Prediction Tools for Unfavourable Outcomes in Clostridium difficile Infection: A Systematic Review

摘要: Context Identifying patients at risk for adverse outcomes of Clostridium difficile infection (CDI), including recurrence and death, will become increasingly important as novel therapies emerge, which are more effective than traditional approaches but very expensive. Clinical prediction rules (CPRs) can improve the accuracy medical decision-making. Several CPRs have been developed CDI, none has gained a widespread acceptance. Methods We systematically reviewed studies describing derivation or validation unfavourable in databases (Medline, Embase, PubMed, Web Science Cochrane) abstracts conferences. Results Of 2945 titles screened, 13 on CPR were identified: two recurrences, five complications (including mortality), mortality alone one response to treatment. Two different severity indices also retrieved. Most secondary analyses using cohorts assembled other purposes. presented several methodological limitations that could explain their limited use clinical practice. Except leukocytosis, albumin age, there was much heterogeneity variables used, most by small sample sizes. Eight models used retrospective design. Only four reported incidence outcome interest, even if this is essential evaluate potential usefulness model populations. performed multivariate adjust confounders. Conclusions The lack weighing variables, validation, calibration measures reproducibility, weak validities performances when assessed, absence sensitivity analyses, all led suboptimal quality debatable utility those CPRs. Evidence-based tools through appropriate prospective would be valuable clinicians empirically-developed