Assessment criteria for compliance with oral hygiene: application of ROC analysis

PURPOSE: Diabetes is an established risk factor for periodontal disease. Management of periodontal disease is highly dependent upon effective oral hygiene. Assessment of plaque and gingivitis has been commonly used and arbitrarily set in clinical practice to evaluate patients' adherence with oral hygiene recommendations. This study aims to determine an objective cut-off criterion for assessing oral hygiene compliance utilising a combination of plaque and bleeding scores. MATERIALS AND METHODS: 161 patients with diabetes, from a prospective clinical trial, provided the clinical periodontal parameters at baseline to be used to determine the oral hygiene compliance criterion in relation to a composite score of pocket depth, subgingival calculus and supragingival calculus. A sequence of different combinations of plaque and gingival bleeding scores were used. Receiver operator characteristic (ROC) curve assessment, sensitivity, specificity, and predictive values were utilised for the determination of the criterion. RESULTS: The combination of 25% plaque scores and 15% gingival bleeding scores obtained the highest ROC value (using a probability cut-off of 0.5) of 0.868 with sensitivity 98.6%, specificity 75.0%, positive predictive value (PPV) 97.3% and negative predictive value (NPV) 85.7%. According to this criterion, amongst the cohort of subjects examined, 145 (90.1%) were categorised as non-compliant, and only 16 (9.9%) were considered compliant with oral hygiene at baseline. CONCLUSIONS: Based upon the clinical periodontal parameters of subjects from this study, a combination of 25% plaque score and 15% bleeding score appears to be a valid target for determining compliance with oral hygiene in oral health programmes.