Table 2.

Variables analyzed in the sensitivity analyses, used more than 5 times in the analyzed sample (N = 678)

Variables	N (%)
Various aspects of quality/risk of bias	90 (13)
Statistics and effect sizes (heterogeneity, effect sizes, imprecise effect estimates, intention-to-treat analysis, different methods for effect size calculations, different results, correlation coefficients, meta-regression, imputation of data, different analysis methods, event rate, standard deviation calculated from standard error)	58 (8.6)
Intervention variations	52 (7.7)
Impact of each individual study (sequential exclusion of single studies)	31 (4.6)
Patients’ characteristics (such as smoking, gender, weight)	24 (3.5)
Type of outcomes (such as different pain scales)	23 (3.4)
Type of included studies (crossover studies, randomized controlled trials, non-randomized studies, non-blinded studies, data from retracted studies, mixed data, peer-reviewed manuscripts)	13 (1.9)
Trial size	7 (1.0)
Publication bias	6 (0.9)
Comparator	6 (0.9)
Covariates	5 (0.7)
Type of funding	5 (0.7)