TY - JOUR
T1 - Preventing overuse of laboratory diagnostics
T2 - a case study into diagnosing anaemia in Dutch general practice
AU - Kip, Michelle M.A.
AU - Oonk, Martijn L.J.
AU - Levin, Mark David
AU - Schop, Annemarie
AU - Bindels, Patrick J.E.
AU - Kusters, Ron
AU - Koffijberg, Hendrik
N1 - Publisher Copyright:
© 2020 The Author(s).
PY - 2020/7/31
Y1 - 2020/7/31
N2 - BACKGROUND: More information is often thought to improve medical decision-making, which may lead to test overuse. This study assesses which out of 15 laboratory tests contribute to diagnosing the underlying cause of anaemia by general practitioners (GPs) and determines a potentially more efficient subset of tests for setting the correct diagnosis.METHODS: Logistic regression was performed to determine the impact of individual tests on the (correct) diagnosis. The statistically optimal test subset for diagnosing a (correct) underlying cause of anaemia by GPs was determined using data from a previous survey including cases of real-world anaemia patients.RESULTS: Only 9 (60%) of the laboratory tests, and patient age, contributed significantly to the GPs' ability to diagnose an underlying cause of anaemia (CRP, ESR, ferritin, folic acid, haemoglobin, leukocytes, eGFR/MDRD, reticulocytes and serum iron). Diagnosing the correct underlying cause may require just five (33%) tests (CRP, ferritin, folic acid, MCV and transferrin), and patient age.CONCLUSIONS: In diagnosing the underlying cause of anaemia a subset of five tests has most added value. The real-world impact of using only this subset should be further investigated. As illustrated in this case study, a statistical approach to assessing the added value of tests may reduce test overuse.
AB - BACKGROUND: More information is often thought to improve medical decision-making, which may lead to test overuse. This study assesses which out of 15 laboratory tests contribute to diagnosing the underlying cause of anaemia by general practitioners (GPs) and determines a potentially more efficient subset of tests for setting the correct diagnosis.METHODS: Logistic regression was performed to determine the impact of individual tests on the (correct) diagnosis. The statistically optimal test subset for diagnosing a (correct) underlying cause of anaemia by GPs was determined using data from a previous survey including cases of real-world anaemia patients.RESULTS: Only 9 (60%) of the laboratory tests, and patient age, contributed significantly to the GPs' ability to diagnose an underlying cause of anaemia (CRP, ESR, ferritin, folic acid, haemoglobin, leukocytes, eGFR/MDRD, reticulocytes and serum iron). Diagnosing the correct underlying cause may require just five (33%) tests (CRP, ferritin, folic acid, MCV and transferrin), and patient age.CONCLUSIONS: In diagnosing the underlying cause of anaemia a subset of five tests has most added value. The real-world impact of using only this subset should be further investigated. As illustrated in this case study, a statistical approach to assessing the added value of tests may reduce test overuse.
KW - Anemia
KW - Data analysis, statistical
KW - Diagnoses and laboratory examinations
KW - General practice
KW - Optimal testing
KW - Overuse
UR - http://www.scopus.com/inward/record.url?scp=85089126002&partnerID=8YFLogxK
U2 - 10.1186/s12911-020-01198-8
DO - 10.1186/s12911-020-01198-8
M3 - Article
C2 - 32736551
AN - SCOPUS:85089126002
SN - 1472-6947
VL - 20
JO - BMC medical informatics and decision making
JF - BMC medical informatics and decision making
IS - 1
M1 - 178
ER -
