Breaking barriers: Validation of a Spanish oral health knowledge tool to enhance patient-provider communication Silvia Spivakovsky, Joyce Figueroa, Ryan Richard Ruff ABSTRACT Objectives: This study aimed to develop and validate the Knowledge Related to Oral Health Literacy Spanish (KROHL- S) instrument to assess oral health knowledge among Spanish-speaking adults in the United States, a population facing significant oral health disparities. Design: A cross-sectional study was conducted at NYU College of Dentistry. A convenience sample of 175 self-identified Spanish-speaking adults (70% female, mean age 49. 79 years) completed the orally administered KROHL- S questionnaire. Participants, mainly born outside the US (91. 9%), also completed the Comprehensive Measure of Oral Health Knowledge (CMOHK) and a single-item literacy screening tool in Spanish (SILS). Psychometric properties of the KROHL- S, including internal consistency (Cronbach’s alpha), discriminant validity (correlation with CMOHK), and known-group validity (comparison across education levels), were evaluated. Confirmatory factor analysis was used to test the original factor structure. Results: The mean KROHL- S score was 8.34 (SD = 5.82), indicating a low level of oral health knowledge in the sample. Internal consistency for the overall KROHL- S was good (Cronbach’s alpha = 0.75), and interrater agreement was high. A moderate positive correlation was found between KROHL- S and CMOHK scores (r = 0.49, p < .0001). Participants with higher education levels showed significantly greater oral health knowledge on the KROHL- S. Confirmatory factor analysis suggested an average fit to the data (RMSEA = 0.064, CFI = 0.86, TLI = 0.83). Conclusion: The KROHL- S could be used to assess oral health knowledge among Spanish-speaking adults and incorporates cultural and linguistic aspects, making it suitable for a wider range of individuals. KROHL-S offers a valuable tool for healthcare providers by not only helping identify individuals’ knowledge gaps to guide customized educational interventions but also helping enhance patient-provider communications. Introduction The Hispanic population is the largest minority group in the US. Although clustered under a single ethnic group, individuals who self-identify as Hispanic include people from at least 19 Latin American countries, with nearly a third being foreign-born [ 1 ]. Despite this demographic significance, Hispanic communities experience disproportionately high rates of dental disease, including caries, periodontal disease, and tooth loss [ 2 ]. These disparities stem from multiple social determinants, including socioeconomic status, language barriers, cultural differences, limited access to healthcare services, and varying levels of health literacy [ 2 – 4 ]. Oral health literacy is defined as one’s ability “to obtain, process, and understand basic oral health information and services needed to make appropriate health decisions” [ 5 ]. Research consistently demonstrates a strong correlation between limited health literacy and poor oral health outcomes [ 2 , 6 ]. Spanish-speaking individuals with low health literacy are significantly more likely to miss routine dental check-ups, report difficulty understanding dental treatment instructions, and are less likely to engage in preventive oral health behaviors [ 2 ]. Different levels of educational attainment and limited access to oral health care in many areas of Latin America are further contributors [ 2 , 7 ]. To better understand the role of health literacy and its contributing factors to the oral health of our Spanish-speaking patients, our team developed a working framework that led to the creation of the KROHL-S model. The model was presented elsewhere [ 8 ]. In brief, the KROHL-S model was designed to help identify clinically relevant factors and relationships that are practical at the point of care, specifically those that can be improved or modified through patient-provider interactions. Cultural factors, self-efficacy, fear, and health literacy all play an important role in the development of effective patient-centered communications, which demonstrated to have a direct impact on oral health outcomes [ 9 ]. After the development of the conceptual model, we proceed to identify those tools that will be used during implementation. Our initial review of existing tools [ 8 ] to evaluate oral health literacy demonstrated that the most widely studied tools rely on word pronunciation (REALD-30, REALD-66, REALM-D, REALMD-20, and TS-REALD). Other instruments use a combination of word recognition and comprehension (OHLIP, OHLA-E/OHLA-S, and MOHLR-K), a combination of comprehension and numeracy (TOHFLID, OHLI, and SOHLS), or decision-making and communication levels are part of the skills evaluated by HelD29/HelD-14 and OHL-AQ, and one tool evaluated conceptual knowledge (CMOHK). The decision to evaluate oral health knowledge rather than oral health literacy was supported by the body