Sarah Safranek, MLIS

Evidence summary

No studies specifically examine the risk of adverse outcomes for women who have mild elevations in blood pressure as a result of taking OCs. However, for the general population cardiovascular risk increases 30% for each 10 mm Hg rise in systolic pressure.¹

A prospective cohort study found an increased incidence of new hypertension developing among women taking OCs. The study, conducted in the US, included 68,297 female nurses aged 25 to 42 years without a previous diagnosis of hypertension, diabetes, coronary heart disease, stroke, or cancer, who were followed for 4 years. Women were excluded from the study if they had not had a physical exam in the last 2 years or were taking antihypertensive medication at study inception. The nurses self-reported their blood pressure readings via questionnaire; medical records were sampled to validate the accuracy of self-reports. After adjusting for risk factors, current oral contraceptive use increased the risk of developing hypertension (relative risk [RR]=1.8; 95% confidence interval [CI], 1.5–2.3; corresponding to 41.5 new cases of hypertension per 10,000 person-years).²

A systematic review of 8 case control studies (with 4907 cases and 13,443 controls) found an increased risk of stroke and MI among hypertensive women taking combined type OCs vs those not taking OCs. Women with hypertension aged 20 to 24 years had an estimated CV event risk of 312 per million woman-years while taking OCs vs 134 per million woman-years not taking OCs. Among hypertensive women aged 40 to 44 years, the estimated risks were 1213 vs 529 per million woman-years, respectively. Primary endpoints varied across the studies and statistical significance was not given.³

Three studies showed that blood pressure elevations due to taking oral contraceptives returned to baseline with discontinuation of the medication. A prospective cohort study followed 13,358 women who were neither pregnant nor postpartum between the ages of 15 and 60. Women who either initiated or resumed using OCs experienced a statistically significant rise of about 4 mm Hg in the systolic pressure and 1 mm Hg in the diastolic pressure. Women who stopped using OCs experienced significant decreases in both systolic and diastolic components (about 5 mm Hg and 1.5 mm Hg, respectively).¹

Similarly, a survey study of 461 women attending family planning clinics found that mean blood pressures were significantly higher for those taking OCs than for those using nonhormonal contraception. Elevated blood pressures correlated with duration of current use of OCs but returned to normal soon after stopping OCs. The mean pressures of those who had stopped OCs more than a month were similar to those of women who had never taken an OC and significantly lower than those of women who were currently taking OCs.⁴

Finally, a prospective study which followed 32 women who had taken combined OCs for 1 to 3 years and then stopped found that blood pressures returned to pretreatment levels at 3 months. Systolic pressure fell by 9.7 mm Hg and diastolic by 2.9 mm Hg compared with measurements made 1 month before stopping. No cardiovascular complications were reported among women during this study.⁵

Recommendations from others

The American College of Obstetricians and Gynecologists says that women with well-controlled and monitored hypertension aged 35 years and younger are appropriate candidates for a trial of combination OCs, provided they are otherwise healthy, have no evidence of end-organ vascular disease, and do not smoke cigarettes.⁶ If blood pressure remains well controlled with careful monitoring several months after initiating OCs, use can be continued.

Evidence summary

No studies specifically examine the risk of adverse outcomes for women who have mild elevations in blood pressure as a result of taking OCs. However, for the general population cardiovascular risk increases 30% for each 10 mm Hg rise in systolic pressure.¹

A prospective cohort study found an increased incidence of new hypertension developing among women taking OCs. The study, conducted in the US, included 68,297 female nurses aged 25 to 42 years without a previous diagnosis of hypertension, diabetes, coronary heart disease, stroke, or cancer, who were followed for 4 years. Women were excluded from the study if they had not had a physical exam in the last 2 years or were taking antihypertensive medication at study inception. The nurses self-reported their blood pressure readings via questionnaire; medical records were sampled to validate the accuracy of self-reports. After adjusting for risk factors, current oral contraceptive use increased the risk of developing hypertension (relative risk [RR]=1.8; 95% confidence interval [CI], 1.5–2.3; corresponding to 41.5 new cases of hypertension per 10,000 person-years).²

A systematic review of 8 case control studies (with 4907 cases and 13,443 controls) found an increased risk of stroke and MI among hypertensive women taking combined type OCs vs those not taking OCs. Women with hypertension aged 20 to 24 years had an estimated CV event risk of 312 per million woman-years while taking OCs vs 134 per million woman-years not taking OCs. Among hypertensive women aged 40 to 44 years, the estimated risks were 1213 vs 529 per million woman-years, respectively. Primary endpoints varied across the studies and statistical significance was not given.³

Three studies showed that blood pressure elevations due to taking oral contraceptives returned to baseline with discontinuation of the medication. A prospective cohort study followed 13,358 women who were neither pregnant nor postpartum between the ages of 15 and 60. Women who either initiated or resumed using OCs experienced a statistically significant rise of about 4 mm Hg in the systolic pressure and 1 mm Hg in the diastolic pressure. Women who stopped using OCs experienced significant decreases in both systolic and diastolic components (about 5 mm Hg and 1.5 mm Hg, respectively).¹

Similarly, a survey study of 461 women attending family planning clinics found that mean blood pressures were significantly higher for those taking OCs than for those using nonhormonal contraception. Elevated blood pressures correlated with duration of current use of OCs but returned to normal soon after stopping OCs. The mean pressures of those who had stopped OCs more than a month were similar to those of women who had never taken an OC and significantly lower than those of women who were currently taking OCs.⁴

Finally, a prospective study which followed 32 women who had taken combined OCs for 1 to 3 years and then stopped found that blood pressures returned to pretreatment levels at 3 months. Systolic pressure fell by 9.7 mm Hg and diastolic by 2.9 mm Hg compared with measurements made 1 month before stopping. No cardiovascular complications were reported among women during this study.⁵

Recommendations from others

The American College of Obstetricians and Gynecologists says that women with well-controlled and monitored hypertension aged 35 years and younger are appropriate candidates for a trial of combination OCs, provided they are otherwise healthy, have no evidence of end-organ vascular disease, and do not smoke cigarettes.⁶ If blood pressure remains well controlled with careful monitoring several months after initiating OCs, use can be continued.

Evidence summary

No studies specifically examine the risk of adverse outcomes for women who have mild elevations in blood pressure as a result of taking OCs. However, for the general population cardiovascular risk increases 30% for each 10 mm Hg rise in systolic pressure.¹

A prospective cohort study found an increased incidence of new hypertension developing among women taking OCs. The study, conducted in the US, included 68,297 female nurses aged 25 to 42 years without a previous diagnosis of hypertension, diabetes, coronary heart disease, stroke, or cancer, who were followed for 4 years. Women were excluded from the study if they had not had a physical exam in the last 2 years or were taking antihypertensive medication at study inception. The nurses self-reported their blood pressure readings via questionnaire; medical records were sampled to validate the accuracy of self-reports. After adjusting for risk factors, current oral contraceptive use increased the risk of developing hypertension (relative risk [RR]=1.8; 95% confidence interval [CI], 1.5–2.3; corresponding to 41.5 new cases of hypertension per 10,000 person-years).²

A systematic review of 8 case control studies (with 4907 cases and 13,443 controls) found an increased risk of stroke and MI among hypertensive women taking combined type OCs vs those not taking OCs. Women with hypertension aged 20 to 24 years had an estimated CV event risk of 312 per million woman-years while taking OCs vs 134 per million woman-years not taking OCs. Among hypertensive women aged 40 to 44 years, the estimated risks were 1213 vs 529 per million woman-years, respectively. Primary endpoints varied across the studies and statistical significance was not given.³

Three studies showed that blood pressure elevations due to taking oral contraceptives returned to baseline with discontinuation of the medication. A prospective cohort study followed 13,358 women who were neither pregnant nor postpartum between the ages of 15 and 60. Women who either initiated or resumed using OCs experienced a statistically significant rise of about 4 mm Hg in the systolic pressure and 1 mm Hg in the diastolic pressure. Women who stopped using OCs experienced significant decreases in both systolic and diastolic components (about 5 mm Hg and 1.5 mm Hg, respectively).¹

Similarly, a survey study of 461 women attending family planning clinics found that mean blood pressures were significantly higher for those taking OCs than for those using nonhormonal contraception. Elevated blood pressures correlated with duration of current use of OCs but returned to normal soon after stopping OCs. The mean pressures of those who had stopped OCs more than a month were similar to those of women who had never taken an OC and significantly lower than those of women who were currently taking OCs.⁴

Finally, a prospective study which followed 32 women who had taken combined OCs for 1 to 3 years and then stopped found that blood pressures returned to pretreatment levels at 3 months. Systolic pressure fell by 9.7 mm Hg and diastolic by 2.9 mm Hg compared with measurements made 1 month before stopping. No cardiovascular complications were reported among women during this study.⁵

Recommendations from others

The American College of Obstetricians and Gynecologists says that women with well-controlled and monitored hypertension aged 35 years and younger are appropriate candidates for a trial of combination OCs, provided they are otherwise healthy, have no evidence of end-organ vascular disease, and do not smoke cigarettes.⁶ If blood pressure remains well controlled with careful monitoring several months after initiating OCs, use can be continued.

Evidence summary

Clinical evaluation with medical history and physical exam, chest radiography, and selected sputum sampling to exclude active tuberculosis are part of the recommended algorithm for all patients who develop a positive PPD (FIGURE).^1-3 These recommendations are derived from expert opinion, and their usefulness has not been evaluated in any population-based study of asymptomatic PPD-positive patients.

A comprehensive review of RCTs from the 1950s and 1960s demonstrated that INH treatment of patients with latent tuberculosis infection is effective in decreasing the progression to active tuberculosis.⁴ A series of double-blinded RCTs performed by the US Public Health Service included 25,923 patients with latent tuberculosis who were randomized to receive either daily INH or placebo for 1 year with 6- to 10-year follow-up. Groups studied included household contacts of patients with active tuberculosis (rate of progression to active disease in placebo group [baseline rate]=27/1000, relative risk with INH [RR]=0.4, number needed to treat [NNT]=63), patients in mental institutions (baseline rate=12/1000, RR=0.3, NNT=121), and patients with x-ray findings of healed tuberculosis (baseline rate=69/1000, RR=0.4, NNT=23).

The optimal length of treatment for PPD-positive patients without active disease was evaluated through 1 double-blinded RCT enrolling 28,000 patients with 5-year follow-up after 12, 24, or 52 weeks of INH or placebo. Active TB developed in 0.35% (24/6919) after 52 weeks of INH compared with 0.49% (34/6965) after 24 weeks (RR=1.4, NNT=708).⁵ Incidence in the placebo group was 1.4%. Subgroup analysis determined that maximum efficacy with fewest side effects was achieved at 9 months.⁶ Nine months of INH is also recommended for HIV-positive patients, based on extrapolations from these and other studies.³

INH monotherapy was compared with combination INH and rifampin in a 2005 meta-analysis of 5 RCTs of variable quality involving 1926 patients.⁷ This meta-analysis found equivalency in risk of active TB and mortality between INH monotherapy for 6 to 12 months and the combination of rifampin and INH for 3 months (pooled risk difference=0%; 95% confidence interval [CI], –1% to 2%). This study also showed similar rates of adverse events in both groups (pooled risk difference=–1%; 95% CI, –7% to 5%). Short-course combination rifampin and pyrazinamide is no longer recommended after an open-label RCT with 589 patients demonstrated severe hepatoxicity in 7.7% (16/207) on a 2-month course of pyrazinamide and rifampin, compared with 1% (2/204) on 6 months of INH (RR=7.9, number needed to harm=15).⁸ Rifampin monotherapy has only been studied in patients with silicosis in a RCT enrolling 652 participants with latent tuberculosis. A 12-week course of rifampin (600 mg daily) was as effective as 6 months of INH in preventing development of active TB over the next 5 years.⁹

FIGURE
Suggested workup of asymptomatic, HIV-negative patients with a positive PPD

Recommendations from others

Centers for Disease Control and Prevention, American Thoracic Society, and Infectious Disease Society of America guidelines recommend targeted screening of high-risk persons followed by further clinical evaluation of all those with a reactive PPD (FIGURE).^2,10 The recommended treatment regimen for latent TB is daily INH for 9 months. Less preferable regimens are daily INH for 6 months, or daily rifampin for 4 months in patients who cannot tolerate INH. A 2-month course of rifampin and pyrazinamide is no longer recommended. The recent meta-analysis supporting a 3-month regimen of combination INH and rifampin has not been incorporated into expert guidelines.⁷

Evidence summary

Clinical evaluation with medical history and physical exam, chest radiography, and selected sputum sampling to exclude active tuberculosis are part of the recommended algorithm for all patients who develop a positive PPD (FIGURE).^1-3 These recommendations are derived from expert opinion, and their usefulness has not been evaluated in any population-based study of asymptomatic PPD-positive patients.

A comprehensive review of RCTs from the 1950s and 1960s demonstrated that INH treatment of patients with latent tuberculosis infection is effective in decreasing the progression to active tuberculosis.⁴ A series of double-blinded RCTs performed by the US Public Health Service included 25,923 patients with latent tuberculosis who were randomized to receive either daily INH or placebo for 1 year with 6- to 10-year follow-up. Groups studied included household contacts of patients with active tuberculosis (rate of progression to active disease in placebo group [baseline rate]=27/1000, relative risk with INH [RR]=0.4, number needed to treat [NNT]=63), patients in mental institutions (baseline rate=12/1000, RR=0.3, NNT=121), and patients with x-ray findings of healed tuberculosis (baseline rate=69/1000, RR=0.4, NNT=23).

