The Diagnostic Accuracy of Digital, Infrared and Mercury-in-Glass Thermometers in Measuring Body Temperature: A Systematic Review and Network Meta-Analysis

Abstract

Background: Not much is known about how accurate and reproducible different thermometers are in diagnosing patients with suspected fever. Methods: We searched Medline, Embase, Scopus, WOS, CENTRAL, and Cinahl to perform: 1) diagnostic accuracy meta-analysis (MA) using rectal mercury-in-glass or digital thermometry as reference and bivariate models for pooling; 2) network MA to estimate differences in mean temperature between devices; 3) Bland-Altman method to estimate devices reproducibility. PROSPERO registration: CRD42020174996. Findings: We included 46 studies enrolling more than 12000 patients, of whom 66% were children. Using 38°C (100·4°F) as cut-off for temperature, temporal infrared thermometry had a sensitivity of 0·76 (95% confidence Interval, 0·65-0·84; low certainty) and specificity of 0·96 (0·92-0·98; moderate certainty); tympanic infrared thermometry had a sensitivity of 0·77 (0·60 to 0·88; low certainty) and a specificity of 0·98 (0·95 – 0·99; moderate certainty). For all other index devices, it was not possible to pool the estimates. Compared to rectal mercury-in-glass thermometer, mean temperature differences were not statistically different from zero for infrared or tympanic infrared thermometry, while axillary temperature was significantly lower; the median coefficient of reproducibility ranged between 0·53°C [0·95°F] for infrared temporal thermometry and 1·2°C [2·16°F] for axillary digital thermometry. Interpretation: Several peripheral thermometers proved specific, but not sensitive to diagnose fever with rectal thermometry as reference standard, meaning that finding a temperature below 38°C does not rule out fever. Fixed differences between temperatures together with random error means facing differences between measurements in the order of 2°C [4·5°F]. Funding Statement: “Fondazione CR Firenze” (#24383). Declaration of Interests: The authors declared no potential conflict of interest