User login
What is the relative cardiovascular benefit of lowering cholesterol, blood pressure, and glucose levels in patients with type 2 diabetes?
ABSTRACT
BACKGROUND: Type 2 diabetes is increasingly recognized as a powerful risk factor for coronary artery disease (CAD) events. In its recommendations for treating cholesterol levels, the Third Adult Treatment Panel of the National Cholesterol Education Program (NCEP) considers diabetes mellitus the equivalent of preexisting CAD.1 The United Kingdom Prospective Diabetes Study (UKPDS) showed that blood pressure control had a greater overall effect on diabetes-related morbidity and mortality than did intensive glucose control.2 The study under consideration examines data from the major trials of cardiovascular risk reduction to determine the relative benefit of controlling blood pressure and cholesterol and glucose levels in patients with type 2 diabetes.
POPULATION STUDIED: Adult patients with diabetes who participated in a variety of studies looking at reduction of risk factors for CAD.
STUDY DESIGN AND VALIDITY: This meta-analysis combined data from previous studies of intensive coronary risk factor reduction in patients with diabetes. The authors searched MEDLINE from 1966 to 2001 for articles published on the topic in English. Studies were included if they were randomized controlled trials of adults that included some patients with diabetes, compared intensive risk factor reduction with drug therapy versus either placebo or routine care, had at least 1 year of follow-up, and reported the requisite cardiovascular outcomes. The studies were independently reviewed by 2 authors for inclusion in the analysis based on these inclusion criteria; disagreement was resolved by consensus. There was no explicit validity assessment of the articles. Data were abstracted in a structured manner. The results were analyzed for heterogeneity and pooled appropriately.
OUTCOMES MEASURED: The outcomes measured included “aggregate cardiac events” (CAD death and nonfatal myocardial infarction [MI]), cardiovascular mortality, MI, and stroke. The results are presented in changes in rates over person-years and as person-years needed to treat. This was done to account for the variable lengths of patient follow-up in these large trials; these findings can be interpreted similarly to standard event rates and numbers needed to treat (NNT). One caveat is that to report an outcome for cholesterol lowering and blood pressure control across a time span of only 1 person-year is artificial, given that most changes in outcomes produced by these therapies take several years to manifest themselves.
RESULTS: Cholesterol lowering (a total of 5 studies of both primary and secondary prevention) reduced aggregate cardiac events (30 vs 41 events per 1000 person-years, NNT for 1 year 106, 95% confidence interval [CI} 62-366). Cholesterol lowering as secondary prevention contributed most to this result (3 trials, 34 vs 44 events per 1000 person-years, NNT for 1 year 120, 95% CI, 61-4856); the results of primary prevention through cholesterol lowering did not reach statistical significance. Blood pressure reduction also reduced aggregate cardiac events (17 vs 23 per 1000 person-years, NNT for 1 year 157, 95% CI, 88-726). Two trials of blood glucose reduction as primary prevention failed to show a significant difference in aggregate cardiac events. The individual cardiac outcomes (cardiovascular mortality and MI each alone) showed results consistent with the aggregate outcomes.
This study reinforces the conclusions of the UKPDS study and the recommendations of the NCEP guidelines that aggressive management of cholesterol and blood pressure in patients with diabetes is essential in preventing CAD. Intensive control of blood sugar levels does not seem to alter CAD events or mortality.
ABSTRACT
BACKGROUND: Type 2 diabetes is increasingly recognized as a powerful risk factor for coronary artery disease (CAD) events. In its recommendations for treating cholesterol levels, the Third Adult Treatment Panel of the National Cholesterol Education Program (NCEP) considers diabetes mellitus the equivalent of preexisting CAD.1 The United Kingdom Prospective Diabetes Study (UKPDS) showed that blood pressure control had a greater overall effect on diabetes-related morbidity and mortality than did intensive glucose control.2 The study under consideration examines data from the major trials of cardiovascular risk reduction to determine the relative benefit of controlling blood pressure and cholesterol and glucose levels in patients with type 2 diabetes.
POPULATION STUDIED: Adult patients with diabetes who participated in a variety of studies looking at reduction of risk factors for CAD.