The optimal length of treatment for PPD-positive patients without active disease was evaluated through 1 double-blinded RCT enrolling 28,000 patients with 5-year follow-up after 12, 24, or 52 weeks of INH or placebo. Active TB developed in 0.35% (24/6919) after 52 weeks of INH compared with 0.49% (34/6965) after 24 weeks (RR=1.4, NNT=708).⁵ Incidence in the placebo group was 1.4%. Subgroup analysis determined that maximum efficacy with fewest side effects was achieved at 9 months.⁶ Nine months of INH is also recommended for HIV-positive patients, based on extrapolations from these and other studies.³

INH monotherapy was compared with combination INH and rifampin in a 2005 meta-analysis of 5 RCTs of variable quality involving 1926 patients.⁷ This meta-analysis found equivalency in risk of active TB and mortality between INH monotherapy for 6 to 12 months and the combination of rifampin and INH for 3 months (pooled risk difference=0%; 95% confidence interval [CI], –1% to 2%). This study also showed similar rates of adverse events in both groups (pooled risk difference=–1%; 95% CI, –7% to 5%). Short-course combination rifampin and pyrazinamide is no longer recommended after an open-label RCT with 589 patients demonstrated severe hepatoxicity in 7.7% (16/207) on a 2-month course of pyrazinamide and rifampin, compared with 1% (2/204) on 6 months of INH (RR=7.9, number needed to harm=15).⁸ Rifampin monotherapy has only been studied in patients with silicosis in a RCT enrolling 652 participants with latent tuberculosis. A 12-week course of rifampin (600 mg daily) was as effective as 6 months of INH in preventing development of active TB over the next 5 years.⁹

FIGURE
Suggested workup of asymptomatic, HIV-negative patients with a positive PPD

Recommendations from others

Centers for Disease Control and Prevention, American Thoracic Society, and Infectious Disease Society of America guidelines recommend targeted screening of high-risk persons followed by further clinical evaluation of all those with a reactive PPD (FIGURE).^2,10 The recommended treatment regimen for latent TB is daily INH for 9 months. Less preferable regimens are daily INH for 6 months, or daily rifampin for 4 months in patients who cannot tolerate INH. A 2-month course of rifampin and pyrazinamide is no longer recommended. The recent meta-analysis supporting a 3-month regimen of combination INH and rifampin has not been incorporated into expert guidelines.⁷

Evidence summary

Clinical evaluation with medical history and physical exam, chest radiography, and selected sputum sampling to exclude active tuberculosis are part of the recommended algorithm for all patients who develop a positive PPD (FIGURE).^1-3 These recommendations are derived from expert opinion, and their usefulness has not been evaluated in any population-based study of asymptomatic PPD-positive patients.

A comprehensive review of RCTs from the 1950s and 1960s demonstrated that INH treatment of patients with latent tuberculosis infection is effective in decreasing the progression to active tuberculosis.⁴ A series of double-blinded RCTs performed by the US Public Health Service included 25,923 patients with latent tuberculosis who were randomized to receive either daily INH or placebo for 1 year with 6- to 10-year follow-up. Groups studied included household contacts of patients with active tuberculosis (rate of progression to active disease in placebo group [baseline rate]=27/1000, relative risk with INH [RR]=0.4, number needed to treat [NNT]=63), patients in mental institutions (baseline rate=12/1000, RR=0.3, NNT=121), and patients with x-ray findings of healed tuberculosis (baseline rate=69/1000, RR=0.4, NNT=23).

The optimal length of treatment for PPD-positive patients without active disease was evaluated through 1 double-blinded RCT enrolling 28,000 patients with 5-year follow-up after 12, 24, or 52 weeks of INH or placebo. Active TB developed in 0.35% (24/6919) after 52 weeks of INH compared with 0.49% (34/6965) after 24 weeks (RR=1.4, NNT=708).⁵ Incidence in the placebo group was 1.4%. Subgroup analysis determined that maximum efficacy with fewest side effects was achieved at 9 months.⁶ Nine months of INH is also recommended for HIV-positive patients, based on extrapolations from these and other studies.³

INH monotherapy was compared with combination INH and rifampin in a 2005 meta-analysis of 5 RCTs of variable quality involving 1926 patients.⁷ This meta-analysis found equivalency in risk of active TB and mortality between INH monotherapy for 6 to 12 months and the combination of rifampin and INH for 3 months (pooled risk difference=0%; 95% confidence interval [CI], –1% to 2%). This study also showed similar rates of adverse events in both groups (pooled risk difference=–1%; 95% CI, –7% to 5%). Short-course combination rifampin and pyrazinamide is no longer recommended after an open-label RCT with 589 patients demonstrated severe hepatoxicity in 7.7% (16/207) on a 2-month course of pyrazinamide and rifampin, compared with 1% (2/204) on 6 months of INH (RR=7.9, number needed to harm=15).⁸ Rifampin monotherapy has only been studied in patients with silicosis in a RCT enrolling 652 participants with latent tuberculosis. A 12-week course of rifampin (600 mg daily) was as effective as 6 months of INH in preventing development of active TB over the next 5 years.⁹

FIGURE
Suggested workup of asymptomatic, HIV-negative patients with a positive PPD

Recommendations from others

Centers for Disease Control and Prevention, American Thoracic Society, and Infectious Disease Society of America guidelines recommend targeted screening of high-risk persons followed by further clinical evaluation of all those with a reactive PPD (FIGURE).^2,10 The recommended treatment regimen for latent TB is daily INH for 9 months. Less preferable regimens are daily INH for 6 months, or daily rifampin for 4 months in patients who cannot tolerate INH. A 2-month course of rifampin and pyrazinamide is no longer recommended. The recent meta-analysis supporting a 3-month regimen of combination INH and rifampin has not been incorporated into expert guidelines.⁷

Evidence summary

An expert guideline recommends a history and physical examination to determine whether an elevated SPL is due to physiologic, pharmacologic, or pathologic causes (TABLE).¹ The fasting morning SPL is least variable and correlates best with a disease state.¹ Clinical correlation is necessary to reveal false positives (due to biologically inactive forms of prolactin) or false negatives (due to very high SPLs that exceed the ability of the assay). If an elevated SPL is suspected despite a normal laboratory report, retesting with serum diluted 1:100 can identify a false-negative value.²

A detailed drug history is important since drug-induced elevated SPL is common.¹ Laboratory evaluation includes thyroid-stimulating hormone, blood urea nitrogen, and creatinine, as well as pregnancy testing when applicable. If no cause of elevated SPL is identified by initial clinical evaluation or if the SPL is greater than 200 ng/mL, experts recommend imaging of the sella turcica with computed tomography or magnetic resonance imaging.¹

Physiologic causes. For patients with a mildly elevated SPL due to a physiologic cause, experts recommend expectant management. Patients should be monitored for symptoms of hypogonadism (amenorrhea, infertility, or sexual dysfunction) and have SPL measured at 6- to 12-month intervals.¹ In cohort studies, treatment of the underlying cause of elevated SPL reverses secondary physiologic changes of low estrogen or testosterone, and hypogonadism.^3-5

Pharmacologic causes. Eliminating a pharmacologic cause may lead to normalization of SPL, although experts recommend psychiatric consultation before discontinuing neuroleptic medications.¹

Pathologic causes. Experts advise treating the underlying cause of a pathologic elevation of SPL. Patients with microadenoma should have SPLs monitored to prevent complications of decreased bone mineral density and sexual dysfunction due to persistently elevated SPL. Patients with a macroadenoma (>1 cm) are at risk for tumor growth and require serial imaging studies in addition to treatment of SPL, according to expert opinion.^1-3

Medical therapy. Medical therapy with a dopamine agonist is indicated for patients with either symptoms of hypogonadism due to elevated SPL, or neurologic symptoms due to the size of a macroadenoma.¹ In a review of 13 cohort studies, bromocriptine improved symptoms and reduced SPLs to normal for 229 of 280 women (82%).⁶ A cohort study of 27 patients with macroadenomas treated with bromocriptine found 10% to 50% reductions of tumor size.⁷ A randomized controlled trial treating 459 women having hyperprolactinemic amenorrhea with either cabergoline or bromocriptine achieved a stable normal SPL in 83% and 59%, respectively (P<.001). Adverse effects were common but were less common with cabergoline (68% vs 78%) and resulted in fewer discontinuations (3% vs 12%).⁸

Surgical therapy. Surgery is indicated for patients unresponsive to or intolerant of medical therapy, or who have visual field loss, cranial nerve palsy, or headache due to macroadenoma.¹ A retrospective review of patients who underwent surgical resection found a 40% recurrence rate.⁹

Recommendations from others

Williams Textbook of Endocrinology includes the recommendations above and advises seeking consultation for patients with mass effects of macroadenomas such as visual field loss, cranial nerve palsy, or headaches; for patients with progressive elevation of SPL despite medical treatment; and for pregnant women.⁴ Conventional antipsychotic agents are commonly associated with elevated prolactin due to dopamine agonist activity. Some atypical antipsychotics may lead to lower levels of elevated prolactin, transient elevations or marked elevations.¹⁰ Experts recommend following serial SPLs, if antipsychotics are truly needed. Psychiatric consultation may assist in making decisions about medication selection. Patients with symptoms (galactorrhea, amenorrhea, headaches, visual disturbances, sexual dysfunction) or levels of 200 or more, should undergo an MRI or CT. Experts recommend monitoring levels every 1 to 3 months.¹

TABLE
Physiologic, pharmacologic, and pathologic causes of an elevated serum prolactin level¹

Evidence summary

An expert guideline recommends a history and physical examination to determine whether an elevated SPL is due to physiologic, pharmacologic, or pathologic causes (TABLE).¹ The fasting morning SPL is least variable and correlates best with a disease state.¹ Clinical correlation is necessary to reveal false positives (due to biologically inactive forms of prolactin) or false negatives (due to very high SPLs that exceed the ability of the assay). If an elevated SPL is suspected despite a normal laboratory report, retesting with serum diluted 1:100 can identify a false-negative value.²

A detailed drug history is important since drug-induced elevated SPL is common.¹ Laboratory evaluation includes thyroid-stimulating hormone, blood urea nitrogen, and creatinine, as well as pregnancy testing when applicable. If no cause of elevated SPL is identified by initial clinical evaluation or if the SPL is greater than 200 ng/mL, experts recommend imaging of the sella turcica with computed tomography or magnetic resonance imaging.¹

Physiologic causes. For patients with a mildly elevated SPL due to a physiologic cause, experts recommend expectant management. Patients should be monitored for symptoms of hypogonadism (amenorrhea, infertility, or sexual dysfunction) and have SPL measured at 6- to 12-month intervals.¹ In cohort studies, treatment of the underlying cause of elevated SPL reverses secondary physiologic changes of low estrogen or testosterone, and hypogonadism.^3-5

Pharmacologic causes. Eliminating a pharmacologic cause may lead to normalization of SPL, although experts recommend psychiatric consultation before discontinuing neuroleptic medications.¹

Pathologic causes. Experts advise treating the underlying cause of a pathologic elevation of SPL. Patients with microadenoma should have SPLs monitored to prevent complications of decreased bone mineral density and sexual dysfunction due to persistently elevated SPL. Patients with a macroadenoma (>1 cm) are at risk for tumor growth and require serial imaging studies in addition to treatment of SPL, according to expert opinion.^1-3

Medical therapy. Medical therapy with a dopamine agonist is indicated for patients with either symptoms of hypogonadism due to elevated SPL, or neurologic symptoms due to the size of a macroadenoma.¹ In a review of 13 cohort studies, bromocriptine improved symptoms and reduced SPLs to normal for 229 of 280 women (82%).⁶ A cohort study of 27 patients with macroadenomas treated with bromocriptine found 10% to 50% reductions of tumor size.⁷ A randomized controlled trial treating 459 women having hyperprolactinemic amenorrhea with either cabergoline or bromocriptine achieved a stable normal SPL in 83% and 59%, respectively (P<.001). Adverse effects were common but were less common with cabergoline (68% vs 78%) and resulted in fewer discontinuations (3% vs 12%).⁸

Surgical therapy. Surgery is indicated for patients unresponsive to or intolerant of medical therapy, or who have visual field loss, cranial nerve palsy, or headache due to macroadenoma.¹ A retrospective review of patients who underwent surgical resection found a 40% recurrence rate.⁹

Recommendations from others

Williams Textbook of Endocrinology includes the recommendations above and advises seeking consultation for patients with mass effects of macroadenomas such as visual field loss, cranial nerve palsy, or headaches; for patients with progressive elevation of SPL despite medical treatment; and for pregnant women.⁴ Conventional antipsychotic agents are commonly associated with elevated prolactin due to dopamine agonist activity. Some atypical antipsychotics may lead to lower levels of elevated prolactin, transient elevations or marked elevations.¹⁰ Experts recommend following serial SPLs, if antipsychotics are truly needed. Psychiatric consultation may assist in making decisions about medication selection. Patients with symptoms (galactorrhea, amenorrhea, headaches, visual disturbances, sexual dysfunction) or levels of 200 or more, should undergo an MRI or CT. Experts recommend monitoring levels every 1 to 3 months.¹

TABLE
Physiologic, pharmacologic, and pathologic causes of an elevated serum prolactin level¹

Evidence summary

An expert guideline recommends a history and physical examination to determine whether an elevated SPL is due to physiologic, pharmacologic, or pathologic causes (TABLE).¹ The fasting morning SPL is least variable and correlates best with a disease state.¹ Clinical correlation is necessary to reveal false positives (due to biologically inactive forms of prolactin) or false negatives (due to very high SPLs that exceed the ability of the assay). If an elevated SPL is suspected despite a normal laboratory report, retesting with serum diluted 1:100 can identify a false-negative value.²

