 years, and 6 years, respectively.^8-10

A cohort study (N=455) in Dundee, Scotland, reported that infants fed solids before 12 weeks were “significantly heavier” at 26 weeks but not at 52 or 104 weeks; this finding has limited validity, however, because the study didn’t control for parental weight.¹¹ The Dundee study also found no significant association between gastrointestinal illness and the timing of introduction of solids.¹¹

Celiac disease: Timing key for high-risk babies

A prospective observational study (N=1560) of infants at increased risk of celiac disease (defined as having either the HLA-DR3 or DR4 allele) determined that the optimum time for introducing gluten-containing foods is 4 to 6 months. Infants exposed in the first 3 months of life had a 5-fold increased risk of developing celiac disease, and babies exposed after the sixth month had nearly a 2-fold increased risk.¹²

Recommendations

Strong consensus exists for withholding solids until 4 to 6 months of age, including recommendations from the American Academy of Pediatrics,¹³ the European Academy of Allergy and Clinical Immunology,¹⁴ and the World Health Organization.¹⁵ The American College of Allergy, Asthma, and Immunology has adopted a more precise recommendation that solids be introduced no earlier than 6 months.¹⁶

Acknowledgements

The views expressed in this article are those of the authors and do not necessarily reflect the official position of the Department of the Navy, Air Force, Department of Defense, or the United States Government.

Evidence summary

PAD affects 7 million to 13 million Americans, or 3% to 18% of the population. Major risk factors include smoking, older age, hyperlipidemia, diabetes mellitus, obesity, cerebrovascular disease, coronary artery disease, hyperhomocysteinemia, and elevated C-reactive protein.¹ PAD may cause claudication, ulcers, impotence, or leg or thigh pain, although 20% to 50% of patients are asymptomatic.²

Femoral artery bruit is better predictor of PAD

Further evaluation of an incidental iliac or femoral artery bruit helps assess the patient’s risk of arterial disease. Auscultation of the femoral arteries for a bruit in asymptomatic patients is a moderately good predictor of PAD (likelihood ratio [LR]=4.80; 95% confidence interval [CI], 2.40-9.50). The absence of a bruit doesn’t exclude disease, however (LR=0.83; 95% CI, 0.73-0.95).³ Auscultation of an iliac artery bruit is a more modest predictor of disease (LR=2.2, no CI provided).⁴

One study of 78 patients showed that a femoral or iliac artery bruit accompanied by either thigh claudication or an abnormal femoral pulse predicted PAD. Patients with 2 out of 3 of these clinical findings had an 83% incidence of aortoiliac disease; the incidence was 100% in patients with all 3 findings.⁵

Another study showed that bruits between the epigastrium and popliteal fossa were found in 63% of 309 patients with arterial disease, but only 7% of 149 patients without PAD diagnosed by AAI or angiogram.⁶

Follow up a bruit with AAI testing

Patients with femoral or iliac artery bruits should undergo AAI testing to assess the severity of disease. The AAI has 95% sensitivity and almost 100% specificity in identifying PAD, compared with angiography.³ An AAI >0.90 is considered normal. An AAI of 0.71 to 0.90 indicates mild disease, 0.41 to 0.70 indicates moderate disease, and ≤0.40 indicates severe disease.

Manage risk factors aggressively

Although no studies show specifically that modifying risk factors in a patient with asymptomatic PAD affects long-term outcomes, aggressive risk factor management is recommended because PAD is highly associated with cerebrovascular and coronary artery disease.¹ No data suggest that treating asymptomatic PAD improves future limb pain or function.

Recommendations

The American College of Cardiology/American Heart Association 2005 Guidelines for the Management of Patients With Peripheral Arterial Disease² make the following recommendations for patients with asymptomatic lower extremity PAD:

• Identify patients with asymptomatic lower extremity PAD by examination or by measuring the ankle-brachial index so therapeutic interventions known to reduce the risk of myocardial infarction, stroke, and death can be offered (level of evidence [LOE]: B).
• Address smoking cessation, lipid lowering, and diabetes and hypertension treatment according to national guidelines (LOE: B).
• Consider antiplatelet therapy to reduce the risk of adverse cardiovascular ischemic events (LOE: C).

The United States Preventive Services Task Force recommends against routine screening for PAD (D recommendation).⁷

Acknowledgements

Special thanks to Felipe Navarro, MD.

Evidence summary

PAD affects 7 million to 13 million Americans, or 3% to 18% of the population. Major risk factors include smoking, older age, hyperlipidemia, diabetes mellitus, obesity, cerebrovascular disease, coronary artery disease, hyperhomocysteinemia, and elevated C-reactive protein.¹ PAD may cause claudication, ulcers, impotence, or leg or thigh pain, although 20% to 50% of patients are asymptomatic.²

Femoral artery bruit is better predictor of PAD

Further evaluation of an incidental iliac or femoral artery bruit helps assess the patient’s risk of arterial disease. Auscultation of the femoral arteries for a bruit in asymptomatic patients is a moderately good predictor of PAD (likelihood ratio [LR]=4.80; 95% confidence interval [CI], 2.40-9.50). The absence of a bruit doesn’t exclude disease, however (LR=0.83; 95% CI, 0.73-0.95).³ Auscultation of an iliac artery bruit is a more modest predictor of disease (LR=2.2, no CI provided).⁴

One study of 78 patients showed that a femoral or iliac artery bruit accompanied by either thigh claudication or an abnormal femoral pulse predicted PAD. Patients with 2 out of 3 of these clinical findings had an 83% incidence of aortoiliac disease; the incidence was 100% in patients with all 3 findings.⁵

Another study showed that bruits between the epigastrium and popliteal fossa were found in 63% of 309 patients with arterial disease, but only 7% of 149 patients without PAD diagnosed by AAI or angiogram.⁶

Follow up a bruit with AAI testing

Patients with femoral or iliac artery bruits should undergo AAI testing to assess the severity of disease. The AAI has 95% sensitivity and almost 100% specificity in identifying PAD, compared with angiography.³ An AAI >0.90 is considered normal. An AAI of 0.71 to 0.90 indicates mild disease, 0.41 to 0.70 indicates moderate disease, and ≤0.40 indicates severe disease.

Manage risk factors aggressively

Although no studies show specifically that modifying risk factors in a patient with asymptomatic PAD affects long-term outcomes, aggressive risk factor management is recommended because PAD is highly associated with cerebrovascular and coronary artery disease.¹ No data suggest that treating asymptomatic PAD improves future limb pain or function.

Recommendations

The American College of Cardiology/American Heart Association 2005 Guidelines for the Management of Patients With Peripheral Arterial Disease² make the following recommendations for patients with asymptomatic lower extremity PAD:

• Identify patients with asymptomatic lower extremity PAD by examination or by measuring the ankle-brachial index so therapeutic interventions known to reduce the risk of myocardial infarction, stroke, and death can be offered (level of evidence [LOE]: B).
• Address smoking cessation, lipid lowering, and diabetes and hypertension treatment according to national guidelines (LOE: B).
• Consider antiplatelet therapy to reduce the risk of adverse cardiovascular ischemic events (LOE: C).

The United States Preventive Services Task Force recommends against routine screening for PAD (D recommendation).⁷

Acknowledgements

Special thanks to Felipe Navarro, MD.

Evidence summary

PAD affects 7 million to 13 million Americans, or 3% to 18% of the population. Major risk factors include smoking, older age, hyperlipidemia, diabetes mellitus, obesity, cerebrovascular disease, coronary artery disease, hyperhomocysteinemia, and elevated C-reactive protein.¹ PAD may cause claudication, ulcers, impotence, or leg or thigh pain, although 20% to 50% of patients are asymptomatic.²

Femoral artery bruit is better predictor of PAD

Further evaluation of an incidental iliac or femoral artery bruit helps assess the patient’s risk of arterial disease. Auscultation of the femoral arteries for a bruit in asymptomatic patients is a moderately good predictor of PAD (likelihood ratio [LR]=4.80; 95% confidence interval [CI], 2.40-9.50). The absence of a bruit doesn’t exclude disease, however (LR=0.83; 95% CI, 0.73-0.95).³ Auscultation of an iliac artery bruit is a more modest predictor of disease (LR=2.2, no CI provided).⁴

One study of 78 patients showed that a femoral or iliac artery bruit accompanied by either thigh claudication or an abnormal femoral pulse predicted PAD. Patients with 2 out of 3 of these clinical findings had an 83% incidence of aortoiliac disease; the incidence was 100% in patients with all 3 findings.⁵

Another study showed that bruits between the epigastrium and popliteal fossa were found in 63% of 309 patients with arterial disease, but only 7% of 149 patients without PAD diagnosed by AAI or angiogram.⁶

Follow up a bruit with AAI testing

Patients with femoral or iliac artery bruits should undergo AAI testing to assess the severity of disease. The AAI has 95% sensitivity and almost 100% specificity in identifying PAD, compared with angiography.³ An AAI >0.90 is considered normal. An AAI of 0.71 to 0.90 indicates mild disease, 0.41 to 0.70 indicates moderate disease, and ≤0.40 indicates severe disease.