STUDY DESIGN AND VALIDITY: This meta-analysis combined data from previous studies of intensive coronary risk factor reduction in patients with diabetes. The authors searched MEDLINE from 1966 to 2001 for articles published on the topic in English. Studies were included if they were randomized controlled trials of adults that included some patients with diabetes, compared intensive risk factor reduction with drug therapy versus either placebo or routine care, had at least 1 year of follow-up, and reported the requisite cardiovascular outcomes. The studies were independently reviewed by 2 authors for inclusion in the analysis based on these inclusion criteria; disagreement was resolved by consensus. There was no explicit validity assessment of the articles. Data were abstracted in a structured manner. The results were analyzed for heterogeneity and pooled appropriately.
OUTCOMES MEASURED: The outcomes measured included “aggregate cardiac events” (CAD death and nonfatal myocardial infarction [MI]), cardiovascular mortality, MI, and stroke. The results are presented in changes in rates over person-years and as person-years needed to treat. This was done to account for the variable lengths of patient follow-up in these large trials; these findings can be interpreted similarly to standard event rates and numbers needed to treat (NNT). One caveat is that to report an outcome for cholesterol lowering and blood pressure control across a time span of only 1 person-year is artificial, given that most changes in outcomes produced by these therapies take several years to manifest themselves.
RESULTS: Cholesterol lowering (a total of 5 studies of both primary and secondary prevention) reduced aggregate cardiac events (30 vs 41 events per 1000 person-years, NNT for 1 year 106, 95% confidence interval [CI} 62-366). Cholesterol lowering as secondary prevention contributed most to this result (3 trials, 34 vs 44 events per 1000 person-years, NNT for 1 year 120, 95% CI, 61-4856); the results of primary prevention through cholesterol lowering did not reach statistical significance. Blood pressure reduction also reduced aggregate cardiac events (17 vs 23 per 1000 person-years, NNT for 1 year 157, 95% CI, 88-726). Two trials of blood glucose reduction as primary prevention failed to show a significant difference in aggregate cardiac events. The individual cardiac outcomes (cardiovascular mortality and MI each alone) showed results consistent with the aggregate outcomes.
This study reinforces the conclusions of the UKPDS study and the recommendations of the NCEP guidelines that aggressive management of cholesterol and blood pressure in patients with diabetes is essential in preventing CAD. Intensive control of blood sugar levels does not seem to alter CAD events or mortality.
ABSTRACT
BACKGROUND: Type 2 diabetes is increasingly recognized as a powerful risk factor for coronary artery disease (CAD) events. In its recommendations for treating cholesterol levels, the Third Adult Treatment Panel of the National Cholesterol Education Program (NCEP) considers diabetes mellitus the equivalent of preexisting CAD.1 The United Kingdom Prospective Diabetes Study (UKPDS) showed that blood pressure control had a greater overall effect on diabetes-related morbidity and mortality than did intensive glucose control.2 The study under consideration examines data from the major trials of cardiovascular risk reduction to determine the relative benefit of controlling blood pressure and cholesterol and glucose levels in patients with type 2 diabetes.
POPULATION STUDIED: Adult patients with diabetes who participated in a variety of studies looking at reduction of risk factors for CAD.
STUDY DESIGN AND VALIDITY: This meta-analysis combined data from previous studies of intensive coronary risk factor reduction in patients with diabetes. The authors searched MEDLINE from 1966 to 2001 for articles published on the topic in English. Studies were included if they were randomized controlled trials of adults that included some patients with diabetes, compared intensive risk factor reduction with drug therapy versus either placebo or routine care, had at least 1 year of follow-up, and reported the requisite cardiovascular outcomes. The studies were independently reviewed by 2 authors for inclusion in the analysis based on these inclusion criteria; disagreement was resolved by consensus. There was no explicit validity assessment of the articles. Data were abstracted in a structured manner. The results were analyzed for heterogeneity and pooled appropriately.
OUTCOMES MEASURED: The outcomes measured included “aggregate cardiac events” (CAD death and nonfatal myocardial infarction [MI]), cardiovascular mortality, MI, and stroke. The results are presented in changes in rates over person-years and as person-years needed to treat. This was done to account for the variable lengths of patient follow-up in these large trials; these findings can be interpreted similarly to standard event rates and numbers needed to treat (NNT). One caveat is that to report an outcome for cholesterol lowering and blood pressure control across a time span of only 1 person-year is artificial, given that most changes in outcomes produced by these therapies take several years to manifest themselves.