A detailed drug history is important since drug-induced elevated SPL is common.¹ Laboratory evaluation includes thyroid-stimulating hormone, blood urea nitrogen, and creatinine, as well as pregnancy testing when applicable. If no cause of elevated SPL is identified by initial clinical evaluation or if the SPL is greater than 200 ng/mL, experts recommend imaging of the sella turcica with computed tomography or magnetic resonance imaging.¹

Physiologic causes. For patients with a mildly elevated SPL due to a physiologic cause, experts recommend expectant management. Patients should be monitored for symptoms of hypogonadism (amenorrhea, infertility, or sexual dysfunction) and have SPL measured at 6- to 12-month intervals.¹ In cohort studies, treatment of the underlying cause of elevated SPL reverses secondary physiologic changes of low estrogen or testosterone, and hypogonadism.^3-5

Pharmacologic causes. Eliminating a pharmacologic cause may lead to normalization of SPL, although experts recommend psychiatric consultation before discontinuing neuroleptic medications.¹

Pathologic causes. Experts advise treating the underlying cause of a pathologic elevation of SPL. Patients with microadenoma should have SPLs monitored to prevent complications of decreased bone mineral density and sexual dysfunction due to persistently elevated SPL. Patients with a macroadenoma (>1 cm) are at risk for tumor growth and require serial imaging studies in addition to treatment of SPL, according to expert opinion.^1-3

Medical therapy. Medical therapy with a dopamine agonist is indicated for patients with either symptoms of hypogonadism due to elevated SPL, or neurologic symptoms due to the size of a macroadenoma.¹ In a review of 13 cohort studies, bromocriptine improved symptoms and reduced SPLs to normal for 229 of 280 women (82%).⁶ A cohort study of 27 patients with macroadenomas treated with bromocriptine found 10% to 50% reductions of tumor size.⁷ A randomized controlled trial treating 459 women having hyperprolactinemic amenorrhea with either cabergoline or bromocriptine achieved a stable normal SPL in 83% and 59%, respectively (P<.001). Adverse effects were common but were less common with cabergoline (68% vs 78%) and resulted in fewer discontinuations (3% vs 12%).⁸

Surgical therapy. Surgery is indicated for patients unresponsive to or intolerant of medical therapy, or who have visual field loss, cranial nerve palsy, or headache due to macroadenoma.¹ A retrospective review of patients who underwent surgical resection found a 40% recurrence rate.⁹

Recommendations from others

Williams Textbook of Endocrinology includes the recommendations above and advises seeking consultation for patients with mass effects of macroadenomas such as visual field loss, cranial nerve palsy, or headaches; for patients with progressive elevation of SPL despite medical treatment; and for pregnant women.⁴ Conventional antipsychotic agents are commonly associated with elevated prolactin due to dopamine agonist activity. Some atypical antipsychotics may lead to lower levels of elevated prolactin, transient elevations or marked elevations.¹⁰ Experts recommend following serial SPLs, if antipsychotics are truly needed. Psychiatric consultation may assist in making decisions about medication selection. Patients with symptoms (galactorrhea, amenorrhea, headaches, visual disturbances, sexual dysfunction) or levels of 200 or more, should undergo an MRI or CT. Experts recommend monitoring levels every 1 to 3 months.¹

TABLE
Physiologic, pharmacologic, and pathologic causes of an elevated serum prolactin level¹

Evidence summary

Progesterone. A Cochrane meta-analysis on the use of progesterone to prevent pregnancy loss looked at a subset of 3 small RCTs that evaluated women with 3 or more pregnancy losses. Patients with primary recurrent spontaneous abortion (RSA) (no prior live births), were not differentiated from those with secondary RSA (previous live birth with subsequent miscarriages).

Progesterone administration resulted in a significant reduction in miscarriage compared with placebo (odds ratio [OR]=0.37; 95% confidence interval [CI], 0.17–0.91), independent of administration routes (oral, vaginal, or intramuscular). This benefit was lost in the larger meta-analysis when studies containing women with fewer than 3 pregnancy losses were included.¹

Human choriogonadotropin. A meta-analysis reviewed 4 trials (n=180 total) of varying methodological quality, which were constructed to determine if women, with at least 2 consecutive miscarriages of unknown cause, derive any protective effect when they receive HCG during the first trimester. Although the overall outcome favored the use of HCG (OR=0.26 compared with placebo; 95% CI, 0.14–0.52), the trials contained major methodological weaknesses (poor description of methods, no power calculations, selection and unclear randomization techniques).²

Immunotherapy. A systematic review of 22 RCTs evaluating 4 different types of immunotherapy for recurrent miscarriage found no significant improvement in live birth rates. All studies were of high quality with a low level of bias. Only one lacked double-blinding.

Immunotherapy types included: paternal leukocyte immunization (PLI) (11 trials, 596 women) (OR=1.05; 95% CI, 0.75–1.47); intravenous immune globulin (IVIG) (OR=0.98; 95% CI, 0.61–1.58); third-party donor cell immunization (3 trials, 156 women) (OR=1.39; 95% CI, 0.68–2.82); and trophoblast membrane infusion (1 trial, 37 women) (OR=0.40; 95% CI, 0.11–1.45).³

A subsequent RCT comparing PLI with placebo among 79 women with primary RSA of unknown cause again found no significant difference in live birth rates (89% vs 71%, respectively).⁴ However, an additional RCT evaluating PLI (32 patients) vs placebo (19 patients) among women with unexplained primary RSA did find significantly higher birth rates with PLI (84% vs 25%; P=.001). This small study used different techniques than previous PLI studies.⁵

A later meta-analysis of 5 RCTs including a total of 246 patients also found that IVIG did not improve the subsequent live birth rate for women with a history of primary or secondary RSA (OR=0.98; 95% CI, 0.45–2.13).⁶

Aspirin. An RCT involving 54 pregnant women (mean age 32.7 years) with a history of primary RSA of unknown cause (negative standard workup) evaluated 50 mg of aspirin daily (n=27) vs placebo (n=27).⁷ The method of blinding was not reported.

The live birth rate was identical for the 2 groups (88%). A second (unblinded) trial randomized 805 women from a large referral center (mean age 34 years) with a history of first-trimester RSA (not differentiated between primary and secondary RSA) of unknown cause to either 75 mg of aspirin daily or no treatment.⁸ There was no significant difference in the live birth rate between those who took aspirin (251/367; 68.4%) and those who did not (278/438; 63.5%; OR=1.24; 95% CI, 0.93–1.67).

Recommendations from others

The American College of Obstetricians and Gynecologists (ACOG) states that “it has not been shown conclusively that progesterone treatment or corpus luteum support (HCG) influences pregnancy outcome for women with recurrent spontaneous abortion.”⁹ ACOG does not recommend immunotherapy, citing a lack of demonstrated efficacy (IVIG and PLI), a lack of standards for cell storage and administration, and a risk profile similar to that of blood transfusion (PLI). They recommend “couples with otherwise unexplained recurrent pregnancy loss should be counseled regarding the potential for successful pregnancy without treatment.”

Evidence summary

Progesterone. A Cochrane meta-analysis on the use of progesterone to prevent pregnancy loss looked at a subset of 3 small RCTs that evaluated women with 3 or more pregnancy losses. Patients with primary recurrent spontaneous abortion (RSA) (no prior live births), were not differentiated from those with secondary RSA (previous live birth with subsequent miscarriages).

Progesterone administration resulted in a significant reduction in miscarriage compared with placebo (odds ratio [OR]=0.37; 95% confidence interval [CI], 0.17–0.91), independent of administration routes (oral, vaginal, or intramuscular). This benefit was lost in the larger meta-analysis when studies containing women with fewer than 3 pregnancy losses were included.¹

Human choriogonadotropin. A meta-analysis reviewed 4 trials (n=180 total) of varying methodological quality, which were constructed to determine if women, with at least 2 consecutive miscarriages of unknown cause, derive any protective effect when they receive HCG during the first trimester. Although the overall outcome favored the use of HCG (OR=0.26 compared with placebo; 95% CI, 0.14–0.52), the trials contained major methodological weaknesses (poor description of methods, no power calculations, selection and unclear randomization techniques).²

Immunotherapy. A systematic review of 22 RCTs evaluating 4 different types of immunotherapy for recurrent miscarriage found no significant improvement in live birth rates. All studies were of high quality with a low level of bias. Only one lacked double-blinding.

Immunotherapy types included: paternal leukocyte immunization (PLI) (11 trials, 596 women) (OR=1.05; 95% CI, 0.75–1.47); intravenous immune globulin (IVIG) (OR=0.98; 95% CI, 0.61–1.58); third-party donor cell immunization (3 trials, 156 women) (OR=1.39; 95% CI, 0.68–2.82); and trophoblast membrane infusion (1 trial, 37 women) (OR=0.40; 95% CI, 0.11–1.45).³

A subsequent RCT comparing PLI with placebo among 79 women with primary RSA of unknown cause again found no significant difference in live birth rates (89% vs 71%, respectively).⁴ However, an additional RCT evaluating PLI (32 patients) vs placebo (19 patients) among women with unexplained primary RSA did find significantly higher birth rates with PLI (84% vs 25%; P=.001). This small study used different techniques than previous PLI studies.⁵

A later meta-analysis of 5 RCTs including a total of 246 patients also found that IVIG did not improve the subsequent live birth rate for women with a history of primary or secondary RSA (OR=0.98; 95% CI, 0.45–2.13).⁶

Aspirin. An RCT involving 54 pregnant women (mean age 32.7 years) with a history of primary RSA of unknown cause (negative standard workup) evaluated 50 mg of aspirin daily (n=27) vs placebo (n=27).⁷ The method of blinding was not reported.

The live birth rate was identical for the 2 groups (88%). A second (unblinded) trial randomized 805 women from a large referral center (mean age 34 years) with a history of first-trimester RSA (not differentiated between primary and secondary RSA) of unknown cause to either 75 mg of aspirin daily or no treatment.⁸ There was no significant difference in the live birth rate between those who took aspirin (251/367; 68.4%) and those who did not (278/438; 63.5%; OR=1.24; 95% CI, 0.93–1.67).

Recommendations from others

The American College of Obstetricians and Gynecologists (ACOG) states that “it has not been shown conclusively that progesterone treatment or corpus luteum support (HCG) influences pregnancy outcome for women with recurrent spontaneous abortion.”⁹ ACOG does not recommend immunotherapy, citing a lack of demonstrated efficacy (IVIG and PLI), a lack of standards for cell storage and administration, and a risk profile similar to that of blood transfusion (PLI). They recommend “couples with otherwise unexplained recurrent pregnancy loss should be counseled regarding the potential for successful pregnancy without treatment.”

Evidence summary

Progesterone. A Cochrane meta-analysis on the use of progesterone to prevent pregnancy loss looked at a subset of 3 small RCTs that evaluated women with 3 or more pregnancy losses. Patients with primary recurrent spontaneous abortion (RSA) (no prior live births), were not differentiated from those with secondary RSA (previous live birth with subsequent miscarriages).

Progesterone administration resulted in a significant reduction in miscarriage compared with placebo (odds ratio [OR]=0.37; 95% confidence interval [CI], 0.17–0.91), independent of administration routes (oral, vaginal, or intramuscular). This benefit was lost in the larger meta-analysis when studies containing women with fewer than 3 pregnancy losses were included.¹

Human choriogonadotropin. A meta-analysis reviewed 4 trials (n=180 total) of varying methodological quality, which were constructed to determine if women, with at least 2 consecutive miscarriages of unknown cause, derive any protective effect when they receive HCG during the first trimester. Although the overall outcome favored the use of HCG (OR=0.26 compared with placebo; 95% CI, 0.14–0.52), the trials contained major methodological weaknesses (poor description of methods, no power calculations, selection and unclear randomization techniques).²

Immunotherapy. A systematic review of 22 RCTs evaluating 4 different types of immunotherapy for recurrent miscarriage found no significant improvement in live birth rates. All studies were of high quality with a low level of bias. Only one lacked double-blinding.

Immunotherapy types included: paternal leukocyte immunization (PLI) (11 trials, 596 women) (OR=1.05; 95% CI, 0.75–1.47); intravenous immune globulin (IVIG) (OR=0.98; 95% CI, 0.61–1.58); third-party donor cell immunization (3 trials, 156 women) (OR=1.39; 95% CI, 0.68–2.82); and trophoblast membrane infusion (1 trial, 37 women) (OR=0.40; 95% CI, 0.11–1.45).³

A subsequent RCT comparing PLI with placebo among 79 women with primary RSA of unknown cause again found no significant difference in live birth rates (89% vs 71%, respectively).⁴ However, an additional RCT evaluating PLI (32 patients) vs placebo (19 patients) among women with unexplained primary RSA did find significantly higher birth rates with PLI (84% vs 25%; P=.001). This small study used different techniques than previous PLI studies.⁵

A later meta-analysis of 5 RCTs including a total of 246 patients also found that IVIG did not improve the subsequent live birth rate for women with a history of primary or secondary RSA (OR=0.98; 95% CI, 0.45–2.13).⁶

Aspirin. An RCT involving 54 pregnant women (mean age 32.7 years) with a history of primary RSA of unknown cause (negative standard workup) evaluated 50 mg of aspirin daily (n=27) vs placebo (n=27).⁷ The method of blinding was not reported.

The live birth rate was identical for the 2 groups (88%). A second (unblinded) trial randomized 805 women from a large referral center (mean age 34 years) with a history of first-trimester RSA (not differentiated between primary and secondary RSA) of unknown cause to either 75 mg of aspirin daily or no treatment.⁸ There was no significant difference in the live birth rate between those who took aspirin (251/367; 68.4%) and those who did not (278/438; 63.5%; OR=1.24; 95% CI, 0.93–1.67).

