Analysis of a Novel Inpatient Acute Limb Ischemia Alert System

Abstract

Acute limb ischemia (ALI) is a surgical emergency that generally develops in the outpatient setting. Hospitalized patients are also at risk for acute limb ischemia, but their presentation may be atypical or altered by medical therapy. Our institution developed an alert system to facilitate the prompt recognition and treatment of ALI that occurs in the inpatient population. We aimed to evaluate the usage of the system after the first 2 years of operation. All ALI alerts from October 2017 to December 2019 were collected from paging records and analyzed for location, timing, and the need for intervention. Alerts undergoing vascular intervention were classified as urgent (within 8 hours) or delayed (after 8 hr). Time and location data were evaluated to determine patterns of usage and true-positive rate of the system. From October 2017 to December 2019, there were 237 ALI alerts obtained from paging records containing time and location information for the alert. More alerts originated from ICUs relative to non-ICU floors (68% vs. 33%, P< 0.001), however a greater proportion of non-ICU floor alerts required intervention compared to ICU alerts (32.0% vs. 5.1%, P < .0001). The highest number of ALI alerts were from the Medical ICU (MRICU) (45.9%) and medical/surgical floors (33.3%), followed by Surgical ICU (20.2%). Alerts were more common within 3 hr of morning and evening nursing shift changes (47.3%, P < 0.001). From the 237 total alerts, the patient was able to be identified retrospectively in 186 cases, and of these 27 resulted in operative interventions (14.5%, positive predictive value), with 11 patients (40.7%) requiring urgent intervention with a median time to intervention of 3.5 hr (range 2.2-4.8), and 16 (59%) alerts undergoing a delayed intervention at a mean of 3 days (range 2-4). A total of 73 (39.2%) alert patients died during their admission, of which 65 (89.0%) were in an ICU, and no deaths were directly related to ALI. The median time to death was 2 days (range 0-95 days), and in 22 cases death occurred <24 hr from time of alert. Our novel hospital-wide ALI alert system demonstrates a 14.5% positive predictive value for ischemia that resulted in an intervention. Alerts were more likely to originate from the ICU setting and during nursing shift changes. Alerts originating from non-ICU floors were 5 times more likely to undergo surgical intervention for ALI. Further analysis is required to assess the effect of this system on patient safety, outcome, and allocation of institutional resources.