SESSION TITLE: Patient Safety
SESSION TYPE: Original Investigation Slide
PRESENTED ON: Tuesday, October 29, 2013 at 02:45 PM - 04:15 PM
PURPOSE: Institutional Harm campaigns have become essential in our improvement of safe practice in critical care, based on Institute for Healthcare improvement initiatives. Our institution embarked on an aggressive project to measure harm on our patients. We hypothesized that critically ill surgical patients were at more risk for harm when compared to medical intensive care patients.
METHODS: Under the approval of the IRB, 3 years of administrative data and Elixhauser score for patients with at least 1 ICU day at an urban tertiary care center was assembled. Harm data was accessed from the Harm campaign. Patients were deemed surgical (SP) if they had at least one trip to the operating room (OR). Univariate analysis was used to compare surgical patients to non-surgical using the occurrence of various harms grouped by category. Logistic regression was used to attempt risk adjustment in predicting a harm event, and in predicting death.
RESULTS: 19844 patients met our criterion, with 7483 surgical patients (37.7%) and an overall mortality of 7.8% (n=1554). More SP experienced harm than non-surgical patients (2923 (39.1%) vs 2798 (22.6%), odds ratio (OR) 2.2). Surgical patients were less likely to die (6.2% vs 8.8% mortality, P<.001) and was true when stratified by harm as well. Largest Odds ratios were African Americans (AA) 1.2, and surgical patients 2.1 (all p values <.001). Regression calculation showed SP patients to be less associated with death (OR .45), as were AA (OR .87), while those that were harmed had an increased OR 3.66 (all p values <0.001).
CONCLUSIONS: Preliminary data shows that critically ill surgical patients experience more harm than non-surgical but the lethality of harm was more significant in medical patients. African Americans were more likely to experience harm but less likely to die from it.
CLINICAL IMPLICATIONS: Our understanding of harm is still developing. Data suggests that harm in surgical patients is procedurally related. Racial discrepancy in harm requires further investigation.
DISCLOSURE: The following authors have nothing to disclose: Ko Un Park, Mathilda Horst, Michael Eichenhorn, Bruno DiGiovine, Jennifer Ritz, Roberta Mooney, Ilan Rubinfeld
No Product/Research Disclosure Information