An overview of methods for network meta-analysis using individual participant data: when do benefits arise?

Network meta-analysis (NMA) is a common approach to summarizing relative treatment effects from randomized trials with different treatment comparisons. Most NMAs are based on published aggregate data (AD) and have limited possibilities for investigating the extent of network consistency and between-study heterogeneity. Given that individual participant data (IPD) are considered the gold standard in evidence synthesis, we explored statistical methods for IPD-NMA and investigated their potential advantages and limitations, compared with AD-NMA. We discuss several one-stage random-effects NMA models that account for within-trial imbalances, treatment effect modifiers, missing response data and longitudinal responses. We illustrate all models in a case study of 18 antidepressant trials with a continuous endpoint (the Hamilton Depression Score). All trials suffered from drop-out; missingness of longitudinal responses ranged from 21 to 41% after 6 weeks follow-up. Our results indicate that NMA based on IPD may lead to increased precision of estimated treatment effects. Furthermore, it can help to improve network consistency and explain between-study heterogeneity by adjusting for participant-level effect modifiers and adopting more advanced models for dealing with missing response data. We conclude that implementation of IPD-NMA should be considered when trials are affected by substantial drop-out rate, and when treatment effects are potentially influenced by participant-level covariates.

Download Full-text

Individual participant data meta-analysis of prognosis studies

Prognosis Research in Health Care ◽

10.1093/med/9780198796619.003.0014 ◽

2019 ◽

pp. 281-297

Author(s):

Richard D Riley ◽

Thomas PA Debray ◽

Karel GM Moons

Keyword(s):

Evidence Synthesis ◽

Meta Analysis ◽

Prognostic Models ◽

Individual Participant Data ◽

Lack Of Information ◽

Alternative Approach ◽

Individual Participant ◽

Prognosis Research ◽

Key Steps ◽

Access Data

An alternative approach to meta-analysis of aggregate data from published prognosis research (as addressed in Chapter 9), with its challenges of heterogeneity and lack of information, is to conduct meta-analysis of individual participant data (IPD), that is, the original raw data of the individuals who are included in the primary prognosis studies. The approach is increasingly feasible as data sharing and open-access data become more popular, and the chapter highlights why they offer enormous advantages for a robust and meaningful evidence synthesis of prognosis studies. In particular, better prognostic models can be developed and directly validated across multiple settings, and power is increased to detect genuine predictors of treatment response. Key steps in such an IPD meta-analysis are explained, including practical guidance on how to obtain, handle, and synthesize data, and what potential challenges may be encountered.

Download Full-text

Compliance With Preferred Reporting Items for Systematic Review and Meta-Analysis Individual Participant Data Statement for Meta-Analyses Published for Stroke Studies

Stroke ◽

10.1161/strokeaha.120.033288 ◽

2021 ◽

Author(s):

Fareed Jumah ◽

Silky Chotai ◽

Omar Ashraf ◽

Michael S. Rallo ◽

Bharath Raju ◽

...

Keyword(s):

Systematic Review ◽

Impact Factor ◽

Evidence Synthesis ◽

Meta Analysis ◽

Interquartile Range ◽

Individual Participant Data ◽

Related Factors ◽

Scoping Reviews ◽

Individual Participant ◽

Meta Analyses

Background and Purpose: Individual-participant data meta-analyses (IPD-MA) are powerful evidence synthesis studies which are considered the gold-standard of MA. The quality of reporting in these studies is guided by the 2015 Preferred Reporting Items for Systematic Review and Meta-Analysis of Individual Participant Data (PRISMA-IPD) guidelines. The growing number of IPD-MA published for stroke studies calls for an assessment of the compliance of these studies with the PRISMA-IPD statement. Methods: PubMed and EMBASE were searched for MA in stroke published between January 1, 2016, and March 30, 2020, in journals with impact factor >2. Literature reviews, scoping reviews, and aggregate MA were excluded. The final articles were scored using the 31-item PRISMA-IPD checklist. Results were depicted using descriptive statistics. Compliance with each item in PRISM-IPD guideline was recorded. The study was defined as compliant to IPD analyses if it satisfied all IPD specific items. Results: From an initial set of 321 articles, 31 met the final eligibility for data extraction. Only 4 (13%) described the use of PRISMA-IPD guidelines in their methodology, while 8/31 (26%) used the old PRISMA guidelines and 19/31 (61%) followed none. Regardless of mention of using IPD specific guidelines, 42% (n=13) of studies were compliant with all 4 IPD specific domains. The poorest areas of compliance were bias assessment within (32%) and across (39%) studies, reporting protocol and registration (42%), and reporting of IPD integrity (48%). The median journal impact factor was similar between the compliant (median, 8.1 [interquartile range, 5.4–39.9]) and noncompliant (median, 6 [interquartile range, 4.5–16.2]) groups ( P =0.24). Similarly, the journal, country of correspondence, number of authors, number of studies included in MA, study sample size, and funding source were statistically similar between the groups. Conclusions: For the published IPD-MA stroke studies, the compliance with PRISMA-IPD statement and compliance with 4 IPD specific items was suboptimal. The journal, author, and study-related factors were not associated with compliance. Additional scrutiny measures to ensure adherence to mandated guidelines might increase the compliance. Several avenues to improve compliance and ensure optimal adherence are discussed.