Recommendations from others

The American College of Obstetricians and Gynecologists (ACOG) states that “it has not been shown conclusively that progesterone treatment or corpus luteum support (HCG) influences pregnancy outcome for women with recurrent spontaneous abortion.”⁹ ACOG does not recommend immunotherapy, citing a lack of demonstrated efficacy (IVIG and PLI), a lack of standards for cell storage and administration, and a risk profile similar to that of blood transfusion (PLI). They recommend “couples with otherwise unexplained recurrent pregnancy loss should be counseled regarding the potential for successful pregnancy without treatment.”

Evidence summary

Though individual studies have suggested that therapy based on endoscopy performed before any other study (early endoscopy) may be superior to empiric antisecretory therapy and as efficacious as a “test and treat” strategy in symptom relief, a Cochrane systematic review of 20 RCTs (11 in primary care settings) provides the best evidence on the role of early endoscopy.¹

A subgroup analysis of 5 RCTs, which compared early endoscopy with empiric antisecretory therapy (typically for 4 weeks), revealed that early endoscopy demonstrated a trend towards improvement in self-reported symptoms and in dyspepsia symptom relief scores, but the difference was not statistically significant (relative risk [RR]=0.89; 95% confidence interval [CI], 0.77–1.1). Because each study used different symptom scores, the relative risk as calculated may under-represent the true benefit of early endoscopy when compared with empiric antisecretory therapy.

When patient satisfaction was evaluated, results were dependent on the location of care. In a primary care setting, patients undergoing early endoscopy were as satisfied as those receiving empiric antisecretory therapy.² In a trial of 414 patients randomized after referral to specialty care, patients in the early endoscopy group were more satisfied with their medical care than those receiving empiric antisecretory therapy (RR=0.13; 95% CI, 0.06–0.29).³

Results from studies comparing the benefits of H pylori “test and treat” strategies to early endoscopy are conflicting. A subgroup analysis reported on 3 RCTs from both primary and secondary settings with 931 patients comparing H pylori “test and treat” to initial endoscopy. It found no significant difference in symptom reduction (RR=1.06; 95% CI, 0.98–1.26).¹ A recent follow-up study of 1 of the trials included in the Cochrane systematic review reported on outcomes of a “test and treat” vs early endoscopy strategy at 6 years. There was no difference in days without symptoms demonstrated between the 2 groups (mean difference=0.05; 95% CI, –0.03 to 0.14 days).⁴ Self-reported symptom tracking and a poor response rate (62%) to patient questionnaires reduces the strength of this study’s conclusions.

Formal cost-effective analyses comparing the “test and treat” with early endoscopy strategy have not been done. A subgroup analysis of 4 trials from the Cochrane review (1 from primary care) demonstrated a significant reduction of the number of endoscopies among patients receiving “test and treat” care vs those receiving early endoscopy (RR=0.23; 95% CI, 0.12–0.44). In the long-term follow-up study, fewer antisecretory medication prescriptions were needed by those patients in the “test and treat” group (P=.047).⁴ These figures are more robust; they were obtained from national registry data rather than personal recall and questionnaire submission.

Recommendations from others

Guidelines from the American Gastroenterological Association for the initial approach to young patients with dyspepsia without alarm symptoms is to first “test and treat” for those testing positive for H pylori, prescribe empiric antisecretory therapy for those testing negative, and proceed with endoscopy for recurrent or persistent dyspepsia at 4 to 8 weeks.⁵ The American Society for Gastrointestinal Endoscopy does not recommend any of initial endoscopy, empiric antisecretory therapy, or “test and treat” over another for the reduction of symptoms.⁶ The British Society of Gastroenterology recommends that initial management of dyspepsia consist of empiric acid suppression and H pylori testing. Persons testing positive for H pylori should undergo endoscopy.⁷ The Institute for Clinical Systems Improvement recommends nonurgent upper endoscopy for those aged 50 years and older with symptoms of uncomplicated dyspepsia. They recommend initial H pylori testing and treating those with positive results, and empiric proton pump inhibitor treatment for 4 weeks for those who are H pylori–negative.⁸

Evidence summary

Though individual studies have suggested that therapy based on endoscopy performed before any other study (early endoscopy) may be superior to empiric antisecretory therapy and as efficacious as a “test and treat” strategy in symptom relief, a Cochrane systematic review of 20 RCTs (11 in primary care settings) provides the best evidence on the role of early endoscopy.¹

A subgroup analysis of 5 RCTs, which compared early endoscopy with empiric antisecretory therapy (typically for 4 weeks), revealed that early endoscopy demonstrated a trend towards improvement in self-reported symptoms and in dyspepsia symptom relief scores, but the difference was not statistically significant (relative risk [RR]=0.89; 95% confidence interval [CI], 0.77–1.1). Because each study used different symptom scores, the relative risk as calculated may under-represent the true benefit of early endoscopy when compared with empiric antisecretory therapy.

When patient satisfaction was evaluated, results were dependent on the location of care. In a primary care setting, patients undergoing early endoscopy were as satisfied as those receiving empiric antisecretory therapy.² In a trial of 414 patients randomized after referral to specialty care, patients in the early endoscopy group were more satisfied with their medical care than those receiving empiric antisecretory therapy (RR=0.13; 95% CI, 0.06–0.29).³

Results from studies comparing the benefits of H pylori “test and treat” strategies to early endoscopy are conflicting. A subgroup analysis reported on 3 RCTs from both primary and secondary settings with 931 patients comparing H pylori “test and treat” to initial endoscopy. It found no significant difference in symptom reduction (RR=1.06; 95% CI, 0.98–1.26).¹ A recent follow-up study of 1 of the trials included in the Cochrane systematic review reported on outcomes of a “test and treat” vs early endoscopy strategy at 6 years. There was no difference in days without symptoms demonstrated between the 2 groups (mean difference=0.05; 95% CI, –0.03 to 0.14 days).⁴ Self-reported symptom tracking and a poor response rate (62%) to patient questionnaires reduces the strength of this study’s conclusions.

Formal cost-effective analyses comparing the “test and treat” with early endoscopy strategy have not been done. A subgroup analysis of 4 trials from the Cochrane review (1 from primary care) demonstrated a significant reduction of the number of endoscopies among patients receiving “test and treat” care vs those receiving early endoscopy (RR=0.23; 95% CI, 0.12–0.44). In the long-term follow-up study, fewer antisecretory medication prescriptions were needed by those patients in the “test and treat” group (P=.047).⁴ These figures are more robust; they were obtained from national registry data rather than personal recall and questionnaire submission.

Recommendations from others

Guidelines from the American Gastroenterological Association for the initial approach to young patients with dyspepsia without alarm symptoms is to first “test and treat” for those testing positive for H pylori, prescribe empiric antisecretory therapy for those testing negative, and proceed with endoscopy for recurrent or persistent dyspepsia at 4 to 8 weeks.⁵ The American Society for Gastrointestinal Endoscopy does not recommend any of initial endoscopy, empiric antisecretory therapy, or “test and treat” over another for the reduction of symptoms.⁶ The British Society of Gastroenterology recommends that initial management of dyspepsia consist of empiric acid suppression and H pylori testing. Persons testing positive for H pylori should undergo endoscopy.⁷ The Institute for Clinical Systems Improvement recommends nonurgent upper endoscopy for those aged 50 years and older with symptoms of uncomplicated dyspepsia. They recommend initial H pylori testing and treating those with positive results, and empiric proton pump inhibitor treatment for 4 weeks for those who are H pylori–negative.⁸

Evidence summary

Though individual studies have suggested that therapy based on endoscopy performed before any other study (early endoscopy) may be superior to empiric antisecretory therapy and as efficacious as a “test and treat” strategy in symptom relief, a Cochrane systematic review of 20 RCTs (11 in primary care settings) provides the best evidence on the role of early endoscopy.¹

A subgroup analysis of 5 RCTs, which compared early endoscopy with empiric antisecretory therapy (typically for 4 weeks), revealed that early endoscopy demonstrated a trend towards improvement in self-reported symptoms and in dyspepsia symptom relief scores, but the difference was not statistically significant (relative risk [RR]=0.89; 95% confidence interval [CI], 0.77–1.1). Because each study used different symptom scores, the relative risk as calculated may under-represent the true benefit of early endoscopy when compared with empiric antisecretory therapy.

When patient satisfaction was evaluated, results were dependent on the location of care. In a primary care setting, patients undergoing early endoscopy were as satisfied as those receiving empiric antisecretory therapy.² In a trial of 414 patients randomized after referral to specialty care, patients in the early endoscopy group were more satisfied with their medical care than those receiving empiric antisecretory therapy (RR=0.13; 95% CI, 0.06–0.29).³

Results from studies comparing the benefits of H pylori “test and treat” strategies to early endoscopy are conflicting. A subgroup analysis reported on 3 RCTs from both primary and secondary settings with 931 patients comparing H pylori “test and treat” to initial endoscopy. It found no significant difference in symptom reduction (RR=1.06; 95% CI, 0.98–1.26).¹ A recent follow-up study of 1 of the trials included in the Cochrane systematic review reported on outcomes of a “test and treat” vs early endoscopy strategy at 6 years. There was no difference in days without symptoms demonstrated between the 2 groups (mean difference=0.05; 95% CI, –0.03 to 0.14 days).⁴ Self-reported symptom tracking and a poor response rate (62%) to patient questionnaires reduces the strength of this study’s conclusions.

Formal cost-effective analyses comparing the “test and treat” with early endoscopy strategy have not been done. A subgroup analysis of 4 trials from the Cochrane review (1 from primary care) demonstrated a significant reduction of the number of endoscopies among patients receiving “test and treat” care vs those receiving early endoscopy (RR=0.23; 95% CI, 0.12–0.44). In the long-term follow-up study, fewer antisecretory medication prescriptions were needed by those patients in the “test and treat” group (P=.047).⁴ These figures are more robust; they were obtained from national registry data rather than personal recall and questionnaire submission.

Recommendations from others

Guidelines from the American Gastroenterological Association for the initial approach to young patients with dyspepsia without alarm symptoms is to first “test and treat” for those testing positive for H pylori, prescribe empiric antisecretory therapy for those testing negative, and proceed with endoscopy for recurrent or persistent dyspepsia at 4 to 8 weeks.⁵ The American Society for Gastrointestinal Endoscopy does not recommend any of initial endoscopy, empiric antisecretory therapy, or “test and treat” over another for the reduction of symptoms.⁶ The British Society of Gastroenterology recommends that initial management of dyspepsia consist of empiric acid suppression and H pylori testing. Persons testing positive for H pylori should undergo endoscopy.⁷ The Institute for Clinical Systems Improvement recommends nonurgent upper endoscopy for those aged 50 years and older with symptoms of uncomplicated dyspepsia. They recommend initial H pylori testing and treating those with positive results, and empiric proton pump inhibitor treatment for 4 weeks for those who are H pylori–negative.⁸

Evidence summary

Thrombotic complications are common in lowerlimb fractures. In 1968, a prospective observational study evaluated the natural history of DVT and pulmonary embolism (PE) in tibial fractures treated with open reduction and internal fixation with early mobilization. Seventy-six consecutive patients with 79 tibial fractures were evaluated with venograms, most within 1 month of injury. The overall incidence of thrombosis was 45%. Half were minor, involving 1 to 3 of the paired deep venous trunks of the lower leg without clinical signs of embolism. Twelve patients (16%) had extensive thrombosis, involving 4 to 6 of the deep venous trunks. Three of these had nonfatal PE diagnosed clinically, and 1 had a fatal PE confirmed at autopsy. The mean age of those with extensive thrombosis or PE was 54 years, and these events were uncommon below age 25 years.¹

Incidence of DVT and PE was also evaluated in a cohort of 102 unselected patients who underwent operative fixation for lower-limb fractures, excluding patella, ankle, and foot fractures. All underwent venography approximately 9 days after fixation and were followed clinically for 6 weeks. The overall incidence of DVT was 28% (40% with femoral shaft, 43% with tibial plateau, 22% with tibial shaft, and 12% with tibial plafond [distal articular tibia]). Four developed clinical evidence of PE during hospitalization but only 1 had objective confirmation. None of the patients showed clinical evidence of PE as outpatients.²

LMWH prophylaxis significantly reduced thrombosis in patients with lower-limb fractures in 3 out of 4 RCTs. The first RCT evaluated 253 patients with lower-limb fractures immobilized in plaster casts after surgical fixation. Half the patients received subcutaneous LMWH (nadroparin [Fraxiparin], a European LMWH similar to enoxaparin), and half received no thrombosis prophylaxis. Based on compression ultrasound at the time of cast removal (17 days postinjury, on average), the overall DVT incidence was 11%. Six patients (5%) receiving LMWH had DVTs vs 21 (17%) in the control group (number needed to treat [NNT]=8 to prevent 1 DVT detectible by compression ultrasound). Two thirds of patients with DVT were asymptomatic. One third had clinical signs of DVT, including 1 patient diagnosed with PE on clinical grounds. There was no difference in bleeding complications between the treatment groups.³

A second RCT evaluated LMWH (Mono-Embolex, a European LMWH) prophylaxis in 328 outpatients with lower limb injuries, which included fractures, severe contusions, and ligamentous injuries. All were treated nonsurgically with cast immobilization (mean=18.8 days, range=2–72 days) and 176 patients used daily LMWH injections. All underwent Doppler evaluation for leg thromboses after cast removal, and positive results were confirmed with venograms. Overall, there were no DVTs among the LMWH prophylaxis group and 7 DVTs (4.3%) in the group without LMWH prophylaxis (P<.006). Among those with fractures, the untreated DVT rate was 5.9% (vs 0% with LMWH prophylaxis). Those over age 40 who did not use LMWH had a DVT rate of 11.4% (vs 1.7% in younger patients). Without LMWH prophylaxis, casting for more than 10 days approximately doubled the risk of DVT compared with less than 10 days (6.1% vs 3.1%). This study did not report on the anatomic location of DVTs or if they were clinically evident.⁴

The third RCT evaluated reviparin (another European LMWH) vs placebo in 440 outpatients with lower limb injuries, of whom 293 had fractures. About half had surgical management and all were treated with a plaster cast or brace for an average of 44 days. Most were ambulatory with crutches. All underwent venography within a week of cast removal. The DVT rate for fracture patients using reviparin was 10.4%, vs 18.2% among those without LMWH prophylaxis (absolute risk reduction=7.8%; NNT=12.8). Three fourths of the DVTs were in distal veins, and 21% of the DVTs in the LMWH patients occurred in deep veins compared with 34% in patients without. Two pulmonary emboli occurred, both in patients without LMWH prophylaxis.⁵

The final RCT evaluated tinzaparin (yet another European LMWH) in 300 adult outpatients immobilized in plaster for at least 3 weeks. Most patients (205 out of 300) underwent venography, and the overall DVT rate was 10% (tinzaparin) vs 17% (controls). Among the 150 fracture patients who underwent venography, the DVT rate was 11% (tinzaparin) vs 13% (controls). This difference was not significant, probably due to insufficient numbers. None of the DVTs was clinically detectable.⁶

In hip fracture and hip arthroplasty, warfarin and LMWH are both effective in preventing thrombosis. No studies have specifically evaluated warfarin prophylaxis in lower extremity fractures or compared it with LMWH.

