Advertisement

Causes of Troponin Elevation and Associated Mortality in Young Patients: Methodological Issues

  • Saeid Safiri
    Affiliations
    Managerial Epidemiology Research Center, Department of Public Health, School of Nursing and Midwifery, Maragheh University of Medical Sciences, Maragheh, Iran

    Department of Epidemiology & Biostatistics, School of Public Health, Tehran University of Medical Sciences, Tehran, Iran
    Search for articles by this author
  • Ahad Ashrafi-Asgarabad
    Affiliations
    Department of Epidemiology, School of Public Health, Shahid Beheshti University of Medical Sciences, Tehran, Iran

    Department of Epidemiology, Bam University of Medical Sciences, Bam, Iran
    Search for articles by this author
      To the Editor:
      We attentively read the paper by Wu and colleagues published in The American Journal of Medicine in 2018.
      • Wu C.
      • Singh A.
      • Collins B.
      • et al.
      Causes of troponin elevation and associated mortality in young patients.
      The study was conducted to examine the various causes of troponin elevation in a large cohort of patients younger than 50 years of age, and to evaluate their relationship with all-cause mortality. Although the study conducted by Wu et al provides valuable findings, it is obvious that in order to elude any misinterpretation, some methodological issues need to be considered.
      First, the authors concluded that most nonmyocardial infarction causes of troponin elevation are associated with higher all-cause mortality compared with acute myocardial infarction, which may be misinterpreted. In fact, the study results have not been validated, which can be done internally using robust statistical methods such as Bootstrapping.
      • Steyerberg E.
      Clinical Prediction Models: A Practical Approach to Development, Validation, and Updating.
      Second, composite outcomes such as all-cause mortality were used to increase the study sample size in the previous studies. Now, more sophisticated statistical methods have been introduced to efficiently examine the relationship between exposures of interest with cause-specific mortalities.
      • Richardson D.B.
      • Hamra G.B.
      • MacLehose R.F.
      • Cole S.R.
      • Chu H.
      Hierarchical regression for analyses of multiple outcomes.
      Hence, combining all of the mortalities in one category, all-cause mortality, is not appropriate, as the strength of relationships may not be homogenous across the different cause-specific mortalities and exposures of interest.
      • Richardson D.B.
      • Hamra G.B.
      • MacLehose R.F.
      • Cole S.R.
      • Chu H.
      Hierarchical regression for analyses of multiple outcomes.
      Finally, as is obvious in their study (Supplementary Figure 2), Kaplan-Meier curves of patients with different causes of troponin elevation are not proportional,
      • Wu C.
      • Singh A.
      • Collins B.
      • et al.
      Causes of troponin elevation and associated mortality in young patients.
      and it means that the proportional hazard assumption may be violated. So using the Cox proportional hazard regression model by Wu et al may be problematic, and extended Cox model is proposed when this assumption is violated.
      • Kleinbaum D.G.
      • Klein M.
      Survival Analysis.

      References

        • Wu C.
        • Singh A.
        • Collins B.
        • et al.
        Causes of troponin elevation and associated mortality in young patients.
        Am J Med. 2018; 131: 284-292
        • Steyerberg E.
        Clinical Prediction Models: A Practical Approach to Development, Validation, and Updating.
        Springer Science & Business Media, New York, NY2008
        • Richardson D.B.
        • Hamra G.B.
        • MacLehose R.F.
        • Cole S.R.
        • Chu H.
        Hierarchical regression for analyses of multiple outcomes.
        Am J Epidemiol. 2015; 182: 459-467
        • Kleinbaum D.G.
        • Klein M.
        Survival Analysis.
        Springer, New York, NY2010