The Journal of Family Practice

EVIDENCE SUMMARY

The TABLE^1-4 shows the major results of the meta-analyses for the various medical therapy trials.

Two systematic reviews with meta-analyses evaluated treatment with INS for CRS with nasal polyps (40 RCTs; 3624 patients, mean age 48 years, 64% male) and without polyps (10 RCTs; 590 patients, mean age 39 years, 51% male).^1,2 Trials reported sinonasal symptom outcomes differently and couldn’t be combined. In addition to reducing rate of polyp occurrence, for both CRS with and without polyps, key findings were:

• Global symptom scores were better for INS than placebo.
• Proportion of patients responding was greater for INS than with placebo.

There was no significant difference between adverse event rates with INS and placebo.

A systematic review and meta-analysis (8 RCTs, 389 patients) compared different SI regimens for CRS.³ The standardized mean difference was used to combine trials using various symptom outcomes. Key findings included the following:

• SI was better than no treatment.
• SI adjunctive therapy (with an antihistamine) improved disease-specific quality-of-life scores.
• SI was less effective than INS therapy for symptom improvement.

Hypertonic and isotonic saline yielded similar symptom scores. No adverse effects were reported.

One meta-analysis evaluated patient-reported outcomes with 12 weeks of macrolide therapy compared to placebo using the results of the SinoNasal Outcome Test (SNOT). The SNOT is a quality-of-life questionnaire that lists symptoms and the social-emotional consequences of CRS; a negative change in the SNOT score, on a 0 to 5 scale, indicates improvement. Overall the SNOT score improved 8% with macrolide therapy—statistically significant, but of uncertain clinical importance.⁴

Surgery improves nasal obstruction, pain, and postnasal discharge

A systematic review of 21 studies (prospective RCTs, prospective controlled clinical trials, cohort studies, case series, and retrospective record reviews) with a total of 2070 patients analyzed the effectiveness of endoscopic sinus surgery alone for improving CRS symptoms.⁵ Mean duration of post-operative follow-up was 14 months. Meta-analysis was performed separately for each symptom and the standard mean difference of the symptom severity score before and after surgery was reported as the effect size (ES) for the outcome measure (an ES of 0.2 is considered small; 0.6, moderate; 1.2, large; and 2, very large).

All symptoms improved compared to their preoperative severity scores. Nasal obstruction improved the most (ES=1.73; 95% CI, 1.45-2.02). Large symptom improvement was also observed for facial pain (ES=1.13; 95% CI, 0.96-1.31) and postnasal discharge (ES=1.19; 95% CI, 0.96-1.43).

Surgery and medical therapy may provide comparable symptom relief

A recent Cochrane review of 4 low-quality RCTs including 378 patients compared surgical with medical interventions for CRS with nasal polyps. Study heterogeneity and selective outcome reporting prevented meta-analysis.

The 3 comparison groups were endoscopic sinus surgery vs systemic steroids + INS; polypectomy vs systemic steroid + INS; and endoscopic surgery + INS vs antibiotic + “high-dose” INS. Overall, neither surgery nor medical therapy was superior in terms of patient-reported symptom scores or quality-of-life measures.⁶

EVIDENCE SUMMARY

The TABLE^1-4 shows the major results of the meta-analyses for the various medical therapy trials.

Two systematic reviews with meta-analyses evaluated treatment with INS for CRS with nasal polyps (40 RCTs; 3624 patients, mean age 48 years, 64% male) and without polyps (10 RCTs; 590 patients, mean age 39 years, 51% male).^1,2 Trials reported sinonasal symptom outcomes differently and couldn’t be combined. In addition to reducing rate of polyp occurrence, for both CRS with and without polyps, key findings were:

• Global symptom scores were better for INS than placebo.
• Proportion of patients responding was greater for INS than with placebo.

There was no significant difference between adverse event rates with INS and placebo.

A systematic review and meta-analysis (8 RCTs, 389 patients) compared different SI regimens for CRS.³ The standardized mean difference was used to combine trials using various symptom outcomes. Key findings included the following:

• SI was better than no treatment.
• SI adjunctive therapy (with an antihistamine) improved disease-specific quality-of-life scores.
• SI was less effective than INS therapy for symptom improvement.

Hypertonic and isotonic saline yielded similar symptom scores. No adverse effects were reported.

One meta-analysis evaluated patient-reported outcomes with 12 weeks of macrolide therapy compared to placebo using the results of the SinoNasal Outcome Test (SNOT). The SNOT is a quality-of-life questionnaire that lists symptoms and the social-emotional consequences of CRS; a negative change in the SNOT score, on a 0 to 5 scale, indicates improvement. Overall the SNOT score improved 8% with macrolide therapy—statistically significant, but of uncertain clinical importance.⁴

Surgery improves nasal obstruction, pain, and postnasal discharge

A systematic review of 21 studies (prospective RCTs, prospective controlled clinical trials, cohort studies, case series, and retrospective record reviews) with a total of 2070 patients analyzed the effectiveness of endoscopic sinus surgery alone for improving CRS symptoms.⁵ Mean duration of post-operative follow-up was 14 months. Meta-analysis was performed separately for each symptom and the standard mean difference of the symptom severity score before and after surgery was reported as the effect size (ES) for the outcome measure (an ES of 0.2 is considered small; 0.6, moderate; 1.2, large; and 2, very large).

All symptoms improved compared to their preoperative severity scores. Nasal obstruction improved the most (ES=1.73; 95% CI, 1.45-2.02). Large symptom improvement was also observed for facial pain (ES=1.13; 95% CI, 0.96-1.31) and postnasal discharge (ES=1.19; 95% CI, 0.96-1.43).

Surgery and medical therapy may provide comparable symptom relief

A recent Cochrane review of 4 low-quality RCTs including 378 patients compared surgical with medical interventions for CRS with nasal polyps. Study heterogeneity and selective outcome reporting prevented meta-analysis.

The 3 comparison groups were endoscopic sinus surgery vs systemic steroids + INS; polypectomy vs systemic steroid + INS; and endoscopic surgery + INS vs antibiotic + “high-dose” INS. Overall, neither surgery nor medical therapy was superior in terms of patient-reported symptom scores or quality-of-life measures.⁶

EVIDENCE SUMMARY

The TABLE^1-4 shows the major results of the meta-analyses for the various medical therapy trials.

Two systematic reviews with meta-analyses evaluated treatment with INS for CRS with nasal polyps (40 RCTs; 3624 patients, mean age 48 years, 64% male) and without polyps (10 RCTs; 590 patients, mean age 39 years, 51% male).^1,2 Trials reported sinonasal symptom outcomes differently and couldn’t be combined. In addition to reducing rate of polyp occurrence, for both CRS with and without polyps, key findings were:

• Global symptom scores were better for INS than placebo.
• Proportion of patients responding was greater for INS than with placebo.

There was no significant difference between adverse event rates with INS and placebo.

A systematic review and meta-analysis (8 RCTs, 389 patients) compared different SI regimens for CRS.³ The standardized mean difference was used to combine trials using various symptom outcomes. Key findings included the following:

• SI was better than no treatment.
• SI adjunctive therapy (with an antihistamine) improved disease-specific quality-of-life scores.
• SI was less effective than INS therapy for symptom improvement.

Hypertonic and isotonic saline yielded similar symptom scores. No adverse effects were reported.

One meta-analysis evaluated patient-reported outcomes with 12 weeks of macrolide therapy compared to placebo using the results of the SinoNasal Outcome Test (SNOT). The SNOT is a quality-of-life questionnaire that lists symptoms and the social-emotional consequences of CRS; a negative change in the SNOT score, on a 0 to 5 scale, indicates improvement. Overall the SNOT score improved 8% with macrolide therapy—statistically significant, but of uncertain clinical importance.⁴

Surgery improves nasal obstruction, pain, and postnasal discharge

A systematic review of 21 studies (prospective RCTs, prospective controlled clinical trials, cohort studies, case series, and retrospective record reviews) with a total of 2070 patients analyzed the effectiveness of endoscopic sinus surgery alone for improving CRS symptoms.⁵ Mean duration of post-operative follow-up was 14 months. Meta-analysis was performed separately for each symptom and the standard mean difference of the symptom severity score before and after surgery was reported as the effect size (ES) for the outcome measure (an ES of 0.2 is considered small; 0.6, moderate; 1.2, large; and 2, very large).

All symptoms improved compared to their preoperative severity scores. Nasal obstruction improved the most (ES=1.73; 95% CI, 1.45-2.02). Large symptom improvement was also observed for facial pain (ES=1.13; 95% CI, 0.96-1.31) and postnasal discharge (ES=1.19; 95% CI, 0.96-1.43).

Surgery and medical therapy may provide comparable symptom relief

A recent Cochrane review of 4 low-quality RCTs including 378 patients compared surgical with medical interventions for CRS with nasal polyps. Study heterogeneity and selective outcome reporting prevented meta-analysis.

The 3 comparison groups were endoscopic sinus surgery vs systemic steroids + INS; polypectomy vs systemic steroid + INS; and endoscopic surgery + INS vs antibiotic + “high-dose” INS. Overall, neither surgery nor medical therapy was superior in terms of patient-reported symptom scores or quality-of-life measures.⁶

EVIDENCE SUMMARY

In 2015, the American College of Rheumatology (ACR) redefined the clinical criteria for diagnosis of gout based on a 3-step system¹ that can be found at: http://goutclassificationcalculator.auckland.ac.nz. The ACR rule was derived from a cross-sectional study of 983 patients in 25 rheumatology centers in 16 countries who presented with a swollen joint.² Of the 983 patients, 509 had gout; the prevalence was 52%. Data from 653 of these patients were used to develop the rule and then validated in the remaining 330 patients.

Compared with the gold standard of monosodium urate crystals in synovial fluid, the ACR rule has a sensitivity of 92% and a specificity of 89%. The rule, designed for the research setting, involves using synovial fluid analysis, ultrasound imaging, and radiography, which makes it less useful in a primary care setting.

The Netherlands rule for primary care

A prospective diagnostic study in 328 family medicine patients (74% male; mean age 57) with monoarthritis tested the ability of multiple clinical variables to diagnose gout using monosodium urate crystals in synovial fluid as the gold standard.³ The prevalence of gout in this population was 57%.

The best diagnostic rule (Netherlands rule) comprised the following predefined variables: male sex, previous patient-reported arthritis attack, onset within one day, joint redness, first metatarsophalangeal joint (MTP1) involvement, hypertension or cardiovascular disease (angina pectoris, myocardial infarction, heart failure, cerebrovascular accident, transient ischemic attack, or peripheral vascular disease), and serum uric acid level above 5.88 mg/dL. The rule gives one point for each item. A score >8 had a positive likelihood ratio for diagnosing gout of 3.48 (TABLE¹) and a higher positive predictive value (PPV) than family physicians’ clinical impressions (83% vs 64%).

The prevalence of gout in patients with scores of <4, 4 to 8, and >8 were 2.8%, 27%, and 80%, respectively. For scores of 4 to 8, the probability of gout is indeterminate, and synovial fluid analysis is recommended.

The Netherlands rule, validated in a secondary care practice of 390 patients with monoarthritis, found that a score >8 had a PPV of 87% and a score <4 had a negative predictive value of 95%.⁴ The probability of gout based on this rule can be calculated at http://www.umcn.nl/goutcalc.

Clinical prediction rules effectively diagnose gout without joint fluid analysis.In the study used to develop the Netherlands rule, no patients with a high probability of gout had septic arthritis. The ability of the rule to differentiate between gout and septic arthritis was tested retrospectively in 33 patients with acute gout (podagra excluded) diagnosed by the presence of monosodium urate joint crystals and 27 patients with septic arthritis diagnosed by positive bacterial culture.⁵ Patients with gout had significantly higher scores than patients with septic arthritis (7.8 ± 1.59 vs 3.4 ± 2.3; P<.001).

A study of 82 Veterans Administration patients compared the American Rheumatology Association (ARA), New York, and Rome prediction rules with regard to their ability to diagnose gout with synovial urate crystals.⁶ The ARA criteria for gout diagnosis require either tophi or monosodium urate crystals in synovial fluid, or 6 out of a list of 12 other criteria.⁷

The New York prediction rule requires that patients meet 2 or more of the following criteria: at least 2 attacks of painful joint swelling with complete resolution within 2 weeks, podagra, tophi, and rapid response to colchicine treatment, defined as a major reduction in the objective signs of inflammation within 48 hours.

The Rome prediction rule requires meeting 2 of 3 criteria: serum uric acid >7 mg/dL in men and >6 mg/dL in women, presence of tophi, and history of attacks of painful joint swelling with abrupt onset and resolution within 2 weeks.

The New York prediction rule had the highest positive likelihood ratio of 4.4 compared with the ARA (1.8) and Rome (4.3) rules.⁶ The utility of the New York and Rome rules, although they have fewer criteria than ARA, is limited by the fact that they include a previous episode of joint swelling and tophi. These criteria increase their specificity but make them less useful in diagnosing a first episode of gout, when tophi are unlikely to have developed.

Prediction rules are more sensitive in established gout

The new ACR prediction rule was compared with the ARA, Rome, and New York clinical prediction rules using urate crystals as the gold standard in early (less than 2 years) and established disease (longer than 2 years).⁸ All clinical prediction rules were more sensitive in established disease than early disease (95.3% vs 84.1%; P<.001) and more specific in early disease than established disease (79.9% vs 52.5%; P<.001).

EVIDENCE SUMMARY

In 2015, the American College of Rheumatology (ACR) redefined the clinical criteria for diagnosis of gout based on a 3-step system¹ that can be found at: http://goutclassificationcalculator.auckland.ac.nz. The ACR rule was derived from a cross-sectional study of 983 patients in 25 rheumatology centers in 16 countries who presented with a swollen joint.² Of the 983 patients, 509 had gout; the prevalence was 52%. Data from 653 of these patients were used to develop the rule and then validated in the remaining 330 patients.

Compared with the gold standard of monosodium urate crystals in synovial fluid, the ACR rule has a sensitivity of 92% and a specificity of 89%. The rule, designed for the research setting, involves using synovial fluid analysis, ultrasound imaging, and radiography, which makes it less useful in a primary care setting.

The Netherlands rule for primary care

A prospective diagnostic study in 328 family medicine patients (74% male; mean age 57) with monoarthritis tested the ability of multiple clinical variables to diagnose gout using monosodium urate crystals in synovial fluid as the gold standard.³ The prevalence of gout in this population was 57%.

The best diagnostic rule (Netherlands rule) comprised the following predefined variables: male sex, previous patient-reported arthritis attack, onset within one day, joint redness, first metatarsophalangeal joint (MTP1) involvement, hypertension or cardiovascular disease (angina pectoris, myocardial infarction, heart failure, cerebrovascular accident, transient ischemic attack, or peripheral vascular disease), and serum uric acid level above 5.88 mg/dL. The rule gives one point for each item. A score >8 had a positive likelihood ratio for diagnosing gout of 3.48 (TABLE¹) and a higher positive predictive value (PPV) than family physicians’ clinical impressions (83% vs 64%).

The prevalence of gout in patients with scores of <4, 4 to 8, and >8 were 2.8%, 27%, and 80%, respectively. For scores of 4 to 8, the probability of gout is indeterminate, and synovial fluid analysis is recommended.

The Netherlands rule, validated in a secondary care practice of 390 patients with monoarthritis, found that a score >8 had a PPV of 87% and a score <4 had a negative predictive value of 95%.⁴ The probability of gout based on this rule can be calculated at http://www.umcn.nl/goutcalc.

Clinical prediction rules effectively diagnose gout without joint fluid analysis.In the study used to develop the Netherlands rule, no patients with a high probability of gout had septic arthritis. The ability of the rule to differentiate between gout and septic arthritis was tested retrospectively in 33 patients with acute gout (podagra excluded) diagnosed by the presence of monosodium urate joint crystals and 27 patients with septic arthritis diagnosed by positive bacterial culture.⁵ Patients with gout had significantly higher scores than patients with septic arthritis (7.8 ± 1.59 vs 3.4 ± 2.3; P<.001).

A study of 82 Veterans Administration patients compared the American Rheumatology Association (ARA), New York, and Rome prediction rules with regard to their ability to diagnose gout with synovial urate crystals.⁶ The ARA criteria for gout diagnosis require either tophi or monosodium urate crystals in synovial fluid, or 6 out of a list of 12 other criteria.⁷

The New York prediction rule requires that patients meet 2 or more of the following criteria: at least 2 attacks of painful joint swelling with complete resolution within 2 weeks, podagra, tophi, and rapid response to colchicine treatment, defined as a major reduction in the objective signs of inflammation within 48 hours.

The Rome prediction rule requires meeting 2 of 3 criteria: serum uric acid >7 mg/dL in men and >6 mg/dL in women, presence of tophi, and history of attacks of painful joint swelling with abrupt onset and resolution within 2 weeks.

The New York prediction rule had the highest positive likelihood ratio of 4.4 compared with the ARA (1.8) and Rome (4.3) rules.⁶ The utility of the New York and Rome rules, although they have fewer criteria than ARA, is limited by the fact that they include a previous episode of joint swelling and tophi. These criteria increase their specificity but make them less useful in diagnosing a first episode of gout, when tophi are unlikely to have developed.

Prediction rules are more sensitive in established gout

The new ACR prediction rule was compared with the ARA, Rome, and New York clinical prediction rules using urate crystals as the gold standard in early (less than 2 years) and established disease (longer than 2 years).⁸ All clinical prediction rules were more sensitive in established disease than early disease (95.3% vs 84.1%; P<.001) and more specific in early disease than established disease (79.9% vs 52.5%; P<.001).

EVIDENCE SUMMARY

In 2015, the American College of Rheumatology (ACR) redefined the clinical criteria for diagnosis of gout based on a 3-step system¹ that can be found at: http://goutclassificationcalculator.auckland.ac.nz. The ACR rule was derived from a cross-sectional study of 983 patients in 25 rheumatology centers in 16 countries who presented with a swollen joint.² Of the 983 patients, 509 had gout; the prevalence was 52%. Data from 653 of these patients were used to develop the rule and then validated in the remaining 330 patients.

Compared with the gold standard of monosodium urate crystals in synovial fluid, the ACR rule has a sensitivity of 92% and a specificity of 89%. The rule, designed for the research setting, involves using synovial fluid analysis, ultrasound imaging, and radiography, which makes it less useful in a primary care setting.

The Netherlands rule for primary care

A prospective diagnostic study in 328 family medicine patients (74% male; mean age 57) with monoarthritis tested the ability of multiple clinical variables to diagnose gout using monosodium urate crystals in synovial fluid as the gold standard.³ The prevalence of gout in this population was 57%.

The best diagnostic rule (Netherlands rule) comprised the following predefined variables: male sex, previous patient-reported arthritis attack, onset within one day, joint redness, first metatarsophalangeal joint (MTP1) involvement, hypertension or cardiovascular disease (angina pectoris, myocardial infarction, heart failure, cerebrovascular accident, transient ischemic attack, or peripheral vascular disease), and serum uric acid level above 5.88 mg/dL. The rule gives one point for each item. A score >8 had a positive likelihood ratio for diagnosing gout of 3.48 (TABLE¹) and a higher positive predictive value (PPV) than family physicians’ clinical impressions (83% vs 64%).

The prevalence of gout in patients with scores of <4, 4 to 8, and >8 were 2.8%, 27%, and 80%, respectively. For scores of 4 to 8, the probability of gout is indeterminate, and synovial fluid analysis is recommended.

The Netherlands rule, validated in a secondary care practice of 390 patients with monoarthritis, found that a score >8 had a PPV of 87% and a score <4 had a negative predictive value of 95%.⁴ The probability of gout based on this rule can be calculated at http://www.umcn.nl/goutcalc.

Clinical prediction rules effectively diagnose gout without joint fluid analysis.In the study used to develop the Netherlands rule, no patients with a high probability of gout had septic arthritis. The ability of the rule to differentiate between gout and septic arthritis was tested retrospectively in 33 patients with acute gout (podagra excluded) diagnosed by the presence of monosodium urate joint crystals and 27 patients with septic arthritis diagnosed by positive bacterial culture.⁵ Patients with gout had significantly higher scores than patients with septic arthritis (7.8 ± 1.59 vs 3.4 ± 2.3; P<.001).

A study of 82 Veterans Administration patients compared the American Rheumatology Association (ARA), New York, and Rome prediction rules with regard to their ability to diagnose gout with synovial urate crystals.⁶ The ARA criteria for gout diagnosis require either tophi or monosodium urate crystals in synovial fluid, or 6 out of a list of 12 other criteria.⁷

The New York prediction rule requires that patients meet 2 or more of the following criteria: at least 2 attacks of painful joint swelling with complete resolution within 2 weeks, podagra, tophi, and rapid response to colchicine treatment, defined as a major reduction in the objective signs of inflammation within 48 hours.

The Rome prediction rule requires meeting 2 of 3 criteria: serum uric acid >7 mg/dL in men and >6 mg/dL in women, presence of tophi, and history of attacks of painful joint swelling with abrupt onset and resolution within 2 weeks.

The New York prediction rule had the highest positive likelihood ratio of 4.4 compared with the ARA (1.8) and Rome (4.3) rules.⁶ The utility of the New York and Rome rules, although they have fewer criteria than ARA, is limited by the fact that they include a previous episode of joint swelling and tophi. These criteria increase their specificity but make them less useful in diagnosing a first episode of gout, when tophi are unlikely to have developed.

Prediction rules are more sensitive in established gout

The new ACR prediction rule was compared with the ARA, Rome, and New York clinical prediction rules using urate crystals as the gold standard in early (less than 2 years) and established disease (longer than 2 years).⁸ All clinical prediction rules were more sensitive in established disease than early disease (95.3% vs 84.1%; P<.001) and more specific in early disease than established disease (79.9% vs 52.5%; P<.001).

EVIDENCE SUMMARY

A post-hoc analysis of a prospective RCT (the Diabetes Prevention Program) comprising 3081 patients with impaired glucose metabolism who received metformin, a lifestyle modification program, or no intervention (placebo) found that metformin reduced the risk of developing diabetes only for patients in the highest risk quartile over 2.8 years. Lifestyle modification reduced diabetes risk in all patients.¹

Investigators stratified patients who met National Cholesterol Education Program Adult Treatment Panel III (NCEP ATP III) criteria for metabolic syndrome into risk quartiles for progression to diabetes using a model they developed based on 7 parameters: fasting plasma glucose, hemoglobin A1c, history of high blood glucose, waist:hip ratio, waist circumference, triglycerides, and height (TABLE¹). The model reasonably fit outcomes—the percentage of patients in each quartile who developed diabetes—with an area under the receiver operator characteristic curve of 0.73 (a measure of diagnostic accuracy where 1 is a perfect predictor and 0.5 is random).

Lifestyle modification reduced risk in all quartiles with progressively greater effect as risk increased (lowest risk quartile: ARR=4.9%, 3-year NNT=20.4; highest risk quartile: ARR=28.3%; 3-year NNT=3.5).

There were 2 key weaknesses of the risk model: It wasn’t validated in a separate population and the true incidence of diabetes among patients taking placebo was higher than predicted. The investigators compared their risk prediction model results with the Framingham Risk Score (FRS) for diabetes and found that they correlated well, although the FRS results were consistently about 6% (absolute) higher when corrected for duration. (The FRS calculator is available online at www.framinghamheartstudy.org/risk-functions/diabetes/.)

Lifestyle change reduces diabetes risk more than metformin

The original Diabetes Prevention Program found that intensive lifestyle intervention and metformin reduced the number of diabetes cases over 2.8 years among 3234 patients at risk for developing diabetes.²

Compared with no intervention, fewer patients developed diabetes with either metformin or lifestyle improvement, although lifestyle change had the larger effect (no intervention: 11 cases per 100 person-years; metformin: 7.8 cases; 95% confidence interval [CI], 6.8-8.8; ARR=3.2% per year vs no intervention; lifestyle improvements: 4.8 cases; 95% CI, 4.1-5.7; ARR=6.2% per year vs no intervention).

A 10-year follow-up study to the Diabetes Prevention Program found that, compared with no intervention, both metformin and lifestyle interventions continued to be associated with a lower incidence of diabetes (no intervention: 7.8 cases per 100 person-years; 95% CI, 4.8-6.5; metformin: 6.4 cases; 95% CI, 4.2-5.7; ARR=1.4% per year; lifestyle interventions: 5.3 cases; 95% CI, 5.1-6.8; ARR=2.5% per year).³

Researchers originally randomized 3234 patients with body mass index ≥24 kg/m², fasting blood sugar 95 to 125 mg/dL, and 2-hour post 75-gm glucose value of 149 to 199 mg/dL to 3 groups: intensive lifestyle modification (weight loss goal of 7%, 150 minutes a week of exercise), metformin (850 mg twice daily), and no inter­vention. After the 2.8-year follow-up period, 2766 patients continued for another 5.7 years of follow-up. Investigators offered group lifestyle counseling to all patients and continued metformin at the same dose in the second group.

