Six-Year Trends in Laboratory Computer Availability

Abstract

Abstract Context.—Failure of a clinical laboratory computer system can disrupt work flow and charge capture and affect patient care. The first comprehensive survey of computer downtime was conducted in 1995 and demonstrated significant interinstitutional variation in system availability. Despite numerous changes in the laboratory and computer industries since 1995, no follow-up study has been reported. Objectives.—To quantify current laboratory computer availability and compare it with 1995 performance. Design.—Ninety-seven laboratories prospectively recorded the frequency and duration of computer downtime during 30 days in 2001. Results were compared with 1995 survey data. Results.—For the median facility, the number of downtime episodes decreased from 8 events per 30 days during 1995 to 3 events per 30 days during 2001 (P < .01). The frequency of unscheduled downtime also improved, from a median of 2 to 0.5 events per 30 days (P < .01). Reduced downtime events were paralleled by reduced cumulative downtime (14.3 vs 4.0 hours per 30 days; P < .01). Improvements were not restricted to the median facility; laboratories performing in the bottom quartile in 2001 recorded substantially less downtime than laboratories in the bottom quartile in 1995. When the comparison was restricted to the 37 institutions that participated in both the 1995 and 2001 surveys, a significant reduction in overall downtime and unscheduled downtime events was still evident (P < .01). More recent installation of vendor software patches was associated with a reduced frequency of downtime events in the 2001 data set. Conclusion.—Laboratory computer downtime was less frequent in 2001 than in 1995; industry performance appears to be improving.