TABLE 7.
Coefficients for fixed effects in models of bacterial density, relative abundance of Enterobacteriaceae and absolute abundance of Enterobacteriaceae
| Coefficient (95% CI) | p-value | |
| Model: bacterial density# | ||
| Hospital day | −0.066 (−0.12–0.0059) | 0.029 |
| Anti-anaerobic treatment | −1.48 (−2.41– −0.59) | 0.0017 |
| Age | 0.026 (−0.0061–0.058) | 0.12 |
| Male | −0.38 (−1.28–0.51) | 0.42 |
| Non-Caucasian | −1.09 (−2.36–0.17) | 0.10 |
| APACHE IV score | −0.073 (−0.22–0.080) | 0.36 |
| Charlson Comorbidity Index | 0.22 (0.0010–0.44) | 0.058 |
| Hospital day×anti-anaerobic treatment | 0.089 (0.020–0.16) | 0.011 |
| Model: relative abundance of Enterobacteriaceae¶ | ||
| Hospital day | −0.19 (−0.41–0.032) | 0.099 |
| Anti-anaerobic treatment | −1.79 (−5.63–2.05) | 0.36 |
| Age | 0.043 (−0.12–0.21) | 0.62 |
| Male | 1.10 (−3.51–5.71) | 0.65 |
| Non-Caucasian | −0.19 (−6.73–6.36) | 0.96 |
| APACHE IV score | 0.36 (−0.43–1.14) | 0.39 |
| Charlson Comorbidity Index | 0.25 (−0.89–1.40) | 0.68 |
| Hospital day×anti-anaerobic treatment | 0.26 (0.0093–0.51) | 0.045 |
| Model: absolute abundance of Enterobacteriaceae+ | ||
| Hospital day | −0.17 (−0.28– −0.066) | 0.0018 |
| Anti-anaerobic treatment | −1.31 (−3.11–0.45) | 0.15 |
| Age | 0.085 (0.0030–0.17) | 0.048 |
| Male | 0.294 (−2.00–2.59) | 0.81 |
| Non-Caucasian | 0.597 (−2.63–3.81) | 0.72 |
| APACHE IV score | −0.033 (−0.44–0.37) | 0.88 |
| Charlson Comorbidity Index | −0.15 (−0.75–0.44) | 0.62 |
| Hospital day×anti-anaerobic treatment | 0.14 (0.0070–0.28) | 0.043 |
APACHE: Acute Physiology and Chronic Health Evaluation. #: the mixed effects regression modelled the bacterial density (log 16S copies per sample) by hospital day. The interaction term between hospital day and anti-anaerobic treatment represents the log change in bacterial density per hospital day in the anti-anaerobic-treated group compared with the nontreated group. ¶: the mixed effects regression modelled the relative abundance of Enterobacteriaceae (% of bacterial community) by hospital day. The interaction term between hospital day and anti-anaerobic treatment represents the relative abundance of Enterobacteriaceae per hospital day in the anti-anaerobic-treated group compared with the nontreated group. +: the mixed effects regression modelled the absolute abundance of Enterobacteriaceae (% of bacterial community) by hospital day. The interaction term between hospital day and anti-anaerobic treatment represents the absolute abundance of Enterobacteriaceae per hospital day in the anti-anaerobic-treated group compared with the nontreated group.