Dataset Information

Comparing statistical methods in assessing the prognostic effect of biomarker variability on time-to-event clinical outcomes.

ABSTRACT:

Background

In recent years there is increasing interest in modeling the effect of early longitudinal biomarker data on future time-to-event or other outcomes. Sometimes investigators are also interested in knowing whether the variability of biomarkers is independently predictive of clinical outcomes. This question in most applications is addressed via a two-stage approach where summary statistics such as variance are calculated in the first stage and then used in models as covariates to predict clinical outcome in the second stage. The objective of this study is to compare the relative performance of various methods in estimating the effect of biomarker variability.

Methods

A joint model and 4 different two-stage approaches (naïve, landmark analysis, time-dependent Cox model, and regression calibration) were illustrated using data from a large multi-center randomized phase III trial, the Ocular Hypertension Treatment Study (OHTS), regarding the association between the variability of intraocular pressure (IOP) and the development of primary open-angle glaucoma (POAG). The model performance was also evaluated in terms of bias using simulated data from the joint model of longitudinal IOP and time to POAG. The parameters for simulation were chosen after OHTS data, and the association between longitudinal and survival data was introduced via underlying, unobserved, and error-free parameters including subject-specific variance.

Results

In the OHTS data, joint modeling and two-stage methods reached consistent conclusion that IOP variability showed no significant association with the risk of POAG. In the simulated data with no association between IOP variability and time-to-POAG, all the two-stage methods (except the naïve approach) provided a reliable estimation. When a moderate effect of IOP variability on POAG was imposed, all the two-stage methods underestimated the true association as compared with the joint modeling while the model-based two-stage method (regression calibration) resulted in the least bias.

Conclusion

Regression calibration and joint modelling are the preferred methods in assessing the effect of biomarker variability. Two-stage methods with sample-based measures should be used with caution unless there exists a relatively long series of longitudinal measurements and/or strong effect size (NCT00000125).

SUBMITTER: Gao F

PROVIDER: S-EPMC9308219 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:The usual method for assessing the reliability of survey data has been to conduct reinterviews a short interval (such as one to two weeks) after an initial interview and to use these data to estimate relatively simple statistics, such as gross difference rates (GDRs). More sophisticated approaches have also been used to estimate reliability. These include estimates from multi-trait, multi-method experiments, models applied to longitudinal data, and latent class analyses. To our knowledge, no prior study has systematically compared these different methods for assessing reliability. The Population Assessment of Tobacco and Health Reliability and Validity (PATH-RV) Study, done on a national probability sample, assessed the reliability of answers to the Wave 4 questionnaire from the PATH Study. Respondents in the PATH-RV were interviewed twice about two weeks apart. We examined whether the classic survey approach yielded different conclusions from the more sophisticated methods. We also examined two ex ante methods for assessing problems with survey questions and item nonresponse rates and response times to see how strongly these related to the different reliability estimates. We found that kappa was highly correlated with both GDRs and over-time correlations, but the latter two statistics were less highly correlated, particularly for adult respondents; estimates from longitudinal analyses of the same items in the main PATH study were also highly correlated with the traditional reliability estimates. The latent class analysis results, based on fewer items, also showed a high level of agreement with the traditional measures. The other methods and indicators had at best weak relationships with the reliability estimates derived from the reinterview data. Although the Question Understanding Aid seems to tap a different factor from the other measures, for adult respondents, it did predict item nonresponse and response latencies and thus may be a useful adjunct to the traditional measures.

Project description:BackgroundThere is substantial interest in the adaptation and application of so-called machine learning approaches to prognostic modelling of censored time-to-event data. These methods must be compared and evaluated against existing methods in a variety of scenarios to determine their predictive performance. A scoping review of how machine learning methods have been compared to traditional survival models is important to identify the comparisons that have been made and issues where they are lacking, biased towards one approach or misleading.MethodsWe conducted a scoping review of research articles published between 1 January 2000 and 2 December 2020 using PubMed. Eligible articles were those that used simulation studies to compare statistical and machine learning methods for risk prediction with a time-to-event outcome in a medical/healthcare setting. We focus on data-generating mechanisms (DGMs), the methods that have been compared, the estimands of the simulation studies, and the performance measures used to evaluate them.ResultsA total of ten articles were identified as eligible for the review. Six of the articles evaluated a method that was developed by the authors, four of which were machine learning methods, and the results almost always stated that this developed method's performance was equivalent to or better than the other methods compared. Comparisons were often biased towards the novel approach, with the majority only comparing against a basic Cox proportional hazards model, and in scenarios where it is clear it would not perform well. In many of the articles reviewed, key information was unclear, such as the number of simulation repetitions and how performance measures were calculated.ConclusionIt is vital that method comparisons are unbiased and comprehensive, and this should be the goal even if realising it is difficult. Fully assessing how newly developed methods perform and how they compare to a variety of traditional statistical methods for prognostic modelling is imperative as these methods are already being applied in clinical contexts. Evaluations of the performance and usefulness of recently developed methods for risk prediction should be continued and reporting standards improved as these methods become increasingly popular.