Earlier study shows an effect for metformin, but with a caveat

An earlier RCT found that metformin reduced the risk of developing diabetes in patients with metabolic syndrome.⁴ Investigators randomized 70 patients to metformin (250 mg 3 times daily) or placebo for a year. Fewer patients developed diabetes with metformin (3% vs 16.2%, P=.011; NNT=7.6) and more had a normal glucose tolerance test result (84.9% vs 51.4%, P=.011; NNT=3). However, by current American Diabetes Association criteria, half of the subjects had early diabetes at baseline.

Metformin lowers fasting blood sugar, but may not reverse metabolic syndrome

A post-hoc analysis of another RCT found that metformin reduced fasting plasma glucose (FPG) levels in patients with upper-body obesity and metabolic syndrome (by 1999 World Health Organization criteria but not NCEP ATP III criteria).⁵

Investigators randomized 457 patients to metformin 850 mg once daily or placebo and followed them for a year. FPG levels decreased with metformin but increased with placebo (reduction FPG 5.9 mg/dL vs increase FPG 12.3 mg/dL; P<.04). The investigators didn’t report whether any patients developed diabetes.

However, another RCT (155 patients) that compared metformin 850 mg twice daily with placebo in subjects with metabolic syndrome but without diabetes found greater normalization of FPG (5% vs 0%; P=.005), but no reversal of metabolic syndrome or change in Framingham 10-year risk score after 12 weeks.⁶

EVIDENCE SUMMARY

A post-hoc analysis of a prospective RCT (the Diabetes Prevention Program) comprising 3081 patients with impaired glucose metabolism who received metformin, a lifestyle modification program, or no intervention (placebo) found that metformin reduced the risk of developing diabetes only for patients in the highest risk quartile over 2.8 years. Lifestyle modification reduced diabetes risk in all patients.¹

Investigators stratified patients who met National Cholesterol Education Program Adult Treatment Panel III (NCEP ATP III) criteria for metabolic syndrome into risk quartiles for progression to diabetes using a model they developed based on 7 parameters: fasting plasma glucose, hemoglobin A1c, history of high blood glucose, waist:hip ratio, waist circumference, triglycerides, and height (TABLE¹). The model reasonably fit outcomes—the percentage of patients in each quartile who developed diabetes—with an area under the receiver operator characteristic curve of 0.73 (a measure of diagnostic accuracy where 1 is a perfect predictor and 0.5 is random).

Lifestyle modification reduced risk in all quartiles with progressively greater effect as risk increased (lowest risk quartile: ARR=4.9%, 3-year NNT=20.4; highest risk quartile: ARR=28.3%; 3-year NNT=3.5).

There were 2 key weaknesses of the risk model: It wasn’t validated in a separate population and the true incidence of diabetes among patients taking placebo was higher than predicted. The investigators compared their risk prediction model results with the Framingham Risk Score (FRS) for diabetes and found that they correlated well, although the FRS results were consistently about 6% (absolute) higher when corrected for duration. (The FRS calculator is available online at www.framinghamheartstudy.org/risk-functions/diabetes/.)

Lifestyle change reduces diabetes risk more than metformin

The original Diabetes Prevention Program found that intensive lifestyle intervention and metformin reduced the number of diabetes cases over 2.8 years among 3234 patients at risk for developing diabetes.²

Compared with no intervention, fewer patients developed diabetes with either metformin or lifestyle improvement, although lifestyle change had the larger effect (no intervention: 11 cases per 100 person-years; metformin: 7.8 cases; 95% confidence interval [CI], 6.8-8.8; ARR=3.2% per year vs no intervention; lifestyle improvements: 4.8 cases; 95% CI, 4.1-5.7; ARR=6.2% per year vs no intervention).

A 10-year follow-up study to the Diabetes Prevention Program found that, compared with no intervention, both metformin and lifestyle interventions continued to be associated with a lower incidence of diabetes (no intervention: 7.8 cases per 100 person-years; 95% CI, 4.8-6.5; metformin: 6.4 cases; 95% CI, 4.2-5.7; ARR=1.4% per year; lifestyle interventions: 5.3 cases; 95% CI, 5.1-6.8; ARR=2.5% per year).³

Researchers originally randomized 3234 patients with body mass index ≥24 kg/m², fasting blood sugar 95 to 125 mg/dL, and 2-hour post 75-gm glucose value of 149 to 199 mg/dL to 3 groups: intensive lifestyle modification (weight loss goal of 7%, 150 minutes a week of exercise), metformin (850 mg twice daily), and no inter­vention. After the 2.8-year follow-up period, 2766 patients continued for another 5.7 years of follow-up. Investigators offered group lifestyle counseling to all patients and continued metformin at the same dose in the second group.

Earlier study shows an effect for metformin, but with a caveat

An earlier RCT found that metformin reduced the risk of developing diabetes in patients with metabolic syndrome.⁴ Investigators randomized 70 patients to metformin (250 mg 3 times daily) or placebo for a year. Fewer patients developed diabetes with metformin (3% vs 16.2%, P=.011; NNT=7.6) and more had a normal glucose tolerance test result (84.9% vs 51.4%, P=.011; NNT=3). However, by current American Diabetes Association criteria, half of the subjects had early diabetes at baseline.

Metformin lowers fasting blood sugar, but may not reverse metabolic syndrome

A post-hoc analysis of another RCT found that metformin reduced fasting plasma glucose (FPG) levels in patients with upper-body obesity and metabolic syndrome (by 1999 World Health Organization criteria but not NCEP ATP III criteria).⁵

Investigators randomized 457 patients to metformin 850 mg once daily or placebo and followed them for a year. FPG levels decreased with metformin but increased with placebo (reduction FPG 5.9 mg/dL vs increase FPG 12.3 mg/dL; P<.04). The investigators didn’t report whether any patients developed diabetes.

However, another RCT (155 patients) that compared metformin 850 mg twice daily with placebo in subjects with metabolic syndrome but without diabetes found greater normalization of FPG (5% vs 0%; P=.005), but no reversal of metabolic syndrome or change in Framingham 10-year risk score after 12 weeks.⁶

EVIDENCE SUMMARY

A post-hoc analysis of a prospective RCT (the Diabetes Prevention Program) comprising 3081 patients with impaired glucose metabolism who received metformin, a lifestyle modification program, or no intervention (placebo) found that metformin reduced the risk of developing diabetes only for patients in the highest risk quartile over 2.8 years. Lifestyle modification reduced diabetes risk in all patients.¹

Investigators stratified patients who met National Cholesterol Education Program Adult Treatment Panel III (NCEP ATP III) criteria for metabolic syndrome into risk quartiles for progression to diabetes using a model they developed based on 7 parameters: fasting plasma glucose, hemoglobin A1c, history of high blood glucose, waist:hip ratio, waist circumference, triglycerides, and height (TABLE¹). The model reasonably fit outcomes—the percentage of patients in each quartile who developed diabetes—with an area under the receiver operator characteristic curve of 0.73 (a measure of diagnostic accuracy where 1 is a perfect predictor and 0.5 is random).

Lifestyle modification reduced risk in all quartiles with progressively greater effect as risk increased (lowest risk quartile: ARR=4.9%, 3-year NNT=20.4; highest risk quartile: ARR=28.3%; 3-year NNT=3.5).

There were 2 key weaknesses of the risk model: It wasn’t validated in a separate population and the true incidence of diabetes among patients taking placebo was higher than predicted. The investigators compared their risk prediction model results with the Framingham Risk Score (FRS) for diabetes and found that they correlated well, although the FRS results were consistently about 6% (absolute) higher when corrected for duration. (The FRS calculator is available online at www.framinghamheartstudy.org/risk-functions/diabetes/.)

Lifestyle change reduces diabetes risk more than metformin

The original Diabetes Prevention Program found that intensive lifestyle intervention and metformin reduced the number of diabetes cases over 2.8 years among 3234 patients at risk for developing diabetes.²

Compared with no intervention, fewer patients developed diabetes with either metformin or lifestyle improvement, although lifestyle change had the larger effect (no intervention: 11 cases per 100 person-years; metformin: 7.8 cases; 95% confidence interval [CI], 6.8-8.8; ARR=3.2% per year vs no intervention; lifestyle improvements: 4.8 cases; 95% CI, 4.1-5.7; ARR=6.2% per year vs no intervention).

A 10-year follow-up study to the Diabetes Prevention Program found that, compared with no intervention, both metformin and lifestyle interventions continued to be associated with a lower incidence of diabetes (no intervention: 7.8 cases per 100 person-years; 95% CI, 4.8-6.5; metformin: 6.4 cases; 95% CI, 4.2-5.7; ARR=1.4% per year; lifestyle interventions: 5.3 cases; 95% CI, 5.1-6.8; ARR=2.5% per year).³

Researchers originally randomized 3234 patients with body mass index ≥24 kg/m², fasting blood sugar 95 to 125 mg/dL, and 2-hour post 75-gm glucose value of 149 to 199 mg/dL to 3 groups: intensive lifestyle modification (weight loss goal of 7%, 150 minutes a week of exercise), metformin (850 mg twice daily), and no inter­vention. After the 2.8-year follow-up period, 2766 patients continued for another 5.7 years of follow-up. Investigators offered group lifestyle counseling to all patients and continued metformin at the same dose in the second group.

Earlier study shows an effect for metformin, but with a caveat

An earlier RCT found that metformin reduced the risk of developing diabetes in patients with metabolic syndrome.⁴ Investigators randomized 70 patients to metformin (250 mg 3 times daily) or placebo for a year. Fewer patients developed diabetes with metformin (3% vs 16.2%, P=.011; NNT=7.6) and more had a normal glucose tolerance test result (84.9% vs 51.4%, P=.011; NNT=3). However, by current American Diabetes Association criteria, half of the subjects had early diabetes at baseline.

Metformin lowers fasting blood sugar, but may not reverse metabolic syndrome

A post-hoc analysis of another RCT found that metformin reduced fasting plasma glucose (FPG) levels in patients with upper-body obesity and metabolic syndrome (by 1999 World Health Organization criteria but not NCEP ATP III criteria).⁵

Investigators randomized 457 patients to metformin 850 mg once daily or placebo and followed them for a year. FPG levels decreased with metformin but increased with placebo (reduction FPG 5.9 mg/dL vs increase FPG 12.3 mg/dL; P<.04). The investigators didn’t report whether any patients developed diabetes.

However, another RCT (155 patients) that compared metformin 850 mg twice daily with placebo in subjects with metabolic syndrome but without diabetes found greater normalization of FPG (5% vs 0%; P=.005), but no reversal of metabolic syndrome or change in Framingham 10-year risk score after 12 weeks.⁶

Since Florida has seen several new cases of local mosquito-borne infection, controlling and preventing Zika infection has great urgency. Zika virus involves an arthropod-borne infection transmitted by Aedes aegypti and Aedes albopictus mosquitoes. Other modes of transmission include the maternal-fetal route, any sexual contact, blood transfusions, organ or tissue transplantation, and laboratory exposure.¹

The first case of Zika infection in the United States and its territories occurred through international travel. According to the Centers for Disease Control and Prevention, as of October 12, 2016, there were 3807 travel-associated cases of Zika infection in the United States and 84 instances in its territories.² As for local transmission, there were 128 people evidencing a Zika infection in the United States and 25,871 in US territories.² Regions between Texas and Florida are at high risk because Aedes mosquitoes primarily inhabit the gulf coast.³ Many cases have occurred despite repellent use and eradication efforts, possibly due to resistance acquired by these mosquitoes.¹

Control measures include using insect repellents, aerial spraying of insecticides, eliminating mosquito breeding sites, covering water tanks, and using mosquito nets or door and window screens. Infection during pregnancy is the greatest concern because of congenital anomalies (including microcephaly) that negatively affect brain development.⁴

Before a possible conception or any sexual contact, women exposed to Zika—with or without symptoms—must wait at least 8 weeks; men with or without symptoms should abstain for 6 months.⁴ Individuals should avoid traveling to areas with Zika infestation, wear long-sleeved clothing treated with permethrin, and minimize outside exposure, especially in evening hours.⁴

The World Health Organization is utilizing genetically modified mosquitoes to diminish Aedes populations; trials conducted in affected areas of Brazil revealed that the number of Aedes mosquitoes was reduced by 90%.⁵ This method of mosquito control is currently being studied in the United States.⁶ Vaccinations to prevent Zika infection are also under investigation.