Recommendations from others

The American College of Chest Physicians (ACCP) says that LMWH prophylaxis reduces the risk of asymptomatic DVTs and is standard of care in Europe. The ACCP does not recommend thromboprophylaxis for isolated lower extremity fractures in the US because of cost and insufficient evidence of clinically important reduction in venous thromboembolism (VTE). However, ACCP lists unspecified “lower extremity or pelvic fracture” as a risk factor for VTE, and does recommend that trauma patients with at least 1 risk factor for VTE receive thromboprophylaxis. They make no recommendation about the use of warfarin.⁷

Evidence summary

Thrombotic complications are common in lowerlimb fractures. In 1968, a prospective observational study evaluated the natural history of DVT and pulmonary embolism (PE) in tibial fractures treated with open reduction and internal fixation with early mobilization. Seventy-six consecutive patients with 79 tibial fractures were evaluated with venograms, most within 1 month of injury. The overall incidence of thrombosis was 45%. Half were minor, involving 1 to 3 of the paired deep venous trunks of the lower leg without clinical signs of embolism. Twelve patients (16%) had extensive thrombosis, involving 4 to 6 of the deep venous trunks. Three of these had nonfatal PE diagnosed clinically, and 1 had a fatal PE confirmed at autopsy. The mean age of those with extensive thrombosis or PE was 54 years, and these events were uncommon below age 25 years.¹

Incidence of DVT and PE was also evaluated in a cohort of 102 unselected patients who underwent operative fixation for lower-limb fractures, excluding patella, ankle, and foot fractures. All underwent venography approximately 9 days after fixation and were followed clinically for 6 weeks. The overall incidence of DVT was 28% (40% with femoral shaft, 43% with tibial plateau, 22% with tibial shaft, and 12% with tibial plafond [distal articular tibia]). Four developed clinical evidence of PE during hospitalization but only 1 had objective confirmation. None of the patients showed clinical evidence of PE as outpatients.²

LMWH prophylaxis significantly reduced thrombosis in patients with lower-limb fractures in 3 out of 4 RCTs. The first RCT evaluated 253 patients with lower-limb fractures immobilized in plaster casts after surgical fixation. Half the patients received subcutaneous LMWH (nadroparin [Fraxiparin], a European LMWH similar to enoxaparin), and half received no thrombosis prophylaxis. Based on compression ultrasound at the time of cast removal (17 days postinjury, on average), the overall DVT incidence was 11%. Six patients (5%) receiving LMWH had DVTs vs 21 (17%) in the control group (number needed to treat [NNT]=8 to prevent 1 DVT detectible by compression ultrasound). Two thirds of patients with DVT were asymptomatic. One third had clinical signs of DVT, including 1 patient diagnosed with PE on clinical grounds. There was no difference in bleeding complications between the treatment groups.³

A second RCT evaluated LMWH (Mono-Embolex, a European LMWH) prophylaxis in 328 outpatients with lower limb injuries, which included fractures, severe contusions, and ligamentous injuries. All were treated nonsurgically with cast immobilization (mean=18.8 days, range=2–72 days) and 176 patients used daily LMWH injections. All underwent Doppler evaluation for leg thromboses after cast removal, and positive results were confirmed with venograms. Overall, there were no DVTs among the LMWH prophylaxis group and 7 DVTs (4.3%) in the group without LMWH prophylaxis (P<.006). Among those with fractures, the untreated DVT rate was 5.9% (vs 0% with LMWH prophylaxis). Those over age 40 who did not use LMWH had a DVT rate of 11.4% (vs 1.7% in younger patients). Without LMWH prophylaxis, casting for more than 10 days approximately doubled the risk of DVT compared with less than 10 days (6.1% vs 3.1%). This study did not report on the anatomic location of DVTs or if they were clinically evident.⁴

The third RCT evaluated reviparin (another European LMWH) vs placebo in 440 outpatients with lower limb injuries, of whom 293 had fractures. About half had surgical management and all were treated with a plaster cast or brace for an average of 44 days. Most were ambulatory with crutches. All underwent venography within a week of cast removal. The DVT rate for fracture patients using reviparin was 10.4%, vs 18.2% among those without LMWH prophylaxis (absolute risk reduction=7.8%; NNT=12.8). Three fourths of the DVTs were in distal veins, and 21% of the DVTs in the LMWH patients occurred in deep veins compared with 34% in patients without. Two pulmonary emboli occurred, both in patients without LMWH prophylaxis.⁵

The final RCT evaluated tinzaparin (yet another European LMWH) in 300 adult outpatients immobilized in plaster for at least 3 weeks. Most patients (205 out of 300) underwent venography, and the overall DVT rate was 10% (tinzaparin) vs 17% (controls). Among the 150 fracture patients who underwent venography, the DVT rate was 11% (tinzaparin) vs 13% (controls). This difference was not significant, probably due to insufficient numbers. None of the DVTs was clinically detectable.⁶

In hip fracture and hip arthroplasty, warfarin and LMWH are both effective in preventing thrombosis. No studies have specifically evaluated warfarin prophylaxis in lower extremity fractures or compared it with LMWH.

Recommendations from others

The American College of Chest Physicians (ACCP) says that LMWH prophylaxis reduces the risk of asymptomatic DVTs and is standard of care in Europe. The ACCP does not recommend thromboprophylaxis for isolated lower extremity fractures in the US because of cost and insufficient evidence of clinically important reduction in venous thromboembolism (VTE). However, ACCP lists unspecified “lower extremity or pelvic fracture” as a risk factor for VTE, and does recommend that trauma patients with at least 1 risk factor for VTE receive thromboprophylaxis. They make no recommendation about the use of warfarin.⁷

Evidence summary

Thrombotic complications are common in lowerlimb fractures. In 1968, a prospective observational study evaluated the natural history of DVT and pulmonary embolism (PE) in tibial fractures treated with open reduction and internal fixation with early mobilization. Seventy-six consecutive patients with 79 tibial fractures were evaluated with venograms, most within 1 month of injury. The overall incidence of thrombosis was 45%. Half were minor, involving 1 to 3 of the paired deep venous trunks of the lower leg without clinical signs of embolism. Twelve patients (16%) had extensive thrombosis, involving 4 to 6 of the deep venous trunks. Three of these had nonfatal PE diagnosed clinically, and 1 had a fatal PE confirmed at autopsy. The mean age of those with extensive thrombosis or PE was 54 years, and these events were uncommon below age 25 years.¹

Incidence of DVT and PE was also evaluated in a cohort of 102 unselected patients who underwent operative fixation for lower-limb fractures, excluding patella, ankle, and foot fractures. All underwent venography approximately 9 days after fixation and were followed clinically for 6 weeks. The overall incidence of DVT was 28% (40% with femoral shaft, 43% with tibial plateau, 22% with tibial shaft, and 12% with tibial plafond [distal articular tibia]). Four developed clinical evidence of PE during hospitalization but only 1 had objective confirmation. None of the patients showed clinical evidence of PE as outpatients.²

LMWH prophylaxis significantly reduced thrombosis in patients with lower-limb fractures in 3 out of 4 RCTs. The first RCT evaluated 253 patients with lower-limb fractures immobilized in plaster casts after surgical fixation. Half the patients received subcutaneous LMWH (nadroparin [Fraxiparin], a European LMWH similar to enoxaparin), and half received no thrombosis prophylaxis. Based on compression ultrasound at the time of cast removal (17 days postinjury, on average), the overall DVT incidence was 11%. Six patients (5%) receiving LMWH had DVTs vs 21 (17%) in the control group (number needed to treat [NNT]=8 to prevent 1 DVT detectible by compression ultrasound). Two thirds of patients with DVT were asymptomatic. One third had clinical signs of DVT, including 1 patient diagnosed with PE on clinical grounds. There was no difference in bleeding complications between the treatment groups.³

A second RCT evaluated LMWH (Mono-Embolex, a European LMWH) prophylaxis in 328 outpatients with lower limb injuries, which included fractures, severe contusions, and ligamentous injuries. All were treated nonsurgically with cast immobilization (mean=18.8 days, range=2–72 days) and 176 patients used daily LMWH injections. All underwent Doppler evaluation for leg thromboses after cast removal, and positive results were confirmed with venograms. Overall, there were no DVTs among the LMWH prophylaxis group and 7 DVTs (4.3%) in the group without LMWH prophylaxis (P<.006). Among those with fractures, the untreated DVT rate was 5.9% (vs 0% with LMWH prophylaxis). Those over age 40 who did not use LMWH had a DVT rate of 11.4% (vs 1.7% in younger patients). Without LMWH prophylaxis, casting for more than 10 days approximately doubled the risk of DVT compared with less than 10 days (6.1% vs 3.1%). This study did not report on the anatomic location of DVTs or if they were clinically evident.⁴

The third RCT evaluated reviparin (another European LMWH) vs placebo in 440 outpatients with lower limb injuries, of whom 293 had fractures. About half had surgical management and all were treated with a plaster cast or brace for an average of 44 days. Most were ambulatory with crutches. All underwent venography within a week of cast removal. The DVT rate for fracture patients using reviparin was 10.4%, vs 18.2% among those without LMWH prophylaxis (absolute risk reduction=7.8%; NNT=12.8). Three fourths of the DVTs were in distal veins, and 21% of the DVTs in the LMWH patients occurred in deep veins compared with 34% in patients without. Two pulmonary emboli occurred, both in patients without LMWH prophylaxis.⁵

The final RCT evaluated tinzaparin (yet another European LMWH) in 300 adult outpatients immobilized in plaster for at least 3 weeks. Most patients (205 out of 300) underwent venography, and the overall DVT rate was 10% (tinzaparin) vs 17% (controls). Among the 150 fracture patients who underwent venography, the DVT rate was 11% (tinzaparin) vs 13% (controls). This difference was not significant, probably due to insufficient numbers. None of the DVTs was clinically detectable.⁶

In hip fracture and hip arthroplasty, warfarin and LMWH are both effective in preventing thrombosis. No studies have specifically evaluated warfarin prophylaxis in lower extremity fractures or compared it with LMWH.

Recommendations from others

The American College of Chest Physicians (ACCP) says that LMWH prophylaxis reduces the risk of asymptomatic DVTs and is standard of care in Europe. The ACCP does not recommend thromboprophylaxis for isolated lower extremity fractures in the US because of cost and insufficient evidence of clinically important reduction in venous thromboembolism (VTE). However, ACCP lists unspecified “lower extremity or pelvic fracture” as a risk factor for VTE, and does recommend that trauma patients with at least 1 risk factor for VTE receive thromboprophylaxis. They make no recommendation about the use of warfarin.⁷

Evidence summary

Studies investigating the role of sputum Gram stain and culture are both difficult to interpret and compare. The difficulty in obtaining an adequate sputum sample, variation in preparation, levels of skill in interpretation, and the lack of a gold standard for the microbiologic diagnosis of pneumonia all contribute to these difficulties.¹

The sole meta-analysis identified 12 studies that met 17 specified study criteria regarding the use of sputum Gram stain for patients with com-munity-acquired pneumococcal pneumonia.¹ Sample sizes ranged from 16 to 404; reference standards were most frequently sputum culture but also included culture of transtracheal and bronchial aspirates. Results revealed that patients with community-acquired pneumococcal pneumonia were able to produce a valid sputum sample (≥20 neutrophils, <10 squamous epithe-lial cells per low-power field) 70% of the time; the sensitivity of sputum Gram stain ranged from 15% to 69% (when reviewed by a lab technician); and specificity ranged from 11% to 100%.

Because of the heterogeneity of test characteristics, interpreter skill levels, study populations, and reference standards among the studies in this meta-analysis, no single estimate of Gram stain sensitivity or specificity could be reached. Similarly, information regarding the sensitivity and specificity of sputum culture is lacking. Small studies (n=13–85) using blood culture, transthoracic aspi-rate, or transtracheal aspirate as reference standards in untreated cases of definite pneumococcal pneumonia demonstrate sensitivities ranging from 36% to 100%.² There are no reliable data regarding the specificity of sputum culture.

