Do out-of-office laboratory tests affect diagnoses in general practice?

• Published in: Scandinavian journal of primary health care Vol. 13; no. 1; pp. 46 - 51
• Main Authors: Zaat, Joost OM, Schellevis, Francois G, van Eijk, Jacques ThM, van der Velden, Koos
• Format: Journal Article
• Language: English
• Published: United States Informa UK Ltd 1995; Taylor & Francis
• Subjects: Clinical Laboratory Techniques; Diagnosis; Diagnostic Errors; diagnostic process; Family Practice; Female; general practice; Humans; laboratory test use; Male; Morbidity; Netherlands; Psychophysiologic Disorders - diagnosis; Sampling Studies; Netherlands
• ISSN: 0281-3432; 1502-7724
• DOI: 10.3109/02813439508996734

Objective - To find out whether the GP diagnosis changed by out-of-office laboratory test results and whether his diagnosis became more certain. Design - Descriptive study. Setting - Dutch survey of morbidity and interventions in general practice: stratified random sample of 161 GPs with a total list of 335 000 patients. Subjects - 2 081 episodes of illness with at least one consultation with clinical chemistry, haematology, or serology tests and at least one follow-up consultation. Main outcome measurements - Change in ICPC component or chapter between the consultation in which a laboratory test was ordered and the follow up contact; change in exact ICPC code in cases with important diseases (infectious diseases, haematological disorders, endocrine abnormalities, auto-immune processes and malignancies (n=330)); change in certainty of a diagnosis and change in somatic/ psychosocial orientation. Results - After laboratory tests done in the first consultation the ICPC component changed in 46% of the diagnoses. of the diagnoses made in first consultations without laboratory tests 41% changed in the follow up consultation. the diagnosis after laboratory tests was the same as before in 51% of the consultations with important diseases. Certainty about a diagnosis increased significantly after laboratory tests (p ≤0.001). An abnormal laboratory result did not affect the clinical certainty of the general practitioner or the percentage of altered diagnoses. Conclusion - the usefulness of tests should be assessed not only in terms of the number of diagnoses changed or of the percentage of abnormal results, but also in terms of the changed certainty concerning a diagnosis.