Association of Inpatient Postpartum Opioid Use with Bedside Nurse.

Abstract

Our objective was to assess the association between the nurse providing bedside care and women's postpartum opioid use. Retrospective study of all women who birthed at a single center (December 2015 to November 2016). Patient, prescriber, and clinical data were abstracted. The postpartum nurse and total opioid use during the first 12-hour postpartum shift after birth were determined. A high amount of opioid use was defined as morphine milligram equivalents greater than or equal to 90% for this population (stratified by vaginal and cesarean births). A logistic regression model was fit with covariates entered in a step-wise manner to identify the extent to which individual nurses were associated with a greater likelihood of high opioid use by establishing one model in which the only covariate was nurse (model 1) and assessing whether the addition of patient (model 2), birth (model 3), and prescriber factors (model 4) altered the association. Kendall rank correlation assessed rank changes between models. Of the 8376 and 2957 women who had vaginal and cesarean births, 17.9% and 10.2%, respectively, had high opioid use. In the vaginal cohort, women cared for by 46 of 200 nurses were significantly less likely to have high opioid use. Following adjustment, patients cared for by 53 of 200 bedside nurses (model 4) had significantly lower odds of having high opioid use. The rank order of nurses, with respect to the likelihood of opioid use, remained similar after adjustment for patient, birth, and prescriber factors (Τ = 0.84). Findings were similar for the cesarean cohort: 35 of 113 nurses were associated with a significantly lower likelihood of their patients having high opioid use, and the rank order remained similar after covariate adjustment (Τ = 0.78). There is significant variation in postpartum women's opioid use based on the nurse that is not explained by patient, birth, or prescriber factors.