Article
Structured reporting to improve transparency of analyses in prognostic biomarker studies
Search Medline for
Authors
-
Willi Sauerbrei
- Institut für Medizinische Biometrie und Informatik, Universitätsklinikum Freiburg, Freiburg, Germany
-
Tim Haeussler
- Institut für Medizinische Biometrie und Informatik, Universitätsklinikum Freiburg, Freiburg, Germany
-
James Balmford
- Institut für Medizinische Biometrie und Informatik, Universitätsklinikum Freiburg, Freiburg, Germany
-
Marianne Huebner
- Department of Statistics and Probability, Michigan State University, East Lansing, United States
| Published: | February 26, 2021 |
|---|
Outline
Text
Background: Reporting guidelines for prognostic tumor biomarker studies and a corresponding ‘explanation and elaboration’ paper have been available for many years [1], [2]. However, a recent review showed that even basic information about study populations and relevant details of statistical analyses are often not provided [3]. In a systematic review Kempf et al. [4] showed that overinterpretation of findings of prognostic biomarker assessment is frequent in high impact journals. Clearly, poorly reported single studies also hinder the conduct of meaningful meta-analyses of prognostic biomarkers [5].
Methods: The two-part REMARK profile, a structured display summarizing key aspects of a study, especially the derivation of the sample and information about the analyses performed, has been proposed to improve completeness and transparency of reporting, specifically of statistical analyses [2]. Created prospectively, it helps authors develop the statistical analysis plan and increases the transparency of the analyses conducted [6]).
We created REMARK profiles for three published biomarker studies with a time-to-event outcome from each of five cancer research journals (BCRT, Cancer, EJC, IJC, JCO). We summarized the analysis steps performed and whether sufficient details of each analysis were provided.
Results: We found that the reporting of analyses was insufficient in nearly all of the studies we reviewed. Concerning the patient population, information about exclusion of patients was incomplete in over half of the studies. Even for the primary outcome, the number of events (the effective sample size) was often not reported, nor was this mentioned for many subgroup analyses.
Conclusions: We argue that the REMARK profile is a suitable instrument to improve the transparency of analyses of prognostic studies. It can also help to reduce the common problem of ‘fishing for significance’ if a statistical analysis plan is registered. These principles can be transferred to many other types of studies.
The authors declare that they have no competing interests.
The authors declare that an ethics committee vote is not required.
References
- 1.
- McShane LM, et al. REporting recommendations for tumor MARKer prognostic studies (REMARK). Journal of the National Cancer Institute. 2005;97:1180-1184.
- 2.
- Altman DG, et al. Reporting recommendations for tumor marker prognostic studies (REMARK): explanation and elaboration. PLoS Med. 2012;9(5): E 1001216.
- 3.
- Sekula P, et al. Did the reporting of prognostic studies of tumour markers improve since the introduction of REMARK guideline? A comparison of reporting in published articles. PLoS ONE. 2017;12(6):e0178531.
- 4.
- Kempf E, et al. Overinterpretation and misreporting of prognostic factor studies in oncology: A systematic review. Br J Cancer. 2018;119(10):1288-1296.
- 5.
- Sauerbrei W, Haeussler T. Comment on “BAG-1 as a biomarker in early breast cancer prognosis: a systematic review with meta-analyses”. British Journal of Cancer. 2018;118:1152-1153.
- 6.
- Winzer KJ, et al. Improving the prognostic ability through better use of standard clinical data – the Nottingham Prognostic Index as an example. PLoS ONE. 2016;11(3):e0149977.