Recent nonrandomized studies and case series have called into question the role of sputum analysis in CAP. In a case-control study of 605 patients hospitalized with CAP diagnosed by chest x-ray and either cough, chest pain, auscultatory findings, or leukocytosis, establishing an etiologic diagnosis did not influence the choice of antibiotic therapy, length of hospital stay, or mortality.³ Of the 482 patients who had microbiological diagnostics performed (Mycoplasma pneumoniae serology, respiratory virus serology, blood culture, or sputum culture), only 132 (27%) had a presumptive etiologic diagnosis made. Therapy was narrowed or focused in 49 of the 132 (37%) patients who had a presumptive eti-ologic diagnosis, while 84 of the 350 (24%) without a presumptive diagnosis had their therapy narrowed (P>.05). There was no difference in in-hospi-tal changes of therapy, the proportion of new regi-mens having a narrower antimicrobial spectrum than the initial one, length of hospital stay, death in hospital, or death within 3 months after admission.

A prospective study of 74 patients suggested sputum studies had little use in a highly selected population aged <65 years with nonsevere, uncomplicated CAP and no comorbidities. In the 74 patients who produced a valid sputum sample, Gram stain failed to identify the causative agent in any patient (sensitivity 0%), and sputum cultures identified a pathogen in only 4 patients (sensitivity 5%). All patients responded similarly and, even with the identification of a pathogen in 4 patients, there were no changes in initial empiric antibi-otics.⁴ In a retrospective case series, 19 of 54 (35%) patients with SCAP did not respond to initial empiric antibiotics and had a change in their antibiotic regimen. There was no difference in mortality between the group that had empiric antibiotic change (11 patients) and the group that had a change based on sputum culture results (3 patients) (relative risk reduction= –0.14; 95% confidence interval, –0.47 to 0.12).⁵ While these studies suggest the need for re-evaluation of routine sputum analysis, the strength of their conclusions are weakened by lack of randomization, small sample size, inadequate blinding, and lack of control group comparison.

Demographic evidence and nonrandomized trials suggest that patients with CAP who have increased risk of infection from multiple-resistant bacteria, such as patients from long-term care facilities, are a unique population that might need to be evaluated differently. However, the only evidence available regarding the utility of either spu-tum Gram stain or culture for patients with NHAP derives from expert opinion. These authors suggest that determining a causative diagnosis of pneumonia in this population is desirable and postulate that sputum examination would permit recognition of multiply resistant organisms that are being isolated with increasing frequency in long-term care facilities.^6,7 However, the same authors acknowledge that the elderly are often too weak or too confused to provide adequate sputum specimens, resulting in a low diagnostic yield, and no data demonstrate that spu-tum evaluation favorably influences the outcome of pneumonia in these patient populations.

Recommendations from others

The Infectious Disease Society of America (IDSA) and the Canadian Infectious Disease Society/Canadian Thoracic Society (CIDS/CTS) recommend routine sputum analysis for all inpa-tients with CAP or NHAP,^8,9 while the American Thoracic Society (ATS)¹⁰ recommends performing sputum analysis only if a drug-resistant pathogen or an organism not covered by usual empiric therapy is suspected. For those with CAP or NHAP treated as outpatients, the ATS, the IDSA, and the CIDS/CTS recommend microbiological testing only if drug-resistant bacteria or an organism not covered by usual empiric therapy is suspected.

Evidence summary

Studies investigating the role of sputum Gram stain and culture are both difficult to interpret and compare. The difficulty in obtaining an adequate sputum sample, variation in preparation, levels of skill in interpretation, and the lack of a gold standard for the microbiologic diagnosis of pneumonia all contribute to these difficulties.¹

The sole meta-analysis identified 12 studies that met 17 specified study criteria regarding the use of sputum Gram stain for patients with com-munity-acquired pneumococcal pneumonia.¹ Sample sizes ranged from 16 to 404; reference standards were most frequently sputum culture but also included culture of transtracheal and bronchial aspirates. Results revealed that patients with community-acquired pneumococcal pneumonia were able to produce a valid sputum sample (≥20 neutrophils, <10 squamous epithe-lial cells per low-power field) 70% of the time; the sensitivity of sputum Gram stain ranged from 15% to 69% (when reviewed by a lab technician); and specificity ranged from 11% to 100%.

Because of the heterogeneity of test characteristics, interpreter skill levels, study populations, and reference standards among the studies in this meta-analysis, no single estimate of Gram stain sensitivity or specificity could be reached. Similarly, information regarding the sensitivity and specificity of sputum culture is lacking. Small studies (n=13–85) using blood culture, transthoracic aspi-rate, or transtracheal aspirate as reference standards in untreated cases of definite pneumococcal pneumonia demonstrate sensitivities ranging from 36% to 100%.² There are no reliable data regarding the specificity of sputum culture.

Recent nonrandomized studies and case series have called into question the role of sputum analysis in CAP. In a case-control study of 605 patients hospitalized with CAP diagnosed by chest x-ray and either cough, chest pain, auscultatory findings, or leukocytosis, establishing an etiologic diagnosis did not influence the choice of antibiotic therapy, length of hospital stay, or mortality.³ Of the 482 patients who had microbiological diagnostics performed (Mycoplasma pneumoniae serology, respiratory virus serology, blood culture, or sputum culture), only 132 (27%) had a presumptive etiologic diagnosis made. Therapy was narrowed or focused in 49 of the 132 (37%) patients who had a presumptive eti-ologic diagnosis, while 84 of the 350 (24%) without a presumptive diagnosis had their therapy narrowed (P>.05). There was no difference in in-hospi-tal changes of therapy, the proportion of new regi-mens having a narrower antimicrobial spectrum than the initial one, length of hospital stay, death in hospital, or death within 3 months after admission.

A prospective study of 74 patients suggested sputum studies had little use in a highly selected population aged <65 years with nonsevere, uncomplicated CAP and no comorbidities. In the 74 patients who produced a valid sputum sample, Gram stain failed to identify the causative agent in any patient (sensitivity 0%), and sputum cultures identified a pathogen in only 4 patients (sensitivity 5%). All patients responded similarly and, even with the identification of a pathogen in 4 patients, there were no changes in initial empiric antibi-otics.⁴ In a retrospective case series, 19 of 54 (35%) patients with SCAP did not respond to initial empiric antibiotics and had a change in their antibiotic regimen. There was no difference in mortality between the group that had empiric antibiotic change (11 patients) and the group that had a change based on sputum culture results (3 patients) (relative risk reduction= –0.14; 95% confidence interval, –0.47 to 0.12).⁵ While these studies suggest the need for re-evaluation of routine sputum analysis, the strength of their conclusions are weakened by lack of randomization, small sample size, inadequate blinding, and lack of control group comparison.

Demographic evidence and nonrandomized trials suggest that patients with CAP who have increased risk of infection from multiple-resistant bacteria, such as patients from long-term care facilities, are a unique population that might need to be evaluated differently. However, the only evidence available regarding the utility of either spu-tum Gram stain or culture for patients with NHAP derives from expert opinion. These authors suggest that determining a causative diagnosis of pneumonia in this population is desirable and postulate that sputum examination would permit recognition of multiply resistant organisms that are being isolated with increasing frequency in long-term care facilities.^6,7 However, the same authors acknowledge that the elderly are often too weak or too confused to provide adequate sputum specimens, resulting in a low diagnostic yield, and no data demonstrate that spu-tum evaluation favorably influences the outcome of pneumonia in these patient populations.

Recommendations from others

The Infectious Disease Society of America (IDSA) and the Canadian Infectious Disease Society/Canadian Thoracic Society (CIDS/CTS) recommend routine sputum analysis for all inpa-tients with CAP or NHAP,^8,9 while the American Thoracic Society (ATS)¹⁰ recommends performing sputum analysis only if a drug-resistant pathogen or an organism not covered by usual empiric therapy is suspected. For those with CAP or NHAP treated as outpatients, the ATS, the IDSA, and the CIDS/CTS recommend microbiological testing only if drug-resistant bacteria or an organism not covered by usual empiric therapy is suspected.

Evidence summary

Studies investigating the role of sputum Gram stain and culture are both difficult to interpret and compare. The difficulty in obtaining an adequate sputum sample, variation in preparation, levels of skill in interpretation, and the lack of a gold standard for the microbiologic diagnosis of pneumonia all contribute to these difficulties.¹

The sole meta-analysis identified 12 studies that met 17 specified study criteria regarding the use of sputum Gram stain for patients with com-munity-acquired pneumococcal pneumonia.¹ Sample sizes ranged from 16 to 404; reference standards were most frequently sputum culture but also included culture of transtracheal and bronchial aspirates. Results revealed that patients with community-acquired pneumococcal pneumonia were able to produce a valid sputum sample (≥20 neutrophils, <10 squamous epithe-lial cells per low-power field) 70% of the time; the sensitivity of sputum Gram stain ranged from 15% to 69% (when reviewed by a lab technician); and specificity ranged from 11% to 100%.

Because of the heterogeneity of test characteristics, interpreter skill levels, study populations, and reference standards among the studies in this meta-analysis, no single estimate of Gram stain sensitivity or specificity could be reached. Similarly, information regarding the sensitivity and specificity of sputum culture is lacking. Small studies (n=13–85) using blood culture, transthoracic aspi-rate, or transtracheal aspirate as reference standards in untreated cases of definite pneumococcal pneumonia demonstrate sensitivities ranging from 36% to 100%.² There are no reliable data regarding the specificity of sputum culture.

Recent nonrandomized studies and case series have called into question the role of sputum analysis in CAP. In a case-control study of 605 patients hospitalized with CAP diagnosed by chest x-ray and either cough, chest pain, auscultatory findings, or leukocytosis, establishing an etiologic diagnosis did not influence the choice of antibiotic therapy, length of hospital stay, or mortality.³ Of the 482 patients who had microbiological diagnostics performed (Mycoplasma pneumoniae serology, respiratory virus serology, blood culture, or sputum culture), only 132 (27%) had a presumptive etiologic diagnosis made. Therapy was narrowed or focused in 49 of the 132 (37%) patients who had a presumptive eti-ologic diagnosis, while 84 of the 350 (24%) without a presumptive diagnosis had their therapy narrowed (P>.05). There was no difference in in-hospi-tal changes of therapy, the proportion of new regi-mens having a narrower antimicrobial spectrum than the initial one, length of hospital stay, death in hospital, or death within 3 months after admission.

A prospective study of 74 patients suggested sputum studies had little use in a highly selected population aged <65 years with nonsevere, uncomplicated CAP and no comorbidities. In the 74 patients who produced a valid sputum sample, Gram stain failed to identify the causative agent in any patient (sensitivity 0%), and sputum cultures identified a pathogen in only 4 patients (sensitivity 5%). All patients responded similarly and, even with the identification of a pathogen in 4 patients, there were no changes in initial empiric antibi-otics.⁴ In a retrospective case series, 19 of 54 (35%) patients with SCAP did not respond to initial empiric antibiotics and had a change in their antibiotic regimen. There was no difference in mortality between the group that had empiric antibiotic change (11 patients) and the group that had a change based on sputum culture results (3 patients) (relative risk reduction= –0.14; 95% confidence interval, –0.47 to 0.12).⁵ While these studies suggest the need for re-evaluation of routine sputum analysis, the strength of their conclusions are weakened by lack of randomization, small sample size, inadequate blinding, and lack of control group comparison.

Demographic evidence and nonrandomized trials suggest that patients with CAP who have increased risk of infection from multiple-resistant bacteria, such as patients from long-term care facilities, are a unique population that might need to be evaluated differently. However, the only evidence available regarding the utility of either spu-tum Gram stain or culture for patients with NHAP derives from expert opinion. These authors suggest that determining a causative diagnosis of pneumonia in this population is desirable and postulate that sputum examination would permit recognition of multiply resistant organisms that are being isolated with increasing frequency in long-term care facilities.^6,7 However, the same authors acknowledge that the elderly are often too weak or too confused to provide adequate sputum specimens, resulting in a low diagnostic yield, and no data demonstrate that spu-tum evaluation favorably influences the outcome of pneumonia in these patient populations.

Recommendations from others

The Infectious Disease Society of America (IDSA) and the Canadian Infectious Disease Society/Canadian Thoracic Society (CIDS/CTS) recommend routine sputum analysis for all inpa-tients with CAP or NHAP,^8,9 while the American Thoracic Society (ATS)¹⁰ recommends performing sputum analysis only if a drug-resistant pathogen or an organism not covered by usual empiric therapy is suspected. For those with CAP or NHAP treated as outpatients, the ATS, the IDSA, and the CIDS/CTS recommend microbiological testing only if drug-resistant bacteria or an organism not covered by usual empiric therapy is suspected.

Evidence summary

In numerous systematic reviews, RCTs, and metaanalyses, 70% of children responded to stimulant medications with short-term improvements in ADHD symptoms (inattention and hyperactivity/ impulsivity) and academic achievement. A fortyyear review looked at 135 trials and 413 RCTs of methylphenidate in over 19,000 children with an average age of 8.8 years (range, 8.3–9.4 years) for an average duration of 6 weeks (range, 3.3–8.0 weeks).^1-3

Study groups included mostly elementary school–aged male children, with few minorities represented. Comorbid conditions, present in 65% of children with ADHD, were often poorly controlled. Outcome measures varied among studies.³

The effect size from stimulant medication in these studies averaged 0.8 for symptom relief and between 0.4 and 0.5 for academic achievement. (Effect size is the difference between the means of the experimental and control groups expressed in standard deviations. An effect size of 0.2 is considered small, 0.5 is medium, and 0.8 is considered moderate to large.)

