Background: UK primary care physicians receive their laboratory test results electronically. This study reports a computerised physician order entry (CPOE) system error in the pathology test request date that went unnoticed in family practices.
Method: We conducted a case study using a causation of risk theoretical framework; comprising interviews with clinicians and the manufacturer to explore the identification of and reaction to the error. The primary outcome was the evolution and recognition of and response to the problem. The secondary outcome was to identify other issues with this system noted by users.
Results: The problem was defined as the incorrect logging of test dates ordered through a CPOE system. The system assigned the test request date to the results, hence a blood test taken after a therapeutic intervention (e.g. an increase in cholesterol-lowering therapy) would appear in the computerised medical record as though it had been tested prior to the increase in treatment. This case demonstrates that: the manufacturers failed to understand family physician workflow; regulation of medical software did not prevent the error; and inherent user trust in technology exacerbated this problem. It took three months before users in two practices independently noted the date errors.
Conclusion: This case illustrates how users take software on trust and suppliers fail to make provision for risks associated with new software. Resulting errors led to inappropriate prescribing, follow-up, costs and risk. The evaluation of such devices should include utilising risk management processes (RMP) to minimise and manage potential risk.