General practitioners (GP’s) often order laboratory tests for their patients. They do this to find diseases, to evaluate (side-)effects of treatments or measure the progression of a condition. GP’s are confronted with more complex care decisions due to an increase in the amount of chronic illnesses, more frequent use of multiple drugs increasing the risk of interactions, and the increase in the amount of laboratory tests. This makes it more difficult to order the most appropriate tests for each patient. Sometimes unnecessary tests are ordered which can lead to results of unknown significance. This in turn can cause unnecessary anxiety and often leads to further testing and costs. On the other hand, some necessary tests may not be ordered, possibly delaying the diagnosis of important conditions or diseases. GP’s order laboratory tests by marking each desired test on a request form, quite often on paper. Based on the reason(s) why the GP orders laboratory tests (the disease to diagnose or follow-up) we will propose necessary tests within a computerised request form. We want to evaluate what influence this new way of ordering laboratory tests has on laboratory test ordering behaviour of GPs.
Summary of results |
---|
Almost 10 000 patients were included in the study by 270 GPs. In the control arm (i.e. GPs not using a clinical decision support system - CDSS), the percentage of appropriate lab tests was 38%. CDSS increased the proportion of appropriate tests by 21% (this was statistically significant) for all tests included in the study. GPs in the CDSS arm ordered 7 tests fewer per panel (the GPs in the control arm ordered 31 tests per panel whereas GPs in the CDSS arm ordered 24 tests per panel). CDSS did not increase diagnostic error: the absolute difference in proportions between the two groups was a very small decrease (<1%) in possible diagnostic error. In conclusion, a CDSS in the form of order sets, integrated within the computerized physician order entry (CPOE) improved appropriateness and decreased volume of laboratory test ordering without increasing diagnostic error. Results expressed in scientific language can be found in the “DOCUMENTS” section. |
Title | The effect of evidence-based order sets within a CPOE (computerised physician order entry) system on the quantity and quality of laboratory test ordering in family practice: a cluster randomised trial (KCE-16011) |
Participants (P) | Patients from primary care physicians (GPs) affiliated to one of three collaborating laboratories in the Leuven, Ghent or Antwerp regions |
Intervention (I) | Evidence-based order sets for laboratory tests based on indication, implemented within a computerised physician order entry (CPOE) system |
Control (C) | CPOE without evidence-based laboratory test order sets |
Outcome (O) | Percentage of appropriate laboratory tests, volume of tests ordered, percentage of missed diagnoses. |
Trial Design | Cluster randomised controlled trial |
Sample Size | 300 physicians randomized, data collected for 12 000 patients |
Trial duration | 2 years |
Budget | 1.8 million € (incl. VAT) |
Sponsor |
UZ Leuven – KU Leuven |
Chief Investigator |
Prof Dr Bert Aertgeerts, |
Trial coordinator |
Dr Nicolas Delvaux, |
External Partners |
Prof Dr An De Sutter, |
Healthdata.be, |
|
Laboratory Anacura, Evergem |
|
Medisch Centrum Huisartsen (MCH), Leuven |
|
Algemeen Medisch Labo (AML), Antwerpen |
Protocol | View protocol Implementation Science |
Results | Download report Download summary View publication Implementation Sciences View publication BMJ Open |
Extra info | View publication JMIR Medical Informatics |
For timelines and recruitment, see dashboard
Commissioning workstream 2016