of evidence linking health literacy, health knowledge, understanding of instructions, and correct disease-specific beliefs in medicine [ 10 – 11 ] as well as in oral health [ 12 – 13 ]. In addition, changes in health knowledge are a parameter widely used to assess public health interventions [ 14 – 16 ]. Moreover, there is no gold standard to measure oral health literacy [ 17 – 19 ], there is evidence questioning the quality of translations and cross-cultural adaptation of oral health literacy tools for non-English speakers [ 19 ], and the uncertain reliability of word recognition tests in general [ 20 ] and specifically in a language with phoneme-grapheme concordance [18,19,20} With health literacy considered content and context specific [ 21 ], the assessment of oral health knowledge was identified as the most useful aspect of personal oral health literacy that could be evaluated at the point of care, moving away from “labeling” a patient’s literacy level and instead identifying concrete patient educational needs. The choice of a tool with non-reading requirements was intentionally selected to address some of the issues reported, leading to poor participation of Hispanics in research [ 22 ] and the scope of the population for whom it may be applicable [ 8 ]. This choice also reflects the shifts in information acquisition, as more individuals rely on digital sources, social media, news websites, and podcasts. [ 23 ]. As such, text-based literacy may not fully capture how individuals obtain information. The study aimed to develop and test the psychometric properties of the Knowledge Related to Oral Health Literacy scale for Spanish speakers. (KROHL S). In the clinical setting, this tool can contribute to patient-centered care by identifying knowledge gaps that can be translated into personalized and linguistically appropriate interventions to improve communication and oral health. Materials and methods Research design and participants This was a cross-sectional study approved by the New York University School of Medicine Institutional Review Board (FY2019-2790). Data were collected from Spanish speakers at the New York University College of Dentistry. Potential participants were recruited from clinical waiting areas within the college. Eligibility criteria required participants to be at least 18 years old and to be Spanish speakers. Individuals who met the inclusion requirements were invited to take part in the study, and the objectives and procedures were explained in detail. Following Institutional Review Board (IRB) approval, oral consent was obtained from everyone who agreed to participate after ensuring their understanding of the study. The consent process was witnessed and documented, and a written copy was provided to the participant. Oral consent was chosen to accommodate potential literacy challenges and to help preserve anonymity. From 06/15 2019 to 03/10/2020, of the 185 individuals approached,175 participants were enrolled and completed the questionnaire. Most participants were patients, although some were parents accompanying patients. The questionnaire was administered through face-to-face interviews in a private setting by trained interviewers who recorded the participants’ responses. A printed copy of the survey was offered as well to view the questions. Participants were instructed that it was perfectly acceptable to respond with “I don’t know” if they were uncertain. Each interview lasted approximately 20 minutes. All data was collected using Qualtrics software. The KROHL-S questionnaire As in the English version, the KROHL-S questionnaire consisted of four questions each for five separate conditions: caries, periodontal disease, oral cancer, tooth loss, and malocclusion. Each question is related to a specific content domain: identification and symptoms, causes, prevention, and treatment. Each question was open-ended and followed a standard format: 1 “What do you see and feel while having ______ (condition options: caries, periodontal disease, oral cancer, tooth loss or malocclusion)?” a “¿Qué es lo que ve o siente cuando tiene ------ (caries, enfermedad de las encías, cáncer oral, perdida de dientes o maloclusión)?” 2 “What causes ______ (condition)?” a “¿ Qué causa -----------(nombre de la condicion)?” 3 “How do you prevent ______ (condition)?” a “¿Qué se puede hacer para prevenir ---------- (nombre de la condicion)?” 