Table 4.
ARIMA models for cefepime-resistant A. baumannii (A) and cefepime use (B). Dynamic regression model for the association between cefepime-resistant A. baumannii and hospital cefepime use (C).
| Estimate | Model Parameter | Standard Error | p-Value |
|---|---|---|---|
| A. A. baumannii resistance | |||
| ar1 | −0.677 | 0.129 | <0.001 |
| ar2 | −0.710 | 0.122 | <0.001 |
| AIC | 100.56 | ||
| R2 | 0.580 | ||
| B. Cefepime use (in DDD/100 PD) | |||
| ar1 | −0.463 | 0.137 | <0.001 |
| ar2 | −0.466 | 0.133 | <0.001 |
| AIC | 139.03 | ||
| R2 | 0.619 | ||
| C. Impact of cefepime use on A. baumannii resistance | |||
| ar1 | −0.576 | 0.210 | 0.006 |
| ar2 | −0.559 | 0.222 | 0.011 |
| cef1 | 0.865 | 0.395 | 0.028 |
| AIC | 85.53 | ||
| R2 | 0.660 | ||
Key: AIC, the estimated Akaike Information Criterion value for the model; ar1, autoregression term with a lag of one month of the ARIMA model; ar2, autoregressive component with lag equal to two months of the ARIMA model; cef1, cefepime use of one month with lag time of one month; R2, the coefficient of determination of the model, DDD, Defined Daily Dose; PD, Patient days.