Manage risk factors aggressively

Although no studies show specifically that modifying risk factors in a patient with asymptomatic PAD affects long-term outcomes, aggressive risk factor management is recommended because PAD is highly associated with cerebrovascular and coronary artery disease.¹ No data suggest that treating asymptomatic PAD improves future limb pain or function.

Recommendations

The American College of Cardiology/American Heart Association 2005 Guidelines for the Management of Patients With Peripheral Arterial Disease² make the following recommendations for patients with asymptomatic lower extremity PAD:

• Identify patients with asymptomatic lower extremity PAD by examination or by measuring the ankle-brachial index so therapeutic interventions known to reduce the risk of myocardial infarction, stroke, and death can be offered (level of evidence [LOE]: B).
• Address smoking cessation, lipid lowering, and diabetes and hypertension treatment according to national guidelines (LOE: B).
• Consider antiplatelet therapy to reduce the risk of adverse cardiovascular ischemic events (LOE: C).

The United States Preventive Services Task Force recommends against routine screening for PAD (D recommendation).⁷

Acknowledgements

Special thanks to Felipe Navarro, MD.

Evidence summary

Significant renal artery stenosis (RAS) is defined anatomically as >50% stenosis of the lumen by renal angiography; stenosis is considered hemodynamically significant (potentially causing renovascular hypertension) if it exceeds 75%.¹

The prevalence of renovascular hypertension among the general hypertensive population varies from 1% to 5%. The prevalence increases to 20% to 40% in the presence of certain clinical criteria:

• hypertension in patients <30 years
• worsening or sudden onset of hypertension in patients >50 years
• hypertension refractory to multiple medications
• malignant hypertension
• worsening renal function after starting an angiotensin-converting enzyme inhibitor (ACE-I).^2,3 (Worsening renal function is defined as >30% decline in estimated glomerular filtration rate or >30% increase in serum creatinine during the first 2 months of ACE-I therapy.¹)

Refractory hypertension associated with generalized atherosclerosis is the most predictive risk factor for RAS.

MRA is usually best, but don’t overlook ultrasound

Among the primary diagnostic tests for RAS (see TABLE W1 on page 216a), MRA is the most consistently accurate⁴ and least operator dependent—which makes it the best choice in most situations. One rare but serious concern with MRA is that contrast agents may cause nephrogenic systemic fibrosis (NSF), a debilitating and sometimes fatal diffuse disease affecting the skin, muscle, and internal organs. In 2006, 25 cases of NSF after exposure to gadolinium-based contrast agents were reported, prompting an FDA warning.⁵

Kidney duplex Doppler ultrasound can rival MRA and CTA in accuracy, but is highly operator dependent. If access to highly skilled, experienced technicians is available, this safe and less expensive option can be considered, especially for patients with chronic kidney disease.

TABLE W1
Diagnostic tests for renovascular hypertension

Recommendations

The American College of Radiology recommends stratifying patients into 3 groups:

• high index of suspicion with normal renal function
• high index of suspicion with diminished renal function
• low index of suspicion.

Recommendations include:

• MRA or CTA for high suspicion with normal renal function
• MRA or ultrasonography for high suspicion with impaired renal function
• All methods equally inappropriate if suspicion is low.³

Cost-effectiveness was not evaluated in the meta-analysis used to derive the guidelines.

The National Kidney Foundation recommends MRA and CTA as accurate, noninvasive, and consistent means of diagnosing RAS. The foundation also recommends duplex ultrasonography as a less invasive and less expensive alternative when local expertise is available. The guidelines include a moderately predictive rule for identifying patients who should be screened for renovascular hypertension—that is, patients with intermediate or high pretest probability (www.kidney.org/professionals/kdoqi/guidelines_bp/guide_4.htm).¹

The American College of Cardiology and the American Heart Association list advantages and disadvantages of each diagnostic method and recommend choosing the one that is best suited to the patient.⁶

Acknowledgements

The opinions and assertions contained herein are the private view of the authors and are not to be construed as official, or as reflecting the view of the US Air Force Medical service or the US Air Force at large.

Evidence summary

Significant renal artery stenosis (RAS) is defined anatomically as >50% stenosis of the lumen by renal angiography; stenosis is considered hemodynamically significant (potentially causing renovascular hypertension) if it exceeds 75%.¹

The prevalence of renovascular hypertension among the general hypertensive population varies from 1% to 5%. The prevalence increases to 20% to 40% in the presence of certain clinical criteria:

• hypertension in patients <30 years
• worsening or sudden onset of hypertension in patients >50 years
• hypertension refractory to multiple medications
• malignant hypertension
• worsening renal function after starting an angiotensin-converting enzyme inhibitor (ACE-I).^2,3 (Worsening renal function is defined as >30% decline in estimated glomerular filtration rate or >30% increase in serum creatinine during the first 2 months of ACE-I therapy.¹)

Refractory hypertension associated with generalized atherosclerosis is the most predictive risk factor for RAS.

MRA is usually best, but don’t overlook ultrasound

Among the primary diagnostic tests for RAS (see TABLE W1 on page 216a), MRA is the most consistently accurate⁴ and least operator dependent—which makes it the best choice in most situations. One rare but serious concern with MRA is that contrast agents may cause nephrogenic systemic fibrosis (NSF), a debilitating and sometimes fatal diffuse disease affecting the skin, muscle, and internal organs. In 2006, 25 cases of NSF after exposure to gadolinium-based contrast agents were reported, prompting an FDA warning.⁵

Kidney duplex Doppler ultrasound can rival MRA and CTA in accuracy, but is highly operator dependent. If access to highly skilled, experienced technicians is available, this safe and less expensive option can be considered, especially for patients with chronic kidney disease.

TABLE W1
Diagnostic tests for renovascular hypertension

Recommendations

The American College of Radiology recommends stratifying patients into 3 groups:

• high index of suspicion with normal renal function
• high index of suspicion with diminished renal function
• low index of suspicion.

Recommendations include:

• MRA or CTA for high suspicion with normal renal function
• MRA or ultrasonography for high suspicion with impaired renal function
• All methods equally inappropriate if suspicion is low.³

Cost-effectiveness was not evaluated in the meta-analysis used to derive the guidelines.

The National Kidney Foundation recommends MRA and CTA as accurate, noninvasive, and consistent means of diagnosing RAS. The foundation also recommends duplex ultrasonography as a less invasive and less expensive alternative when local expertise is available. The guidelines include a moderately predictive rule for identifying patients who should be screened for renovascular hypertension—that is, patients with intermediate or high pretest probability (www.kidney.org/professionals/kdoqi/guidelines_bp/guide_4.htm).¹

The American College of Cardiology and the American Heart Association list advantages and disadvantages of each diagnostic method and recommend choosing the one that is best suited to the patient.⁶

Acknowledgements

The opinions and assertions contained herein are the private view of the authors and are not to be construed as official, or as reflecting the view of the US Air Force Medical service or the US Air Force at large.

Evidence summary

Significant renal artery stenosis (RAS) is defined anatomically as >50% stenosis of the lumen by renal angiography; stenosis is considered hemodynamically significant (potentially causing renovascular hypertension) if it exceeds 75%.¹

The prevalence of renovascular hypertension among the general hypertensive population varies from 1% to 5%. The prevalence increases to 20% to 40% in the presence of certain clinical criteria:

• hypertension in patients <30 years
• worsening or sudden onset of hypertension in patients >50 years
• hypertension refractory to multiple medications
• malignant hypertension
• worsening renal function after starting an angiotensin-converting enzyme inhibitor (ACE-I).^2,3 (Worsening renal function is defined as >30% decline in estimated glomerular filtration rate or >30% increase in serum creatinine during the first 2 months of ACE-I therapy.¹)

Refractory hypertension associated with generalized atherosclerosis is the most predictive risk factor for RAS.