Physicians should educate patients regarding the clinical manifestations and complications of Zika virus infection; people need to know that the Zika virus can be sexually transmitted. Doctors should also counsel patients to curtail travel to areas that have Zika infestations, or to at least wear protective clothing while in such areas to minimize mosquito bite risk. Educating travelers about appropriate postponement of sexual contact after any exposure to the Zika virus is also essential.⁴

Hema Madhuri Mekala, MD
Priyanga Jayakumar, MD
Rajashekar Reddy Yeruva, MD
Steven Lippmann, MD
Louisville, KY

Since Florida has seen several new cases of local mosquito-borne infection, controlling and preventing Zika infection has great urgency. Zika virus involves an arthropod-borne infection transmitted by Aedes aegypti and Aedes albopictus mosquitoes. Other modes of transmission include the maternal-fetal route, any sexual contact, blood transfusions, organ or tissue transplantation, and laboratory exposure.¹

The first case of Zika infection in the United States and its territories occurred through international travel. According to the Centers for Disease Control and Prevention, as of October 12, 2016, there were 3807 travel-associated cases of Zika infection in the United States and 84 instances in its territories.² As for local transmission, there were 128 people evidencing a Zika infection in the United States and 25,871 in US territories.² Regions between Texas and Florida are at high risk because Aedes mosquitoes primarily inhabit the gulf coast.³ Many cases have occurred despite repellent use and eradication efforts, possibly due to resistance acquired by these mosquitoes.¹

Control measures include using insect repellents, aerial spraying of insecticides, eliminating mosquito breeding sites, covering water tanks, and using mosquito nets or door and window screens. Infection during pregnancy is the greatest concern because of congenital anomalies (including microcephaly) that negatively affect brain development.⁴

Before a possible conception or any sexual contact, women exposed to Zika—with or without symptoms—must wait at least 8 weeks; men with or without symptoms should abstain for 6 months.⁴ Individuals should avoid traveling to areas with Zika infestation, wear long-sleeved clothing treated with permethrin, and minimize outside exposure, especially in evening hours.⁴

The World Health Organization is utilizing genetically modified mosquitoes to diminish Aedes populations; trials conducted in affected areas of Brazil revealed that the number of Aedes mosquitoes was reduced by 90%.⁵ This method of mosquito control is currently being studied in the United States.⁶ Vaccinations to prevent Zika infection are also under investigation.

Physicians should educate patients regarding the clinical manifestations and complications of Zika virus infection; people need to know that the Zika virus can be sexually transmitted. Doctors should also counsel patients to curtail travel to areas that have Zika infestations, or to at least wear protective clothing while in such areas to minimize mosquito bite risk. Educating travelers about appropriate postponement of sexual contact after any exposure to the Zika virus is also essential.⁴

Hema Madhuri Mekala, MD
Priyanga Jayakumar, MD
Rajashekar Reddy Yeruva, MD
Steven Lippmann, MD
Louisville, KY

Since Florida has seen several new cases of local mosquito-borne infection, controlling and preventing Zika infection has great urgency. Zika virus involves an arthropod-borne infection transmitted by Aedes aegypti and Aedes albopictus mosquitoes. Other modes of transmission include the maternal-fetal route, any sexual contact, blood transfusions, organ or tissue transplantation, and laboratory exposure.¹

The first case of Zika infection in the United States and its territories occurred through international travel. According to the Centers for Disease Control and Prevention, as of October 12, 2016, there were 3807 travel-associated cases of Zika infection in the United States and 84 instances in its territories.² As for local transmission, there were 128 people evidencing a Zika infection in the United States and 25,871 in US territories.² Regions between Texas and Florida are at high risk because Aedes mosquitoes primarily inhabit the gulf coast.³ Many cases have occurred despite repellent use and eradication efforts, possibly due to resistance acquired by these mosquitoes.¹

Control measures include using insect repellents, aerial spraying of insecticides, eliminating mosquito breeding sites, covering water tanks, and using mosquito nets or door and window screens. Infection during pregnancy is the greatest concern because of congenital anomalies (including microcephaly) that negatively affect brain development.⁴

Before a possible conception or any sexual contact, women exposed to Zika—with or without symptoms—must wait at least 8 weeks; men with or without symptoms should abstain for 6 months.⁴ Individuals should avoid traveling to areas with Zika infestation, wear long-sleeved clothing treated with permethrin, and minimize outside exposure, especially in evening hours.⁴

The World Health Organization is utilizing genetically modified mosquitoes to diminish Aedes populations; trials conducted in affected areas of Brazil revealed that the number of Aedes mosquitoes was reduced by 90%.⁵ This method of mosquito control is currently being studied in the United States.⁶ Vaccinations to prevent Zika infection are also under investigation.

Physicians should educate patients regarding the clinical manifestations and complications of Zika virus infection; people need to know that the Zika virus can be sexually transmitted. Doctors should also counsel patients to curtail travel to areas that have Zika infestations, or to at least wear protective clothing while in such areas to minimize mosquito bite risk. Educating travelers about appropriate postponement of sexual contact after any exposure to the Zika virus is also essential.⁴

Hema Madhuri Mekala, MD
Priyanga Jayakumar, MD
Rajashekar Reddy Yeruva, MD
Steven Lippmann, MD
Louisville, KY

THE CASE

A 47-year-old African American woman was admitted to the hospital with pulmonary edema revealed on a computed tomography (CT) scan. She had a history of systemic lupus erythematosus (SLE), hypertension, and end-stage renal disease (ESRD). The patient had been hospitalized one month earlier for lupus nephritis with a hypertensive emergency that led to a seizure. During this earlier hospitalization, she was given a diagnosis of posterior reversible encephalopathy syndrome.

Two weeks into her more recent hospitalization, the patient developed a fever that was accompanied by cough and fatigue. By the third week, there was no identified cause of the fever, and the patient met the criteria for fever of unknown origin (FUO).

Her medications included cyclophosphamide, prednisone, nebivolol, clonidine, phenytoin, and epoetin alfa. The patient was also receiving dialysis every other day. Chest x-ray findings suggested pneumonia, and the patient was treated with vancomycin and piperacillin/tazobactam. However, her fever persisted after completing the antibiotics. Central line sepsis was high in the differential, as the patient was on dialysis, but blood and catheter tip cultures were negative. Chest and abdominal CT scans showed no new disease process. Urine and sputum cultures were collected and were negative for infection. Drug-induced fever was then suspected, but was ruled out when the fever persisted after the removal of potential offending agents (phenytoin, nebivolol, and cyclophosphamide).

THE DIAGNOSIS

We then followed the American Academy of Family Physicians’ diagnostic protocol for FUO.¹

Initial labs included a complete blood count (CBC), 2 blood cultures, a urine culture, erythrocyte sedimentation rate (ESR), a purified protein derivative skin test, chest and abdominal CT scans, and double-stranded DNA (dsDNA) levels (since this patient had known SLE). The patient’s hemoglobin level and mean corpuscular volume were consistent with normocytic anemia, which was attributed to the ESRD. The ESR was mildly elevated at 46 mm/hr, but dsDNA was not, ruling out a lupus flare. Thrombocytopenia (platelet count, 82 K/mcL) and lymphocytopenia (absolute lymphocyte count, 0.2 K/mcL) were assumed to be secondary to cyclophosphamide use.

Because the initial labs were non-diagnostic, we proceeded with a sputum stain and culture, human immunodeficiency virus testing, a hepatitis panel, and a peripheral blood smear.¹ All were negative except for the peripheral blood smear, which showed hemophagocytic cells. This was the first finding that brought hemophagocytic lymphohistiocytosis (HLH) into the differential.

We then performed a bone marrow biopsy (FIGURE), which also revealed hemophagocytic cells, so we ordered HLH-specific labs (more on those in a bit). Liver enzymes were elevated to 3 times their normal value. Triglycerides (414 mg/dL), ferritin (>15,000 ng/mL), and interleukin-2 (IL-2) receptor levels (>20,000 pg/m) were also elevated.

The patient was tested for herpes simplex virus, Epstein-Barr virus (EBV), and cytomegalovirus (CMV), since these viruses are associated with HLH. She had 3.1 million copies/mL of CMV, leading to the diagnosis of secondary HLH. This diagnosis might not have been made if not for a persistent fever investigation.

HLH is a life-threatening syndrome of excessive immune activation that results in tissue damage.² There are primary and secondary forms, but they share the same mechanism of impaired regulation of cytotoxic granules and cytokines. Primary HLH results from a congenital gene mutation,³ while secondary HLH is triggered by an autoimmune or inflammatory disease or an infection.⁴ EBV is the most common viral etiology, followed closely by CMV.⁵

The diagnosis may be established genetically (based on mutations of the genes loci PRF1, UNC13D, or STX11) or by fulfillment of 5 out of 8 criteria: fever; splenomegaly; cytopenia; hypertriglyceridemia; hypofibrinogenemia; hemophagocytosis in the bone marrow, spleen, or lymph nodes; low or absent natural killer cell activity; and an elevated ferritin level (>500 ng/mL). Elevated soluble CD25 and IL-2 receptor markers are HLH-specific markers.³ This patient had fever, cytopenia, hypertriglyceridemia, hemophagocytosis, and elevated ferritin with elevated IL-2, meeting the criteria for secondary HLH.

First treat the underlying condition, then the HLH

Treatment for HLH includes treating the underlying condition (such as EBV or CMV) with antiretroviral medications, and using immunosuppressive agents such as chemotherapy drugs and steroids for the HLH.

Our patient was treated with valganciclovir 900 mg/d for 2 weeks for the CMV and an etoposide/prednisone taper for 3 months for HLH chemotherapy and suppression. Within one month, her CMV viral load decreased to <300 copies/mL and her fever resolved. Ferritin, triglycerides, and liver enzyme levels returned to normal within 3 months.

THE TAKEAWAY

FUO can be frustrating for both the physician and the patient. Not only is the differential large, but testing is extensive. It is important to get a thorough history and to consider medications as the cause. Testing should be patient-specific and systematic. Persistent investigation is critical to saving the patient’s life.

THE CASE

A 47-year-old African American woman was admitted to the hospital with pulmonary edema revealed on a computed tomography (CT) scan. She had a history of systemic lupus erythematosus (SLE), hypertension, and end-stage renal disease (ESRD). The patient had been hospitalized one month earlier for lupus nephritis with a hypertensive emergency that led to a seizure. During this earlier hospitalization, she was given a diagnosis of posterior reversible encephalopathy syndrome.

Two weeks into her more recent hospitalization, the patient developed a fever that was accompanied by cough and fatigue. By the third week, there was no identified cause of the fever, and the patient met the criteria for fever of unknown origin (FUO).

Her medications included cyclophosphamide, prednisone, nebivolol, clonidine, phenytoin, and epoetin alfa. The patient was also receiving dialysis every other day. Chest x-ray findings suggested pneumonia, and the patient was treated with vancomycin and piperacillin/tazobactam. However, her fever persisted after completing the antibiotics. Central line sepsis was high in the differential, as the patient was on dialysis, but blood and catheter tip cultures were negative. Chest and abdominal CT scans showed no new disease process. Urine and sputum cultures were collected and were negative for infection. Drug-induced fever was then suspected, but was ruled out when the fever persisted after the removal of potential offending agents (phenytoin, nebivolol, and cyclophosphamide).

THE DIAGNOSIS

We then followed the American Academy of Family Physicians’ diagnostic protocol for FUO.¹

Initial labs included a complete blood count (CBC), 2 blood cultures, a urine culture, erythrocyte sedimentation rate (ESR), a purified protein derivative skin test, chest and abdominal CT scans, and double-stranded DNA (dsDNA) levels (since this patient had known SLE). The patient’s hemoglobin level and mean corpuscular volume were consistent with normocytic anemia, which was attributed to the ESRD. The ESR was mildly elevated at 46 mm/hr, but dsDNA was not, ruling out a lupus flare. Thrombocytopenia (platelet count, 82 K/mcL) and lymphocytopenia (absolute lymphocyte count, 0.2 K/mcL) were assumed to be secondary to cyclophosphamide use.

Because the initial labs were non-diagnostic, we proceeded with a sputum stain and culture, human immunodeficiency virus testing, a hepatitis panel, and a peripheral blood smear.¹ All were negative except for the peripheral blood smear, which showed hemophagocytic cells. This was the first finding that brought hemophagocytic lymphohistiocytosis (HLH) into the differential.

We then performed a bone marrow biopsy (FIGURE), which also revealed hemophagocytic cells, so we ordered HLH-specific labs (more on those in a bit). Liver enzymes were elevated to 3 times their normal value. Triglycerides (414 mg/dL), ferritin (>15,000 ng/mL), and interleukin-2 (IL-2) receptor levels (>20,000 pg/m) were also elevated.

The patient was tested for herpes simplex virus, Epstein-Barr virus (EBV), and cytomegalovirus (CMV), since these viruses are associated with HLH. She had 3.1 million copies/mL of CMV, leading to the diagnosis of secondary HLH. This diagnosis might not have been made if not for a persistent fever investigation.