RESULTS: Cholesterol lowering (a total of 5 studies of both primary and secondary prevention) reduced aggregate cardiac events (30 vs 41 events per 1000 person-years, NNT for 1 year 106, 95% confidence interval [CI} 62-366). Cholesterol lowering as secondary prevention contributed most to this result (3 trials, 34 vs 44 events per 1000 person-years, NNT for 1 year 120, 95% CI, 61-4856); the results of primary prevention through cholesterol lowering did not reach statistical significance. Blood pressure reduction also reduced aggregate cardiac events (17 vs 23 per 1000 person-years, NNT for 1 year 157, 95% CI, 88-726). Two trials of blood glucose reduction as primary prevention failed to show a significant difference in aggregate cardiac events. The individual cardiac outcomes (cardiovascular mortality and MI each alone) showed results consistent with the aggregate outcomes.
This study reinforces the conclusions of the UKPDS study and the recommendations of the NCEP guidelines that aggressive management of cholesterol and blood pressure in patients with diabetes is essential in preventing CAD. Intensive control of blood sugar levels does not seem to alter CAD events or mortality.
Do antibiotics improve neonatal outcomes in preterm premature rupture of membranes (PPROM)?
BACKGROUND: PPROM is the leading cause of preterm birth, occurring in up to 3.5% of all pregnancies. A recent Cochrane review of several smaller studies was unable to show a statistically significant benefit of antibiotic use for PPROM in improving neonatal mortality or morbidity.1 Another study by these researchers found these antibiotics had no significant effect on outcomes in preterm labor without prelabor rupture of membranes.2 The ORACLE Collaborative Group conducted this large multicenter trial to better address the issue.
POPULATION STUDIED: This was a multicenter international study (though it did not include sites in the United States) intended to be reflective of clinical practice. This researchers recruited 4826 pregnant women with PPROM who were at less than 37 weeks’ gestation. Women were excluded if they were previously taking antibiotics, had a known infection requiring antibiotics, or had a contraindication to the study medications. Approximately half of the patients were beyond 32 weeks’ gestation. There were no significant differences between the study groups after randomization in terms of maternal age, gestational age, cervical dilatation, or other medication use.
STUDY DESIGN AND VALIDITY: This was a randomized double-blinded placebo-controlled study. The women were assigned to 1 of 4 treatment groups: erythromycin 250 mg, amoxicillin/clavulanate 250/125 mg, both antibiotics, or placebo. The medications were taken orally 4 times daily for 10 days or until delivery. The strengths of this study include a large sample size, adequate allocation concealment, multiple checks to ensure validity in data collection and treatment assignment, and high follow-up rate (>99.6%). Data were analyzed by intention to treat. This was an overall very well-designed study, though one weakness is its use of a definition of “chronic lung disease” that is simply a neonate’s requirement of oxygen beyond a 36 weeks’ gestational age.
OUTCOMES MEASURED: The composite primary outcome included the overall rate of neonatal death, chronic lung disease, or major cerebral abnormality by ultrasound examination before discharge. Secondary outcomes included delivery within 48 hours, delivery within 7 days, mode of delivery, neonatal infection, number of days in the hospital, and the development of suspected or proven necrotizing enterocolitis.
RESULTS: For the composite primary outcome, there was no statistically significant difference among any of the treatments or placebo. However, there was a nonsignificant trend toward benefit for this outcome in the erythromycin-only group. This outcome may reach statistical significance with a larger sample size. No benefit was found with amoxicillin/clavulanate. In a subgroup analysis of only singleton pregnancies, the researchers found a statistically significant reduction in the composite primary outcome in the erythromycin-only group versus placebo (11.2% vs 14.4%, number needed to treat [NNT]=33).
The important secondary outcomes measured that were of statistical significance included a delay in delivery by at least 7 days (63.3% vs 57.7%, NNT=18) for women taking any antibiotic, as well as reductions in some disease-oriented short-term pulmonary outcomes with erythromycin. However, infants of women receiving amoxicillin/clavulanate were more likely to develop necrotizing enterocolitis (1.8% vs 0.7%, number needed to harm=91), without receiving any benefit from the treatment.
