Researchers are interested in determining whether oral corticosteroid administration is a more effective treatment for patients with multiple sclerosis exacerbation than intravenous corticosteroid administration. They plan to conduct a systematic review and meta-analysis to compare the efficacy of these 2 treatments for acute exacerbations in this population. The primary outcome is defined as the proportion of patients with improvement on the Expanded Disability Status Scale at 4 weeks. The table lists the studies that meet inclusion criteria for the meta-analysis: The researchers included only studies 2, 3, and 5 in their meta-analysis. Which of the following statements concerning the possible results of the meta-analysis is most likely true?
A)Compared with intravenous administration, oral corticosteroid administration will be significantly associated with improvement in scores on the Expanded Disability Status Scale
B)Oral corticosteroid administration will not have a statistically significant benefit over intravenous administration for treating acute relapses
C)Patients on oral corticosteroids will have 70% increased odds of improvement compared with patients on intravenous corticosteroids
D)Results may overestimate the true effect size as they are based on a biased sample of the target population of studies
E)The sample size in the meta-analysis will be too small to detect a significant difference between the 2 treatments

Question

Researchers are interested in determining whether oral corticosteroid administration is a more effective treatment for patients with multiple sclerosis exacerbation than intravenous corticosteroid administration.  They plan to conduct a systematic review and meta-analysis to compare the efficacy of these 2 treatments for acute exacerbations in this population.  The primary outcome is defined as the proportion of patients with improvement on the Expanded Disability Status Scale at 4 weeks.  The table lists the studies that meet inclusion criteria for the meta-analysis: The researchers included only studies 2, 3, and 5 in their meta-analysis.  Which of the following statements concerning the possible results of the meta-analysis is most likely true?
A)Compared with intravenous administration, oral corticosteroid administration will be significantly associated with improvement in scores on the Expanded Disability Status Scale
B)Oral corticosteroid administration will not have a statistically significant benefit over intravenous administration for treating acute relapses
C)Patients on oral corticosteroids will have 70% increased odds of improvement compared with patients on intravenous corticosteroids
D)Results may overestimate the true effect size as they are based on a biased sample of the target population of studies
E)The sample size in the meta-analysis will be too small to detect a significant difference between the 2 treatments

Quizplus · Accepted Answer

Explanation: In meta-analysis studies, individual study results are assessed and weighted to obtain a pooled (ie, mean) estimate for the effect. A meta-analysis should produce an accurate synthesis of the studies included in the analysis; however, if these studies are a biased sample of all relevant studies, the estimated mean effect reflects this bias. Publication bias occurs when studies with significant, positive results are published or considered for inclusion in systematic reviews and meta-analyses, but studies with negative or nonsignificant results are not. In this example, of the 5 studies meeting the inclusion criteria for the study, the researchers included in the meta-analysis only the 3 studies showing a positive association between oral corticosteroid administration and improvement in disability scale scores (ie, with odds ratios [ORs] >1). Consequently, the selected studies represent a biased sample of the population of studies, and any estimate of effect derived from the meta-analysis is likely to be biased. The inclusion of studies with positive associations most likely overestimates the true mean effect of oral corticosteroid administration on disability scale scores in patients with multiple sclerosis. (Choices A, B, and C) A statistically significant improvement in disability scores among patients on oral corticosteroids when compared with patients on intravenous corticosteroids was reported only in study 5 (OR = 1.76, with a confidence interval that does not include the null value of 1.0). The other 4 studies did not reach statistical significance. Only 3 studies, including study 5, were included in the meta-analysis. Therefore, any evidence in favor of or against oral corticosteroid administration on disability scale scores reported in the meta-analysis comes from a biased sample of all relevant studies and is therefore also biased. (Choice E) A meta-analysis combines the results of several studies to increase statistical power through an increased sample size. Ideally, the results are the same as conducting a single study with a larger sample size. Therefore, conducting a meta-analysis will resolve the problem of a small sample size in individual studies. However, in this case, it is unknown whether the meta-analysis will have sufficient sample size to detect statistical significance. Educational objective: Publication bias occurs when studies with positive results are published but studies with negative/null results are not. Meta-analysis yields an accurate synthesis of the studies included in the analysis, but a biased sample of all relevant studies in the literature distorts the true mean effect computed by the meta-analysis and leads to publication bias. References: What is publication bias in a meta-analysis? (http://www.ncbi.nlm.nih.gov/pubmed/26276792)

Researchers Are Interested in Determining Whether Oral Corticosteroid Administration Is