MRA is usually best, but don’t overlook ultrasound

Among the primary diagnostic tests for RAS (see TABLE W1 on page 216a), MRA is the most consistently accurate⁴ and least operator dependent—which makes it the best choice in most situations. One rare but serious concern with MRA is that contrast agents may cause nephrogenic systemic fibrosis (NSF), a debilitating and sometimes fatal diffuse disease affecting the skin, muscle, and internal organs. In 2006, 25 cases of NSF after exposure to gadolinium-based contrast agents were reported, prompting an FDA warning.⁵

Kidney duplex Doppler ultrasound can rival MRA and CTA in accuracy, but is highly operator dependent. If access to highly skilled, experienced technicians is available, this safe and less expensive option can be considered, especially for patients with chronic kidney disease.

TABLE W1
Diagnostic tests for renovascular hypertension

Recommendations

The American College of Radiology recommends stratifying patients into 3 groups:

• high index of suspicion with normal renal function
• high index of suspicion with diminished renal function
• low index of suspicion.

Recommendations include:

• MRA or CTA for high suspicion with normal renal function
• MRA or ultrasonography for high suspicion with impaired renal function
• All methods equally inappropriate if suspicion is low.³

Cost-effectiveness was not evaluated in the meta-analysis used to derive the guidelines.

The National Kidney Foundation recommends MRA and CTA as accurate, noninvasive, and consistent means of diagnosing RAS. The foundation also recommends duplex ultrasonography as a less invasive and less expensive alternative when local expertise is available. The guidelines include a moderately predictive rule for identifying patients who should be screened for renovascular hypertension—that is, patients with intermediate or high pretest probability (www.kidney.org/professionals/kdoqi/guidelines_bp/guide_4.htm).¹

The American College of Cardiology and the American Heart Association list advantages and disadvantages of each diagnostic method and recommend choosing the one that is best suited to the patient.⁶

Acknowledgements

The opinions and assertions contained herein are the private view of the authors and are not to be construed as official, or as reflecting the view of the US Air Force Medical service or the US Air Force at large.

Evidence summary

Parent education reduces clinic visits for asthmatic children

A 2001 trial randomized 81 families with a smoking parent and an asthmatic child between 3 and 12 years of age to 3 sessions of behavioral and educational counseling or usual care at an outpatient asthma clinic.¹ Parental education included information on second-hand smoke, basic asthma education, and feedback about urine cotinine levels (a marker of nicotine absorption). Behavioral counseling focused on reducing second-hand smoke exposure by caregivers.

The education group had a significantly reduced risk of 2 or more asthma-related clinic visits in the following 12 months compared with usual care (odds ratio=0.32; P=.03; number needed to treat=5). No significant decrease was noted in mean urine cotinine levels between groups (adjusted mean difference=-0.38 ng/mg favoring education; P=.26).

A similar trial that measured changes in urine cotinine randomized 91 families with a smoking parent and an asthmatic child into 3 groups:²

• A control group received usual care (regular office visits at an asthma clinic and medication management)
• A monitoring group used a parental smoking diary and a children’s asthma symptom diary
• A counseling group received 5 counseling sessions and also kept diaries. An environmental monitor in the home was used to assess exposure to secondhand smoke.

In the counseling group, 21.4% of patients (6 of 28) maintained 0% exposure throughout the 30-month trial period compared with 3.6% and 3.8% in the monitoring and control groups, respectively (P<.05 for comparison of counseling group to monitoring and control).

Banning indoor smoking sharply cuts nicotine exposure

No data are available on education about second-hand smoke in families with nonasthmatic children. However, strong evidence suggests that smoking outside the house reduces exposure generally.

A 2003 cross-sectional survey of 164 households in the United Kingdom with at least 1 smoking parent and 1 bottle-fed infant looked for a correlation between strategies to reduce second-hand smoke and urine cotinine-to-creatinine ratios in the infants.³ Parents were classified into 3 groups according to whether they maintained a strict ban on smoking in the home, a less strict ban (smoking at home but not near the infant), or no ban.

The mean infant urinary cotinine-to-creatinine ratio was 2.43 in the no-ban group and 2.61 in the less-strict ban group (difference not significant). The combined mean for these 2 groups—2.58—was significantly higher than the mean of 1.26 in the strictest group (P<.001).

A later study recruited a convenience sample of 49 interested families with a smoking mother and a nonbreastfeeding infant between 2 and 12 months of age.⁴ Families were classified by smoking history into one of 3 groups: nonsmoking households, smoking households where efforts were made to limit smoke exposure, and smoking households where no efforts were made to limit exposure. Urine samples were obtained 3 times over 1 week. Urine cotinine levels in infants averaged 0.33 ng/mL in nonsmoking households, 2.47 ng/mL in smoking households with limited exposure, and 15.47 ng/mL in smoking households with unlimited exposure (P<.001 for all comparisons).

A case-control study that recruited families with asthmatic and nonasthmatic children assessed the effectiveness of parental behaviors to reduce second-hand smoke in 182 households with 1 smoking parent and a child between 6 and 12 years of age.⁵ Researchers measured room air nicotine and salivary cotinine concentrations.

The nicotine levels on children’s belts and in their bedrooms and the family room were approximately 3 log units lower in houses with strict smoking bans compared with households with any degree of indoor smoking (P<.0001). Similarly, salivary cotinine levels were approximately 4 log units lower in children of households with indoor smoking bans (P<.0001).

Recommendations

The United States Preventive Services Task Force (USPSTF) strongly recommends that physicians help all smoking adults to quit.⁶ The American Academy of Family Physicians endorses the USPSTF position and further advises that smoking parents be counseled about the health effects of environmental tobacco smoke on their children.⁷

The American Academy of Pediatrics⁸ and the Veterans Administration⁹ recommend urging parents to stop smoking to prevent serious health implications for their children; they further encourage pediatric clinicians to offer parents advice on quitting in order to limit children’s exposure to second-hand smoke.

Evidence summary

Parent education reduces clinic visits for asthmatic children

A 2001 trial randomized 81 families with a smoking parent and an asthmatic child between 3 and 12 years of age to 3 sessions of behavioral and educational counseling or usual care at an outpatient asthma clinic.¹ Parental education included information on second-hand smoke, basic asthma education, and feedback about urine cotinine levels (a marker of nicotine absorption). Behavioral counseling focused on reducing second-hand smoke exposure by caregivers.

The education group had a significantly reduced risk of 2 or more asthma-related clinic visits in the following 12 months compared with usual care (odds ratio=0.32; P=.03; number needed to treat=5). No significant decrease was noted in mean urine cotinine levels between groups (adjusted mean difference=-0.38 ng/mg favoring education; P=.26).

A similar trial that measured changes in urine cotinine randomized 91 families with a smoking parent and an asthmatic child into 3 groups:²

• A control group received usual care (regular office visits at an asthma clinic and medication management)
• A monitoring group used a parental smoking diary and a children’s asthma symptom diary
• A counseling group received 5 counseling sessions and also kept diaries. An environmental monitor in the home was used to assess exposure to secondhand smoke.

In the counseling group, 21.4% of patients (6 of 28) maintained 0% exposure throughout the 30-month trial period compared with 3.6% and 3.8% in the monitoring and control groups, respectively (P<.05 for comparison of counseling group to monitoring and control).

Banning indoor smoking sharply cuts nicotine exposure

No data are available on education about second-hand smoke in families with nonasthmatic children. However, strong evidence suggests that smoking outside the house reduces exposure generally.

A 2003 cross-sectional survey of 164 households in the United Kingdom with at least 1 smoking parent and 1 bottle-fed infant looked for a correlation between strategies to reduce second-hand smoke and urine cotinine-to-creatinine ratios in the infants.³ Parents were classified into 3 groups according to whether they maintained a strict ban on smoking in the home, a less strict ban (smoking at home but not near the infant), or no ban.

The mean infant urinary cotinine-to-creatinine ratio was 2.43 in the no-ban group and 2.61 in the less-strict ban group (difference not significant). The combined mean for these 2 groups—2.58—was significantly higher than the mean of 1.26 in the strictest group (P<.001).

