B40-15 Quantifying Nurse and Clinician Contributions to Racial Disparities in Severity of Illness Among Mechanically Ventilated Patients

Rationale

Population-attributable differences in acute disease severity may impact racial clinical outcome disparities. However, no study has quantified clinician contributions to between-race disparities in disease severity trajectory. Our objective was to quantify and compare independent contributions of nurses and physicians to disease severity trajectories for Black and White patients undergoing mechanical ventilation.

Methods

In a retrospective cohort, we identified mechanically ventilated patients across five hospitals at an academic health system between 2018 and 2022 who identified as having Black or White race. Our outcome was the change in Laboratory Acute Physiology Score 2 during intensive care unit (ICU) admission (ΔLAPS) (higher scores indicate higher mortality risk). Electronic health record data was used to abstract patient race, nurse and physician assignment during 4-hour intervals, patient age, sex, Elixhauser score, admission source, medical or surgical admission, hospital duration prior to mechanical ventilation, day of week, day or night shift, hospital admission LAPS, and study ICU. We adapted value-added modeling to quantify an individual clinician’s value-add as the race-specific incremental change in disease severity independently attributable to that clinician. Clinician value-add was extracted from linear models (with clustered variance) that regressed ΔLAPS against patient race, clinician assignments; clinician assignment-race interaction terms; and all confounding variables. Marginal associations of value-add with ΔLAPS were determined using a linear regression of ΔLAPS on extracted nurse and physician value-add, patient race, and confounders.

Results

Black patients accounted for 38% and 39% of 5,983 distinct patients and 8,259 distinct encounters, respectively. Black race was associated with a more positive ΔLAPS during ICU admission compared to White patients (+4 [0.5, 8], p = 0.026). The highest value-add nurse was associated with ΔLAPS of -54 (-66, -42) for Black patients and -39 (-40, -37) for White patients; the lowest value-add nurse was associated with -29 (-37, -22) for Black patients and -38 (-39, -36) for White patients. The highest value-add physician was associated with -46 (-53, -39) for Black patients and -39 (-41, -38) for White patients; the lowest value-add physician was associated with -31 (-38, -12) for Black patients and -39 (-40, -37) for White patients.

Conclusions

Mechanically ventilated Black patients are independently associated with less pronounced disease severity improvements after ICU admission compared to White patients. Nurses and physicians are independently associated with a range of disease severity trajectory. These findings may guide future interventions to mitigate disparities by optimizing clinical team structure based on clinician performance.

This abstract is funded by: NHLBI