Economic modeling to understand the relationship between clinicians and outcomes of mechanically ventilated patients

点击翻译按钮获取中文摘要

PROJECT SUMMARY Up to one million Americans undergo mechanical ventilation (MV) each year and are subject to high rates of morbidity and mortality. Despite the availability of high-quality evidence for therapies that improve outcomes, adherence to best practices and patient outcomes remain variable. MV patients are nearly universally cared for in intensive care units (ICUs), where interprofessional team-based care – involving physicians, respiratory therapists, and nurses, among others – is the norm. Evidence for the benefits of team-based care has driven recommendations from international societies to support this model; however, best practices for team collaboration remain unknown. We have shown that physicians contribute to variability in outcomes of MV patients, and other studies have demonstrated that specific nursing factors are associated with patient outcomes. However, previous studies have not accounted for the entire team in assessment of individual members. An improved understanding of the roles and contributions of different members of an ICU team would inform practices and interventions that could improve patient outcomes in this high risk population. The primary objective of this proposal, therefore, is to better understand the relative contributions of physicians, nurses, and respiratory therapists individually and as teams to outcomes of patients who undergo MV. We will use “value-added modeling” (VAM), an economic approach developed for empirical individual performance measurement when multiple individuals work in a single process. VAM has been validated for measuring performance of acute care nurses and is uniquely suited to this question, in that it accounts for the multiple clinicians that influence outcomes of individual patients. We will achieve our objective through four aims. First, we will apply VAM to model the relationship between clinicians and patient acuity in a broad population of MV patients, in order to estimate relative clinician performance within each professional group. Second, we will compare the contributions of physician, nurse, and respiratory therapist performance based on VAM across a broad range of patient and economic outcomes. We will quantify the contributions of clinician performance relative to patient factors using the omega statistic, a method for estimating the relative variance explained by different variable groups in a multivariable model. Third, we will evaluate the performance of VAM in a second patient population of MV patients admitted to a national sample of Veterans Administration acute care hospitals, and in patients who do not undergo MV, to evaluate the generalizability of VAM methods to estimate clinician performance. Fourth, we will identify clinician and team factors associated with higher performance using a mixed-methods approach. This study will further our understanding of the relative clinician contributions to patient outcomes using innovative modeling techniques, will contribute to the evidence base for team-based care of critically ill patients, will advance the field of team effectiveness, and will directly lead to future work to reduce care variability among the high-risk population of patients who undergo MV.