A later study recruited a convenience sample of 49 interested families with a smoking mother and a nonbreastfeeding infant between 2 and 12 months of age.⁴ Families were classified by smoking history into one of 3 groups: nonsmoking households, smoking households where efforts were made to limit smoke exposure, and smoking households where no efforts were made to limit exposure. Urine samples were obtained 3 times over 1 week. Urine cotinine levels in infants averaged 0.33 ng/mL in nonsmoking households, 2.47 ng/mL in smoking households with limited exposure, and 15.47 ng/mL in smoking households with unlimited exposure (P<.001 for all comparisons).

A case-control study that recruited families with asthmatic and nonasthmatic children assessed the effectiveness of parental behaviors to reduce second-hand smoke in 182 households with 1 smoking parent and a child between 6 and 12 years of age.⁵ Researchers measured room air nicotine and salivary cotinine concentrations.

The nicotine levels on children’s belts and in their bedrooms and the family room were approximately 3 log units lower in houses with strict smoking bans compared with households with any degree of indoor smoking (P<.0001). Similarly, salivary cotinine levels were approximately 4 log units lower in children of households with indoor smoking bans (P<.0001).

Recommendations

The United States Preventive Services Task Force (USPSTF) strongly recommends that physicians help all smoking adults to quit.⁶ The American Academy of Family Physicians endorses the USPSTF position and further advises that smoking parents be counseled about the health effects of environmental tobacco smoke on their children.⁷

The American Academy of Pediatrics⁸ and the Veterans Administration⁹ recommend urging parents to stop smoking to prevent serious health implications for their children; they further encourage pediatric clinicians to offer parents advice on quitting in order to limit children’s exposure to second-hand smoke.

Evidence summary

Parent education reduces clinic visits for asthmatic children

A 2001 trial randomized 81 families with a smoking parent and an asthmatic child between 3 and 12 years of age to 3 sessions of behavioral and educational counseling or usual care at an outpatient asthma clinic.¹ Parental education included information on second-hand smoke, basic asthma education, and feedback about urine cotinine levels (a marker of nicotine absorption). Behavioral counseling focused on reducing second-hand smoke exposure by caregivers.

The education group had a significantly reduced risk of 2 or more asthma-related clinic visits in the following 12 months compared with usual care (odds ratio=0.32; P=.03; number needed to treat=5). No significant decrease was noted in mean urine cotinine levels between groups (adjusted mean difference=-0.38 ng/mg favoring education; P=.26).

A similar trial that measured changes in urine cotinine randomized 91 families with a smoking parent and an asthmatic child into 3 groups:²

• A control group received usual care (regular office visits at an asthma clinic and medication management)
• A monitoring group used a parental smoking diary and a children’s asthma symptom diary
• A counseling group received 5 counseling sessions and also kept diaries. An environmental monitor in the home was used to assess exposure to secondhand smoke.

In the counseling group, 21.4% of patients (6 of 28) maintained 0% exposure throughout the 30-month trial period compared with 3.6% and 3.8% in the monitoring and control groups, respectively (P<.05 for comparison of counseling group to monitoring and control).

Banning indoor smoking sharply cuts nicotine exposure

No data are available on education about second-hand smoke in families with nonasthmatic children. However, strong evidence suggests that smoking outside the house reduces exposure generally.

A 2003 cross-sectional survey of 164 households in the United Kingdom with at least 1 smoking parent and 1 bottle-fed infant looked for a correlation between strategies to reduce second-hand smoke and urine cotinine-to-creatinine ratios in the infants.³ Parents were classified into 3 groups according to whether they maintained a strict ban on smoking in the home, a less strict ban (smoking at home but not near the infant), or no ban.

The mean infant urinary cotinine-to-creatinine ratio was 2.43 in the no-ban group and 2.61 in the less-strict ban group (difference not significant). The combined mean for these 2 groups—2.58—was significantly higher than the mean of 1.26 in the strictest group (P<.001).

A later study recruited a convenience sample of 49 interested families with a smoking mother and a nonbreastfeeding infant between 2 and 12 months of age.⁴ Families were classified by smoking history into one of 3 groups: nonsmoking households, smoking households where efforts were made to limit smoke exposure, and smoking households where no efforts were made to limit exposure. Urine samples were obtained 3 times over 1 week. Urine cotinine levels in infants averaged 0.33 ng/mL in nonsmoking households, 2.47 ng/mL in smoking households with limited exposure, and 15.47 ng/mL in smoking households with unlimited exposure (P<.001 for all comparisons).

A case-control study that recruited families with asthmatic and nonasthmatic children assessed the effectiveness of parental behaviors to reduce second-hand smoke in 182 households with 1 smoking parent and a child between 6 and 12 years of age.⁵ Researchers measured room air nicotine and salivary cotinine concentrations.

The nicotine levels on children’s belts and in their bedrooms and the family room were approximately 3 log units lower in houses with strict smoking bans compared with households with any degree of indoor smoking (P<.0001). Similarly, salivary cotinine levels were approximately 4 log units lower in children of households with indoor smoking bans (P<.0001).

Recommendations

The United States Preventive Services Task Force (USPSTF) strongly recommends that physicians help all smoking adults to quit.⁶ The American Academy of Family Physicians endorses the USPSTF position and further advises that smoking parents be counseled about the health effects of environmental tobacco smoke on their children.⁷

The American Academy of Pediatrics⁸ and the Veterans Administration⁹ recommend urging parents to stop smoking to prevent serious health implications for their children; they further encourage pediatric clinicians to offer parents advice on quitting in order to limit children’s exposure to second-hand smoke.

Evidence summary

More than 4 million babies are delivered each year in the United States, and amniotomy is one of the most common obstetric procedures. It’s typically employed to accelerate labor and was originally thought to decrease cesarean section rates. However, the extent to which amniotomy alone shortens labor varies widely from study to study, and no clear consensus exists concerning the potential harms or unintended effects of this practice.^1-7

A 2007 Cochrane review of 14 trials (4893 women) investigated the risks and benefits of routine amniotomy vs intention to leave membranes intact. All trials included only very-low-risk women in spontaneous labor at term with a singleton fetus in vertex presentation.⁸ Because of the strict inclusion criteria, up to 80% of women giving birth in participating centers were excluded. Plus, many of the women in control groups underwent amniotomy at some stage of labor, because most trials allowed clinicians to perform amniotomy if clinically indicated.^1-7

Little effect on first-stage labor

Five of the trials (1127 women) reported length of first-stage labor. No statistically significant difference was found between amniotomy and control groups (weighted mean difference [WMD]=–20.43 minutes; 95% confidence interval [CI], –95.93 to 55.06). Furthermore, subgroup analysis found no statistically significant reduction in length of first-stage labor for nulliparous (WMD=–57.93 min; 95% CI, –152.66 to 36.80) or multiparous women (WMD= 23.10 min; 95% CI, –50.89 to 97.09).

Seven trials (1237 women) in the Cochrane review reported length of second-stage labor. No statistically significant difference was noted between amniotomy and control groups (WMD=–2.38 minutes; 95% CI, –5.27 to 0.50). However, subgroup analysis of primiparous women showed a statistically significant reduction in length of second-stage labor in the amniotomy group (WMD=–6.59 minutes; 95% CI, –12.34 to –0.84).⁸

More cesareans

Nine trials (4370 women) included in the Cochrane review reported cesarean section rates. Women in the amniotomy group had an increased risk of cesarean delivery compared with the control group, but the difference did not reach statistical significance (relative risk=1.26; 95% CI, 0.98-1.62).⁸ Because cesarean section was surprisingly rare in this low-risk patient population compared with the national average, the studies were not powered to show statistical significance in this secondary outcome.

What about neonatal outcomes?

No significant differences between the amniotomy and intact groups were noted in less uniformly reported maternal outcomes, including need for oxytocin to augment labor, rate of infection, serious morbidity, or death.⁸ Likewise, differences in neonatal outcomes—such as sepsis, respiratory failure, admission to the special care unit, and death—weren’t statistically significant. Notably, however, these secondary outcomes occurred too rarely to measure the effect precisely.

Because of the relatively small sample sizes and rarity of complications, the studies have limited ability to address the effect of routine amniotomy on maternal and neonatal morbidity in the general population. Larger studies, with a wider variety of patients, would improve clarity.