4 “How do you treat _____ (condition)?” a “¿ Cual es o son los tratamientos de -------(nombre de la condicion)?” The questionnaire was developed in Spanish by a native Spanish speaker. Five non-team native Spanish speakers from different countries then reviewed the questionnaire for readability and provided recommendations on included terminology. In addition to domain-specific items, demographic information (e.g., age, sex, place of birth, years living within the United States, level of education, insurance status, and oral health status) was also collected. Following data collection, trained evaluators who were native Spanish speakers evaluated each response for scoring. The scoring system was built following the same approach as with the English version of the KROHL [ 24 ]. Team members agreed on the content areas that should be used as proof of optimal knowledge for each question, and from that starting point, a set of parameters was developed and tested until reaching consensus. Based on the pre-established set of parameters, the most commonly used words and their synonyms provided by the participants when answering to a particular question were used to develop the detailed scoring criteria. For example, when assessing caries prevention, three content areas were assessed: oral self-care practices, dietary considerations, and professional care. The highest score was given to an answer that include brushing teeth after a meal (“cepillarse despues de comer”), don’t eat sweets (“no comer dulces”) and go to the dentist (“ir seguido al dentista”). Responses to each question were then graded using a 4-point scale: good (2 points), fair (1 point), minimal (0.5 point), or no points (0) if the response was ‘I don’t know’ or incorrect. The final KROHL-S score was the sum of the 4 individual questions for each of the 5 conditions. The question on see/feel for tooth loss was excluded from the final calculation because, as with the English version, patients only rephrased the concept of tooth loss. Additional measures Data were also collected using the Comprehensive Measure of Oral Health Knowledge (CMOHK) scale and the Single-item literacy screening tool in Spanish (SILS). The CMOHK [ 25 ] consists of 23 multiple-choice questions assessing general knowledge of oral health, including the prevention and management of dental caries, periodontal disease, and oral cancer. The CMOHK, which is also available in Spanish, has shown strong psychometric reliability across various settings, including Spanish speakers [ 26 ]. Scores are calculated by totaling the number of correct responses. It categorized individuals as having poor (0–11), fair (12–14), or good levels (15–23) of oral health knowledge. A modified scoring system is also available with 2 categories: low (0–14) or good (15–23) [ 26 ]. The single-item literacy screening (SILS) tool in Spanish uses a 5-point Likert scale to answer: “How confident are you filling out medical forms by yourself? (¿Con qué frecuencia puede usted completar formularios de salud sin recibir ayuda?), with 1 indicating never (nunca) and 5 indicating always (siempre). SILS has effectively identified individuals with inadequate health literacy among Spanish speakers in various settings [ 27 , 28 ]. Statistical analysis Descriptive statistics were computed for the full KROHL-S scale and each individual subscale, and histograms were produced. Internal consistency was evaluated using Cronbach’s alpha, and discriminant validity was assessed using Pearson’s correlation comparing KROHL-S scores to CMOK scores. Interrater agreement was calculated using the Intraclass Correlation Coefficient. Known-group validity was explored by comparing scores by level of education using one-way analysis of variance with post-hoc tests using the Tukey multiple comparisons of means method. The preexisting factor structure of the English-language version of the KROHL was then tested in the new sample using confirmatory factor analysis. Sample size rationale. Conventional sample size requirements for factor analyses were followed [ 29 ]. There were no instances of missing data and no convergence issues in the analysis. Results Of the 175 enrolled participants, 70.85% were female, with an average age of 49.79 (ranging between 18 and 85 years). The majority (92%) were born outside the US, representing 13 Latin American countries, and had lived in the US for an average of 22.72 years (range from 1 to 60 years). Spanish was the primary language for 97.17% of participants. Approximately 37.14% of participants reported not having insurance to cover for their dental treatment. Regarding dental visits, 70.85% had visited a dentist within the past year, 18.28% had done so between 1 and 2 years, and 10.85% had not visited a dentist for more than 2 years. Notably, 2 of the participants were attending their first-ever dental visit ( Table 1 ). 10.1371/journal.pone.0350384.t001 Table 1 Participants characteristics. Age 18-29 22 12.57% 30-39 27 15.43% 40-49 36 20.57% 50-59 37 21.14% 60-69 32 18.28% 70 and older 21 12.00% Sex Female 124 70.85% Male 51 29.14% Country of birth US 14 8% Other 161 92% Years in the US Average 22.72 years from 1 to 60 Main language Spanish 170 97.17% English 5 2.82% Highest level of education post College 9 5.14% College 56 32.00% HS 72 41.14% Elementary 27 15.43% Incomplete Elementary 12 6.85% Oral health status self-reported Very poor 14 8.00% Poor 38 21.71% Good 97 55.42% Very good 26 14.85% Dental insurance Yes 110 62.85% No 65 37.14% Last dental visit <1 year 124 70.85% Between 1 and 2 years 32 18.28% >2 but