This study confirms the lack of benefit of antibiotics in women with PPROM. There might be a neonatal morbidity and mortality benefit from erythromycin in singleton pregnancies, but further study with longer-term patient-oriented morbidity outcomes (such as clinical diagnosis of chronic lung disease or neurologic abnormalities) is necessary to confirm this observation. This study suggests that the use of any antibiotics after PPROM is associated with delaying delivery, which may be useful in allowing time for full administration of parenteral corticosteroids before delivery. Amoxicillin/clavulanate should not be used for PPROM because of its lack of effectiveness and risk of harm.
BACKGROUND: PPROM is the leading cause of preterm birth, occurring in up to 3.5% of all pregnancies. A recent Cochrane review of several smaller studies was unable to show a statistically significant benefit of antibiotic use for PPROM in improving neonatal mortality or morbidity.1 Another study by these researchers found these antibiotics had no significant effect on outcomes in preterm labor without prelabor rupture of membranes.2 The ORACLE Collaborative Group conducted this large multicenter trial to better address the issue.
POPULATION STUDIED: This was a multicenter international study (though it did not include sites in the United States) intended to be reflective of clinical practice. This researchers recruited 4826 pregnant women with PPROM who were at less than 37 weeks’ gestation. Women were excluded if they were previously taking antibiotics, had a known infection requiring antibiotics, or had a contraindication to the study medications. Approximately half of the patients were beyond 32 weeks’ gestation. There were no significant differences between the study groups after randomization in terms of maternal age, gestational age, cervical dilatation, or other medication use.
STUDY DESIGN AND VALIDITY: This was a randomized double-blinded placebo-controlled study. The women were assigned to 1 of 4 treatment groups: erythromycin 250 mg, amoxicillin/clavulanate 250/125 mg, both antibiotics, or placebo. The medications were taken orally 4 times daily for 10 days or until delivery. The strengths of this study include a large sample size, adequate allocation concealment, multiple checks to ensure validity in data collection and treatment assignment, and high follow-up rate (>99.6%). Data were analyzed by intention to treat. This was an overall very well-designed study, though one weakness is its use of a definition of “chronic lung disease” that is simply a neonate’s requirement of oxygen beyond a 36 weeks’ gestational age.
OUTCOMES MEASURED: The composite primary outcome included the overall rate of neonatal death, chronic lung disease, or major cerebral abnormality by ultrasound examination before discharge. Secondary outcomes included delivery within 48 hours, delivery within 7 days, mode of delivery, neonatal infection, number of days in the hospital, and the development of suspected or proven necrotizing enterocolitis.
RESULTS: For the composite primary outcome, there was no statistically significant difference among any of the treatments or placebo. However, there was a nonsignificant trend toward benefit for this outcome in the erythromycin-only group. This outcome may reach statistical significance with a larger sample size. No benefit was found with amoxicillin/clavulanate. In a subgroup analysis of only singleton pregnancies, the researchers found a statistically significant reduction in the composite primary outcome in the erythromycin-only group versus placebo (11.2% vs 14.4%, number needed to treat [NNT]=33).
The important secondary outcomes measured that were of statistical significance included a delay in delivery by at least 7 days (63.3% vs 57.7%, NNT=18) for women taking any antibiotic, as well as reductions in some disease-oriented short-term pulmonary outcomes with erythromycin. However, infants of women receiving amoxicillin/clavulanate were more likely to develop necrotizing enterocolitis (1.8% vs 0.7%, number needed to harm=91), without receiving any benefit from the treatment.
This study confirms the lack of benefit of antibiotics in women with PPROM. There might be a neonatal morbidity and mortality benefit from erythromycin in singleton pregnancies, but further study with longer-term patient-oriented morbidity outcomes (such as clinical diagnosis of chronic lung disease or neurologic abnormalities) is necessary to confirm this observation. This study suggests that the use of any antibiotics after PPROM is associated with delaying delivery, which may be useful in allowing time for full administration of parenteral corticosteroids before delivery. Amoxicillin/clavulanate should not be used for PPROM because of its lack of effectiveness and risk of harm.
