Effect of reducing the number of items of the Oral Health Impact Profile on responsiveness, validity and reliability in edentulous populations

Abstract – Background: The 49‐item Oral Health Impact Profile (OHIP) has shown strong responsiveness, reliability and validity. However, the large number of items included may limit its use in clinical trials, clinical practice and surveys. Objective: The main objective of this study is to assess the effect of reducing the number of items in each domain, one at a time, on responsiveness, reliability and validity of the OHIP in edentulous populations. Materials and methods: Data used in this study were obtained from two randomized clinical trials comparing mandibular implant overdentures and conventional dentures among 102 subjects between 35 and 65 years of age, and 60 subjects over the age of 65 years. Participants were edentulous individuals who wished to replace their current prostheses. Subjects in both trials were asked to complete the 49‐item OHIP prior to treatment and at 2 months post‐treatment. Within the study, effect sizes were computed at each stage of item reduction using the impact method. Intraclass correlation coefficients and Pearson's correlation coefficients were also assessed at each stage of item reduction. In addition, receiver‐operating characteristic (ROC) curves were used to indicate the accuracy with which measurement changes corresponded to judgements of important changes in Oral Health Related Quality of Life (OHRQL). Results: The results indicated that, in general, domain responsiveness was not affected by the reduction of the number of items used per domain. However, there was a decrease in reliability, especially within the ‘psychological’ and ‘social’ disabilities and ‘handicap’ domains (35‐ to 65‐year group). In addition, there was a decrease in construct validity of the ‘physical pain’, ‘psychological’ and ‘social disabilities’ domains (35‐ to 65‐year group), as well as on ‘physical pain’, ‘psychological discomfort’, ‘physical’ and ‘psychological’ disabilities in the 65‐year and older group. This occurred primarily, when reducing from two to one item per domain. Among the 35‐ to 65‐year group, there were consistencies in patients’ ratings of the importance of similarly measured changes in oral health. Conclusion: The results indicate that although the 49‐item OHIP responsiveness could be maintained with item reduction, this will lead to compromises in reliability and validity.