Is Psychosocial Impact of Dental Aesthetics Questionnaire (PIDAQ) Valid for the Indian Population?—A Psychometric Study

Abstract

INTRODUCTION The orofacial region is an area of prime concern to individuals as it draws major attention during interpersonal communications.[1] Quality of life (QoL) is an individual’s perception of well-being that results from satisfaction/dissatisfaction with the aspects of life that are of greater significance to him/her.[2] Oral health forms a vital part of general health which is crucial for well-being and is a determining factor of QoL; whereas Oral Health-Related Quality of Life (OHRQoL) is a subset of QoL pertaining to the health of oral tissues. Malocclusion comprises a broad array of dental malalignments such as protruded teeth, crowding, spacing, rotations, etc. It is one of the most prevalent oral conditions, next to dental caries and periodontal disease, and ranks third among the worldwide dental health problems. Most patients seek orthodontic care for aesthetic reasons which are reflected in their psychosocial well-being. This indicates a need for psychometric scales to assess the impact of malocclusion on OHRQoL. The Psychosocial Impact of Dental Aesthetics Questionnaire (PIDAQ) developed by Klages et al.[3] is a condition-specific (CS) scale that evaluates the psychosocial impacts of dental aesthetics on OHRQoL similar to the Orthognathic Quality of Life Questionnaire (OQLQ)[4] for subjects undergoing orthognathic surgery and the Malocclusion Impact Questionnaire (MIQ)[5] for young adolescents seeking orthodontic treatment. The final version of PIDAQ[3] contains 23 questions in four domains, viz., Dental Self-Confidence (DSC) with six items that measure the influence of dental aesthetics on an optimistic self-image; Social Impact (SI) comprises eight items which weigh stress levels of subjects toward others’ reaction when the individual’s teeth get exposed; Psychological Impact (PI) domain consists of six items that quantify negative feelings toward one’s own dental appearance; and Aesthetic Concern (AC) with three items that evaluate the subject’s discontent toward his/her dentition getting exposed in different situations. Initially drafted in German for use among young adults aged between 18 and 30 years, it was soon professionally translated into English for its worldwide application[3] and subsequently adapted for younger adolescents.[6] The use of a CS scale in a different geographical area calls for its translation, cross-cultural adaptation, and validation prior to use. This ensures regional comparison of the psychosocial impacts of malocclusions across populations using a standardized scale. PIDAQ has been previously translated into and validated in various languages.[7891011121314151617] However, the scale is not available in any of the native languages of the Indian subcontinent except Hindi.[18] Hence, the present study aimed to derive a culturally and semantically equivalent South-Indian version of PIDAQ (PIDAQMal) and to compare the perceptions among subjects of previously validated groups. MATERIALS AND METHODS The study commenced after obtaining Institution Ethics Committee approval (IEC/M/14/2017/DCK) and was conducted at the Department of Orthodontics, Government Dental College, Kottayam, Kerala, India. Individuals aged between 18 and 25 years, who reported for orthodontic treatment, were included in the study after obtaining informed consent. Subjects undergoing or with a previous history of orthodontic treatment and those with structural, functional, or cognitive abnormalities were excluded. The method involved deriving a regional version of the scale (PIDAQMal) followed by its application among the South-Indian population to assess the psychosocial impacts of malocclusion. The process of development of PIDAQMal included validation of both linguistic and psychometric properties. The linguistic validation process comprised a preliminary translation of the questionnaire into Malayalam, a pretest of this version in a sample of 15 subjects followed by back-translation to derive the draft questionnaire. The psychometric validation of the draft scale was performed to ensure its reproducibility, internal consistency, reliability, and construct validity. TRANSLATIONAL VALIDATION AND CROSS-CULTURAL ADAPTATION OF PIDAQMAL The English version of PIDAQ[3] was translated into the target language by a team of three independent bilingual translators. The team of translators included a public health dentist who was an expert in OHRQoL measures, an orthodontist competent in the translation and validation process, and a college professor. The translations were analyzed with respect to diction and content, taking care of the item and conceptual equivalences between the published and translated versions. Special consideration was given to vernacular expressions and verb tenses. The best among the three translations was selected by consensus as the draft scale. This was pretested on a sample of 15 subjects who reported to the orthodontic outpatient department and the questionnaire was dictated to them individually for clarity of language, ambiguity, and need for any cultural modifications. Responses were collected and discussed with subjects to assess difficulties in understanding each item and the need for any modifications as a means to optimize face validity. The