Electronic Laboratory Medicine ordering with evidence-based Order sets in primary care (ELMO study): protocol for a cluster randomised trial
Delvaux et al. Implementation Science
Electronic Laboratory Medicine ordering with evidence-based Order sets in primary care (ELMO study): protocol for a cluster randomised trial
Nicolas Delvaux 0
An De Sutter 2
Stijn Van de Velde 0 1
Dirk Ramaekers 0
Steffen Fieuws 0
Bert Aertgeerts 0
0 Department of Public Health and Primary Care, KU Leuven , Kapucijnenvoer 33, Blok J, PB 7001, B-3000 Leuven , Belgium
1 Centre for Informed Health Choices, Norwegian Institute of Public Health , Oslo , Norway
2 Department of Family Medicine and Primary Health Care, Ghent University , Ghent , Belgium
Background: Laboratory testing is an important clinical act with a valuable role in screening, diagnosis, management and monitoring of diseases or therapies. However, inappropriate laboratory test ordering is frequent, burdening health care spending and negatively influencing quality of care. Inappropriate tests may also result in false-positive results and potentially cause excessive downstream activities. Clinical decision support systems (CDSSs) have shown promising results to influence the test-ordering behaviour of physicians and to improve appropriateness. Order sets, a form of CDSS where a limited set of evidence-based tests are proposed for a series of indications, integrated in a computerised physician order entry (CPOE) have been shown to be effective in reducing the volume of ordered laboratory tests but convincing evidence that they influence appropriateness is lacking. The aim of this study is to evaluate the effect of order sets on the quality and quantity of laboratory test orders by physicians. We also aim to evaluate the effect of order sets on diagnostic error and explore the effect on downstream or cascade activities. Methods: We will conduct a cluster randomised controlled trial in Belgian primary care practices. The study is powered to measure two outcomes. We will primarily measure the influence of our CDSS on the appropriateness of laboratory test ordering. Additionally, we will also measure the influence on diagnostic error. We will also explore the effects of our intervention on cascade activities due to altered results of inappropriate tests. Discussion: We have designed a study that should be able to demonstrate whether the CDSS aimed at diagnostic testing is not only able to influence appropriateness but also safe with respect to diagnostic error. These findings will influence a lager, nationwide implementation of this CDSS. Trial registration: ClinicalTrials.gov, NCT02950142.
Background
Laboratory testing is an important clinical act with a
valuable role in screening, diagnosis, management and
monitoring of diseases or therapies. Thirty percent of
patient contacts in primary care result in ordering one
or more laboratory tests [
1, 2
]. With 370 million tests
annually, laboratory testing is the most frequent medical
activity in Belgium [3]. There is a large variation in the
appropriateness of these orders [
4–7
]. Inappropriate
laboratory test ordering has been estimated to be as high
as 30% [
8, 9
]. Besides the burden this poses on health
care spending, it also negatively influences quality of
care. Inappropriate tests may also result in false-positive
results and potentially cause excessive downstream
activities. Downstream or cascade activities are those
medical acts which result from altered or deviant tests.
This phenomenon is often referred to as the Ulysses
effect, and it is generally assumed that the effects of
inappropriate test ordering are larger on the downstream
activities than on the tests themselves [
10
]. To date, little
research has been done on these cascades and the true
extent of this Ulysses effect in primary care remains
unclear [
11
].
Education-based interventions, feedback-based
interventions and clinical decision support systems (CDSSs)
have shown promising results to influence the
testordering behaviour of physicians and to improve
appropriateness [
1, 9, 12, 13
]. These findings, however, tend
not to be generalisable because many studies either
focus on very limited indications or measure testing
volume rather than appropriateness. Order sets, a form of
CDSS where a limited set of evidence-based tests are
proposed for a series of indications, integrated in a
computerised physician order entry (CPOE) have been
shown to be effective in reducing the volume of ordered
laboratory tests [
14, 15
]. However, good evidence that
the use of order sets aimed at multiple indications
improves the appropriateness of laboratory test ordering is
still lacking. A barrier to adhering to evidence-based
policy is the fear for missing important pathology and
the liability this may create [
2
]. There is currently no
evidence that increasing appropriateness of laboratory
testing influences morbidity through diagnostic error or
delay. The aim of this study is to evaluate the effect of
order sets on the quality and quantity of laboratory test
orders by physicians. We also (...truncated)