HLH is a life-threatening syndrome of excessive immune activation that results in tissue damage.² There are primary and secondary forms, but they share the same mechanism of impaired regulation of cytotoxic granules and cytokines. Primary HLH results from a congenital gene mutation,³ while secondary HLH is triggered by an autoimmune or inflammatory disease or an infection.⁴ EBV is the most common viral etiology, followed closely by CMV.⁵

The diagnosis may be established genetically (based on mutations of the genes loci PRF1, UNC13D, or STX11) or by fulfillment of 5 out of 8 criteria: fever; splenomegaly; cytopenia; hypertriglyceridemia; hypofibrinogenemia; hemophagocytosis in the bone marrow, spleen, or lymph nodes; low or absent natural killer cell activity; and an elevated ferritin level (>500 ng/mL). Elevated soluble CD25 and IL-2 receptor markers are HLH-specific markers.³ This patient had fever, cytopenia, hypertriglyceridemia, hemophagocytosis, and elevated ferritin with elevated IL-2, meeting the criteria for secondary HLH.

First treat the underlying condition, then the HLH

Treatment for HLH includes treating the underlying condition (such as EBV or CMV) with antiretroviral medications, and using immunosuppressive agents such as chemotherapy drugs and steroids for the HLH.

Our patient was treated with valganciclovir 900 mg/d for 2 weeks for the CMV and an etoposide/prednisone taper for 3 months for HLH chemotherapy and suppression. Within one month, her CMV viral load decreased to <300 copies/mL and her fever resolved. Ferritin, triglycerides, and liver enzyme levels returned to normal within 3 months.

THE TAKEAWAY

FUO can be frustrating for both the physician and the patient. Not only is the differential large, but testing is extensive. It is important to get a thorough history and to consider medications as the cause. Testing should be patient-specific and systematic. Persistent investigation is critical to saving the patient’s life.

THE CASE

A 47-year-old African American woman was admitted to the hospital with pulmonary edema revealed on a computed tomography (CT) scan. She had a history of systemic lupus erythematosus (SLE), hypertension, and end-stage renal disease (ESRD). The patient had been hospitalized one month earlier for lupus nephritis with a hypertensive emergency that led to a seizure. During this earlier hospitalization, she was given a diagnosis of posterior reversible encephalopathy syndrome.

Two weeks into her more recent hospitalization, the patient developed a fever that was accompanied by cough and fatigue. By the third week, there was no identified cause of the fever, and the patient met the criteria for fever of unknown origin (FUO).

Her medications included cyclophosphamide, prednisone, nebivolol, clonidine, phenytoin, and epoetin alfa. The patient was also receiving dialysis every other day. Chest x-ray findings suggested pneumonia, and the patient was treated with vancomycin and piperacillin/tazobactam. However, her fever persisted after completing the antibiotics. Central line sepsis was high in the differential, as the patient was on dialysis, but blood and catheter tip cultures were negative. Chest and abdominal CT scans showed no new disease process. Urine and sputum cultures were collected and were negative for infection. Drug-induced fever was then suspected, but was ruled out when the fever persisted after the removal of potential offending agents (phenytoin, nebivolol, and cyclophosphamide).

THE DIAGNOSIS

We then followed the American Academy of Family Physicians’ diagnostic protocol for FUO.¹

Initial labs included a complete blood count (CBC), 2 blood cultures, a urine culture, erythrocyte sedimentation rate (ESR), a purified protein derivative skin test, chest and abdominal CT scans, and double-stranded DNA (dsDNA) levels (since this patient had known SLE). The patient’s hemoglobin level and mean corpuscular volume were consistent with normocytic anemia, which was attributed to the ESRD. The ESR was mildly elevated at 46 mm/hr, but dsDNA was not, ruling out a lupus flare. Thrombocytopenia (platelet count, 82 K/mcL) and lymphocytopenia (absolute lymphocyte count, 0.2 K/mcL) were assumed to be secondary to cyclophosphamide use.

Because the initial labs were non-diagnostic, we proceeded with a sputum stain and culture, human immunodeficiency virus testing, a hepatitis panel, and a peripheral blood smear.¹ All were negative except for the peripheral blood smear, which showed hemophagocytic cells. This was the first finding that brought hemophagocytic lymphohistiocytosis (HLH) into the differential.

We then performed a bone marrow biopsy (FIGURE), which also revealed hemophagocytic cells, so we ordered HLH-specific labs (more on those in a bit). Liver enzymes were elevated to 3 times their normal value. Triglycerides (414 mg/dL), ferritin (>15,000 ng/mL), and interleukin-2 (IL-2) receptor levels (>20,000 pg/m) were also elevated.

The patient was tested for herpes simplex virus, Epstein-Barr virus (EBV), and cytomegalovirus (CMV), since these viruses are associated with HLH. She had 3.1 million copies/mL of CMV, leading to the diagnosis of secondary HLH. This diagnosis might not have been made if not for a persistent fever investigation.

HLH is a life-threatening syndrome of excessive immune activation that results in tissue damage.² There are primary and secondary forms, but they share the same mechanism of impaired regulation of cytotoxic granules and cytokines. Primary HLH results from a congenital gene mutation,³ while secondary HLH is triggered by an autoimmune or inflammatory disease or an infection.⁴ EBV is the most common viral etiology, followed closely by CMV.⁵

The diagnosis may be established genetically (based on mutations of the genes loci PRF1, UNC13D, or STX11) or by fulfillment of 5 out of 8 criteria: fever; splenomegaly; cytopenia; hypertriglyceridemia; hypofibrinogenemia; hemophagocytosis in the bone marrow, spleen, or lymph nodes; low or absent natural killer cell activity; and an elevated ferritin level (>500 ng/mL). Elevated soluble CD25 and IL-2 receptor markers are HLH-specific markers.³ This patient had fever, cytopenia, hypertriglyceridemia, hemophagocytosis, and elevated ferritin with elevated IL-2, meeting the criteria for secondary HLH.

First treat the underlying condition, then the HLH

Treatment for HLH includes treating the underlying condition (such as EBV or CMV) with antiretroviral medications, and using immunosuppressive agents such as chemotherapy drugs and steroids for the HLH.

Our patient was treated with valganciclovir 900 mg/d for 2 weeks for the CMV and an etoposide/prednisone taper for 3 months for HLH chemotherapy and suppression. Within one month, her CMV viral load decreased to <300 copies/mL and her fever resolved. Ferritin, triglycerides, and liver enzyme levels returned to normal within 3 months.

THE TAKEAWAY

FUO can be frustrating for both the physician and the patient. Not only is the differential large, but testing is extensive. It is important to get a thorough history and to consider medications as the cause. Testing should be patient-specific and systematic. Persistent investigation is critical to saving the patient’s life.

CASE › Ms. B, a 26-year-old woman, presents to your office with her 3-year-old son for a well-child examination. During the course of the conversation, she asks you if she should be giving her child probiotics to improve his general health. Many of her friends, who also have their children in day care, have told her that probiotics, “are nature’s way of fighting infection.” Her son currently takes no medications, and has no history of asthma or recent gastrointestinal disturbances. He was treated for 2 ear infections last winter, approximately 3 months apart. His physical exam is normal and, after today, his immunizations will be up to date. How should you respond?

The use of probiotics as over-the-counter treatments for a variety of conditions continues to grow, with retail sales of functional probiotic foods and supplements topping $35 billion worldwide in 2014.¹ In children, claims of benefit for gastrointestinal (GI) disorders, colic, and allergy prevention, as well as prevention and treatment of upper respiratory infections (URIs) have existed for over 10 years.^2-4 The human gut flora develops rapidly after birth and is known to be influenced by route of delivery (vaginal vs cesarean), type of feeding (breast vs formula), and other environmental factors.⁵ The use of probiotics to influence the types of bacteria in a child’s intestinal tract continues to be an area of active research. (For more on probiotic formulations, see TABLE 1.)

This article summarizes recent research on probiotic use in infants and children. New data support the use of probiotics for the treatment of colic and atopic eczema; however, the data on using probiotics in the management of URIs is less robust and mixed. And while probiotics improve irritable bowel syndrome (IBS) stomach pain, they do not help with related diarrhea or constipation. All of these data are summarized in TABLE 2.^6-29

L reuteri improves symptoms in breastfed infants with colic

Infant colic is a relatively common condition known to negatively impact maternal mental health and the mother/child relationship.⁶ Numerous randomized controlled trials (RCTs) over the years have demonstrated mixed results with using probiotics to decrease crying times, with differences noted between infants who are solely breastfed and those who are not.⁷

In the most recent meta-analysis of 6 studies (n=427) that focused only on the probiotic Lactobacillus reuteri, breastfed infants with colic receiving a daily dose of 10⁸ colony forming units (CFU) cried an average of 56 fewer minutes/day than those in the control group (95% confidence interval [CI], -64.4 to -47.3; P=.001) at day 21 of treatment.⁸ Although 2 studies in this meta-analysis included a small number of mixed-fed and formula-fed infants, the majority of trials do not show benefit for these infants. Trials assessing the use of L reuteri for prevention of colic have not shown positive results.⁷

Probiotics may help prevent and shorten the course of URIs

The mechanisms by which probiotics may prevent or shorten the course of URIs are not obvious. Current theories include boosting the immune function of the respiratory mucosa, acting as a competitive inhibitor for viruses, and secreting antiviral compounds.⁹ Multiple reviews published in the last 3 years, however, add to the evidence that the apparent benefit is real.

A 2013 meta-analysis assessed data from 4 RCTs (N=1805), which used Lactobacillus rhamnosus as the sole probiotic for prevention of URIs. In treated children, otitis media incidence was reduced by 24% (relative risk [RR] 0.76; 95% CI, 0.64-0.91) and risk of URI was reduced by 38% (RR 0.62; 95% CI, 0.50-0.78).¹⁰ The number needed to treat (NNT) was 4 for URI prevention, and the authors noted that adverse events were similar in the treatment and control groups.

A 2014 systematic review and meta-analysis of 20 RCTs examining duration of illness included 10 studies dedicated to pediatric subjects (age 12 months to 12 years).¹¹ There were significantly fewer days of illness per person (standardized mean difference -0.31; 95% CI, -0.41 to -0.11) and each illness episode was shorter by three-quarters of a day (weighted mean difference -0.77; 95% CI, -1.5 to -0.04) in participants who received a probiotic vs those who received a placebo. Probiotics used in these studies belonged to the Lactobacillus and Bifidobacterium genera.

A 2015 systematic review of 14 RCTs assessing the benefits of probiotics, particularly Lactobacillus and Bifidobacterium strains, on URI occurrence and symptoms, showed mixed results.¹² Seven of 12 studies found lowered rates of URI and otitis media incidence, 7 of 11 RCTs reported a significant reduction in severity scores for URI, and 4 of 8 RCTs reported significant reductions in school absenteeism between the probiotic and control groups. In a summary statement, the authors noted that “at least one beneficial effect of prophylactic probiotics was observed in the majority of RCTs,” and that “none of the studies reported any serious adverse events.”

Perinatal probiotics: No benefit for allergic conditions—except eczema

Allergic disease is on the rise and continues to plague children with reduced quality of life, potentially life-threatening reactions, and missed activities, including school. The gut microbiome likely influences a child’s allergic propensity through its effects on T-helper cells, transforming growth factor (TGF), and immunoglobulin A (IgA)—all known components of the allergic response. As the hygiene hypothesis suggests, the quantity and types of bacteria that inhabit the GI tract early in life play a significant role in determining a person’s later allergic responses.¹³

In a 2013 meta-analysis of 20 trials (N=4866), researchers looked specifically at probiotic use and the diagnosis of asthma and incident wheezing. Single and combination products of Lactobacillus and Bifidobacterium given prenatally and/or postnatally were included in the studies. The authors found no evidence to support a protective association between perinatal use of probiotics and diagnosed asthma (RR=0.99; 95% CI, 0.81-0.21) or childhood incident wheezing (RR=0.97; 95% CI, 0.87-1.09; 9 trials, 1949 infants).¹⁴

In a more recent meta-analysis (2015) conducted to inform the World Allergy Organization, 29 studies were evaluated to assess the impact of probiotics on allergic symptoms of the skin, respiratory system, and GI tract.¹⁵ No significant benefit was noted for any allergic condition except for eczema. Probiotics reduced the risk of eczema when given during the last trimester of pregnancy (RR=0.71; 95% CI, 0.60-0.84), when used by breastfeeding mothers (RR=0.57; 95% CI, 0.47-0.69), and when given to infants (RR=0.80; 95% CI, 0.68-0.94).