BACKGROUND: PPROM is the leading cause of preterm birth, occurring in up to 3.5% of all pregnancies. A recent Cochrane review of several smaller studies was unable to show a statistically significant benefit of antibiotic use for PPROM in improving neonatal mortality or morbidity.1 Another study by these researchers found these antibiotics had no significant effect on outcomes in preterm labor without prelabor rupture of membranes.2 The ORACLE Collaborative Group conducted this large multicenter trial to better address the issue.
POPULATION STUDIED: This was a multicenter international study (though it did not include sites in the United States) intended to be reflective of clinical practice. This researchers recruited 4826 pregnant women with PPROM who were at less than 37 weeks’ gestation. Women were excluded if they were previously taking antibiotics, had a known infection requiring antibiotics, or had a contraindication to the study medications. Approximately half of the patients were beyond 32 weeks’ gestation. There were no significant differences between the study groups after randomization in terms of maternal age, gestational age, cervical dilatation, or other medication use.
STUDY DESIGN AND VALIDITY: This was a randomized double-blinded placebo-controlled study. The women were assigned to 1 of 4 treatment groups: erythromycin 250 mg, amoxicillin/clavulanate 250/125 mg, both antibiotics, or placebo. The medications were taken orally 4 times daily for 10 days or until delivery. The strengths of this study include a large sample size, adequate allocation concealment, multiple checks to ensure validity in data collection and treatment assignment, and high follow-up rate (>99.6%). Data were analyzed by intention to treat. This was an overall very well-designed study, though one weakness is its use of a definition of “chronic lung disease” that is simply a neonate’s requirement of oxygen beyond a 36 weeks’ gestational age.
OUTCOMES MEASURED: The composite primary outcome included the overall rate of neonatal death, chronic lung disease, or major cerebral abnormality by ultrasound examination before discharge. Secondary outcomes included delivery within 48 hours, delivery within 7 days, mode of delivery, neonatal infection, number of days in the hospital, and the development of suspected or proven necrotizing enterocolitis.
RESULTS: For the composite primary outcome, there was no statistically significant difference among any of the treatments or placebo. However, there was a nonsignificant trend toward benefit for this outcome in the erythromycin-only group. This outcome may reach statistical significance with a larger sample size. No benefit was found with amoxicillin/clavulanate. In a subgroup analysis of only singleton pregnancies, the researchers found a statistically significant reduction in the composite primary outcome in the erythromycin-only group versus placebo (11.2% vs 14.4%, number needed to treat [NNT]=33).
The important secondary outcomes measured that were of statistical significance included a delay in delivery by at least 7 days (63.3% vs 57.7%, NNT=18) for women taking any antibiotic, as well as reductions in some disease-oriented short-term pulmonary outcomes with erythromycin. However, infants of women receiving amoxicillin/clavulanate were more likely to develop necrotizing enterocolitis (1.8% vs 0.7%, number needed to harm=91), without receiving any benefit from the treatment.
This study confirms the lack of benefit of antibiotics in women with PPROM. There might be a neonatal morbidity and mortality benefit from erythromycin in singleton pregnancies, but further study with longer-term patient-oriented morbidity outcomes (such as clinical diagnosis of chronic lung disease or neurologic abnormalities) is necessary to confirm this observation. This study suggests that the use of any antibiotics after PPROM is associated with delaying delivery, which may be useful in allowing time for full administration of parenteral corticosteroids before delivery. Amoxicillin/clavulanate should not be used for PPROM because of its lack of effectiveness and risk of harm.
In children with asthma, do inhaled steroids reduce linear growth (height)?
BACKGROUND: Inhaled corticosteroids have been recommended as an important part of asthma therapy in children; however, there have been concerns about long-term side effects of these medications. In 1994, Allen1 published a meta-analysis of trials of inhaled steroids that suggested that inhaled corticosteroids (beclomethasone dipropionate) were not associated with growth delay. Because of some methodologic criticisms of that meta-analysis and because of 3 newer studies on the topic, the authors of this Cochrane review decided to reexamine the literature.
