An Adjudication Protocol for Severe Bacterial and Viral Pneumonia

Rationale Clinical endpoints that constitute successful treatment in severe pneumonia are difficult to ascertain and vulnerable to bias. Utility of a protocolized adjudication procedure to determine meaningful endpoints in severe pneumonia is not well described. Objectives To develop and validate a protocol for classification and adjudication of clinical endpoints in severe bacterial and viral pneumonia in a prospective cohort of critically ill, mechanically ventilated patients. Methods Each episode of pneumonia was independently reviewed by two of six pulmonary and critical care physician adjudicators. If a discrepancy in at least one critical answer occurred between the two adjudicators, a third adjudicator reviewed the case and answered the specific question(s) for which there was a lack of agreement. If discrepancy remained after all three adjudications, consensus was achieved through committee review. Results Evaluation of 784 pneumonia episodes during 593 hospitalizations achieved a 79% crude rate of interobserver agreement defined as agreement between 2 of 3 reviewers. Culture-negative pneumonia was associated with increased interobserver agreement. Multiple episodes of pneumonia and bacterial and viral co-infection in the initial episode of pneumonia were associated with decreased interobserver agreement. For bacterial pneumonia, patients with an adjudicated day 7-8 clinical impression of cure for the initial episode of pneumonia were more likely to be discharged alive compared to patients with a day 7-8 clinical impression of indeterminate (p < 0.01), superinfection (p = 0.03), or a combined impression of persistence and superinfection (p = 0.04). In viral pneumonia, patients with an adjudicated clinical impression of cure for an initial episode of viral pneumonia were more likely to be discharged alive compared to patients with an adjudicated clinical impression of persistence (p < 0.01), indeterminate (p < 0.01), or bacterial superinfection (p < 0.01). Conclusions We developed and validated a protocol for classification and adjudication of clinical endpoints in severe pneumonia. This protocol can be applied to cohorts of patients with severe pneumonia to provide uniform assessment of patient-centered endpoints.