Mixed-effects Poisson regression: association between staffing and high acuity missed observations with (A) and without (B) inclusion of a linear interaction term between RN and NA staffing levels
| Model | A | B | ||||
| IRR | 95% CI | P value | IRR | 95% CI | P value | |
| RN staffing | 0.982 | 0.972 to 0.992 | <0.001 | 0.982 | 0.972 to 0.992 | <0.001 |
| NA staffing | 1 | 0.990 to 1.01 | 0.822 | 1 | 0.991 to 1.01 | 0.791 |
| Patient turnover | 0.997 | 0.988 to 1.01 | 0.59 | 0.997 | 0.988 to 1.01 | 0.582 |
| Observations in higher acuity patients | 1.01 | 0.936 to 1.09 | 0.769 | 1.01 | 0.937 to 1.09 | 0.747 |
| RN staffing × NA staffing | 0.999 | 0.994 to 1.00 | 0.64 | |||
Model A: AIC 76747; BIC 76796. Model B: AIC 76749; BIC 76806.
AIC, Akaike information criterion; BIC, Bayesian information criterion; IRR, incidence rate ratio; NA, nursing assistant; RN, registered nurse.