suggested changes were discussed among a panel of specialists from different fields and incorporated into the draft scale based on their recommendations. The draft questionnaire was back-translated into English by another team of three bilinguals who had not seen the original version beforehand. The team of back-translators included a dentist efficient in the translation and validation process, an expert in QoL measures, and an English-teaching college professor. The three versions (published English, regional version, and back-translated English) were assessed by two postgraduate dental professionals trained in the validation process. Inconsistencies between the back-translated version and the published one were analyzed and rectified, ensuring that the final questionnaire was semantically and conceptually equivalent to the original version. RELIABILITY TESTING Test–retest reliability of PIDAQMal was ensured by conducting a pilot test on a class of 30 first-year undergraduate dental students. The same exercise was repeated after 15 days without prior intimation. An insight into the average time needed to complete the questionnaire was also obtained. CONSTRUCT VALIDATION Validation was subsequently carried out in 288 subjects who fulfilled the inclusion criteria. The sample size was calculated using the rule of thumb suggested by Plichta and Kelvin[19]: N= number of items in the scale × 10 subjects; hence 23 × 10= 230 was the required number of participants. Accounting for 25% data loss due to incomplete questionnaires, the final sample size was decided to be 288. Sociodemographic data of participants were also collected. The draft questionnaire was administered to subjects in a separate room and adequate time was given to fill the response sheet. Response to each item was rated using a five-point Likert scale as in the original version: 0 = “not at all,” 1 = “a little,” 2 = “somewhat,” 3 = “strongly,” and 4 = “very strongly.”[3] All items in the DSC domain which were positively worded were reverse-scored to bring the direction of scoring in line with other subscales.[9151617] EVALUATION OF TREATMENT NEED Normative treatment need was assessed using Dental Health Component (DHC) and subjective need using Aesthetic Component (AC) of the Index of Orthodontic Treatment Need (IOTN) and Perception of Occlusion Scale (POS). The sample was divided into three IOTN-DHC groups (Group 1 with grades 1 and 2 denoting no need for treatment, Group 2 included grade 3 representing moderate need, and Group 3 with grades 4 and 5 denoting definite treatment need). For assessing IOTN-AC, 10 photographs of anterior teeth depicting different levels of malocclusion were presented to subjects and asked to select the one that most closely simulated their dentition. The photographs were graded on a 10-point scale from the most attractive to the least. The POS includes six items that describe the upper and lower anterior occlusal traits. The response was rated on a five-point Likert scale ranging from 0—“not at all” to 4—“very strongly.” Convergent validation was performed by comparing PIDAQMal scores with self-rated IOTN-AC and POS scores and discriminant validity by comparing PIDAQMal with the normative assessment of malocclusion status. Mean domain scores of the available translated versions of PIDAQ were extracted and tabulated for each domain and compared with domain scores of the present study to bring out the sociocultural differences in perceptions. Data were analyzed using SPSS software (version 16.0, Chicago, IL, USA). Internal consistency was analyzed using Cronbach’s alpha coefficient and test–retest reliability using Intraclass Correlation Coefficient (ICC). Psychometric properties were assessed by performing Exploratory Factor Analysis (EFA) using Principal Component Analysis (PCA) with varimax rotation and Kaiser normalization.[20] One-way analysis of variance (ANOVA) with post hoc Tukey test and Spearman’s correlation were used to validate the scale. Independent t-test compared the difference in perceptions between genders in the sample. The level of significance was set at 5% (P < 0.05). RESULTS The study included 288 participants (31.2% males and 68.8% females) with an age range of 18–25 years (mean = 20.43, SD = 2.185). Three questionnaires (1.04%) were incomplete, resulting in a final count of 285 samples. The sociodemographic data are shown in Table 1.Table 1: Sociodemographic data of participantsDuring the pretest, few participants had difficulty in understanding the item “I hold myself back when I smile so my teeth don’t show so much.” Following discussion with experts, minor linguistic modifications were made, which were incorporated into the final version. Cronbach’s alpha coefficient of the scale was 0.926 and that of the domains ranged from 0.827 for PI to 0.883 for SI [Table 2]. Alpha if item deleted for 23 items and corrected item total correlations are shown in Table 2. The test–retest reliability estimate (ICC) ranged from 0.74 to 0.91.Table 2: Item-wise factor loading after Principal Component Analysis and varimax rotation with Kaiser normalization, amount and percentages of explained variance, Cronbach’s alpha if item deleted, and reliability of each subscaleThe Kaiser–Meyer–Olkin measure of sampling adequacy (0.89) ensured the sample size to be adequate. Data were found to be factorizable as