Table 5.

Effects of Organization of Hospital Nursing on 30-day Readmissions (N=220,914)

	Unadjusted, separate			Adjusted, separate			Adjusted, joint*
	OR	95% CI	P-value	OR	95% CI	P-value	OR	95% CI	P-value
Work environment	0.95	0.92-0.97	0.000	0.97	0.95-0.99	0.003	0.97	0.95-1.00	0.030
Nurse staffing†	1.03	1.00-1.05	0.027	1.03	1.00-1.05	0.015	1.01	0.99-1.04	0.225
Nurse education	1.22	1.01-1.48	0.039	1.06	0.90-1.24	0.504	-	-	-
Work environment attributes
Hospital affairs participation	0.96	0.94-0.99	0.007	0.98	0.96-1.00	0.062	0.99	0.96-1.01	0.233
Foundations for care quality	0.95	0.93-0.98	0.000	0.97	0.95-0.99	0.010	0.98	0.96-1.00	0.110
Administrative support	0.93	0.91-0.96	0.000	0.96	0.94-0.99	0.003	0.98	0.99-0.99	0.009
RN-MD relations	0.95	0.93-0.98	0.000	0.97	0.95-0.99	0.002	0.97	0.95-1.00	0.031

Notes: RN, registered nurse; MD, medical doctor.

^*ROC (a measure of model performance) ranged from 0.6576-0.6577, which are similar to other reports.

Control variables in adjusted models (adjusted separate and adjusted joint) included patient characteristics (age, gender, race/ethnicity, number of hospitalizations within 180 days prior to index admission, 27 comorbid conditions, and surgery DRGs) and hospital characteristics (bed size, teaching status, ownership, technology level, and hospital geographic area population density).

The adjusted separate models examined each hospital nursing variable separately when controlling for patient and hospital covariates. The adjusted joint models included both work environment (overall work environment measure or each attribute of the work environment respectively) and staffing in the same model when controlling for patient and hospital covariates.

^†The estimates of the effect of nurse staffing on readmission in adjusted joint models were similar when we included either the overall work environment measure or each subscale measure. Therefore we only reported the results from the adjusted join model with both nurse staffing and overall work environment measure.