Article
Statistical aspects of the allocation of patients for remote patient management via biomarker criteria: a simulation study
Search Medline for
Authors
-
Mareen Pigorsch
- Institute of Biometry and clinical Epidemiology, Charité – Universitätsmedizin Berlin, Berlin, Germany
-
Martin Möckel
- Division of Emergency and Acute Medicine, Cardiovascular Process Research, Charité – Universitätsmedizin Berlin, Berlin, Germany
-
Stefan Gehrig
- Clinical Diagnostics, Thermo Fisher Scientific, Henningsdorf, Germany
-
Jan C. Wiemer
- Clinical Diagnostics, Thermo Fisher Scientific, Henningsdorf, Germany
-
Friedrich Köhler
- Centre for Cardiovascular Telemedicine, Department of Cardiology and Angiology, Charité – Universitätsmedizin Berlin, Berlin, Germany
-
Geraldine Rauch
- Institute of Biometry and clinical Epidemiology, Charité – Universitätsmedizin Berlin, Berlin, Germany
| Published: | February 26, 2021 |
|---|
Outline
Text
Remote patient management (RPM) can be useful to observe early signs of upcoming medical events and enable appropriate care thereby ideally preventing unfavorable events.
While low-risk patients are unlikely to benefit from RPM, the telemedicine-intervention is likely to be most useful for high-risk patients. Since telemedicine is too expensive to be offered to all patients, an efficient allocation is desirable.
Möckel et al. [1] used the following approach to identify patients that should be recommended to RPM: First, all patients including high-risk and low-risk patients were randomly assigned to RPM or standard of care (SOC) within the TIM-HF2 Trial [2], as at the study inclusion the risk profile of the patients is still unknown. The actual population of interest was then a subpopulation of the original one, namely the high-risk patients “recommended for RPM”. At the baseline assessment, biomarkers were recorded in addition, which are highly associated with the outcome event. However, at the time of randomization, there existed no established cut-off values for the biomarkers for allocation of patients for telemedicine. Thus, in principle a diagnostic biomarker trial would have been necessary before using the biomarkers as identifier for the population of interest. As a separate biomarker study was not feasible, Möckel et al. [1] determined the cut-off values for the biomarkers based on outcomes within the SOC-group. These cut-offs were subsequently applied to the RPM- and the SOC-group to identify the subgroup “recommended for RPM”, including only the high-risk patients. The final efficacy analysis was performed in this subgroup by comparing SOC and RPM. Note that this approach is different from a common subgroup-analysis, as the subgroup defines the population of interest here.
In this talk, we examine whether the approach of Möckel et al. [1] leads to biased treatment effect estimates because of defining and applying the cut-off values within the same study. The design by Möckel et al. [1] can be interpreted as a diagnostic-efficacy-combination design. To evaluate the performance of the design, we set up a simulation study that mimics the above approach.
The authors declare that they have no competing interests.
The authors declare that an ethics committee vote is not required.
References
- 1.
- Möckel M, et al. Biomarker Guidance allows a more personalized allocation of patients for Remote Patient Management in Heart Failure Results from the TIM-HF2 Trial. European journal of heart failure. 2019.
- 2.
- Koehler F, et al. Efficacy of telemedical interventional management in patients with heart failure (TIM-HF2): a randomised, controlled, parallel-group, unmasked trial. Lancet. 2018;392(10152): 1047-1057.
