Examining Test Order Appropriateness Using Mean Abnormal Result Rate in Inpatient and Outpatient Care: A Retrospective Cross‐Sectional Study
Parastou Gorovanchi, Parham Panahi, Zahra Farrokhi, Rasoul Hossein Zadeh, Maryam Baradaran‐Binazir, Rasoul EstakhriABSTRACT
Background and Aims
Patients may experience more stress and the healthcare system may incur greater expenditures as a result of doctors ordering too many tests. In order to assess the suitability of testing, the Mean Abnormal Result Rate (MARR) metric—defined as the percentage of tests having abnormal results—was created. Data on how the COVID‐19 epidemic changed test ordering patterns and the corresponding MARR are limited, despite the fact that it greatly increased the demand for laboratory testing, including several tests with no evidence to justify their usage. This study aims to examine MARR in the internal medicine and surgical departments, as well as the emergency and outpatient departments, of Sina Hospital in Tabriz, Iran during the COVID‐19 pandemic.
Methods
Laboratory test results from Sina Hospital's internal medicine and surgical departments, as well as the emergency and outpatient departments, were examined in this cross‐sectional study between March 21, 2021, and March 20, 2022.
Results
The emergency department's mean MARR was 30% (95% CI: 29.8%–30.2%) versus 27% (95% CI: 26.9%–27.1%) in the outpatient department; this difference was not statistically significant ( p = 0.9). MARR showed wide monthly fluctuations, and a time‐series analysis revealed a negative but non‐significant association between monthly test volume and MARR ( p = 0.114). After the emergence of new viral variants, MARR decreased from a median of 41% (months 1–6) to 27% (months 7–12), with a borderline significant difference (exact p = 0.065).
Conclusion
No significant difference in MARR was observed between clinical settings or specialties. Although MARR and test volume exhibited an inverse temporal pattern, the association was not statistically significant after accounting for autocorrelation. The decline in MARR during the variant wave was borderline significant. MARR may serve as a supplementary metric for monitoring test selectivity, but individual‐level data are needed to assess physician experience or clinical appropriateness directly.