Recommendations

The American College of Obstetricians and Gynecologists (ACOG) hasn’t issued a statement on the use of routine amniotomy in normal labor. With regard to labor dystocia, ACOG states that “amniotomy may enhance progress in the active phase and negate the need for oxytocin augmentation, but it may increase the risk of chorioamnionitis.”⁹

And the ACOG bulletin on induction of labor reports that “the potential risks associated with amniotomy include prolapse of the umbilical cord, chorioamnionitis, significant umbilical cord compression, and rupture of vasa previa.”¹⁰

Evidence summary

More than 4 million babies are delivered each year in the United States, and amniotomy is one of the most common obstetric procedures. It’s typically employed to accelerate labor and was originally thought to decrease cesarean section rates. However, the extent to which amniotomy alone shortens labor varies widely from study to study, and no clear consensus exists concerning the potential harms or unintended effects of this practice.^1-7

A 2007 Cochrane review of 14 trials (4893 women) investigated the risks and benefits of routine amniotomy vs intention to leave membranes intact. All trials included only very-low-risk women in spontaneous labor at term with a singleton fetus in vertex presentation.⁸ Because of the strict inclusion criteria, up to 80% of women giving birth in participating centers were excluded. Plus, many of the women in control groups underwent amniotomy at some stage of labor, because most trials allowed clinicians to perform amniotomy if clinically indicated.^1-7

Little effect on first-stage labor

Five of the trials (1127 women) reported length of first-stage labor. No statistically significant difference was found between amniotomy and control groups (weighted mean difference [WMD]=–20.43 minutes; 95% confidence interval [CI], –95.93 to 55.06). Furthermore, subgroup analysis found no statistically significant reduction in length of first-stage labor for nulliparous (WMD=–57.93 min; 95% CI, –152.66 to 36.80) or multiparous women (WMD= 23.10 min; 95% CI, –50.89 to 97.09).

Seven trials (1237 women) in the Cochrane review reported length of second-stage labor. No statistically significant difference was noted between amniotomy and control groups (WMD=–2.38 minutes; 95% CI, –5.27 to 0.50). However, subgroup analysis of primiparous women showed a statistically significant reduction in length of second-stage labor in the amniotomy group (WMD=–6.59 minutes; 95% CI, –12.34 to –0.84).⁸

More cesareans

Nine trials (4370 women) included in the Cochrane review reported cesarean section rates. Women in the amniotomy group had an increased risk of cesarean delivery compared with the control group, but the difference did not reach statistical significance (relative risk=1.26; 95% CI, 0.98-1.62).⁸ Because cesarean section was surprisingly rare in this low-risk patient population compared with the national average, the studies were not powered to show statistical significance in this secondary outcome.

What about neonatal outcomes?

No significant differences between the amniotomy and intact groups were noted in less uniformly reported maternal outcomes, including need for oxytocin to augment labor, rate of infection, serious morbidity, or death.⁸ Likewise, differences in neonatal outcomes—such as sepsis, respiratory failure, admission to the special care unit, and death—weren’t statistically significant. Notably, however, these secondary outcomes occurred too rarely to measure the effect precisely.

Because of the relatively small sample sizes and rarity of complications, the studies have limited ability to address the effect of routine amniotomy on maternal and neonatal morbidity in the general population. Larger studies, with a wider variety of patients, would improve clarity.

Recommendations

The American College of Obstetricians and Gynecologists (ACOG) hasn’t issued a statement on the use of routine amniotomy in normal labor. With regard to labor dystocia, ACOG states that “amniotomy may enhance progress in the active phase and negate the need for oxytocin augmentation, but it may increase the risk of chorioamnionitis.”⁹

And the ACOG bulletin on induction of labor reports that “the potential risks associated with amniotomy include prolapse of the umbilical cord, chorioamnionitis, significant umbilical cord compression, and rupture of vasa previa.”¹⁰

Evidence summary

More than 4 million babies are delivered each year in the United States, and amniotomy is one of the most common obstetric procedures. It’s typically employed to accelerate labor and was originally thought to decrease cesarean section rates. However, the extent to which amniotomy alone shortens labor varies widely from study to study, and no clear consensus exists concerning the potential harms or unintended effects of this practice.^1-7

A 2007 Cochrane review of 14 trials (4893 women) investigated the risks and benefits of routine amniotomy vs intention to leave membranes intact. All trials included only very-low-risk women in spontaneous labor at term with a singleton fetus in vertex presentation.⁸ Because of the strict inclusion criteria, up to 80% of women giving birth in participating centers were excluded. Plus, many of the women in control groups underwent amniotomy at some stage of labor, because most trials allowed clinicians to perform amniotomy if clinically indicated.^1-7

Little effect on first-stage labor

Five of the trials (1127 women) reported length of first-stage labor. No statistically significant difference was found between amniotomy and control groups (weighted mean difference [WMD]=–20.43 minutes; 95% confidence interval [CI], –95.93 to 55.06). Furthermore, subgroup analysis found no statistically significant reduction in length of first-stage labor for nulliparous (WMD=–57.93 min; 95% CI, –152.66 to 36.80) or multiparous women (WMD= 23.10 min; 95% CI, –50.89 to 97.09).

Seven trials (1237 women) in the Cochrane review reported length of second-stage labor. No statistically significant difference was noted between amniotomy and control groups (WMD=–2.38 minutes; 95% CI, –5.27 to 0.50). However, subgroup analysis of primiparous women showed a statistically significant reduction in length of second-stage labor in the amniotomy group (WMD=–6.59 minutes; 95% CI, –12.34 to –0.84).⁸

More cesareans

Nine trials (4370 women) included in the Cochrane review reported cesarean section rates. Women in the amniotomy group had an increased risk of cesarean delivery compared with the control group, but the difference did not reach statistical significance (relative risk=1.26; 95% CI, 0.98-1.62).⁸ Because cesarean section was surprisingly rare in this low-risk patient population compared with the national average, the studies were not powered to show statistical significance in this secondary outcome.

What about neonatal outcomes?

No significant differences between the amniotomy and intact groups were noted in less uniformly reported maternal outcomes, including need for oxytocin to augment labor, rate of infection, serious morbidity, or death.⁸ Likewise, differences in neonatal outcomes—such as sepsis, respiratory failure, admission to the special care unit, and death—weren’t statistically significant. Notably, however, these secondary outcomes occurred too rarely to measure the effect precisely.

Because of the relatively small sample sizes and rarity of complications, the studies have limited ability to address the effect of routine amniotomy on maternal and neonatal morbidity in the general population. Larger studies, with a wider variety of patients, would improve clarity.

Recommendations

The American College of Obstetricians and Gynecologists (ACOG) hasn’t issued a statement on the use of routine amniotomy in normal labor. With regard to labor dystocia, ACOG states that “amniotomy may enhance progress in the active phase and negate the need for oxytocin augmentation, but it may increase the risk of chorioamnionitis.”⁹

And the ACOG bulletin on induction of labor reports that “the potential risks associated with amniotomy include prolapse of the umbilical cord, chorioamnionitis, significant umbilical cord compression, and rupture of vasa previa.”¹⁰

Evidence summary

NSAIDs: Benefits with limits

A Cochrane systematic review evaluating the efficacy of topical and oral NSAIDs to treat lateral epicondylitis found that topically applied diclofenac gel was more effective than placebo, as measured by overall patient satisfaction (relative risk [RR]=0.39; 95% confidence interval [CI], 0.23-0.66; number needed to treat [NNT]=3).¹ Topical diclofenac or benzydamine gel had a significant effect on the patient’s perception of pain compared with placebo, but not beyond 4 weeks of therapy (weighted mean difference [WMD] on a 10-point scale=-1.88 points; 95% CI, –2.54 to –1.21). However, no difference was noted in functional outcomes, measured by grip or wrist extension strength.

Patients who used topical NSAIDs reported more adverse events than those using placebo, including minor skin irritation (RR=2.26; 95% CI, 1.04-4.94).¹

Oral NSAIDs relieve pain, but not as much as steroids

In the same review, oral diclofenac reduced pain scores at 4 weeks compared with placebo (WMD on 100-point scale=-13.9 points; 95% CI, -23.21 to -4.59).¹ Adequate studies are lacking to show a benefit of oral NSAIDs past 4 weeks. Significantly more complaints of abdominal pain occurred with oral diclofenac than placebo (RR=3.17; 95% CI, 1.35-7.41; number needed to harm [NNH]=5).¹

One study that directly compared diflunisal with naproxen for lateral epicondylitis found no difference between the therapies in patients’ subjective perception of pain on a 5-point scale (RR=0.24; 95% CI, 0.03-1.89).¹ When oral NSAIDs were compared with steroid injections, patients receiving an injection reported more improvement in pain than patients who took an oral NSAID (RR=3.06; 95% CI, 1.55-6.06; NNT=4).¹

Corticosteroids more effective in short term than long term

A subanalysis of 4 studies in another systematic review found corticosteroid injections to be superior to other conservative treatments such as elbow supports, oral NSAIDs, and physiotherapy at 2 to 6 weeks (RR=0.50; 95% CI, 0.36-0.70).² The positive effects weren’t maintained at 6 weeks.

