BACKGROUND: Decision support systems (DSSs) are being developed to use events entered in anesthesia information management systems (AIMS) for quality of care, compliance, billing, documentation, and management purposes. DSS performance is impacted by latency from the actual time an event occurs to when it is written to the database, as well as how often the database is queried. Such latencies may result in poor DSS recommendations. METHODS: We analyzed approximately 48,000 cases at Hospital A for latency of two DSS prototype events, Surgery Begin and Surgery End. Each latency was measured from 1) the time that the event was recorded in the AIMS database as having taken place to 2) the time when the first DSS query would have been executed after the documentation of that event by the provider. The effects on latency of 1, 5, and 10 min query intervals were determined. Latencies for Surgery Begin and Surgery End were compared with those of Hospital B, where a different AIMS was used. RESULTS: Network delays and the event processing time of the AIMS contributed <1 s and 30 s, respectively, to latency. Average latencies for the two studied events were approximately half of the query interval, the expected value if the events occurred randomly within each interval. However, the longest 5% of latencies exceeded the query interval. This was not due to providers editing the times of the Begin or End Surgery events, as each occurred in only 0.7% of cases. Although the median latencies for the two events were longer at Hospital B than Hospital A by a few minutes, the 90th and 95th percentiles of the latencies were much longer at Hospital B (8-30 min, depending on the query interval and the percentile). CONCLUSIONS: DSS performance is influenced by the timeliness of documentation, the incidence of missing documentation and the query interval. Facilities using a DSS, including electronic whiteboards showing patient status, should assess the latencies of the measured events and critique the influence of the latencies on clinical and managerial decisions.
BACKGROUND: Decision support systems (DSSs) are being developed to use events entered in anesthesia information management systems (AIMS) for quality of care, compliance, billing, documentation, and management purposes. DSS performance is impacted by latency from the actual time an event occurs to when it is written to the database, as well as how often the database is queried. Such latencies may result in poor DSS recommendations. METHODS: We analyzed approximately 48,000 cases at Hospital A for latency of two DSS prototype events, Surgery Begin and Surgery End. Each latency was measured from 1) the time that the event was recorded in the AIMS database as having taken place to 2) the time when the first DSS query would have been executed after the documentation of that event by the provider. The effects on latency of 1, 5, and 10 min query intervals were determined. Latencies for Surgery Begin and Surgery End were compared with those of Hospital B, where a different AIMS was used. RESULTS: Network delays and the event processing time of the AIMS contributed <1 s and 30 s, respectively, to latency. Average latencies for the two studied events were approximately half of the query interval, the expected value if the events occurred randomly within each interval. However, the longest 5% of latencies exceeded the query interval. This was not due to providers editing the times of the Begin or End Surgery events, as each occurred in only 0.7% of cases. Although the median latencies for the two events were longer at Hospital B than Hospital A by a few minutes, the 90th and 95th percentiles of the latencies were much longer at Hospital B (8-30 min, depending on the query interval and the percentile). CONCLUSIONS:DSS performance is influenced by the timeliness of documentation, the incidence of missing documentation and the query interval. Facilities using a DSS, including electronic whiteboards showing patient status, should assess the latencies of the measured events and critique the influence of the latencies on clinical and managerial decisions.
Authors: Jesse M Ehrenfeld; Luke M Funk; Johan Van Schalkwyk; Alan F Merry; Warren S Sandberg; Atul Gawande Journal: Can J Anaesth Date: 2010-07-31 Impact factor: 5.063
Authors: Jesse M Ehrenfeld; Matthew D McEvoy; William R Furman; Dylan Snyder; Warren S Sandberg Journal: Anesthesiology Date: 2014-01 Impact factor: 7.892
Authors: Jesse M Ehrenfeld; Richard H Epstein; Stephen Bader; Sachin Kheterpal; Warren S Sandberg Journal: Anesth Analg Date: 2011-03-17 Impact factor: 5.108