POPULATION STUDIED: The authors performed an exhaustive search of the literature using the methods typically employed by Cochrane reviewers, including searching the Cochrane Airways Group Asthma Trials registry, searching bibliographies of trials on the subject, and contacting colleagues and researchers in the field. The authors selected only studies involving children (aged <18 years) with asthma who had not been taking inhaled or oral steroids for at least 3 months. These studies had to be randomized controlled trials comparing beclomethasone with nonsteroidal medication and had to have data from which linear growth velocity could be calculated. Interestingly, there was no overlap between the studies evaluated (regardless of inclusion) for this review and the studies included in the meta-analysis by Allen. Only 3 studies were found that met the inclusion criteria. The patients in these studies were diagnosed with clinically stable asthma in the mild to moderate category. No information on the ages of the subjects was available from the primary authors. All of these trials used 200 mg of beclomethasone delivered by diskhaler (a dry powder inhaler) for 7 to 12 months.
STUDY DESIGN AND VALIDITY: The review methodology used was standard for the Cochrane Collaboration. The authors assessed study quality by examining randomization adequacy, allocation concealment, blinding, and description of withdrawals. They performed the final study selection independently and resolved disagreement by consensus. They both abstracted the data and contacted the primary authors of the original studies to fill in data where needed. The appropriate subgroup and sensitivity analyses were planned and performed when necessary.
OUTCOMES MEASURED: The main outcome measure was change in growth velocity (measured in cm/yr).
RESULTS: There were some important differences between the trials that could have implications for the generalizability and validity of the review, such as the definition of asthma and the 10% to 25% dropout rates. The dropout rates were adequately explained in each study, but only one study used intention-to-treat analysis to compensate for the dropout rates. There was a mean reduction in growth velocity of 1.54 cm per year (95% confidence interval, 1.15-1.94 cm/yr), corresponding to a reduction in growth velocity of 25%. The sensitivity analyses performed for methodologic quality, publication bias, and statistical model did not reveal any significant concerns for the validity of the meta-analysis, and there was no significant heterogeneity between the studies.
The authors of this study found a decrease in growth velocity with chronic administration of inhaled beclomethasone in children with asthma. The lack of published data about other inhaled steroids makes it difficult to generalize this finding. This reduction does seem, however, to be independent of the usual confounders of height (eg, severity of asthma, parental height, and so forth). In addition, there were no data in this review concerning final adult height to address the clinical impact of steroids on the age-related change in growth velocity. Given the concerning results of this study, clinicians should counsel families about the beneficial effects of inhaled steroids on controlling asthma and their possible negative effects on growth and should minimize the dose of any required inhaled steroid therapy.
BACKGROUND: Inhaled corticosteroids have been recommended as an important part of asthma therapy in children; however, there have been concerns about long-term side effects of these medications. In 1994, Allen1 published a meta-analysis of trials of inhaled steroids that suggested that inhaled corticosteroids (beclomethasone dipropionate) were not associated with growth delay. Because of some methodologic criticisms of that meta-analysis and because of 3 newer studies on the topic, the authors of this Cochrane review decided to reexamine the literature.
POPULATION STUDIED: The authors performed an exhaustive search of the literature using the methods typically employed by Cochrane reviewers, including searching the Cochrane Airways Group Asthma Trials registry, searching bibliographies of trials on the subject, and contacting colleagues and researchers in the field. The authors selected only studies involving children (aged <18 years) with asthma who had not been taking inhaled or oral steroids for at least 3 months. These studies had to be randomized controlled trials comparing beclomethasone with nonsteroidal medication and had to have data from which linear growth velocity could be calculated. Interestingly, there was no overlap between the studies evaluated (regardless of inclusion) for this review and the studies included in the meta-analysis by Allen. Only 3 studies were found that met the inclusion criteria. The patients in these studies were diagnosed with clinically stable asthma in the mild to moderate category. No information on the ages of the subjects was available from the primary authors. All of these trials used 200 mg of beclomethasone delivered by diskhaler (a dry powder inhaler) for 7 to 12 months.
STUDY DESIGN AND VALIDITY: The review methodology used was standard for the Cochrane Collaboration. The authors assessed study quality by examining randomization adequacy, allocation concealment, blinding, and description of withdrawals. They performed the final study selection independently and resolved disagreement by consensus. They both abstracted the data and contacted the primary authors of the original studies to fill in data where needed. The appropriate subgroup and sensitivity analyses were planned and performed when necessary.
OUTCOMES MEASURED: The main outcome measure was change in growth velocity (measured in cm/yr).