In a randomized study with 3 treatment arms, 185 patients were treated with a corticosteroid injection, physiotherapy, or a wait-and-see approach (ergonomic advice, rest, and oral anti-inflammatory medication). Corticosteroid injections were significantly more effective for the patients’ main complaint at 6 weeks compared with wait-and-see (mean difference in improvement [MDI] on a 100-point scale=24; 95% CI, 14-35; NNT=2) or physiotherapy (MDI=20; 95% CI, 10-31; NNT=2).³ By contrast, at 26 and 52 weeks’ follow-up, physiotherapy was more effective than steroid injections (MDI=15; 95% CI, 5-25) but statistically equivalent to a wait-and-see approach (MDI=7; 95% CI, –4 to 17).

Physiotherapy, exercise, acupuncture bring short-term relief

In a separate RCT, physiotherapy and exercise were significantly better than a wait-and-see approach at 6 weeks for pain-free grip force, rating of pain severity, and global improvement (RR=0.5; 99% CI, 0.2-0.8; NNT=3), but by 52 weeks the outcomes were statistically equal.⁴

An individual RCT, cited in a Cochrane review, showed acupuncture had a very short-term benefit for pain relief compared with placebo (WMD=18.8 hours; 95% CI, 10.1-27.5).⁵ Another individual RCT, which was not included in the meta-analysis because of methodologic problems in the other studies, found that a short course of 10 acupuncture treatments resulted in an excellent or good outcome (as reported by participants) compared with placebo (RR=0.09; 95% CI, 0.01-0.64; NNT=4).⁵ No benefit was noted after 3 or 12 months.

Physiotherapy techniques, orthotics are hard to evaluate

Systematic reviews of specific physiotherapy or orthotic (bracing) treatments are hampered by the large number of treatment options available and the heterogeneity of the available studies, which prevent statistically useful evaluation.^6,7

Shock wave, ultrasound, massage offer little or no benefit

In a meta-analysis of 3 trials, shock wave therapy provided no significant benefit at 4 to 6 weeks compared with placebo (WMD on a 100-point scale=-9.42; 95% CI, -20.70 to 1.86).⁸ Pooling 2 studies in a different systematic review showed weak evidence that ultrasound reduced pain at 13 weeks compared with placebo (standardized mean difference=–0.98; 95% CI, –1.64 to –0.33).⁶ Another Cochrane review found no added benefit in function from combining deep transverse friction massage with ultrasound or a placebo ointment (RR=3.3; 95% CI, 0.4-24.3).⁹

Recommendations

The Work Loss Data Institute recommends ice, rest, ergonomic modifications, and short-term topical or oral NSAIDs. Progressive physical or occupational therapy may follow if no improvement is seen in 2 weeks.

Splinting, acupuncture, and corticosteroid administration by injection or iontophoresis may reduce pain for as long as 2 to 6 weeks. If these conservative measures fail, surgical treatment is recommended as a last resort.¹⁰

Evidence summary

NSAIDs: Benefits with limits

A Cochrane systematic review evaluating the efficacy of topical and oral NSAIDs to treat lateral epicondylitis found that topically applied diclofenac gel was more effective than placebo, as measured by overall patient satisfaction (relative risk [RR]=0.39; 95% confidence interval [CI], 0.23-0.66; number needed to treat [NNT]=3).¹ Topical diclofenac or benzydamine gel had a significant effect on the patient’s perception of pain compared with placebo, but not beyond 4 weeks of therapy (weighted mean difference [WMD] on a 10-point scale=-1.88 points; 95% CI, –2.54 to –1.21). However, no difference was noted in functional outcomes, measured by grip or wrist extension strength.

Patients who used topical NSAIDs reported more adverse events than those using placebo, including minor skin irritation (RR=2.26; 95% CI, 1.04-4.94).¹

Oral NSAIDs relieve pain, but not as much as steroids

In the same review, oral diclofenac reduced pain scores at 4 weeks compared with placebo (WMD on 100-point scale=-13.9 points; 95% CI, -23.21 to -4.59).¹ Adequate studies are lacking to show a benefit of oral NSAIDs past 4 weeks. Significantly more complaints of abdominal pain occurred with oral diclofenac than placebo (RR=3.17; 95% CI, 1.35-7.41; number needed to harm [NNH]=5).¹

One study that directly compared diflunisal with naproxen for lateral epicondylitis found no difference between the therapies in patients’ subjective perception of pain on a 5-point scale (RR=0.24; 95% CI, 0.03-1.89).¹ When oral NSAIDs were compared with steroid injections, patients receiving an injection reported more improvement in pain than patients who took an oral NSAID (RR=3.06; 95% CI, 1.55-6.06; NNT=4).¹

Corticosteroids more effective in short term than long term

A subanalysis of 4 studies in another systematic review found corticosteroid injections to be superior to other conservative treatments such as elbow supports, oral NSAIDs, and physiotherapy at 2 to 6 weeks (RR=0.50; 95% CI, 0.36-0.70).² The positive effects weren’t maintained at 6 weeks.

In a randomized study with 3 treatment arms, 185 patients were treated with a corticosteroid injection, physiotherapy, or a wait-and-see approach (ergonomic advice, rest, and oral anti-inflammatory medication). Corticosteroid injections were significantly more effective for the patients’ main complaint at 6 weeks compared with wait-and-see (mean difference in improvement [MDI] on a 100-point scale=24; 95% CI, 14-35; NNT=2) or physiotherapy (MDI=20; 95% CI, 10-31; NNT=2).³ By contrast, at 26 and 52 weeks’ follow-up, physiotherapy was more effective than steroid injections (MDI=15; 95% CI, 5-25) but statistically equivalent to a wait-and-see approach (MDI=7; 95% CI, –4 to 17).

Physiotherapy, exercise, acupuncture bring short-term relief

In a separate RCT, physiotherapy and exercise were significantly better than a wait-and-see approach at 6 weeks for pain-free grip force, rating of pain severity, and global improvement (RR=0.5; 99% CI, 0.2-0.8; NNT=3), but by 52 weeks the outcomes were statistically equal.⁴

An individual RCT, cited in a Cochrane review, showed acupuncture had a very short-term benefit for pain relief compared with placebo (WMD=18.8 hours; 95% CI, 10.1-27.5).⁵ Another individual RCT, which was not included in the meta-analysis because of methodologic problems in the other studies, found that a short course of 10 acupuncture treatments resulted in an excellent or good outcome (as reported by participants) compared with placebo (RR=0.09; 95% CI, 0.01-0.64; NNT=4).⁵ No benefit was noted after 3 or 12 months.

Physiotherapy techniques, orthotics are hard to evaluate

Systematic reviews of specific physiotherapy or orthotic (bracing) treatments are hampered by the large number of treatment options available and the heterogeneity of the available studies, which prevent statistically useful evaluation.^6,7

Shock wave, ultrasound, massage offer little or no benefit

In a meta-analysis of 3 trials, shock wave therapy provided no significant benefit at 4 to 6 weeks compared with placebo (WMD on a 100-point scale=-9.42; 95% CI, -20.70 to 1.86).⁸ Pooling 2 studies in a different systematic review showed weak evidence that ultrasound reduced pain at 13 weeks compared with placebo (standardized mean difference=–0.98; 95% CI, –1.64 to –0.33).⁶ Another Cochrane review found no added benefit in function from combining deep transverse friction massage with ultrasound or a placebo ointment (RR=3.3; 95% CI, 0.4-24.3).⁹

Recommendations

The Work Loss Data Institute recommends ice, rest, ergonomic modifications, and short-term topical or oral NSAIDs. Progressive physical or occupational therapy may follow if no improvement is seen in 2 weeks.

Splinting, acupuncture, and corticosteroid administration by injection or iontophoresis may reduce pain for as long as 2 to 6 weeks. If these conservative measures fail, surgical treatment is recommended as a last resort.¹⁰

Evidence summary

NSAIDs: Benefits with limits

A Cochrane systematic review evaluating the efficacy of topical and oral NSAIDs to treat lateral epicondylitis found that topically applied diclofenac gel was more effective than placebo, as measured by overall patient satisfaction (relative risk [RR]=0.39; 95% confidence interval [CI], 0.23-0.66; number needed to treat [NNT]=3).¹ Topical diclofenac or benzydamine gel had a significant effect on the patient’s perception of pain compared with placebo, but not beyond 4 weeks of therapy (weighted mean difference [WMD] on a 10-point scale=-1.88 points; 95% CI, –2.54 to –1.21). However, no difference was noted in functional outcomes, measured by grip or wrist extension strength.

