Diagnostic accuracy of patient-reported dry mouth as a predictor for oral dryness in terminally ill cancer patients.

ABSTRACT

PURPOSE: The aim of the study was to determine the diagnostic accuracy of patient-reported dry mouth using an oral moisture-checking device in terminally ill cancer patients. METHODS: The study was conducted following the STARD guidelines, and the participants were recruited prospectively from the Palliative Care Unit, Kyoto Medical Center, Japan, between 1 January 2017 and 30 November 2018. Patients reporting dry mouth were asked to rate oral dryness on a 5-point rating scale. The outcome was oral dryness at the lingual mucosa, measured using an oral moisture-checking device. Receiver operating characteristic (ROC) curves were plotted, and the sensitivity, specificity, positive and negative predictive values (PPV and NPV), positive and negative likelihood ratios (LR), and overall diagnostic accuracy were calculated. RESULTS: Of 103 participants, the prevalence of oral dryness was 65.0%. ROC analysis indicated that patient-reported dry mouth was a poor predictor of oral dryness, with an area under the curve of 0.616 (95% confidence interval 0.508-0.723), a sensitivity of 46.3%, a specificity of 75.8%, a PPV of 55.9%, an NPV of 68.1, a positive LR of 1.9, a negative LR of 0.7, and an overall diagnostic accuracy of 64.1%, with a cut-off value of 3 points. CONCLUSION: In conclusion, patient-reported dry mouth is not a useful parameter for the assessment of oral dryness in terminally ill cancer patients.