A large randomized trial of 579 children with ADHD (20% girls) aged 7 to 9.9 years compared outcomes of 4 treatment strategies: stimulant medication, intensive behavioral treatment, combined stimulant medication and behavioral interventions, and standard community care.⁴ All children met the DSM-IV. criteria for ADHD Combined Type (the most common type of ADHD in this age group). The stimulant medication strategy included an initial dose titration period followed by monthly 30-minute visits. Intensive behavioral treatment involved child, parent, and school personnel components of therapy. Combination therapy added the regimens for medication and behavioral treatment together. Standard community care consisted of usual (nonsystematic) care, evaluated at 6 different sites.

After 14 months of treatment, children in the medication group and the combined treatment groups showed more improvement in ADHD symptoms than children given intensive behavioral treatment or those who received standard community care. When combined with medication, those treated with behavioral therapy showed slight improvement in social skills, anxiety, aggression, oppositional behavior, and academic achievement over medication alone. At the conclusion of the study, 74% of the 212 children on medication were successfully maintained on methylphenidate alone, 10% required dextroamphetamine, and no children required more than one medication. This study found that higher doses of medication with more frequent office follow-up and regular school contact were important features of successful treatment. Only 40% of families were able to complete the intensive behavioral therapy.

Several short-term reviews and meta-analyses show that side effects from stimulant medications are mild and have short duration.⁵ More long-term studies are required to evaluate effects on growth. RCTs have limited power to detect rare adverse events that may be better detected by large observational studies.⁶

Atomoxetine, a specific norepinephrine reuptake inhibitor, is an FDA-approved alternative to stimulants for ADHD treatment in children and adolescents. Based on 3 RCTs⁷ of 588 children between the ages of 7 and 18 years, atomoxetine showed dose-related improvement in ADHD rating scales. Side effects of atomoxetine are similar to stimulants and include mild but significant increases in blood pressure and pulse.⁷

A meta-analysis of 11 non-randomized trials using clonidine for ADHD showed a smaller effect size compared with stimulants.⁸ One RCT of 136 children with ADHD and tics showed improvement of both problems with the use of methylphenidate and clonidine, particularly in combination.⁹ Second-line medications such as clonidine, pemoline (Cylert), and tricyclic antidepressants have more potential serious side effects and are not well studied.¹⁰

Recommendations from others

The American Academy of Pediatrics recommends that clinicians: 1) manage ADHD as a chronic illness, 2) collaborate with parents, the child, and school personnel to define specific desired outcomes, 3) use stimulant or behavioral therapy to improve these outcomes; if one stimulant is not effective at the highest feasible dose, try another, 4) reevaluate the diagnosis, treatment options, adherence, and possible coexisting conditions if treatment is not achieving the desired outcomes, and 5) follow-up regularly with parents, child, and teachers to monitor for progress and adverse effects.¹¹

TABLE
Commonly used medications for ADHD

Evidence summary

In numerous systematic reviews, RCTs, and metaanalyses, 70% of children responded to stimulant medications with short-term improvements in ADHD symptoms (inattention and hyperactivity/ impulsivity) and academic achievement. A fortyyear review looked at 135 trials and 413 RCTs of methylphenidate in over 19,000 children with an average age of 8.8 years (range, 8.3–9.4 years) for an average duration of 6 weeks (range, 3.3–8.0 weeks).^1-3

Study groups included mostly elementary school–aged male children, with few minorities represented. Comorbid conditions, present in 65% of children with ADHD, were often poorly controlled. Outcome measures varied among studies.³

The effect size from stimulant medication in these studies averaged 0.8 for symptom relief and between 0.4 and 0.5 for academic achievement. (Effect size is the difference between the means of the experimental and control groups expressed in standard deviations. An effect size of 0.2 is considered small, 0.5 is medium, and 0.8 is considered moderate to large.)

A large randomized trial of 579 children with ADHD (20% girls) aged 7 to 9.9 years compared outcomes of 4 treatment strategies: stimulant medication, intensive behavioral treatment, combined stimulant medication and behavioral interventions, and standard community care.⁴ All children met the DSM-IV. criteria for ADHD Combined Type (the most common type of ADHD in this age group). The stimulant medication strategy included an initial dose titration period followed by monthly 30-minute visits. Intensive behavioral treatment involved child, parent, and school personnel components of therapy. Combination therapy added the regimens for medication and behavioral treatment together. Standard community care consisted of usual (nonsystematic) care, evaluated at 6 different sites.

After 14 months of treatment, children in the medication group and the combined treatment groups showed more improvement in ADHD symptoms than children given intensive behavioral treatment or those who received standard community care. When combined with medication, those treated with behavioral therapy showed slight improvement in social skills, anxiety, aggression, oppositional behavior, and academic achievement over medication alone. At the conclusion of the study, 74% of the 212 children on medication were successfully maintained on methylphenidate alone, 10% required dextroamphetamine, and no children required more than one medication. This study found that higher doses of medication with more frequent office follow-up and regular school contact were important features of successful treatment. Only 40% of families were able to complete the intensive behavioral therapy.

Several short-term reviews and meta-analyses show that side effects from stimulant medications are mild and have short duration.⁵ More long-term studies are required to evaluate effects on growth. RCTs have limited power to detect rare adverse events that may be better detected by large observational studies.⁶

Atomoxetine, a specific norepinephrine reuptake inhibitor, is an FDA-approved alternative to stimulants for ADHD treatment in children and adolescents. Based on 3 RCTs⁷ of 588 children between the ages of 7 and 18 years, atomoxetine showed dose-related improvement in ADHD rating scales. Side effects of atomoxetine are similar to stimulants and include mild but significant increases in blood pressure and pulse.⁷

A meta-analysis of 11 non-randomized trials using clonidine for ADHD showed a smaller effect size compared with stimulants.⁸ One RCT of 136 children with ADHD and tics showed improvement of both problems with the use of methylphenidate and clonidine, particularly in combination.⁹ Second-line medications such as clonidine, pemoline (Cylert), and tricyclic antidepressants have more potential serious side effects and are not well studied.¹⁰

Recommendations from others

The American Academy of Pediatrics recommends that clinicians: 1) manage ADHD as a chronic illness, 2) collaborate with parents, the child, and school personnel to define specific desired outcomes, 3) use stimulant or behavioral therapy to improve these outcomes; if one stimulant is not effective at the highest feasible dose, try another, 4) reevaluate the diagnosis, treatment options, adherence, and possible coexisting conditions if treatment is not achieving the desired outcomes, and 5) follow-up regularly with parents, child, and teachers to monitor for progress and adverse effects.¹¹

TABLE
Commonly used medications for ADHD

Evidence summary

In numerous systematic reviews, RCTs, and metaanalyses, 70% of children responded to stimulant medications with short-term improvements in ADHD symptoms (inattention and hyperactivity/ impulsivity) and academic achievement. A fortyyear review looked at 135 trials and 413 RCTs of methylphenidate in over 19,000 children with an average age of 8.8 years (range, 8.3–9.4 years) for an average duration of 6 weeks (range, 3.3–8.0 weeks).^1-3

Study groups included mostly elementary school–aged male children, with few minorities represented. Comorbid conditions, present in 65% of children with ADHD, were often poorly controlled. Outcome measures varied among studies.³

The effect size from stimulant medication in these studies averaged 0.8 for symptom relief and between 0.4 and 0.5 for academic achievement. (Effect size is the difference between the means of the experimental and control groups expressed in standard deviations. An effect size of 0.2 is considered small, 0.5 is medium, and 0.8 is considered moderate to large.)

A large randomized trial of 579 children with ADHD (20% girls) aged 7 to 9.9 years compared outcomes of 4 treatment strategies: stimulant medication, intensive behavioral treatment, combined stimulant medication and behavioral interventions, and standard community care.⁴ All children met the DSM-IV. criteria for ADHD Combined Type (the most common type of ADHD in this age group). The stimulant medication strategy included an initial dose titration period followed by monthly 30-minute visits. Intensive behavioral treatment involved child, parent, and school personnel components of therapy. Combination therapy added the regimens for medication and behavioral treatment together. Standard community care consisted of usual (nonsystematic) care, evaluated at 6 different sites.

After 14 months of treatment, children in the medication group and the combined treatment groups showed more improvement in ADHD symptoms than children given intensive behavioral treatment or those who received standard community care. When combined with medication, those treated with behavioral therapy showed slight improvement in social skills, anxiety, aggression, oppositional behavior, and academic achievement over medication alone. At the conclusion of the study, 74% of the 212 children on medication were successfully maintained on methylphenidate alone, 10% required dextroamphetamine, and no children required more than one medication. This study found that higher doses of medication with more frequent office follow-up and regular school contact were important features of successful treatment. Only 40% of families were able to complete the intensive behavioral therapy.

Several short-term reviews and meta-analyses show that side effects from stimulant medications are mild and have short duration.⁵ More long-term studies are required to evaluate effects on growth. RCTs have limited power to detect rare adverse events that may be better detected by large observational studies.⁶

Atomoxetine, a specific norepinephrine reuptake inhibitor, is an FDA-approved alternative to stimulants for ADHD treatment in children and adolescents. Based on 3 RCTs⁷ of 588 children between the ages of 7 and 18 years, atomoxetine showed dose-related improvement in ADHD rating scales. Side effects of atomoxetine are similar to stimulants and include mild but significant increases in blood pressure and pulse.⁷

A meta-analysis of 11 non-randomized trials using clonidine for ADHD showed a smaller effect size compared with stimulants.⁸ One RCT of 136 children with ADHD and tics showed improvement of both problems with the use of methylphenidate and clonidine, particularly in combination.⁹ Second-line medications such as clonidine, pemoline (Cylert), and tricyclic antidepressants have more potential serious side effects and are not well studied.¹⁰

Recommendations from others

The American Academy of Pediatrics recommends that clinicians: 1) manage ADHD as a chronic illness, 2) collaborate with parents, the child, and school personnel to define specific desired outcomes, 3) use stimulant or behavioral therapy to improve these outcomes; if one stimulant is not effective at the highest feasible dose, try another, 4) reevaluate the diagnosis, treatment options, adherence, and possible coexisting conditions if treatment is not achieving the desired outcomes, and 5) follow-up regularly with parents, child, and teachers to monitor for progress and adverse effects.¹¹

TABLE
Commonly used medications for ADHD

Evidence summary

Acetaminophen has a different mechanism of action from other antipyretics. It halts the production of prostaglandin in the brain but not in the periphery, solely lowering fever. Aspirin and other nonsteroidal anti-inflammatory agents inhibit both central and peripheral cyclooxygenase and may cause multiple effects in addition to temperature reduction. Clinical outcome studies of their antipyretic effects are inconclusive.¹

A randomized controlled trial involving 60 volunteers given intranasal rhinovirus type 2 monitored the effect of aspirin, acetaminophen, ibuprofen, or placebo on virus shedding, immune response, and clinical status. There was no difference in duration of illness. There was a trend toward longer duration of virus shedding in the aspirin and acetaminophen groups, but serum neutralizing antibody response was suppressed (P<.05 vs placebo). Aspirin and acetaminophen worsened symptoms of turbinate edema and nasal obstruction (P<.05 vs placebo).²

In 2 double-blind trials, 45 adults infected with rhinovirus were given aspirin or placebo for 5 days, beginning on the day after viral exposure (as opposed to the typical use in response to symptoms). Aspirin treatment improved symptoms of conjunctivitis significantly, but did not change the duration of illness. Other symptoms (headache, sneezing, chills, malaise, nasal discharge) were not significantly different. Aspirin increased the amount of viral shedding by 36% in 1 trial and 17% in the other (P<.01), potentially increasing risk of spread.³

In a randomized controlled trial evaluating antipyretic effects on the duration or severity of childhood varicella, 31 children received placebo and 37 received acetaminophen for 4 days. There was no difference in itching, appetite, activity, or overall condition between the 2 groups. Children treated with acetaminophen took 1.1 days longer to total scabbing (P<.05), although the number of days until the appearance of the last new vesicle and the time to total healing were unchanged. The duration of viral shedding was not measured, but it is possible that the delay in healing of lesions would prolong viral shedding as well.⁴

A retrospective observational study of 54 volunteers demonstrated prolonged illness in subjects infected with influenza A that received antipyretic therapy. Patients who got antipyretics were sick 3.5 days longer than those who did not (8.8 ± 2.3 days vs 5.3 ± 3.0 days; P<.001). Only patients with temperatures >38.9°C on 2 readings 6 hours apart received antipyretics, indicating that the longer course correlated with greater severity of illness as well as with antipyretic use.

In the same study, antipyretics were associated with a trend towards prolonged duration of illness in a group of 21 patients infected with S sonnei (4.6 ± 2.1 days with antipyretics vs 1.9 ± 1.6 days without; P=not significant).⁵

A Cochrane review examined 3 trials of acetaminophen vs placebo for fever in 128 adults and children with P falciparum malaria. Although fever clearance varied between the trials, the malaria cure rate was similar in all, and the review concluded that data were insufficient to evaluate an effect on parasitemia.⁶

Recommendations from others

We found no recommendations regarding the use of antipyretics and their effect on the duration of febrile illness.