Patients who used topical NSAIDs reported more adverse events than those using placebo, including minor skin irritation (RR=2.26; 95% CI, 1.04-4.94).¹

Oral NSAIDs relieve pain, but not as much as steroids

In the same review, oral diclofenac reduced pain scores at 4 weeks compared with placebo (WMD on 100-point scale=-13.9 points; 95% CI, -23.21 to -4.59).¹ Adequate studies are lacking to show a benefit of oral NSAIDs past 4 weeks. Significantly more complaints of abdominal pain occurred with oral diclofenac than placebo (RR=3.17; 95% CI, 1.35-7.41; number needed to harm [NNH]=5).¹

One study that directly compared diflunisal with naproxen for lateral epicondylitis found no difference between the therapies in patients’ subjective perception of pain on a 5-point scale (RR=0.24; 95% CI, 0.03-1.89).¹ When oral NSAIDs were compared with steroid injections, patients receiving an injection reported more improvement in pain than patients who took an oral NSAID (RR=3.06; 95% CI, 1.55-6.06; NNT=4).¹

Corticosteroids more effective in short term than long term

A subanalysis of 4 studies in another systematic review found corticosteroid injections to be superior to other conservative treatments such as elbow supports, oral NSAIDs, and physiotherapy at 2 to 6 weeks (RR=0.50; 95% CI, 0.36-0.70).² The positive effects weren’t maintained at 6 weeks.

In a randomized study with 3 treatment arms, 185 patients were treated with a corticosteroid injection, physiotherapy, or a wait-and-see approach (ergonomic advice, rest, and oral anti-inflammatory medication). Corticosteroid injections were significantly more effective for the patients’ main complaint at 6 weeks compared with wait-and-see (mean difference in improvement [MDI] on a 100-point scale=24; 95% CI, 14-35; NNT=2) or physiotherapy (MDI=20; 95% CI, 10-31; NNT=2).³ By contrast, at 26 and 52 weeks’ follow-up, physiotherapy was more effective than steroid injections (MDI=15; 95% CI, 5-25) but statistically equivalent to a wait-and-see approach (MDI=7; 95% CI, –4 to 17).

Physiotherapy, exercise, acupuncture bring short-term relief

In a separate RCT, physiotherapy and exercise were significantly better than a wait-and-see approach at 6 weeks for pain-free grip force, rating of pain severity, and global improvement (RR=0.5; 99% CI, 0.2-0.8; NNT=3), but by 52 weeks the outcomes were statistically equal.⁴

An individual RCT, cited in a Cochrane review, showed acupuncture had a very short-term benefit for pain relief compared with placebo (WMD=18.8 hours; 95% CI, 10.1-27.5).⁵ Another individual RCT, which was not included in the meta-analysis because of methodologic problems in the other studies, found that a short course of 10 acupuncture treatments resulted in an excellent or good outcome (as reported by participants) compared with placebo (RR=0.09; 95% CI, 0.01-0.64; NNT=4).⁵ No benefit was noted after 3 or 12 months.

Physiotherapy techniques, orthotics are hard to evaluate

Systematic reviews of specific physiotherapy or orthotic (bracing) treatments are hampered by the large number of treatment options available and the heterogeneity of the available studies, which prevent statistically useful evaluation.^6,7

Shock wave, ultrasound, massage offer little or no benefit

In a meta-analysis of 3 trials, shock wave therapy provided no significant benefit at 4 to 6 weeks compared with placebo (WMD on a 100-point scale=-9.42; 95% CI, -20.70 to 1.86).⁸ Pooling 2 studies in a different systematic review showed weak evidence that ultrasound reduced pain at 13 weeks compared with placebo (standardized mean difference=–0.98; 95% CI, –1.64 to –0.33).⁶ Another Cochrane review found no added benefit in function from combining deep transverse friction massage with ultrasound or a placebo ointment (RR=3.3; 95% CI, 0.4-24.3).⁹

Recommendations

The Work Loss Data Institute recommends ice, rest, ergonomic modifications, and short-term topical or oral NSAIDs. Progressive physical or occupational therapy may follow if no improvement is seen in 2 weeks.

Splinting, acupuncture, and corticosteroid administration by injection or iontophoresis may reduce pain for as long as 2 to 6 weeks. If these conservative measures fail, surgical treatment is recommended as a last resort.¹⁰

Evidence summary

An estimated 1.7 million people in the United States are living with limb loss. The number is expected to increase because of ongoing military conflicts.¹ The incidence of PLP is 60% to 80% among amputees.¹

A multidisciplinary approach

A lack of comparative clinical trials of therapies for PLP has led health-care providers to adopt a multidisciplinary approach that combines evaluative management, desensitization, psychotherapy, and pharmacotherapy (FIGURE).

Evaluative management, based largely on expert opinion, includes assessing the fit of the prosthesis, treating referred pain, and assessing aggravating factors. Because residual limb pain can exacerbate PLP, adjusting a poorly fitting prosthesis or providing the patient with NSAIDs when there is evidence of stump inflammation may adequately control pain.^2,3 Anatomically distant pain syndromes, such as hip or lower back pain, can also aggravate PLP and should be managed to provide optimal pain relief.²

Desensitization, using TENS, has reduced PLP in multiple placebo-controlled trials and epidemiologic surveys.^2-5 TENS is an easy-to-use, low-cost, noninvasive, first-line therapy.⁵ Its long-term effectiveness in alleviating PLP remains unknown.² Some experts suggest that pain reductions after 1 year of treatment are comparable to placebo.² Other forms of desensitization (percussion and massage) are supported only by anecdotal reports.

Psychotherapy, including biofeedback, has been found in several case studies to effectively treat chronic PLP.^2,5 Psychotherapy can reportedly reveal the underlying mechanisms (muscle spasm, vascular insufficiency) and therefore direct therapeutic interventions by biofeedback or other focus techniques.²

FIGURE Management of phantom limb pain^1-10

Pharmacotherapy is best used as an adjunct to other treatments.² Although PLP is typically treated as neuropathic pain, only a few medications have been critically evaluated for treating it.⁶ Morphine (number needed to treat [NNT]=2.5; 95% confidence interval [CI], 1.9-3.4) and other opioids, including tramadol (NNT=3.9; 95% CI, 2.7-6.7 in neuropathic pain) help some patients.^6,7 Despite the proven benefit of tricyclic antidepressants (TCAs) in other neuropathic pain conditions, a recent RCT demonstrated no benefit of TCAs over placebo in PLP.⁸ Anticonvulsants, including gabapentin, have documented benefit in neuropathic pain modalities and are often used for PLP.⁶ However, their value in reducing PLP is still under investigation.⁶ One 2002 RCT showed benefit regarding an improvement of the visual analog scale by an average of 3 points (on a 10-point scale) after 6 weeks of gabapentin therapy.⁹ A similarly designed 2006 RCT of gabapentin, however did not identify significant pain reductions.¹⁰

Promising adjuvant therapies use mirroring techniques

Of the adjuvant treatments mentioned previously, only mirror box therapy has shown promise. This technique allows the amputee to perceive the missing limb by focusing on the reflection of the remaining limb during specific movements and activities. Theoretically, this perception allows reconfiguration of the amputee’s sensory cortex.

Virtual reality therapy employs similar techniques based on the idea that the brain can be deceived. Initial case studies are promising and have prompted further research.¹¹

Recommendations

The US Department of Veterans Affairs and Department of Defense recently issued clinical guidelines for rehabilitating lower-limb amputees that include a segment on pain management.¹² The guidelines stress the importance of an interdisciplinary team approach that addresses each pathology plaguing the amputee.

They recommend narcotics during the immediate postoperative period, followed by transition to a non-narcotic medical regimen during the rehabilitation process. The guidelines don’t support a single, specific pain control method over others; they recommend the following approaches to PLP:

• pharmacologic treatment, which may include antiseizure medications, tricyclic antidepressants, selective serotonin reuptake inhibitors, NSAIDs, dextromethorathane, or long-acting narcotics
• epidural analgesia, patient-controlled analgesia, or regional analgesia
• nonpharmacologic therapies, including TENS, desensitization, scar mobilization, relaxation, and biofeedback.

