For the present study, data from a cohort of patients followed at a Portuguese university dental clinic (Egas Moniz Dental Clinic, Almada, Portugal) was used. Data collection followed a consecutive sampling of incoming participants, from January 2022 until June 2024, that seek a first appointment. The study was approved by the Egas Moniz Ethics Committee (1050/2022) and all participants gave their informed consent. The study is reported according to the Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD) for model development (10).

To be included in this study, participants had to: have 18 years of age or older; being able to read, understand, and sign the informed consent form; and were seeking initial triage at the university dental clinic. Participants were invited to participate voluntarily and anonymously. Given the lack of studies integrating clinical oral conditions with patient reported outcomes, we did not estimate a minimum sample size and opted for study design based on a consecutive random sample obtained over a pre-defined timescale (18-month).

The proposed index (MOHi) was developed under the guidelines of the “Handbook on Constructing Composite Indicators: Methodology and User Guide” (17). MOHi was developed as a composite multi-dimension instrument, being built as an equally weighted linear combination of the normalized OHVS, OHIP-14 and DMFT scores. It is expressed in a continuous scale from 0 to 3, with increasing values progressively representing a more degraded oral health condition. After validation and optimization, the MOHi distribution, dynamics and sensitivity was evaluated, not only as a function of patient sociodemographic and behavioral characteristics, but also when considering the individual periodontal (self-reported) status.

Data analyses included descriptive, inferential and modelling procedures, which were performed using IBM SPSS Statistics v.30 software (Armonk, NY, USA). Continuous data were expressed as mean and standard deviation (SD), while categorical variables were expressed as frequency and percentages (%). For continuous variables, the adequacy to normal distribution was evaluated prior to inferential inter-group comparisons (t-Student's test and ANOVA one-way were applied). Stepwise multivariate logistic regression methodologies were used to model MOHi data and identify significant predictors towards a degraded oral health condition (MOHi score >= 1.5). An alternative model was performed with DMFT as a single item, with a cutoff of 14 or higher (18). These two reduced models were compared to verify the comprehensiveness of predictors and performance. Performance was evaluated by Receiver Operator Curve (ROC)/Area under the Curve (AUC) analysis. A significance level of 5% (p < 0.05) was used for all statistical inferential analyses.

For this study, 1,127 participants were invited to participate. A total of sixty-three refused to participate and undergo clinical observation, thirty had incomplete questionnaire responses, resulting in a final sample of 1,034 participants (Figure 1). The characteristics of this population are detailed in Table 1). This sample had a majority of female participants (58.1%), with a mean age of 46.8 (±18.6) years. Most participants were employed (67.9%), with either middle (33.5%) or higher education (33.8%), and 24.7% of active smokers. The average caries experience was 12.8 (±8.4) teeth, with an average of 6.3 (±7.5) missing teeth. About 9.0% of participants had severe tooth loss. Overall, 42.0% of participants had periodontitis according to their self-report. When estimating the oral health value, participants had a mean score of 31.1 (±6.3), and whose percentage scores 64.8% (±13.2). The average OHRQoL was 12.1 (±12.0) and most participants reported a frequently affected quality of life (57.8%).

The distribution characteristics indicate OHVS and DMFT scores were relatively symmetrically distributed with slight platy kurtosis (Table 2), and OHIP-14 scores were more positively skewed, highlighting a tendency for a larger number of participants to report lower impact levels on their oral health-related quality of life.

The new score composite regarded three normalized algebraic components (Table 3). The MOHi score distribution is fairly symmetrical with a slight right skew (Figure 2 and Table 4), suggesting that most participants score around the mean, but some report higher values.

The MOHi scores varied significantly across self-reported periodontal status and sociodemographic and behavioral factors (Table 5). Participants with periodontitis, older adults, individuals with lower education levels, smokers, retirees, and those who were divorced or widowed exhibited higher MOHi scores, indicating poorer oral health outcomes (p < 0.001 for all). A distinct age-related trend was observed, with MOHi scores rising from 1.02 in individuals aged 18–44 to 1.47 in those 65 and older. Additionally, lower educational attainment and current smoking habits were linked to poorer outcomes. Although women exhibited slightly higher MOHi scores compared to men, this variance did not reach statistical significance (p = 0.066). These results highlight the substantial impact of factors such as age, education level, tobacco use, and socioeconomic conditions on individuals’ perceptions and outcomes related to oral health.

The results of a multivariate logistic regression model show age, self-reported periodontitis, being female, lower education levels, and smoking were all significant predictors of degraded oral health status (MOHi ≥ 1.5) (Table 6). Active smoking and lower education emerged as the strongest risk factors, highlighting the critical role of modifiable behaviors and socioeconomic conditions in influencing oral health outcomes. The results of the reduced model for high risk of dental caries experience (DMFT ≥ 14) revealed only age, sex and smoking habits as associated risk factors (Supplementary Table S1) compared to the MOHi model.

While the MOHi assessment represents a promising innovation, several limitations warrant consideration. First, the self-reported components may be subject to reporting bias, potentially affecting the accuracy of risk categorization. Also, the association with self-reported periodontitis warrants due to being a modest prediction screening tool (80%) that does not replace clinical diagnosis (12). Future studies might explore whether periodontal clinical diagnosis links with MOHi remains. Second, the applicability of the MOHi assessment across diverse populations remains to be tested. Factors such as cultural differences in oral health values and access to care could influence its generalizability. Additional research is needed to refine the tool and evaluate its performance in varied contexts. Third, while the MOHi assessment provides valuable insights into risk stratification, it does not account for broader determinants of oral health, such as other oral diseases (e.g., xerostomia, temporomandibular disorder, bruxism), systemic conditions, environmental factors, and genetic predispositions. The integration of these variables into future iterations of the tool could potentially enhance its predictive capability.