Lactobacillus reuteri decreased crying in breastfed infants with colic by nearly an hour a day.A 2014 systematic review and meta-analysis (N=2797) explored probiotic use specifically for the prevention of eczema.¹⁶ The pooled relative risk for all the studies was 0.74 (95% CI, 0.67-0.82). Evidence was strongest for probiotics containing the Lactobacillus species rhamnosus and paracasei, as well as for Bifidobacterium lactis. No benefit was noted with Lactobacillus acidophilus or other Bifidobacterium species. These newer reviews on eczema prevention contrast with an older Cochrane review published in 2008 (12 RCTs, N=781), which did not show significant benefit for the treatment of eczema.¹⁷

Probiotics improve IBS stomach pain, but not diarrhea or constipation

IBS is a functional disorder of the GI tract that affects up to 20% of children and teenagers and leads to a significant decrease in quality of life.¹⁸ Current theories of causation include bacterial overgrowth and neuronal hyperactivity, which may be amenable to change with supplemental probiotics.

A 2015 systematic review of non-pharmacological treatments for functional abdominal pain disorders identified 4 studies dedicated to IBS in children.¹⁹ A subgroup analysis of 3 RCTs (n=309) that looked at giving L rhamnosus to 5- to 17-year-olds with IBS showed improved abdominal pain (according to various pain scales) compared to the placebo group. Study participants received at least 3 x 10⁹ CFU twice a day for 4 to 8 weeks. Relative risk for improvement was 1.7 (95% CI, 1.27-2.27) with an NNT of 4. None of these studies showed significant improvement in either frequency or severity of diarrhea or constipation.

A separate crossover RCT (N=59) compared placebo to VSL#3, a product containing 8 probiotics (Bifidobacterium breve, Bifidobacterium longum, Bifidobacterium infantis, L acidophilus, Lactobacillus plantarum, L paracasei, Lactobacillus bulgaricus, and Streptococcus hermophiles), given in age-dependent doses for 6 weeks to children aged 4 to 18 years.²⁰ The frequency and intensity of abdominal pain were measured on a 5-point Likert scale. The group treated with VSL#3 dropped 1.0 ± 0.2 points vs 0.5 ± 0.2 points in the control group (P<.05) and reported an improved quality of life.

These agents reduce antibiotic-associated diarrhea

Antibiotic-associated diarrhea (AAD) occurs in 5% to 30% of children who receive antibiotic therapy.²¹ It occurs most frequently with the use of cephalosporins, penicillin, fluoroquinolones, and clindamycin, and is likely caused by an alteration of the normal gut flora. Colitis caused by Clostridium difficile remains the most serious antibiotic-associated GI complication.

A systematic review of the specific probiotic Saccharomyces boulardii conducted in 2015 analyzed data from 6 RCTs (n=1653) to determine the effect of co-administration of this probiotic with antibiotics.²² The pooled relative risk for AAD in children receiving the probiotic was 0.43 (95% CI, 0.3-0.6) compared to antibiotics alone. The absolute risk of AAD dropped from 20.9% to 8.8%, translating to a NNT of 8. Two of the RCTs specifically looked at rates of C difficile infection (n=579). C difficile infection rates dropped by 75% (RR=.25; 95% CI, 0.08-0.73) in the treatment group. This dramatic treatment effect was not seen in studies involving adults.

A similar systematic review focusing on L rhamnosus conducted in 2015 pooled data from 5 RCTs (n=445) to see if the probiotic would decrease AAD in children if it was co-administered with antibiotics.²³ The relative risk for AAD in this treatment group was 0.48 (95% CI, 0.26-0.89) with an absolute risk reduction of 13.4% (23% compared to 9.6%), translating to an NNT of 7.

A Cochrane review published in 2015 included 23 studies (N=3938) and found similar results with an RR for AAD of 0.46 for treated children (95% CI, 0.35-0.61).²⁴ Doses of probiotics ranged from 5 to 40 billion CFU/day. Although many probiotic species were used in these studies, S boulardii and L rhamnosus were cited as having the strongest data to support use in this context.

Probiotics reduce the duration, frequency of acute infectious diarrhea

Diarrhea remains the second leading cause of death among children one to 59 months of age worldwide.²⁵ Current World Health Organization recommendations include oral rehydration salts, continued feeding to avoid dehydration, and zinc to decrease the duration and severity of illness.²⁶ Multiple studies in adults confirm that a variety of probiotics decrease both the duration and severity of diarrhea in acute gastroenteritis.²⁷

Lactobacillus rhamnosus reduced the incidence of otitis media and upper respiratory infections by 24% and 38%, respectively.The authors of a 2013 systematic review of probiotics for the treatment of community-acquired acute diarrhea in children less than 5 years of age analyzed data from 8 RCTs (N=1755).²⁸ Various probiotics were used including Lactobacillus species, Streptococcus thermophilus, Bifidobacterium species, and Saccharomyces boulardii for between 4 and 10 days. Six of these studies (n=1164) measured diarrhea duration and found a 14% reduction (95% CI, 3.8%-24.2%) in days of illness for those children treated vs those receiving placebo. Five studies (n=925) measured the difference in stool frequency on Day 2 of illness and reported a reduction of 13.1% (95% CI, 0.8%-5.3%) in the number of stools in the treated group vs the placebo group.

This review augments a Cochrane meta-analysis of 63 studies (N=8014) published in 2010.²⁷ Fifty-six of these studies included infants and children. Pooled analysis of the varied probiotic treatments showed a mean reduction in duration of diarrhea of just over a day (24.76 hours; 95% CI, 15.9-33.6 hours; n=4555, trials=35) and decreased stool frequency on Day 2 of treatment (mean difference 0.80; 95% CI, 0.45-1.14; n=2751, trials=20). The authors concluded that probiotics “have clear beneficial effects in shortening the duration and reducing stool frequency in acute infectious diarrhea.”

Pediatric society weighs in. In 2014, the European Society for Paediatric Gastroenterology, Hepatology, and Nutrition issued guidelines regarding probiotic use for the treatment of acute gastroenteritis.²⁹ In addition to rehydration therapy, these guidelines recommend the use of L rhamnosus and/or S boulardii as first-line treatments. Lower quality evidence is available for the use of L reuteri.

CASE › In response to Ms. B’s query about starting her young son on probiotics, you tell her that studies have shown that probiotics are safe for children when given in appropriate doses. They have been shown to help children recover from diarrheal illnesses and can help reduce the number of colds and ear infections when taken regularly. The reason you are giving them determines which strains you should use. You recommend giving her child a formulation of probiotic that contains Lactobacillus or Bifidobacterium with a dose range of 2 to 10 billion CFUs taken daily to reduce the risk of her child getting another ear infection.

CORRESPONDENCE
Paul Dassow, MD, MSPH, 1100 E. 3rd St, Chattanooga, TN 37403; [email protected].

CASE › Ms. B, a 26-year-old woman, presents to your office with her 3-year-old son for a well-child examination. During the course of the conversation, she asks you if she should be giving her child probiotics to improve his general health. Many of her friends, who also have their children in day care, have told her that probiotics, “are nature’s way of fighting infection.” Her son currently takes no medications, and has no history of asthma or recent gastrointestinal disturbances. He was treated for 2 ear infections last winter, approximately 3 months apart. His physical exam is normal and, after today, his immunizations will be up to date. How should you respond?

The use of probiotics as over-the-counter treatments for a variety of conditions continues to grow, with retail sales of functional probiotic foods and supplements topping $35 billion worldwide in 2014.¹ In children, claims of benefit for gastrointestinal (GI) disorders, colic, and allergy prevention, as well as prevention and treatment of upper respiratory infections (URIs) have existed for over 10 years.^2-4 The human gut flora develops rapidly after birth and is known to be influenced by route of delivery (vaginal vs cesarean), type of feeding (breast vs formula), and other environmental factors.⁵ The use of probiotics to influence the types of bacteria in a child’s intestinal tract continues to be an area of active research. (For more on probiotic formulations, see TABLE 1.)

This article summarizes recent research on probiotic use in infants and children. New data support the use of probiotics for the treatment of colic and atopic eczema; however, the data on using probiotics in the management of URIs is less robust and mixed. And while probiotics improve irritable bowel syndrome (IBS) stomach pain, they do not help with related diarrhea or constipation. All of these data are summarized in TABLE 2.^6-29

L reuteri improves symptoms in breastfed infants with colic

Infant colic is a relatively common condition known to negatively impact maternal mental health and the mother/child relationship.⁶ Numerous randomized controlled trials (RCTs) over the years have demonstrated mixed results with using probiotics to decrease crying times, with differences noted between infants who are solely breastfed and those who are not.⁷

In the most recent meta-analysis of 6 studies (n=427) that focused only on the probiotic Lactobacillus reuteri, breastfed infants with colic receiving a daily dose of 10⁸ colony forming units (CFU) cried an average of 56 fewer minutes/day than those in the control group (95% confidence interval [CI], -64.4 to -47.3; P=.001) at day 21 of treatment.⁸ Although 2 studies in this meta-analysis included a small number of mixed-fed and formula-fed infants, the majority of trials do not show benefit for these infants. Trials assessing the use of L reuteri for prevention of colic have not shown positive results.⁷

Probiotics may help prevent and shorten the course of URIs

The mechanisms by which probiotics may prevent or shorten the course of URIs are not obvious. Current theories include boosting the immune function of the respiratory mucosa, acting as a competitive inhibitor for viruses, and secreting antiviral compounds.⁹ Multiple reviews published in the last 3 years, however, add to the evidence that the apparent benefit is real.

A 2013 meta-analysis assessed data from 4 RCTs (N=1805), which used Lactobacillus rhamnosus as the sole probiotic for prevention of URIs. In treated children, otitis media incidence was reduced by 24% (relative risk [RR] 0.76; 95% CI, 0.64-0.91) and risk of URI was reduced by 38% (RR 0.62; 95% CI, 0.50-0.78).¹⁰ The number needed to treat (NNT) was 4 for URI prevention, and the authors noted that adverse events were similar in the treatment and control groups.

A 2014 systematic review and meta-analysis of 20 RCTs examining duration of illness included 10 studies dedicated to pediatric subjects (age 12 months to 12 years).¹¹ There were significantly fewer days of illness per person (standardized mean difference -0.31; 95% CI, -0.41 to -0.11) and each illness episode was shorter by three-quarters of a day (weighted mean difference -0.77; 95% CI, -1.5 to -0.04) in participants who received a probiotic vs those who received a placebo. Probiotics used in these studies belonged to the Lactobacillus and Bifidobacterium genera.

A 2015 systematic review of 14 RCTs assessing the benefits of probiotics, particularly Lactobacillus and Bifidobacterium strains, on URI occurrence and symptoms, showed mixed results.¹² Seven of 12 studies found lowered rates of URI and otitis media incidence, 7 of 11 RCTs reported a significant reduction in severity scores for URI, and 4 of 8 RCTs reported significant reductions in school absenteeism between the probiotic and control groups. In a summary statement, the authors noted that “at least one beneficial effect of prophylactic probiotics was observed in the majority of RCTs,” and that “none of the studies reported any serious adverse events.”

Perinatal probiotics: No benefit for allergic conditions—except eczema

Allergic disease is on the rise and continues to plague children with reduced quality of life, potentially life-threatening reactions, and missed activities, including school. The gut microbiome likely influences a child’s allergic propensity through its effects on T-helper cells, transforming growth factor (TGF), and immunoglobulin A (IgA)—all known components of the allergic response. As the hygiene hypothesis suggests, the quantity and types of bacteria that inhabit the GI tract early in life play a significant role in determining a person’s later allergic responses.¹³

In a 2013 meta-analysis of 20 trials (N=4866), researchers looked specifically at probiotic use and the diagnosis of asthma and incident wheezing. Single and combination products of Lactobacillus and Bifidobacterium given prenatally and/or postnatally were included in the studies. The authors found no evidence to support a protective association between perinatal use of probiotics and diagnosed asthma (RR=0.99; 95% CI, 0.81-0.21) or childhood incident wheezing (RR=0.97; 95% CI, 0.87-1.09; 9 trials, 1949 infants).¹⁴

In a more recent meta-analysis (2015) conducted to inform the World Allergy Organization, 29 studies were evaluated to assess the impact of probiotics on allergic symptoms of the skin, respiratory system, and GI tract.¹⁵ No significant benefit was noted for any allergic condition except for eczema. Probiotics reduced the risk of eczema when given during the last trimester of pregnancy (RR=0.71; 95% CI, 0.60-0.84), when used by breastfeeding mothers (RR=0.57; 95% CI, 0.47-0.69), and when given to infants (RR=0.80; 95% CI, 0.68-0.94).

