Identifying low value pathology test ordering in hospitalised patients: a retrospective cohort study across two hospitals

Alexis Hure; Kerrin Palazzi; Roseanne Peel; David Geraghty; Phillip Collard; Theo De Malmanche; Huy Tran; Penny Reeves; Andrew Searles; Louisa Jorm; John Attia

doi:10.1016/j.pathol.2019.06.003

Identifying low value pathology test ordering in hospitalised patients: a retrospective cohort study across two hospitals

Pathology. 2019 Oct;51(6):621-627. doi: 10.1016/j.pathol.2019.06.003. Epub 2019 Aug 29.

Authors

Affiliations

¹ School of Medicine and Public Health, University of Newcastle, Newcastle, NSW, Australia.
² Hunter Medical Research Institute, Newcastle, NSW, Australia.
³ Applications Development ICT Services Hunter New England Local Health District, Newcastle, NSW, Australia.
⁴ New South Wales Health Pathology, Immunology, Hunter New England Local Health District, Newcastle, NSW, Australia.
⁵ School of Medicine and Public Health, University of Newcastle, Newcastle, NSW, Australia; Hunter Medical Research Institute, Newcastle, NSW, Australia.
⁶ Centre for Big Data Research in Health, University of New South Wales, Sydney, NSW, Australia.
⁷ School of Medicine and Public Health, University of Newcastle, Newcastle, NSW, Australia; Hunter Medical Research Institute, Newcastle, NSW, Australia. Electronic address: john.attia@newcastle.edu.au.

PMID: 31472983
DOI: 10.1016/j.pathol.2019.06.003

Abstract

The push to identify low value care has led to scrutiny of pathology test re-ordering. The objective of this study was to identify the patterns of ordering pathology tests among inpatients in teaching hospitals and model strategies to reduce unnecessary testing. This was a retrospective cohort study of all adult medical and surgical inpatients admitted to one major teaching hospital and one rural hospital in the same health district over 2 years. Obstetric, gynaecological, intensive care, elective/day procedures and dialysis admissions were excluded. Orders for electrolytes, urea and creatinine (EUC), full blood count (FBC), thyroid stimulating hormone (TSH), glycated haemoglobin (HbA1c), vitamin D, and troponin, date of order, and value of the resulting test, were obtained from a health district data warehouse. Pathology results were mapped to each inpatient day. EUC and FBC constituted over 90% of all inpatient pathology requests for these six tests. Between 40-45% of inpatients had EUC and/or FBC performed daily. After the first couple of tests, the retest interval was consistently around 24 hours, regardless of the previous value of the test, consistent with a culture of routine ordering. This was less pronounced in the rural hospital compared to the urban teaching hospital. Lockouts (applied when previous tests normal) or minimum retest intervals (applied to previously normal and abnormal tests) of various lengths were tested on the data to find optimal combinations that reduced unnecessary tests without missing too many very abnormal tests. A lockout of 48 hours for EUC and 48 hour lockout combined with a 12 hour minimum retest interval for FBC appear optimal to reduce over ordering and could save approximately AU$400/inpatient bed per year at a single teaching hospital. There is evidence of low value re-ordering of EUC and FBC pathology tests. Implementation of a computerised physician order entry system with inbuilt prompts to restrict unnecessary re-ordering of pathology tests may be a practical solution.

Keywords: Haematology; adults; biochemistry; economic evaluation; inpatients; pathology tests.

MeSH terms

Diagnostic Tests, Routine / economics*
Health Care Costs*
Hospitals
Humans
Medical Order Entry Systems
Practice Patterns, Physicians' / economics*
Retrospective Studies