Acknowledgements

The opinions and assertions contained herein are the private views of the authors and not to be construed as official, or as reflecting the views of the US Air Force Medical Service or the US Air Force at large.

Evidence summary

An estimated 1.7 million people in the United States are living with limb loss. The number is expected to increase because of ongoing military conflicts.¹ The incidence of PLP is 60% to 80% among amputees.¹

A multidisciplinary approach

A lack of comparative clinical trials of therapies for PLP has led health-care providers to adopt a multidisciplinary approach that combines evaluative management, desensitization, psychotherapy, and pharmacotherapy (FIGURE).

Evaluative management, based largely on expert opinion, includes assessing the fit of the prosthesis, treating referred pain, and assessing aggravating factors. Because residual limb pain can exacerbate PLP, adjusting a poorly fitting prosthesis or providing the patient with NSAIDs when there is evidence of stump inflammation may adequately control pain.^2,3 Anatomically distant pain syndromes, such as hip or lower back pain, can also aggravate PLP and should be managed to provide optimal pain relief.²

Desensitization, using TENS, has reduced PLP in multiple placebo-controlled trials and epidemiologic surveys.^2-5 TENS is an easy-to-use, low-cost, noninvasive, first-line therapy.⁵ Its long-term effectiveness in alleviating PLP remains unknown.² Some experts suggest that pain reductions after 1 year of treatment are comparable to placebo.² Other forms of desensitization (percussion and massage) are supported only by anecdotal reports.

Psychotherapy, including biofeedback, has been found in several case studies to effectively treat chronic PLP.^2,5 Psychotherapy can reportedly reveal the underlying mechanisms (muscle spasm, vascular insufficiency) and therefore direct therapeutic interventions by biofeedback or other focus techniques.²

FIGURE Management of phantom limb pain^1-10

Pharmacotherapy is best used as an adjunct to other treatments.² Although PLP is typically treated as neuropathic pain, only a few medications have been critically evaluated for treating it.⁶ Morphine (number needed to treat [NNT]=2.5; 95% confidence interval [CI], 1.9-3.4) and other opioids, including tramadol (NNT=3.9; 95% CI, 2.7-6.7 in neuropathic pain) help some patients.^6,7 Despite the proven benefit of tricyclic antidepressants (TCAs) in other neuropathic pain conditions, a recent RCT demonstrated no benefit of TCAs over placebo in PLP.⁸ Anticonvulsants, including gabapentin, have documented benefit in neuropathic pain modalities and are often used for PLP.⁶ However, their value in reducing PLP is still under investigation.⁶ One 2002 RCT showed benefit regarding an improvement of the visual analog scale by an average of 3 points (on a 10-point scale) after 6 weeks of gabapentin therapy.⁹ A similarly designed 2006 RCT of gabapentin, however did not identify significant pain reductions.¹⁰

Promising adjuvant therapies use mirroring techniques

Of the adjuvant treatments mentioned previously, only mirror box therapy has shown promise. This technique allows the amputee to perceive the missing limb by focusing on the reflection of the remaining limb during specific movements and activities. Theoretically, this perception allows reconfiguration of the amputee’s sensory cortex.

Virtual reality therapy employs similar techniques based on the idea that the brain can be deceived. Initial case studies are promising and have prompted further research.¹¹

Recommendations

The US Department of Veterans Affairs and Department of Defense recently issued clinical guidelines for rehabilitating lower-limb amputees that include a segment on pain management.¹² The guidelines stress the importance of an interdisciplinary team approach that addresses each pathology plaguing the amputee.

They recommend narcotics during the immediate postoperative period, followed by transition to a non-narcotic medical regimen during the rehabilitation process. The guidelines don’t support a single, specific pain control method over others; they recommend the following approaches to PLP:

• pharmacologic treatment, which may include antiseizure medications, tricyclic antidepressants, selective serotonin reuptake inhibitors, NSAIDs, dextromethorathane, or long-acting narcotics
• epidural analgesia, patient-controlled analgesia, or regional analgesia
• nonpharmacologic therapies, including TENS, desensitization, scar mobilization, relaxation, and biofeedback.

Acknowledgements

The opinions and assertions contained herein are the private views of the authors and not to be construed as official, or as reflecting the views of the US Air Force Medical Service or the US Air Force at large.

Evidence summary

An estimated 1.7 million people in the United States are living with limb loss. The number is expected to increase because of ongoing military conflicts.¹ The incidence of PLP is 60% to 80% among amputees.¹

A multidisciplinary approach

A lack of comparative clinical trials of therapies for PLP has led health-care providers to adopt a multidisciplinary approach that combines evaluative management, desensitization, psychotherapy, and pharmacotherapy (FIGURE).

Evaluative management, based largely on expert opinion, includes assessing the fit of the prosthesis, treating referred pain, and assessing aggravating factors. Because residual limb pain can exacerbate PLP, adjusting a poorly fitting prosthesis or providing the patient with NSAIDs when there is evidence of stump inflammation may adequately control pain.^2,3 Anatomically distant pain syndromes, such as hip or lower back pain, can also aggravate PLP and should be managed to provide optimal pain relief.²

Desensitization, using TENS, has reduced PLP in multiple placebo-controlled trials and epidemiologic surveys.^2-5 TENS is an easy-to-use, low-cost, noninvasive, first-line therapy.⁵ Its long-term effectiveness in alleviating PLP remains unknown.² Some experts suggest that pain reductions after 1 year of treatment are comparable to placebo.² Other forms of desensitization (percussion and massage) are supported only by anecdotal reports.

Psychotherapy, including biofeedback, has been found in several case studies to effectively treat chronic PLP.^2,5 Psychotherapy can reportedly reveal the underlying mechanisms (muscle spasm, vascular insufficiency) and therefore direct therapeutic interventions by biofeedback or other focus techniques.²

FIGURE Management of phantom limb pain^1-10

Pharmacotherapy is best used as an adjunct to other treatments.² Although PLP is typically treated as neuropathic pain, only a few medications have been critically evaluated for treating it.⁶ Morphine (number needed to treat [NNT]=2.5; 95% confidence interval [CI], 1.9-3.4) and other opioids, including tramadol (NNT=3.9; 95% CI, 2.7-6.7 in neuropathic pain) help some patients.^6,7 Despite the proven benefit of tricyclic antidepressants (TCAs) in other neuropathic pain conditions, a recent RCT demonstrated no benefit of TCAs over placebo in PLP.⁸ Anticonvulsants, including gabapentin, have documented benefit in neuropathic pain modalities and are often used for PLP.⁶ However, their value in reducing PLP is still under investigation.⁶ One 2002 RCT showed benefit regarding an improvement of the visual analog scale by an average of 3 points (on a 10-point scale) after 6 weeks of gabapentin therapy.⁹ A similarly designed 2006 RCT of gabapentin, however did not identify significant pain reductions.¹⁰

Promising adjuvant therapies use mirroring techniques

Of the adjuvant treatments mentioned previously, only mirror box therapy has shown promise. This technique allows the amputee to perceive the missing limb by focusing on the reflection of the remaining limb during specific movements and activities. Theoretically, this perception allows reconfiguration of the amputee’s sensory cortex.

Virtual reality therapy employs similar techniques based on the idea that the brain can be deceived. Initial case studies are promising and have prompted further research.¹¹

Recommendations

The US Department of Veterans Affairs and Department of Defense recently issued clinical guidelines for rehabilitating lower-limb amputees that include a segment on pain management.¹² The guidelines stress the importance of an interdisciplinary team approach that addresses each pathology plaguing the amputee.

They recommend narcotics during the immediate postoperative period, followed by transition to a non-narcotic medical regimen during the rehabilitation process. The guidelines don’t support a single, specific pain control method over others; they recommend the following approaches to PLP:

• pharmacologic treatment, which may include antiseizure medications, tricyclic antidepressants, selective serotonin reuptake inhibitors, NSAIDs, dextromethorathane, or long-acting narcotics
• epidural analgesia, patient-controlled analgesia, or regional analgesia
• nonpharmacologic therapies, including TENS, desensitization, scar mobilization, relaxation, and biofeedback.

Acknowledgements

The opinions and assertions contained herein are the private views of the authors and not to be construed as official, or as reflecting the views of the US Air Force Medical Service or the US Air Force at large.