Dataset Information

Untapped potential of multicenter studies: a review of cardiovascular risk prediction models revealed inappropriate analyses and wide variation in reporting.

ABSTRACT:

Background

Clinical prediction models are often constructed using multicenter databases. Such a data structure poses additional challenges for statistical analysis (clustered data) but offers opportunities for model generalizability to a broad range of centers. The purpose of this study was to describe properties, analysis, and reporting of multicenter studies in the Tufts PACE Clinical Prediction Model Registry and to illustrate consequences of common design and analyses choices.

Methods

Fifty randomly selected studies that are included in the Tufts registry as multicenter and published after 2000 underwent full-text screening. Simulated examples illustrate some key concepts relevant to multicenter prediction research.

Results

Multicenter studies differed widely in the number of participating centers (range 2 to 5473). Thirty-nine of 50 studies ignored the multicenter nature of data in the statistical analysis. In the others, clustering was resolved by developing the model on only one center, using mixed effects or stratified regression, or by using center-level characteristics as predictors. Twenty-three of 50 studies did not describe the clinical settings or type of centers from which data was obtained. Four of 50 studies discussed neither generalizability nor external validity of the developed model.

Conclusions

Regression methods and validation strategies tailored to multicenter studies are underutilized. Reporting on generalizability and potential external validity of the model lacks transparency. Hence, multicenter prediction research has untapped potential.

Registration

This review was not registered.

SUBMITTER: Wynants L

PROVIDER: S-EPMC6460661 | biostudies-literature | 2019

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Untapped potential of multicenter studies: a review of cardiovascular risk prediction models revealed inappropriate analyses and wide variation in reporting.

Wynants L L Kent D M DM Timmerman D D Lundquist C M CM Van Calster B B

Diagnostic and prognostic research 20190222

<h4>Background</h4>Clinical prediction models are often constructed using multicenter databases. Such a data structure poses additional challenges for statistical analysis (clustered data) but offers opportunities for model generalizability to a broad range of centers. The purpose of this study was to describe properties, analysis, and reporting of multicenter studies in the Tufts PACE Clinical Prediction Model Registry and to illustrate consequences of common design and analyses choices.<h4>Met ...[more]

PMID: 31093576

Similar Datasets

Project description:IntroductionThe use of cardiovascular magnetic resonance (CMR) for diagnosis and management of a broad range of cardiac and vascular conditions has quickly expanded worldwide. It is essential to understand how CMR is utilized in different regions around the world and the potential practice differences between high-volume and low-volume centers.MethodsCMR practitioners and developers from around the world were electronically surveyed by the Society for Cardiovascular Magnetic Resonance (SCMR) twice, requesting data from 2017. Both surveys were carefully merged, and the data were curated professionally by a data expert using cross-references in key questions and the specific media access control IP address. According to the United Nations classification, responses were analyzed by region and country and interpreted in the context of practice volumes and demography.ResultsFrom 70 countries and regions, 1092 individual responses were included. CMR was performed more often in academic (695/1014, 69%) and hospital settings (522/606, 86%), with adult cardiologists being the primary referring providers (680/818, 83%). Evaluation of cardiomyopathy was the top indication in high-volume and low-volume centers (p = 0.06). High-volume centers were significantly more likely to list evaluation of ischemic heart disease (e.g., stress CMR) as a primary indicator compared to low-volume centers (p < 0.001), while viability assessment was more commonly listed as a primary referral reason in low-volume centers (p = 0.001). Both developed and developing countries noted cost and competing technologies as top barriers to CMR growth. Access to scanners was listed as the most common barrier in developed countries (30% of responders), while lack of training (22% of responders) was the most common barrier in developing countries.ConclusionThis is the most extensive global assessment of CMR practice to date and provides insights from different regions worldwide. We identified CMR as heavily hospital-based, with referral volumes driven primarily by adult cardiology. Indications for CMR utilization varied by center volume. Efforts to improve the adoption and utilization of CMR should include growth beyond the traditional academic, hospital-based location and an emphasis on cardiomyopathy and viability assessment in community centers.

Dataset Information

Untapped potential of multicenter studies: a review of cardiovascular risk prediction models revealed inappropriate analyses and wide variation in reporting.

Background

Methods

Results

Conclusions

Registration

Publications

Untapped potential of multicenter studies: a review of cardiovascular risk prediction models revealed inappropriate analyses and wide variation in reporting.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets