Background Excessive use of diagnostic laboratory does not only constitute a waste of resources, it may also result in an increased rate of false positives which may lead to further unnecessary testing, unnecessary treatment, and increased anxiety in both patients and physicians. Still, many clinicians order irrelevant laboratory tests despite the available information about unnecessary test utilization.
Objectives To examine the effects of an educational meeting and subsequent computer reminders on the number of ordered laboratory tests.
Methods Using interrupted time series analysis we assessed whether trends in the number of laboratory tests ordered by rheumatologists between September 2012 and September 2015 at the Sint Maartenskliniek (the Netherlands) changed following an educational meeting (September 2013) and introduction of computer reminders into the Computerized Physician Order Entry System (July 2014). The reminders functioned as follows: whenever a clinician tried to order one of the intervention tests a pop-up message appeared explaining in which specific rheumatology-related situation the test was indicated or not, and asked for a reason for ordering the test. The analyses were done for the set of tests on which both interventions had focussed (intervention tests; complement, cryoglobulins, immunoglobins, M protein) and for a set of tests unrelated to the interventions, included as a control (alanine transferase, anti-cyclic citrullinated peptide, C-reactive protein, creatine, haemoglobin, leukocytes, mean corpuscular volume, rheumatoid factor and thrombocytes). Secondary outcomes were the percentage of abnormal intervention test results and the percentage of valid reasons provided with the intervention test orders, as judged by two experts.
Results At study start 101 intervention tests and 7660 control tests were ordered per month by the rheumatologists. After the educational meeting both the level and trend of ordered intervention and control tests did not change significantly. After implementation of the reminders, the level of ordered intervention tests decreased with 85 tests (95%>CI -133 to -37, p<0.01), though the trend did not change significantly (1.06, 95%>CI -6.2 to 8.3, p=0.77). The level and trend of control tests did not change following the introduction of reminders. Of the reasons given for ordering the intervention tests after the introduction of reminders, only 34% was deemed valid by two experts. Interestingly, the percentage of abnormal results in the intervention tests did not change significantly after the introduction of the reminders (17.4% before introduction of reminders compared 18.4% after introduction (p=0.73)), nor was it different in those cases were the reason for ordering the test was judged as valid (18.8%).
Conclusions The educational meeting alone was not effective in decreasing the number of ordered intervention tests, but subsequent introduction of computer reminders did result in a large decrease of those tests. Therefore, we recommend using computer reminders additionally to education if reduction of inappropriate test use is aimed for.
Disclosure of Interest None declared