Limits...
An empirical study using permutation-based resampling in meta-regression.

Gagnier JJ, Moher D, Boon H, Bombardier C, Beyene J - Syst Rev (2012)

Bottom Line: We then compared final models and final P values between methods.When applying large sample methods and permutation-based methods in our backwards stepwise regression the covariates in the final models were identical in all cases.The P values for the covariates in the final model were larger in 78% (7/9) of the cases for permutation and identical for 22% (2/9) of the cases.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Orthopaedic Surgery, University of Michigan, Ann Arbor, MI, USA. jgagnier@umich.edu

ABSTRACT

Background: In meta-regression, as the number of trials in the analyses decreases, the risk of false positives or false negatives increases. This is partly due to the assumption of normality that may not hold in small samples. Creation of a distribution from the observed trials using permutation methods to calculate P values may allow for less spurious findings. Permutation has not been empirically tested in meta-regression. The objective of this study was to perform an empirical investigation to explore the differences in results for meta-analyses on a small number of trials using standard large sample approaches verses permutation-based methods for meta-regression.

Methods: We isolated a sample of randomized controlled clinical trials (RCTs) for interventions that have a small number of trials (herbal medicine trials). Trials were then grouped by herbal species and condition and assessed for methodological quality using the Jadad scale, and data were extracted for each outcome. Finally, we performed meta-analyses on the primary outcome of each group of trials and meta-regression for methodological quality subgroups within each meta-analysis. We used large sample methods and permutation methods in our meta-regression modeling. We then compared final models and final P values between methods.

Results: We collected 110 trials across 5 intervention/outcome pairings and 5 to 10 trials per covariate. When applying large sample methods and permutation-based methods in our backwards stepwise regression the covariates in the final models were identical in all cases. The P values for the covariates in the final model were larger in 78% (7/9) of the cases for permutation and identical for 22% (2/9) of the cases.

Conclusions: We present empirical evidence that permutation-based resampling may not change final models when using backwards stepwise regression, but may increase P values in meta-regression of multiple covariates for relatively small amount of trials.

Show MeSH

Related in: MedlinePlus

Flow chart of inclusion of trials.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3351721&req=5

Figure 1: Flow chart of inclusion of trials.

Mentions: A total of 110 trials were included in our analyses (see Figure 1). The intervention outcome pairings are described in Table 1. All other intervention outcome pairings had less than 20 trials and thus were not included. The primary meta-analyses for each of the pairings are listed in Table 2. For all analyses there was a statistically significant effect in favor of the herbal medicine intervention. We also ran a sensitivity analysis removing those trials for which standard deviations were imputed. Results of the sensitivity analyses did not appear to differ from the analyses with all trials included. In the primary analysis there was statistically significant heterogeneity for all analyses with I2 values ranging from 65.7% to 99.1%.


An empirical study using permutation-based resampling in meta-regression.

Gagnier JJ, Moher D, Boon H, Bombardier C, Beyene J - Syst Rev (2012)

Flow chart of inclusion of trials.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3351721&req=5

Figure 1: Flow chart of inclusion of trials.
Mentions: A total of 110 trials were included in our analyses (see Figure 1). The intervention outcome pairings are described in Table 1. All other intervention outcome pairings had less than 20 trials and thus were not included. The primary meta-analyses for each of the pairings are listed in Table 2. For all analyses there was a statistically significant effect in favor of the herbal medicine intervention. We also ran a sensitivity analysis removing those trials for which standard deviations were imputed. Results of the sensitivity analyses did not appear to differ from the analyses with all trials included. In the primary analysis there was statistically significant heterogeneity for all analyses with I2 values ranging from 65.7% to 99.1%.

Bottom Line: We then compared final models and final P values between methods.When applying large sample methods and permutation-based methods in our backwards stepwise regression the covariates in the final models were identical in all cases.The P values for the covariates in the final model were larger in 78% (7/9) of the cases for permutation and identical for 22% (2/9) of the cases.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Orthopaedic Surgery, University of Michigan, Ann Arbor, MI, USA. jgagnier@umich.edu

ABSTRACT

Background: In meta-regression, as the number of trials in the analyses decreases, the risk of false positives or false negatives increases. This is partly due to the assumption of normality that may not hold in small samples. Creation of a distribution from the observed trials using permutation methods to calculate P values may allow for less spurious findings. Permutation has not been empirically tested in meta-regression. The objective of this study was to perform an empirical investigation to explore the differences in results for meta-analyses on a small number of trials using standard large sample approaches verses permutation-based methods for meta-regression.

Methods: We isolated a sample of randomized controlled clinical trials (RCTs) for interventions that have a small number of trials (herbal medicine trials). Trials were then grouped by herbal species and condition and assessed for methodological quality using the Jadad scale, and data were extracted for each outcome. Finally, we performed meta-analyses on the primary outcome of each group of trials and meta-regression for methodological quality subgroups within each meta-analysis. We used large sample methods and permutation methods in our meta-regression modeling. We then compared final models and final P values between methods.

Results: We collected 110 trials across 5 intervention/outcome pairings and 5 to 10 trials per covariate. When applying large sample methods and permutation-based methods in our backwards stepwise regression the covariates in the final models were identical in all cases. The P values for the covariates in the final model were larger in 78% (7/9) of the cases for permutation and identical for 22% (2/9) of the cases.

Conclusions: We present empirical evidence that permutation-based resampling may not change final models when using backwards stepwise regression, but may increase P values in meta-regression of multiple covariates for relatively small amount of trials.

Show MeSH
Related in: MedlinePlus