RESULTS: There were some important differences between the trials that could have implications for the generalizability and validity of the review, such as the definition of asthma and the 10% to 25% dropout rates. The dropout rates were adequately explained in each study, but only one study used intention-to-treat analysis to compensate for the dropout rates. There was a mean reduction in growth velocity of 1.54 cm per year (95% confidence interval, 1.15-1.94 cm/yr), corresponding to a reduction in growth velocity of 25%. The sensitivity analyses performed for methodologic quality, publication bias, and statistical model did not reveal any significant concerns for the validity of the meta-analysis, and there was no significant heterogeneity between the studies.
The authors of this study found a decrease in growth velocity with chronic administration of inhaled beclomethasone in children with asthma. The lack of published data about other inhaled steroids makes it difficult to generalize this finding. This reduction does seem, however, to be independent of the usual confounders of height (eg, severity of asthma, parental height, and so forth). In addition, there were no data in this review concerning final adult height to address the clinical impact of steroids on the age-related change in growth velocity. Given the concerning results of this study, clinicians should counsel families about the beneficial effects of inhaled steroids on controlling asthma and their possible negative effects on growth and should minimize the dose of any required inhaled steroid therapy.
BACKGROUND: Inhaled corticosteroids have been recommended as an important part of asthma therapy in children; however, there have been concerns about long-term side effects of these medications. In 1994, Allen1 published a meta-analysis of trials of inhaled steroids that suggested that inhaled corticosteroids (beclomethasone dipropionate) were not associated with growth delay. Because of some methodologic criticisms of that meta-analysis and because of 3 newer studies on the topic, the authors of this Cochrane review decided to reexamine the literature.
POPULATION STUDIED: The authors performed an exhaustive search of the literature using the methods typically employed by Cochrane reviewers, including searching the Cochrane Airways Group Asthma Trials registry, searching bibliographies of trials on the subject, and contacting colleagues and researchers in the field. The authors selected only studies involving children (aged <18 years) with asthma who had not been taking inhaled or oral steroids for at least 3 months. These studies had to be randomized controlled trials comparing beclomethasone with nonsteroidal medication and had to have data from which linear growth velocity could be calculated. Interestingly, there was no overlap between the studies evaluated (regardless of inclusion) for this review and the studies included in the meta-analysis by Allen. Only 3 studies were found that met the inclusion criteria. The patients in these studies were diagnosed with clinically stable asthma in the mild to moderate category. No information on the ages of the subjects was available from the primary authors. All of these trials used 200 mg of beclomethasone delivered by diskhaler (a dry powder inhaler) for 7 to 12 months.
STUDY DESIGN AND VALIDITY: The review methodology used was standard for the Cochrane Collaboration. The authors assessed study quality by examining randomization adequacy, allocation concealment, blinding, and description of withdrawals. They performed the final study selection independently and resolved disagreement by consensus. They both abstracted the data and contacted the primary authors of the original studies to fill in data where needed. The appropriate subgroup and sensitivity analyses were planned and performed when necessary.
OUTCOMES MEASURED: The main outcome measure was change in growth velocity (measured in cm/yr).
RESULTS: There were some important differences between the trials that could have implications for the generalizability and validity of the review, such as the definition of asthma and the 10% to 25% dropout rates. The dropout rates were adequately explained in each study, but only one study used intention-to-treat analysis to compensate for the dropout rates. There was a mean reduction in growth velocity of 1.54 cm per year (95% confidence interval, 1.15-1.94 cm/yr), corresponding to a reduction in growth velocity of 25%. The sensitivity analyses performed for methodologic quality, publication bias, and statistical model did not reveal any significant concerns for the validity of the meta-analysis, and there was no significant heterogeneity between the studies.
The authors of this study found a decrease in growth velocity with chronic administration of inhaled beclomethasone in children with asthma. The lack of published data about other inhaled steroids makes it difficult to generalize this finding. This reduction does seem, however, to be independent of the usual confounders of height (eg, severity of asthma, parental height, and so forth). In addition, there were no data in this review concerning final adult height to address the clinical impact of steroids on the age-related change in growth velocity. Given the concerning results of this study, clinicians should counsel families about the beneficial effects of inhaled steroids on controlling asthma and their possible negative effects on growth and should minimize the dose of any required inhaled steroid therapy.