Article type
Abstract
Background:Current methods for estimating effects of treatments in different subgroups in systematic reviews (SRs) of randomised clinical trials (RCTs) may produce aggregation bias, which is defined as the mistaken assumption that trends seen in aggregated data apply to subgroups that compose the overall sample—without considering the heterogeneity between studies. This could obscure the estimated efficacy/safety within specific subgroups,skewing interpretation and applicability.A within-trial framework to estimate interactions and subgroup specific effects has been developed and showed promising results.
HIV is a major public health concern.There were 39M people living with HIV in 2022, with 1.3M who acquired the virus in the same year, and 630,000 deaths from HIV-related causes.
Overall outcomes in patients with HIV are better when care is delivered by clinicians with HIV expertise.Due to the diversity of clinical categories,stages and comorbidities in HIV,it is relevant to reassess the recommendations made in clinical practice guidelines estimating specific subgroup effects to determine whether or not are differences in the British HIV association (BHIVA) recommendations.
Aim:To assess whether estimating Interactions And Subgroup-specific treatment effects in meta-analysis of RCTs affects the recommendations for initiating antiretroviral therapy(ART) and selecting ART regimens,in the 2022 BHIVA recommendations.
Methods:We retrieved eligible SRs of RCTs cited in the BHIV CPG to retrieve the corresponding recommendations, considering the GRADE certainty of evidence provided by these SRs.Then we made a full text revision of the SRs looking for the presence of subgroups reported on the review .We are extracting data on:Intervention characteristics, results of the covariate of interest and its categories, treatment effect estimators and covariate data. Finally, we will perform within-trial framework analyses, employing the metafloat STATA package to finally evaluate if the novel analyses affect CoE retrieved from SRs and the recommendations.
Results:Preliminarily, five SRs were identified,two of them were discarded because they were based on observational studies. Three SRs included only RCTs, with a total amount including 32 RCTs.All three SRs had subgroup analysis.
Conclusions:The within-trial framework is a paradigm-shifting approach to explore heterogeneity which may affect current recommendations that disregard subgroup-specific treatment effects.We expect our results to be aligned with our hypothesis in BHIVA recommendations.
HIV is a major public health concern.There were 39M people living with HIV in 2022, with 1.3M who acquired the virus in the same year, and 630,000 deaths from HIV-related causes.
Overall outcomes in patients with HIV are better when care is delivered by clinicians with HIV expertise.Due to the diversity of clinical categories,stages and comorbidities in HIV,it is relevant to reassess the recommendations made in clinical practice guidelines estimating specific subgroup effects to determine whether or not are differences in the British HIV association (BHIVA) recommendations.
Aim:To assess whether estimating Interactions And Subgroup-specific treatment effects in meta-analysis of RCTs affects the recommendations for initiating antiretroviral therapy(ART) and selecting ART regimens,in the 2022 BHIVA recommendations.
Methods:We retrieved eligible SRs of RCTs cited in the BHIV CPG to retrieve the corresponding recommendations, considering the GRADE certainty of evidence provided by these SRs.Then we made a full text revision of the SRs looking for the presence of subgroups reported on the review .We are extracting data on:Intervention characteristics, results of the covariate of interest and its categories, treatment effect estimators and covariate data. Finally, we will perform within-trial framework analyses, employing the metafloat STATA package to finally evaluate if the novel analyses affect CoE retrieved from SRs and the recommendations.
Results:Preliminarily, five SRs were identified,two of them were discarded because they were based on observational studies. Three SRs included only RCTs, with a total amount including 32 RCTs.All three SRs had subgroup analysis.
Conclusions:The within-trial framework is a paradigm-shifting approach to explore heterogeneity which may affect current recommendations that disregard subgroup-specific treatment effects.We expect our results to be aligned with our hypothesis in BHIVA recommendations.