Lactobacillus reuteri decreased crying in breastfed infants with colic by nearly an hour a day.A 2014 systematic review and meta-analysis (N=2797) explored probiotic use specifically for the prevention of eczema.¹⁶ The pooled relative risk for all the studies was 0.74 (95% CI, 0.67-0.82). Evidence was strongest for probiotics containing the Lactobacillus species rhamnosus and paracasei, as well as for Bifidobacterium lactis. No benefit was noted with Lactobacillus acidophilus or other Bifidobacterium species. These newer reviews on eczema prevention contrast with an older Cochrane review published in 2008 (12 RCTs, N=781), which did not show significant benefit for the treatment of eczema.¹⁷

Probiotics improve IBS stomach pain, but not diarrhea or constipation

IBS is a functional disorder of the GI tract that affects up to 20% of children and teenagers and leads to a significant decrease in quality of life.¹⁸ Current theories of causation include bacterial overgrowth and neuronal hyperactivity, which may be amenable to change with supplemental probiotics.

A 2015 systematic review of non-pharmacological treatments for functional abdominal pain disorders identified 4 studies dedicated to IBS in children.¹⁹ A subgroup analysis of 3 RCTs (n=309) that looked at giving L rhamnosus to 5- to 17-year-olds with IBS showed improved abdominal pain (according to various pain scales) compared to the placebo group. Study participants received at least 3 x 10⁹ CFU twice a day for 4 to 8 weeks. Relative risk for improvement was 1.7 (95% CI, 1.27-2.27) with an NNT of 4. None of these studies showed significant improvement in either frequency or severity of diarrhea or constipation.

A separate crossover RCT (N=59) compared placebo to VSL#3, a product containing 8 probiotics (Bifidobacterium breve, Bifidobacterium longum, Bifidobacterium infantis, L acidophilus, Lactobacillus plantarum, L paracasei, Lactobacillus bulgaricus, and Streptococcus hermophiles), given in age-dependent doses for 6 weeks to children aged 4 to 18 years.²⁰ The frequency and intensity of abdominal pain were measured on a 5-point Likert scale. The group treated with VSL#3 dropped 1.0 ± 0.2 points vs 0.5 ± 0.2 points in the control group (P<.05) and reported an improved quality of life.

These agents reduce antibiotic-associated diarrhea

Antibiotic-associated diarrhea (AAD) occurs in 5% to 30% of children who receive antibiotic therapy.²¹ It occurs most frequently with the use of cephalosporins, penicillin, fluoroquinolones, and clindamycin, and is likely caused by an alteration of the normal gut flora. Colitis caused by Clostridium difficile remains the most serious antibiotic-associated GI complication.

A systematic review of the specific probiotic Saccharomyces boulardii conducted in 2015 analyzed data from 6 RCTs (n=1653) to determine the effect of co-administration of this probiotic with antibiotics.²² The pooled relative risk for AAD in children receiving the probiotic was 0.43 (95% CI, 0.3-0.6) compared to antibiotics alone. The absolute risk of AAD dropped from 20.9% to 8.8%, translating to a NNT of 8. Two of the RCTs specifically looked at rates of C difficile infection (n=579). C difficile infection rates dropped by 75% (RR=.25; 95% CI, 0.08-0.73) in the treatment group. This dramatic treatment effect was not seen in studies involving adults.

A similar systematic review focusing on L rhamnosus conducted in 2015 pooled data from 5 RCTs (n=445) to see if the probiotic would decrease AAD in children if it was co-administered with antibiotics.²³ The relative risk for AAD in this treatment group was 0.48 (95% CI, 0.26-0.89) with an absolute risk reduction of 13.4% (23% compared to 9.6%), translating to an NNT of 7.

A Cochrane review published in 2015 included 23 studies (N=3938) and found similar results with an RR for AAD of 0.46 for treated children (95% CI, 0.35-0.61).²⁴ Doses of probiotics ranged from 5 to 40 billion CFU/day. Although many probiotic species were used in these studies, S boulardii and L rhamnosus were cited as having the strongest data to support use in this context.

Probiotics reduce the duration, frequency of acute infectious diarrhea

Diarrhea remains the second leading cause of death among children one to 59 months of age worldwide.²⁵ Current World Health Organization recommendations include oral rehydration salts, continued feeding to avoid dehydration, and zinc to decrease the duration and severity of illness.²⁶ Multiple studies in adults confirm that a variety of probiotics decrease both the duration and severity of diarrhea in acute gastroenteritis.²⁷

Lactobacillus rhamnosus reduced the incidence of otitis media and upper respiratory infections by 24% and 38%, respectively.The authors of a 2013 systematic review of probiotics for the treatment of community-acquired acute diarrhea in children less than 5 years of age analyzed data from 8 RCTs (N=1755).²⁸ Various probiotics were used including Lactobacillus species, Streptococcus thermophilus, Bifidobacterium species, and Saccharomyces boulardii for between 4 and 10 days. Six of these studies (n=1164) measured diarrhea duration and found a 14% reduction (95% CI, 3.8%-24.2%) in days of illness for those children treated vs those receiving placebo. Five studies (n=925) measured the difference in stool frequency on Day 2 of illness and reported a reduction of 13.1% (95% CI, 0.8%-5.3%) in the number of stools in the treated group vs the placebo group.

This review augments a Cochrane meta-analysis of 63 studies (N=8014) published in 2010.²⁷ Fifty-six of these studies included infants and children. Pooled analysis of the varied probiotic treatments showed a mean reduction in duration of diarrhea of just over a day (24.76 hours; 95% CI, 15.9-33.6 hours; n=4555, trials=35) and decreased stool frequency on Day 2 of treatment (mean difference 0.80; 95% CI, 0.45-1.14; n=2751, trials=20). The authors concluded that probiotics “have clear beneficial effects in shortening the duration and reducing stool frequency in acute infectious diarrhea.”

Pediatric society weighs in. In 2014, the European Society for Paediatric Gastroenterology, Hepatology, and Nutrition issued guidelines regarding probiotic use for the treatment of acute gastroenteritis.²⁹ In addition to rehydration therapy, these guidelines recommend the use of L rhamnosus and/or S boulardii as first-line treatments. Lower quality evidence is available for the use of L reuteri.

CASE › In response to Ms. B’s query about starting her young son on probiotics, you tell her that studies have shown that probiotics are safe for children when given in appropriate doses. They have been shown to help children recover from diarrheal illnesses and can help reduce the number of colds and ear infections when taken regularly. The reason you are giving them determines which strains you should use. You recommend giving her child a formulation of probiotic that contains Lactobacillus or Bifidobacterium with a dose range of 2 to 10 billion CFUs taken daily to reduce the risk of her child getting another ear infection.

CORRESPONDENCE
Paul Dassow, MD, MSPH, 1100 E. 3rd St, Chattanooga, TN 37403; [email protected].

CASE › Ms. B, a 26-year-old woman, presents to your office with her 3-year-old son for a well-child examination. During the course of the conversation, she asks you if she should be giving her child probiotics to improve his general health. Many of her friends, who also have their children in day care, have told her that probiotics, “are nature’s way of fighting infection.” Her son currently takes no medications, and has no history of asthma or recent gastrointestinal disturbances. He was treated for 2 ear infections last winter, approximately 3 months apart. His physical exam is normal and, after today, his immunizations will be up to date. How should you respond?

The use of probiotics as over-the-counter treatments for a variety of conditions continues to grow, with retail sales of functional probiotic foods and supplements topping $35 billion worldwide in 2014.¹ In children, claims of benefit for gastrointestinal (GI) disorders, colic, and allergy prevention, as well as prevention and treatment of upper respiratory infections (URIs) have existed for over 10 years.^2-4 The human gut flora develops rapidly after birth and is known to be influenced by route of delivery (vaginal vs cesarean), type of feeding (breast vs formula), and other environmental factors.⁵ The use of probiotics to influence the types of bacteria in a child’s intestinal tract continues to be an area of active research. (For more on probiotic formulations, see TABLE 1.)

This article summarizes recent research on probiotic use in infants and children. New data support the use of probiotics for the treatment of colic and atopic eczema; however, the data on using probiotics in the management of URIs is less robust and mixed. And while probiotics improve irritable bowel syndrome (IBS) stomach pain, they do not help with related diarrhea or constipation. All of these data are summarized in TABLE 2.^6-29

L reuteri improves symptoms in breastfed infants with colic

Infant colic is a relatively common condition known to negatively impact maternal mental health and the mother/child relationship.⁶ Numerous randomized controlled trials (RCTs) over the years have demonstrated mixed results with using probiotics to decrease crying times, with differences noted between infants who are solely breastfed and those who are not.⁷

In the most recent meta-analysis of 6 studies (n=427) that focused only on the probiotic Lactobacillus reuteri, breastfed infants with colic receiving a daily dose of 10⁸ colony forming units (CFU) cried an average of 56 fewer minutes/day than those in the control group (95% confidence interval [CI], -64.4 to -47.3; P=.001) at day 21 of treatment.⁸ Although 2 studies in this meta-analysis included a small number of mixed-fed and formula-fed infants, the majority of trials do not show benefit for these infants. Trials assessing the use of L reuteri for prevention of colic have not shown positive results.⁷

Probiotics may help prevent and shorten the course of URIs

The mechanisms by which probiotics may prevent or shorten the course of URIs are not obvious. Current theories include boosting the immune function of the respiratory mucosa, acting as a competitive inhibitor for viruses, and secreting antiviral compounds.⁹ Multiple reviews published in the last 3 years, however, add to the evidence that the apparent benefit is real.

A 2013 meta-analysis assessed data from 4 RCTs (N=1805), which used Lactobacillus rhamnosus as the sole probiotic for prevention of URIs. In treated children, otitis media incidence was reduced by 24% (relative risk [RR] 0.76; 95% CI, 0.64-0.91) and risk of URI was reduced by 38% (RR 0.62; 95% CI, 0.50-0.78).¹⁰ The number needed to treat (NNT) was 4 for URI prevention, and the authors noted that adverse events were similar in the treatment and control groups.

A 2014 systematic review and meta-analysis of 20 RCTs examining duration of illness included 10 studies dedicated to pediatric subjects (age 12 months to 12 years).¹¹ There were significantly fewer days of illness per person (standardized mean difference -0.31; 95% CI, -0.41 to -0.11) and each illness episode was shorter by three-quarters of a day (weighted mean difference -0.77; 95% CI, -1.5 to -0.04) in participants who received a probiotic vs those who received a placebo. Probiotics used in these studies belonged to the Lactobacillus and Bifidobacterium genera.

A 2015 systematic review of 14 RCTs assessing the benefits of probiotics, particularly Lactobacillus and Bifidobacterium strains, on URI occurrence and symptoms, showed mixed results.¹² Seven of 12 studies found lowered rates of URI and otitis media incidence, 7 of 11 RCTs reported a significant reduction in severity scores for URI, and 4 of 8 RCTs reported significant reductions in school absenteeism between the probiotic and control groups. In a summary statement, the authors noted that “at least one beneficial effect of prophylactic probiotics was observed in the majority of RCTs,” and that “none of the studies reported any serious adverse events.”

Perinatal probiotics: No benefit for allergic conditions—except eczema

Allergic disease is on the rise and continues to plague children with reduced quality of life, potentially life-threatening reactions, and missed activities, including school. The gut microbiome likely influences a child’s allergic propensity through its effects on T-helper cells, transforming growth factor (TGF), and immunoglobulin A (IgA)—all known components of the allergic response. As the hygiene hypothesis suggests, the quantity and types of bacteria that inhabit the GI tract early in life play a significant role in determining a person’s later allergic responses.¹³

In a 2013 meta-analysis of 20 trials (N=4866), researchers looked specifically at probiotic use and the diagnosis of asthma and incident wheezing. Single and combination products of Lactobacillus and Bifidobacterium given prenatally and/or postnatally were included in the studies. The authors found no evidence to support a protective association between perinatal use of probiotics and diagnosed asthma (RR=0.99; 95% CI, 0.81-0.21) or childhood incident wheezing (RR=0.97; 95% CI, 0.87-1.09; 9 trials, 1949 infants).¹⁴

In a more recent meta-analysis (2015) conducted to inform the World Allergy Organization, 29 studies were evaluated to assess the impact of probiotics on allergic symptoms of the skin, respiratory system, and GI tract.¹⁵ No significant benefit was noted for any allergic condition except for eczema. Probiotics reduced the risk of eczema when given during the last trimester of pregnancy (RR=0.71; 95% CI, 0.60-0.84), when used by breastfeeding mothers (RR=0.57; 95% CI, 0.47-0.69), and when given to infants (RR=0.80; 95% CI, 0.68-0.94).