Download Full-text

Predicting personalised absolute treatment effects in individual participant data meta‐analysis: an introduction to splines

Research Synthesis Methods ◽

10.1002/jrsm.1546 ◽

2022 ◽

Author(s):

Michail Belias ◽

Maroeska M. Rovers ◽

Jeroen Hoogland ◽

Johannes B. Reitsma ◽

Thomas P. A. Debray ◽

...

Keyword(s):

Treatment Effects ◽

Meta Analysis ◽

Individual Participant Data ◽

Individual Participant

Download Full-text

Impact of BMI on Growth Hormone Stimulation Tests in Children and Adolescents: A Systematic Review and Meta-Analysis

Journal of the Endocrine Society ◽

10.1210/jendso/bvab048.1382 ◽

2021 ◽

Vol 5 (Supplement_1) ◽

pp. A678-A678

Author(s):

Ozair Abawi ◽

Dieuwertje Augustijn ◽

Sanne Hoeks ◽

Yolanda B de Rijke ◽

Erica L T van den Akker

Keyword(s):

Systematic Review ◽

Growth Hormone ◽

Short Stature ◽

Meta Analysis ◽

Overweight And Obesity ◽

Individual Participant Data ◽

Study Heterogeneity ◽

Individual Participant ◽

Stimulation Tests ◽

The Impact

Abstract Background: Peak stimulated growth hormone (GH) levels are known to decrease with increasing BMI, possibly leading to overdiagnosis of GH deficiency (GHD) in children with overweight and obesity. However, current guidelines do not provide guidance how to interpret peak GH values of these children, nor has this been assessed systematically. The aim of this systematic review and meta-analysis was to study the effect of BMI on stimulated peak GH values in children, and to quantify to which extent peak GH values in children with obesity are decreased. Methods: We searched the Medline, Embase, Cochrane, Web of Science, and Google Scholar databases (13 July 2020) for studies reporting impact of BMI on peak GH in children. Where possible, individual participant data was extracted and/or obtained from the authors. Primary outcome was the association between peak GH values and BMI standard deviation score (SDS). Pooled correlation coefficients were calculated under a random effects model, and exploratory moderator analyses and meta-regression were performed. Study heterogeneity was assessed using the I2 statistic. For studies with available individual participant data, linear mixed-models regression analysis was performed with BMI SDS as predictor and ln(peak GH) as outcome, accounting for used GH stimulation agent (fixed effect) and study (random effect). This systematic review was performed in accordance to the PRISMA guidelines. Results: In total, 56 studies were included, providing data on n=5100 children (1346 with individual participant data). Across all studies, a pooled r of -0.37 (95% CI -0.44 to -0.31, n=2785) was found. Study heterogeneity was large (I2=58%). Pubertal status, sex, presence of syndromic obesity, and mean age and BMI SDS of the population did not significantly moderate the pooled r (all p>0.05). Individual participant data analysis revealed a beta of -0.11 (95% CI -0.08 to -0.15, p<0.001), i.e., per 1 point increase in BMI SDS, peak GH decreases by 11% (95% CI 7 to 14%). In the 8 studies performed in children referred for short stature, obesity was present in 27/893 (3.02%) children without GHD and in 36/615 (5.85%) children with GHD (p=0.0069). This corresponds to a RR of 1.43 (95% CI 1.14 to 1.78, p=0.002) for a diagnosis of GHD in children with short stature with obesity compared to children without obesity. Discussion: To our knowledge, this is the first systematic review and meta-analysis to investigate the impact of BMI on peak GH values in children, showing a significant negative correlation and risk of overdiagnosis of GHD in children with obesity. All in all, with ever-rising prevalence of pediatric obesity, our study highlights the urgent need for BMI (SDS)-specific cut-off values for GH stimulation tests in children.

Download Full-text