Search

Abstract Background Prognostic factors are associated with the risk of a subsequent outcome in people with a given disease or health condition. Meta-analysis using individual participant data (IPD), where the raw data are synthesised from multiple studies, has been championed as the gold-standard for synthesising prognostic factor studies. We assessed the feasibility and conduct of this approach. Methods A systematic review to identify published IPD meta-analyses of prognostic factors studies, followed by detailed assessment of a random sample of 20 articles published from 2006. Six of these 20 articles were from the IMPACT (International Mission for Prognosis and Analysis of Clinical Trials in traumatic brain injury) collaboration, for which additional information was also used from simultaneously published companion papers. Results Forty-eight published IPD meta-analyses of prognostic factors were identified up to March 2009. Only three were published before 2000 but thereafter a median of four articles exist per year, with traumatic brain injury the most active research field. Availability of IPD offered many advantages, such as checking modelling assumptions; analysing variables on their continuous scale with the possibility of assessing for non-linear relationships; and obtaining results adjusted for other variables. However, researchers also faced many challenges, such as large cost and time required to obtain and clean IPD; unavailable IPD for some studies; different sets of prognostic factors in each study; and variability in study methods of measurement. The IMPACT initiative is a leading example, and had generally strong design, methodological and statistical standards. Elsewhere, standards are not always as high and improvements in the conduct of IPD meta-analyses of prognostic factor studies are often needed; in particular, continuous variables are often categorised without reason; publication bias and availability bias are rarely examined; and important methodological details and summary results are often inadequately reported. Conclusions IPD meta-analyses of prognostic factors are achievable and offer many advantages, as displayed most expertly by the IMPACT initiative. However such projects face numerous logistical and methodological obstacles, and their conduct and reporting can often be substantially improved.

Non-linear exposure-outcome relationships such as between body mass index (BMI) and mortality are common. They are best explored as continuous functions using individual participant data from multiple studies. We explore two two-stage methods for meta-analysis of such relationships, where the confounder-adjusted relationship is first estimated in a non-linear regression model in each study, then combined across studies. The "metacurve" approach combines the estimated curves using multiple meta-analyses of the relative effect between a given exposure level and a reference level. The "mvmeta" approach combines the estimated model parameters in a single multivariate meta-analysis. Both methods allow the exposure-outcome relationship to differ across studies. Using theoretical arguments, we show that the methods differ most when covariate distributions differ across studies; using simulated data, we show that mvmeta gains precision but metacurve is more robust to model mis-specification. We then compare the two methods using data from the Emerging Risk Factors Collaboration on BMI, coronary heart disease events, and all-cause mortality (>80 cohorts, >18 000 events). For each outcome, we model BMI using fractional polynomials of degree 2 in each study, with adjustment for confounders. For metacurve, the powers defining the fractional polynomials may be study-specific or common across studies. For coronary heart disease, metacurve with common powers and mvmeta correctly identify a small increase in risk in the lowest levels of BMI, but metacurve with study-specific powers does not. For all-cause mortality, all methods identify a steep U-shape. The metacurve and mvmeta methods perform well in combining complex exposure-disease relationships across studies. [ABSTRACT FROM AUTHOR]

The identification and assessment of prognostic markers constitutes one of the major tasks in clinical research. Despite huge research effort, the prognostic value of most traditional factors under discussion is uncertain and the usefulness of many specific markers, prognostic indices, and classification schemes is still unproven. Results from different studies are often contradictory, and a general assessment of the usefulness of a specific marker is very difficult. One reason is that systematic reviews of prognostic marker studies have received rather little attention in the literature. It is obvious that a clinically useful and sensible systematic review of a prognostic marker is only possible if the published studies reflect the true nature of the marker and if sufficient details are given in each report. An important goal of a systematic review is to produce a quantitative summary of an effect of interest by a meta-analysis, a statistical approach that combines the results of individual primary studies by a weighted average. For observational studies, an estimate from a univariate model is only of limited interest; a multivariable approach is absolutely essential to derive an estimate that is adjusted for other factors. However, even when “adjusted estimates” are presented, it is common for different studies to use different variables for adjustment, and specific “adjustment” variables may be measured in different ways or may be used with different scales. These difficulties are partly caused by the large variety of available statistical methods of analyzing prognostic marker studies. In three related papers published in a proceedings volume, Holländer and Sauerbrei (2006), Riley et al. (2006), and Altman et al. (2006) discuss statistical approaches for multivariable analysis, issues of reporting of primary studies, and the feasibility of obtaining individual patient data from multiple studies on prognosis. Holländer and Sauerbrei (2006) show that the specific statistical method can have a strong influence on the final multivariable model and on the interpretation of the effect of a specific factor. Possible approaches to help improve reporting standards are discussed in the paper by Riley et al. (2006), which also considers other important issues such as how to improve the design and clinical relevance of primary prognostic studies. For a sensible summary assessment, individual patient data (IPD) and a close collaboration between different study groups seem to be essential. However, Altman et al. (2006) discuss in their paper practical problems in using the IPD approach to evaluate evidence relating to a prognostic marker. Here the three papers are summarized with the aim of demonstrating difficulties and making some recommendations to improve future research in evidence-based assessment of prognostic markers. For many more details we refer to the original papers. [ABSTRACT FROM AUTHOR]

Comments on a meta-analysis of the tumor suppressor protein TP53 which is used as a prognostic marker for head and neck cancer. View that selective reporting biases are a major barrier to conducting meaningful meta-analyses of prognostic marker studies; Association between TP53 status and mortality; Randomized controlled trial.