24-Hour Intensivist Staffing Is Not Beneficial for Patients.
24-Hour Intensivist Staffing Is Not Beneficial for Patients.
复制标题
24 小时重症监护室人员配置对患者没有好处。
DOI：
10.1097/ccm.0000000000002805
复制
发表时间：
2018
影响因子：
8.8
通讯作者：
Kerlin,MeetaPrasad
中科院分区：
文献类型：
--
作者：
Kerlin,MeetaPrasad
Critical Care Medicine www. ccmjournal. org 153 trial performed by Garland et al (21) in one academic and one community ICU in Canada. Both compared staffing by an inhospital daytime intensivist who took call from home at night (control) with staffing by an in-hospital daytime intensivist and a different in-hospital intensivist at night (intervention), and both showed no mortality difference. A recent systematic review and meta-analysis synthesizing all these studies concluded that no association exists between nighttime intensivist staffing and patient outcomes (22). Meanwhile, as the literature for 24-hour intensivist staffing grew, more evidence emerged that calls into question whether intensivists affect patient outcomes at all in the modern era of critical care delivery. A meta-analysis in 2013 reported a pooled odds ratio for hospital mortality of 0.83 in favor of highintensity intensivist staffing. However, most of the 34 studies that were included were single-center and used before-after designs; study heterogeneity was high; and the association was not consistent across ICU types or time periods (23). Multiple larger, multicenter studies have failed to corroborate the association between intensivists and patient outcomes. Levy et al (24) performed a study in 123 ICUs, adjusting for patient severity of illness and propensity to receive care by an intensivist, and found that exposure to an intensivist was actually associated with increased odds of death. Though I suspect this finding was due at least partly to residual confounding, it certainly does not support the prior studies. Costa et al (25) found that high-intensity daytime intensivist staffing was not associated with lower mortality in a modern cohort after adjustment for other organizational factors. Most recently, Nagendran et al (26) used an elegant difference-in-differences approach to address both patient-and ICU-specific confounding in a large multicenter cohort and demonstrated that ICUs that transitioned from low-to high-intensity staffing models did not have any improvements in mortality among Medicare beneficiaries. Collectively, the results of these multicenter cohort studies suggest that observed differences in patient outcomes in prior studies may have been due to other factors, rather than intensivists themselves.Whether 24-hour intensivist staffing is associated with other patient outcomes is even less clear. Several studies have evaluated ICU and hospital length of stay (LOS), with highly inconsistent results. Both of the aforementioned experimental studies showed no significant differences in LOS outcomes (20, 21). Even in studies where statistically significant associations were present, the differences were almost always very small, with questionable clinical significance (7–9, 12–15, 19, 27). A few studies have examined ICU readmission rates (7, 15, 28), ICU complications (7, 13, 15, 27), and patient and family satisfaction (7, 21). For all outcomes, again most studies were either null or fraught with methodological issues that limit any conclusions.