Lactobacillus reuteri decreased crying in breastfed infants with colic by nearly an hour a day.A 2014 systematic review and meta-analysis (N=2797) explored probiotic use specifically for the prevention of eczema.¹⁶ The pooled relative risk for all the studies was 0.74 (95% CI, 0.67-0.82). Evidence was strongest for probiotics containing the Lactobacillus species rhamnosus and paracasei, as well as for Bifidobacterium lactis. No benefit was noted with Lactobacillus acidophilus or other Bifidobacterium species. These newer reviews on eczema prevention contrast with an older Cochrane review published in 2008 (12 RCTs, N=781), which did not show significant benefit for the treatment of eczema.¹⁷

Probiotics improve IBS stomach pain, but not diarrhea or constipation

IBS is a functional disorder of the GI tract that affects up to 20% of children and teenagers and leads to a significant decrease in quality of life.¹⁸ Current theories of causation include bacterial overgrowth and neuronal hyperactivity, which may be amenable to change with supplemental probiotics.

A 2015 systematic review of non-pharmacological treatments for functional abdominal pain disorders identified 4 studies dedicated to IBS in children.¹⁹ A subgroup analysis of 3 RCTs (n=309) that looked at giving L rhamnosus to 5- to 17-year-olds with IBS showed improved abdominal pain (according to various pain scales) compared to the placebo group. Study participants received at least 3 x 10⁹ CFU twice a day for 4 to 8 weeks. Relative risk for improvement was 1.7 (95% CI, 1.27-2.27) with an NNT of 4. None of these studies showed significant improvement in either frequency or severity of diarrhea or constipation.

A separate crossover RCT (N=59) compared placebo to VSL#3, a product containing 8 probiotics (Bifidobacterium breve, Bifidobacterium longum, Bifidobacterium infantis, L acidophilus, Lactobacillus plantarum, L paracasei, Lactobacillus bulgaricus, and Streptococcus hermophiles), given in age-dependent doses for 6 weeks to children aged 4 to 18 years.²⁰ The frequency and intensity of abdominal pain were measured on a 5-point Likert scale. The group treated with VSL#3 dropped 1.0 ± 0.2 points vs 0.5 ± 0.2 points in the control group (P<.05) and reported an improved quality of life.

These agents reduce antibiotic-associated diarrhea

Antibiotic-associated diarrhea (AAD) occurs in 5% to 30% of children who receive antibiotic therapy.²¹ It occurs most frequently with the use of cephalosporins, penicillin, fluoroquinolones, and clindamycin, and is likely caused by an alteration of the normal gut flora. Colitis caused by Clostridium difficile remains the most serious antibiotic-associated GI complication.

A systematic review of the specific probiotic Saccharomyces boulardii conducted in 2015 analyzed data from 6 RCTs (n=1653) to determine the effect of co-administration of this probiotic with antibiotics.²² The pooled relative risk for AAD in children receiving the probiotic was 0.43 (95% CI, 0.3-0.6) compared to antibiotics alone. The absolute risk of AAD dropped from 20.9% to 8.8%, translating to a NNT of 8. Two of the RCTs specifically looked at rates of C difficile infection (n=579). C difficile infection rates dropped by 75% (RR=.25; 95% CI, 0.08-0.73) in the treatment group. This dramatic treatment effect was not seen in studies involving adults.

A similar systematic review focusing on L rhamnosus conducted in 2015 pooled data from 5 RCTs (n=445) to see if the probiotic would decrease AAD in children if it was co-administered with antibiotics.²³ The relative risk for AAD in this treatment group was 0.48 (95% CI, 0.26-0.89) with an absolute risk reduction of 13.4% (23% compared to 9.6%), translating to an NNT of 7.

A Cochrane review published in 2015 included 23 studies (N=3938) and found similar results with an RR for AAD of 0.46 for treated children (95% CI, 0.35-0.61).²⁴ Doses of probiotics ranged from 5 to 40 billion CFU/day. Although many probiotic species were used in these studies, S boulardii and L rhamnosus were cited as having the strongest data to support use in this context.

Probiotics reduce the duration, frequency of acute infectious diarrhea

Diarrhea remains the second leading cause of death among children one to 59 months of age worldwide.²⁵ Current World Health Organization recommendations include oral rehydration salts, continued feeding to avoid dehydration, and zinc to decrease the duration and severity of illness.²⁶ Multiple studies in adults confirm that a variety of probiotics decrease both the duration and severity of diarrhea in acute gastroenteritis.²⁷

Lactobacillus rhamnosus reduced the incidence of otitis media and upper respiratory infections by 24% and 38%, respectively.The authors of a 2013 systematic review of probiotics for the treatment of community-acquired acute diarrhea in children less than 5 years of age analyzed data from 8 RCTs (N=1755).²⁸ Various probiotics were used including Lactobacillus species, Streptococcus thermophilus, Bifidobacterium species, and Saccharomyces boulardii for between 4 and 10 days. Six of these studies (n=1164) measured diarrhea duration and found a 14% reduction (95% CI, 3.8%-24.2%) in days of illness for those children treated vs those receiving placebo. Five studies (n=925) measured the difference in stool frequency on Day 2 of illness and reported a reduction of 13.1% (95% CI, 0.8%-5.3%) in the number of stools in the treated group vs the placebo group.

This review augments a Cochrane meta-analysis of 63 studies (N=8014) published in 2010.²⁷ Fifty-six of these studies included infants and children. Pooled analysis of the varied probiotic treatments showed a mean reduction in duration of diarrhea of just over a day (24.76 hours; 95% CI, 15.9-33.6 hours; n=4555, trials=35) and decreased stool frequency on Day 2 of treatment (mean difference 0.80; 95% CI, 0.45-1.14; n=2751, trials=20). The authors concluded that probiotics “have clear beneficial effects in shortening the duration and reducing stool frequency in acute infectious diarrhea.”

Pediatric society weighs in. In 2014, the European Society for Paediatric Gastroenterology, Hepatology, and Nutrition issued guidelines regarding probiotic use for the treatment of acute gastroenteritis.²⁹ In addition to rehydration therapy, these guidelines recommend the use of L rhamnosus and/or S boulardii as first-line treatments. Lower quality evidence is available for the use of L reuteri.

CASE › In response to Ms. B’s query about starting her young son on probiotics, you tell her that studies have shown that probiotics are safe for children when given in appropriate doses. They have been shown to help children recover from diarrheal illnesses and can help reduce the number of colds and ear infections when taken regularly. The reason you are giving them determines which strains you should use. You recommend giving her child a formulation of probiotic that contains Lactobacillus or Bifidobacterium with a dose range of 2 to 10 billion CFUs taken daily to reduce the risk of her child getting another ear infection.

CORRESPONDENCE
Paul Dassow, MD, MSPH, 1100 E. 3rd St, Chattanooga, TN 37403; [email protected].

ILLUSTRATIVE CASE

A 26-year-old G2P1001 at 35 weeks, 2 days of gestation presents with leakage of clear fluid for the last 2 hours. There is obvious pooling in the vaginal vault, and rupture of membranes is confirmed with appropriate testing. Her cervix is closed, she is not in labor, and tests of fetal well-being are reassuring. She had an uncomplicated vaginal delivery with her first child. How should you manage this situation?

Preterm premature rupture of membranes (PPROM)—when rupture of membranes occurs before 37 weeks’ gestation—affects about 3% of all pregnancies in the United States, and is a major contributor to perinatal morbidity and mortality.^2,3 PPROM management remains controversial, especially during the late preterm stage (ie, 34 weeks to 36 weeks, 6 days). Non-reassuring fetal status, clinical chorioamnionitis, cord prolapse, and significant placental abruption are clear indications for delivery. In the absence of those factors, delivery vs expectant management is determined by gestational age. Between 23 and 34 weeks’ gestation, when the fetus is at or close to viability, expectant management is recommended, provided there are no signs of infection or maternal or fetal compromise.⁴ This is because of the significant morbidity and mortality associated with births before 34 weeks’ gestation.⁴

The American College of Obstetricians and Gynecologists (ACOG) currently recommends delivery for all women with rupture of membranes after 34 weeks’ gestation, while acknowledging that this recommendation is based on “limited and inconsistent scientific evidence.”⁵ The recommendation for delivery after 34 weeks is predicated on the belief that disability-free survival is high in late preterm infants. However, there is a growing body of evidence that shows negative short- and long-term effects for these children, including medical concerns, academic difficulties, and more frequent hospital admissions in early childhood.^6,7

Higher birth weights, fewer C-sections, and no increased sepsis with wait-and-see

The Preterm Pre-labour Rupture Of the Membranes close to Term (PPROMT) trial was a multicenter (65 institutions across 11 countries), randomized controlled trial (RCT) that included 1839 women with singleton pregnancies and confirmed rupture of membranes between 34 weeks and 36 weeks, 6 days’ gestation.¹ Conducted from May 2004 to June 2013, participants were randomized to expectant management (915 women) vs immediate delivery by induction (924 women). Patients and care providers were not masked to treatment allocation, but those determining the primary outcome were masked to group allocation.

One woman in each group was lost to follow-up, and 2 additional women withdrew from the immediate birth group. Women already in active labor or with clinical indications for delivery (chorioamnionitis, abruption, cord prolapse, fetal distress) were excluded. The baseline characteristics of the 2 groups were similar.

Women in the induction group had delivery scheduled as soon as possible after randomization. Women in the expectant management group were allowed to go into spontaneous labor and were only induced if they reached term or the clinician identified other indications for immediate delivery.

The primary outcome was probable or confirmed neonatal sepsis. Secondary infant outcomes included a composite neonatal morbidity and mortality indicator (sepsis, mechanical ventilation ≥24 hours, still birth, or neonatal death), respiratory distress syndrome, any mechanical ventilation, birth weight, and duration of stay in a neonatal intensive care unit (NICU) or special care nursery. Secondary maternal outcomes included antepartum or intrapartum hemorrhage, intrapartum fever, mode of delivery, duration of hospital stay, and development of chorioamnionitis in the expectant management group.

The primary outcome of neonatal sepsis occurred in 2% of the neonates assigned to immediate delivery and 3% of neonates assigned to expectant management (relative risk [RR]=0.8; 95% confidence interval [CI], 0.5-1.3; P=.37). There was also no statistically significant difference in composite neonatal morbidity and mortality (RR=1.2; 95% CI, 0.9-1.6; P=.32). However, infants born in the immediate delivery group had significantly lower birth weights (2574.7 g vs 2673.2 g; absolute difference= -125 g; P<.0001), a higher incidence of respiratory distress (RR=1.6; 95% CI, 1.1-2.3; P=.008; number needed to treat [NNT]=32), and spent more time in the NICU/special care nursery (4 days vs 2 days; P<.0001).

This study is the largest to show that immediate birth increases the risk of respiratory distress and duration of NICU stay.

Compared to immediate delivery, expectant management was associated with a higher likelihood of antepartum or intrapartum hemorrhage (RR=0.6; 95% CI, 0.4-0.9; P=.02; number needed to harm [NNH]=50) and intrapartum fever (RR=0.4; 95% CI, 0.2-0.9; P=.02; NNH=100). In the women assigned to immediate delivery, 26% had a cesarean section, compared to 19% in the expectant management group (RR=1.4; 95% CI, 1.2-1.7, P=.0001; NNT=14). A total of 56 women (6%) assigned to the expectant management group developed clinically significant chorioamnionitis requiring delivery. All other secondary maternal and neonatal outcomes were equivalent with no significant differences between the 2 groups.

WHAT'S NEW?

Largest study to show no increased sepsis with expectant management

Two prior RCTs (the PPROMEXIL trial⁸ and PPROMEXIL-2⁹), involving a total of 736 women, evaluated expectant management vs induction in the late preterm stage of pregnancy. There was no increased risk of neonatal sepsis with expectant management in either study. However, those studies did not have sufficient power to show a statistically significant change in any of the outcomes.

The PPROMT study is the largest one to show that immediate birth increases the risk of respiratory distress and duration of NICU/special care stay for the baby and increases the risk of cesarean section for the mother. It also showed that the risk of neonatal sepsis was not higher in the expectant management group.

CAVEATS

Findings only apply to singleton pregnancies

Delivery of the infants in the expectant management group was not by specified protocol; each birth was managed according to the policies of the local center and clinician judgment. Because of this, there was variation in fetal and maternal monitoring. The vast majority of women in both groups (92% to 93%) received intrapartum antibiotics. Expectant management should include careful monitoring for infection and hemorrhage and may need to be changed to immediate delivery if one of these occurs.

The study participants all had singleton pregnancies; this recommendation cannot be extended to non-singleton pregnancies. However, a prior cesarean section was not an exclusion criterion for the study, and these recommendations would be valid for that group of women, too.

CHALLENGES TO IMPLEMENTATION

Going against the tide of ACOG

The most recent ACOG guidelines, updated October 2016, recommend induction of labor for women with ruptured membranes in the late preterm stages.⁵ This may present a challenge to widespread acceptance of expectant management for PPROM.

ACKNOWLEDGEMENT

The PURLs Surveillance System was supported in part by Grant Number UL1RR024999 from the National Center For Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center For Research Resources or the National Institutes of Health.