This article is only available in the PDF format. Download the PDF to view the article, as well as its associated figures and tables.
To the Editor:—
In reference to the article by Drs. Dimond and Andrews on the accuracy of clinical thermometers and urinometers, in The Journal for Sept. 11, 1954, page 125, we do not believe that their results are representative. With reference to clinical thermometers, the state law in Connecticut requires that the bureau of laboratories of the state department of health test representative samples of the products of each manufacturer who sells thermometers in the state. Permissible tolerances are ±0.2 F at 98 F, ±0.2 F at 102 F, and ±0.3 F at 106 F. In 1951, the last year for which published figures are available, only 2.1% of 34,161 thermometers tested were rejected for reasons that impaired the reliability of the instruments. When this is compared with about 65% exceeding 0.2 F tolerance at a single temperature, the figure derived from the chart in Dr. Dimond's article, the discrepancy
Barnett RN, Katz J. CLINICAL THERMOMETERS AND URINOMETERS. JAMA. 1954;156(15):1427. doi:10.1001/jama.1954.02950150049019