Evidence summary

Acetaminophen has a different mechanism of action from other antipyretics. It halts the production of prostaglandin in the brain but not in the periphery, solely lowering fever. Aspirin and other nonsteroidal anti-inflammatory agents inhibit both central and peripheral cyclooxygenase and may cause multiple effects in addition to temperature reduction. Clinical outcome studies of their antipyretic effects are inconclusive.¹

A randomized controlled trial involving 60 volunteers given intranasal rhinovirus type 2 monitored the effect of aspirin, acetaminophen, ibuprofen, or placebo on virus shedding, immune response, and clinical status. There was no difference in duration of illness. There was a trend toward longer duration of virus shedding in the aspirin and acetaminophen groups, but serum neutralizing antibody response was suppressed (P<.05 vs placebo). Aspirin and acetaminophen worsened symptoms of turbinate edema and nasal obstruction (P<.05 vs placebo).²

In 2 double-blind trials, 45 adults infected with rhinovirus were given aspirin or placebo for 5 days, beginning on the day after viral exposure (as opposed to the typical use in response to symptoms). Aspirin treatment improved symptoms of conjunctivitis significantly, but did not change the duration of illness. Other symptoms (headache, sneezing, chills, malaise, nasal discharge) were not significantly different. Aspirin increased the amount of viral shedding by 36% in 1 trial and 17% in the other (P<.01), potentially increasing risk of spread.³

In a randomized controlled trial evaluating antipyretic effects on the duration or severity of childhood varicella, 31 children received placebo and 37 received acetaminophen for 4 days. There was no difference in itching, appetite, activity, or overall condition between the 2 groups. Children treated with acetaminophen took 1.1 days longer to total scabbing (P<.05), although the number of days until the appearance of the last new vesicle and the time to total healing were unchanged. The duration of viral shedding was not measured, but it is possible that the delay in healing of lesions would prolong viral shedding as well.⁴

A retrospective observational study of 54 volunteers demonstrated prolonged illness in subjects infected with influenza A that received antipyretic therapy. Patients who got antipyretics were sick 3.5 days longer than those who did not (8.8 ± 2.3 days vs 5.3 ± 3.0 days; P<.001). Only patients with temperatures >38.9°C on 2 readings 6 hours apart received antipyretics, indicating that the longer course correlated with greater severity of illness as well as with antipyretic use.

In the same study, antipyretics were associated with a trend towards prolonged duration of illness in a group of 21 patients infected with S sonnei (4.6 ± 2.1 days with antipyretics vs 1.9 ± 1.6 days without; P=not significant).⁵

A Cochrane review examined 3 trials of acetaminophen vs placebo for fever in 128 adults and children with P falciparum malaria. Although fever clearance varied between the trials, the malaria cure rate was similar in all, and the review concluded that data were insufficient to evaluate an effect on parasitemia.⁶

Recommendations from others

We found no recommendations regarding the use of antipyretics and their effect on the duration of febrile illness.

Evidence summary

Acetaminophen has a different mechanism of action from other antipyretics. It halts the production of prostaglandin in the brain but not in the periphery, solely lowering fever. Aspirin and other nonsteroidal anti-inflammatory agents inhibit both central and peripheral cyclooxygenase and may cause multiple effects in addition to temperature reduction. Clinical outcome studies of their antipyretic effects are inconclusive.¹

A randomized controlled trial involving 60 volunteers given intranasal rhinovirus type 2 monitored the effect of aspirin, acetaminophen, ibuprofen, or placebo on virus shedding, immune response, and clinical status. There was no difference in duration of illness. There was a trend toward longer duration of virus shedding in the aspirin and acetaminophen groups, but serum neutralizing antibody response was suppressed (P<.05 vs placebo). Aspirin and acetaminophen worsened symptoms of turbinate edema and nasal obstruction (P<.05 vs placebo).²

In 2 double-blind trials, 45 adults infected with rhinovirus were given aspirin or placebo for 5 days, beginning on the day after viral exposure (as opposed to the typical use in response to symptoms). Aspirin treatment improved symptoms of conjunctivitis significantly, but did not change the duration of illness. Other symptoms (headache, sneezing, chills, malaise, nasal discharge) were not significantly different. Aspirin increased the amount of viral shedding by 36% in 1 trial and 17% in the other (P<.01), potentially increasing risk of spread.³

In a randomized controlled trial evaluating antipyretic effects on the duration or severity of childhood varicella, 31 children received placebo and 37 received acetaminophen for 4 days. There was no difference in itching, appetite, activity, or overall condition between the 2 groups. Children treated with acetaminophen took 1.1 days longer to total scabbing (P<.05), although the number of days until the appearance of the last new vesicle and the time to total healing were unchanged. The duration of viral shedding was not measured, but it is possible that the delay in healing of lesions would prolong viral shedding as well.⁴

A retrospective observational study of 54 volunteers demonstrated prolonged illness in subjects infected with influenza A that received antipyretic therapy. Patients who got antipyretics were sick 3.5 days longer than those who did not (8.8 ± 2.3 days vs 5.3 ± 3.0 days; P<.001). Only patients with temperatures >38.9°C on 2 readings 6 hours apart received antipyretics, indicating that the longer course correlated with greater severity of illness as well as with antipyretic use.

In the same study, antipyretics were associated with a trend towards prolonged duration of illness in a group of 21 patients infected with S sonnei (4.6 ± 2.1 days with antipyretics vs 1.9 ± 1.6 days without; P=not significant).⁵

A Cochrane review examined 3 trials of acetaminophen vs placebo for fever in 128 adults and children with P falciparum malaria. Although fever clearance varied between the trials, the malaria cure rate was similar in all, and the review concluded that data were insufficient to evaluate an effect on parasitemia.⁶

Recommendations from others

We found no recommendations regarding the use of antipyretics and their effect on the duration of febrile illness.

Evidence summary

Standardized clinical criteria (Table) exist to determine the risk of serious bacterial infection, which includes meningitis; of particular note, these criteria do not require cerebrospinal fluid examination. Infants aged <3 months who fall into the “high-risk” category or appear toxic have 21% probability of a serious bacterial infection, 10% probability of bacteremia, and 2% probability of bacterial meningitis.¹ The “low-risk” infants have a correspondingly lower incidence of serious bacterial infection: the negative predictive value of the Rochester classification is 98.9% (95% confidence interval [CI], 97.2–99.6%).²

The negative predictive value for bacterial meningitis (a subset of serious bacterial infection) is even greater. Five studies applied the standardized criteria to febrile infants and monitored them for the development of serious bacterial infection, including meningitis. Two prospective cohort studies of outpatients aged 0 to 2 months used the Rochester criteria to assign infants to risk groups. They studied a total of 1294 infants; 659 (51%) were low-risk. None of the low-risk infants developed bacterial meningitis.^2,3

One prospective cohort study of infants aged <1 month hospitalized for fever used a similar method for assessing risk, but added a C-reactive protein value <20 mg/L to criteria for low-risk. Of 250 infants studied, 131 (52%) were low-risk; none of these developed bacterial meningitis.⁴

A retrospective chart review of 492 infants aged <3 months who were hospitalized due to fever included 108 infants aged <1 month. Thirty percent (114) of the infants aged 1 to 3 months and 67% (72) of the younger infants underwent lumbar puncture at the discretion of the treating physician. All infants were retrospectively assigned to low- or high-risk groups for serious bacterial infection using the Rochester criteria. Of the 296 infants rated “low-risk,” none developed bacterial meningitis. Ten of these infants subsequently developed evidence of another bacterial focus (predominantly urinary tract infection).⁵

RECOMMENDATIONS FROM OTHERS

The American Academy of Pediatrics has not issued a clinical practice guideline or clinical report addressing this issue. An evidence-based guideline developed at Cincinnati Children’s Hospital Medical Center in 1998 recommends hospitalization and a full sepsis workup (including lumbar puncture) for infants aged <1 month, or infants aged 1 to 2 months who are high-risk.⁶

A clinical review-based guideline published in 1993 gives the same recommendations.⁷ The expert panel that devised this guideline emphasized a full sepsis evaluation (including cerebrospinal fluid cultures) for infants <28 days of age “despite the low probability of serious bacterial infections in this age group and the favorable outcome of the children managed to date with careful observation.” For low-risk infants aged 1 to 2 months, lumbar puncture is not necessary unless empiric antibiotics are given; having a cerebrospinal fluid culture prior to empiric antibiotics reduces the concern of partially treated meningitis in the case of clinical deterioration after hospital discharge.^6,7

TABLE
How to identify infants at low risk of serious bacterial infection: Rochester Classification

Evidence summary

Standardized clinical criteria (Table) exist to determine the risk of serious bacterial infection, which includes meningitis; of particular note, these criteria do not require cerebrospinal fluid examination. Infants aged <3 months who fall into the “high-risk” category or appear toxic have 21% probability of a serious bacterial infection, 10% probability of bacteremia, and 2% probability of bacterial meningitis.¹ The “low-risk” infants have a correspondingly lower incidence of serious bacterial infection: the negative predictive value of the Rochester classification is 98.9% (95% confidence interval [CI], 97.2–99.6%).²

The negative predictive value for bacterial meningitis (a subset of serious bacterial infection) is even greater. Five studies applied the standardized criteria to febrile infants and monitored them for the development of serious bacterial infection, including meningitis. Two prospective cohort studies of outpatients aged 0 to 2 months used the Rochester criteria to assign infants to risk groups. They studied a total of 1294 infants; 659 (51%) were low-risk. None of the low-risk infants developed bacterial meningitis.^2,3

One prospective cohort study of infants aged <1 month hospitalized for fever used a similar method for assessing risk, but added a C-reactive protein value <20 mg/L to criteria for low-risk. Of 250 infants studied, 131 (52%) were low-risk; none of these developed bacterial meningitis.⁴

A retrospective chart review of 492 infants aged <3 months who were hospitalized due to fever included 108 infants aged <1 month. Thirty percent (114) of the infants aged 1 to 3 months and 67% (72) of the younger infants underwent lumbar puncture at the discretion of the treating physician. All infants were retrospectively assigned to low- or high-risk groups for serious bacterial infection using the Rochester criteria. Of the 296 infants rated “low-risk,” none developed bacterial meningitis. Ten of these infants subsequently developed evidence of another bacterial focus (predominantly urinary tract infection).⁵

RECOMMENDATIONS FROM OTHERS

The American Academy of Pediatrics has not issued a clinical practice guideline or clinical report addressing this issue. An evidence-based guideline developed at Cincinnati Children’s Hospital Medical Center in 1998 recommends hospitalization and a full sepsis workup (including lumbar puncture) for infants aged <1 month, or infants aged 1 to 2 months who are high-risk.⁶

A clinical review-based guideline published in 1993 gives the same recommendations.⁷ The expert panel that devised this guideline emphasized a full sepsis evaluation (including cerebrospinal fluid cultures) for infants <28 days of age “despite the low probability of serious bacterial infections in this age group and the favorable outcome of the children managed to date with careful observation.” For low-risk infants aged 1 to 2 months, lumbar puncture is not necessary unless empiric antibiotics are given; having a cerebrospinal fluid culture prior to empiric antibiotics reduces the concern of partially treated meningitis in the case of clinical deterioration after hospital discharge.^6,7

TABLE
How to identify infants at low risk of serious bacterial infection: Rochester Classification

Evidence summary

Standardized clinical criteria (Table) exist to determine the risk of serious bacterial infection, which includes meningitis; of particular note, these criteria do not require cerebrospinal fluid examination. Infants aged <3 months who fall into the “high-risk” category or appear toxic have 21% probability of a serious bacterial infection, 10% probability of bacteremia, and 2% probability of bacterial meningitis.¹ The “low-risk” infants have a correspondingly lower incidence of serious bacterial infection: the negative predictive value of the Rochester classification is 98.9% (95% confidence interval [CI], 97.2–99.6%).²

The negative predictive value for bacterial meningitis (a subset of serious bacterial infection) is even greater. Five studies applied the standardized criteria to febrile infants and monitored them for the development of serious bacterial infection, including meningitis. Two prospective cohort studies of outpatients aged 0 to 2 months used the Rochester criteria to assign infants to risk groups. They studied a total of 1294 infants; 659 (51%) were low-risk. None of the low-risk infants developed bacterial meningitis.^2,3

One prospective cohort study of infants aged <1 month hospitalized for fever used a similar method for assessing risk, but added a C-reactive protein value <20 mg/L to criteria for low-risk. Of 250 infants studied, 131 (52%) were low-risk; none of these developed bacterial meningitis.⁴

A retrospective chart review of 492 infants aged <3 months who were hospitalized due to fever included 108 infants aged <1 month. Thirty percent (114) of the infants aged 1 to 3 months and 67% (72) of the younger infants underwent lumbar puncture at the discretion of the treating physician. All infants were retrospectively assigned to low- or high-risk groups for serious bacterial infection using the Rochester criteria. Of the 296 infants rated “low-risk,” none developed bacterial meningitis. Ten of these infants subsequently developed evidence of another bacterial focus (predominantly urinary tract infection).⁵

RECOMMENDATIONS FROM OTHERS

The American Academy of Pediatrics has not issued a clinical practice guideline or clinical report addressing this issue. An evidence-based guideline developed at Cincinnati Children’s Hospital Medical Center in 1998 recommends hospitalization and a full sepsis workup (including lumbar puncture) for infants aged <1 month, or infants aged 1 to 2 months who are high-risk.⁶

A clinical review-based guideline published in 1993 gives the same recommendations.⁷ The expert panel that devised this guideline emphasized a full sepsis evaluation (including cerebrospinal fluid cultures) for infants <28 days of age “despite the low probability of serious bacterial infections in this age group and the favorable outcome of the children managed to date with careful observation.” For low-risk infants aged 1 to 2 months, lumbar puncture is not necessary unless empiric antibiotics are given; having a cerebrospinal fluid culture prior to empiric antibiotics reduces the concern of partially treated meningitis in the case of clinical deterioration after hospital discharge.^6,7

TABLE
How to identify infants at low risk of serious bacterial infection: Rochester Classification