Bartlett’s test of sphericity was significant (P < 0.001). PCA extracted four factors in accordance with the Kaiser–Guttman criterion, each with an eigenvalue more than 1.0. The factor loading of each item following varimax orthogonal rotation along with the amount and percentages of explained variance is listed in Table 2. The first factor contained items belonging to the SI subscale (items 7–14 and 19) which explained 38.86% of the variance. Factor 2 embodied items 1–6 of the DSC subscale and explained 11.12% of the variance. Factor 3 contained items of the PI subscale (15–18 and 20) explaining 6.42% of the variance. Finally, the fourth factor contained items 21–23 of the AC subscale explaining 5.61% of the variance. The four components together accounted for 62.01% of the total variance. A statistically significant difference (P < 0.001) in the total PIDAQMal scores (mean = 60.42, SE = 1.057) among the IOTN-DHC subgroups tested using ANOVA with post hoc Tukey test ensured discriminant validity of PIDAQMal [Table 3].Table 3: Comparison of scores among the DHC of the IOTN categorized groups using one-way ANOVA with post hoc Tukey testA significant correlation (P < 0.001) between PIDAQMal scores and self-rated IOTN-AC (ρ = 0.317) and POS scores (ρ = 0.354) ensured convergent validity [Table 4].Table 4: Comparison and correlation of the domain and total scores in subjects with different self-rated ACs of the IOTN scores and POS scores using one-way ANOVA and Spearman’s correlationThere were no statistically significant differences (P > 0.05) in the mean PIDAQMal scores between males (60.48, SE = 1.71) and females (60.39, SE = 1.33). Data extracted from the currently available versions of PIDAQ for PI, SI, AC, and DSC domains are presented in Figure 1. A notable difference in psychosocial impact was evident across the European, Asian, and South American populations. DSC trait, being a positive protective factor against negative impacts of malocclusion, also showed similar variations.Figure 1: Comparison of PI, AC, SI, and DSC among the currently available versions of the questionnaire and the present studyDISCUSSION The importance given to patient-reported outcome measures in clinical studies by international agencies like the World Health Organization paved way for a rise in QoL research over the last two decades. Several OHRQoL assessment scales have been introduced to assess the impact of oral conditions on QoL, which require cross-cultural adaptation and validation prior to use.[45,2122] PIDAQ is a CS scale that evaluates the self-reported influence of dental aesthetics on OHRQoL.[3] Translation into the native language was performed meticulously, following the guidelines of Guillemin et al.[23] The original and back-translated versions were made similar following minor modifications in the initial translation, which was subsequently pilot-tested to assess test–retest reliability. The sample size selected based on the rule of thumb was similar to other studies.[78,1215] As PIDAQ was primarily developed for young adults, subjects in the age group of 18–25 years were selected.[712,13] The process of cross-cultural adaptation involves ensuring equivalence between the original and translated versions. It was done according to the steps outlined by Herdman et al.,[24] the same protocol being used in other versions.[716,17] The team of experts confirmed the presence of conceptual equivalence between the two versions. Semantic equivalence was attained by transferring the meaning of concepts from the original version to the translated one, thereby achieving similar responses from individuals belonging to different ethnic backgrounds. With respect to item equivalence, there was difficulty in understanding the item “I hold myself back when I smile so my teeth don’t show so much” during the pretest. This was because an exact linguistic equivalence could not be obtained for the phrase “hold myself back”; hence, words that convey similar meaning were used. This was discussed among the panel of translators. As there were differences among the three translators, the item was presented to another team of two experts who analyzed the three versions and derived a final one by consensus. Malay PIDAQ depicted a similar problem.[16] Recently, when MIQ was cross-culturally adapted to Chinese, an east-west difference in culture was attributed to such issues.[25] Operational equivalence was ensured using pilot test which confirmed that the mode of administration, format, and response options were acceptable. The questionnaire was effortless to answer and suitable for self-administration with an average response time of 14.5 min per subject. The response time for the Moroccan Arabic version of MIQ which had fewer items than PIDAQ was 10–17 min.[26] Internal consistency of PIDAQMal was good with Cronbach’s alpha ranging from 0.83 to 0.88, which was higher than the Malaysian English (0.56–0.84)[17] and French PIDAQ (0.67–0.87)[10] while similar to other versions.[37,1314,16] Cronbach’s α of 0.70 or greater is considered acceptable for a new scale and above 0.80 for established scales.[20] Alpha if item deleted remained constant for all items, ensuring the importance of each item contributing to the internal consistency of the scale. The test–retest reliability was ensured by a repeated measure analysis (ICC) with a 15 days interval. ICC of PIDAQMal (0.74–0.91) was marginally greater than that of Malay PIDAQ (0.72–0.89)[16] and equivalent to the Spanish version (0.76–0.90).[11] The time interval suggested for test–retest reliability ranges from few days to a month.[27] Most previously validated versions have used a similar time interval.[57,916,17] When subjected to PCA, PIDAQMal was found to have a similar structure to the original version with four common factors extracted, accounting for 62.01% of the total variance. This is above the minimum value (60%) suggested for the factor solution to be stable.[20] Unlike the present study, the Chinese[9] and Italian[14] versions had three and Nepali PIDAQ[12] had five common factors extracted. The item “I feel bad when I think about what my teeth look like” originally included in the PI domain showed good factor loading under the SI domain in the current study. This might be due to differences in understanding and perception of concepts by the study population compared with other groups. Such cross-loading of items was also reported in other versions.[1215] The item that faced problems during the pretest, however, showed factor loading under its original domain. But, the loading was The translated scale good convergent validity its with self-rated IOTN-AC and POS groups with subjects higher domain and total scores higher IOTN-AC and POS were in other versions of of the of IOTN-AC is that it has no depicting malocclusion, which of the subjects in the present study. it has been used in studies The current study had a number of subjects in the moderate and definite treatment need However, the results of one-way ANOVA showed that PIDAQMal was to individuals with no need for treatment from those with moderate and definite need, thereby good discriminant validity. But, it to individuals with moderate need from with the results being statistically not significant (P > 0.05). This the between normative and psychometric of to of participants in this study. This could be attributed to the that females are more about their and dental aesthetics than the last showed a to in the region other in the were no statistically significant differences in the domain and total PIDAQMal which is similar to the Nepali version (P > This however, not in accordance with other studies in which females more concern and about their dental than A of this study was an to a comparison of the mean scores among the previously validated versions of PIDAQ with the current study. This is the of a scale and it across populations than using different scales to measure similar The impact due to malocclusion among those in the definite malocclusion was found to be the among the followed by the Malay population and among The previously validated version and the present study showed similar PI regional AC was found to be the among the and with the The current study showed a level of concern similar to that of the Spanish population items under PI were under the Aesthetics domain in the hence, a was The level of impact due to malocclusion was also among the Malay version followed by the present and and However, the SI was with the English version The of PIDAQ was its to a measure of positive due to one’s own dentition. the items in this domain were a indicates subjects with a level of due to their dental and Moroccan Arabic and Nepali version showed the DSC among participants. The current study and had similar DSC domain ranges cultural among However, DSC being a subscale that the in this domain among versions. The Malay reported was not evident from the scores subjects in the malocclusion showed DSC this was an in scoring or the of OHRQoL when compared with normative assessment is A of while DSC not be in perceptions due to malocclusion across different population groups have been reported in the et in their found cultural differences in the between malocclusion and OHRQoL. The difference both in the perception of malocclusion and in the of OHRQoL. The reasons attributed were the differences in the of other oral conditions and the importance given to aesthetics in different and in the which are also to the present study. PIDAQ has been validated for the South-Indian population more than and those in this across the PIDAQMal be available as a scale to assess the psychosocial impacts due to malocclusion and the after treatment for the regional in the psychosocial impacts be made using the translated scale. However, the present study was not without The sample selected not for a of the population because the of subjects with definite treatment need was The of the scale could not be due to the of the study. It a study to ensure this A of PI have been more to bring out the perceptions across however, this was not due to of in the published versions. The English version of PIDAQ was adapted for the Indian population following linguistic and cultural The translated and cross-culturally adapted version was found to good reliability and validity. A geographical comparison across population groups their psychosocial impacts due to malocclusion significant The study for the of PIDAQMal as a condition-specific OHRQoL measure to assess the psychosocial impacts of malocclusion among young AND OF data draft and AND OF Data used in the current study be made available on to the to the of Health for the research and for the final

Topics & Concepts