Background
Both in high and low income countries, outcomes in dentistry have relied on clinical measures considered important from a clinician’s point of view [
1]. Although informative, the clinical approach has been criticized because it does not consider functional and psychosocial aspects of oral health. Thus, a shift towards patient-important outcomes has been promoted [
1‐
3]. Concerns that clinical measures alone may not be adequate for assessing people’s oral health needs have sparked the development of Oral Health-Related Quality of Life (OHRQoL) measures. Tested in a variety of populations, OHRQoL measures are increasingly used to complement clinical indicators [
1‐
3]. Increased use of OHRQoL measures warrants cross-cultural adaptation of the existing instruments. Efforts are ongoing to translate and adapt OHRQoL measures for use in non-western cultural settings [
4,
5].
OHRQoL indices should be simple to use, reliable, valid, precise, sensitive for change, and amenable to statistical analysis. Two indices come close to meeting those criteria: the eight-item Oral Impacts on Daily Performance (OIDP) scale and the 14-item Oral Health Impact Profile (OHIP-14) [
6‐
8]. Both measures are based on the conceptual framework of the World Health Organization’s International Classification of Impairments, Disabilities and Handicaps, (ICIDH) [
9], amended for dentistry by Locker [
10]. The OIDP focuses on measuring the most severe oral impacts, namely disability and handicap [
6]. The OHIP-14 is derived from the original 49-item OHIP questionnaire. It assesses seven dimensions of impact, including functional limitations, pain, psychological discomfort, physical disability, psychological disability, social disability, and handicap [
7,
8]. In terms of respondent burden, both the OIDP and OHIP-14 inventories are relatively short and thus suitable for use in population surveys. Both measures seem to perform well using un-weighted, rather than weighted, scores, although the individually sensitive weighting system of the OIDP inventory gives prominence and increased validity to respondent views [
11,
12].
The OIDP and its specific child version, the Child OIDP, appear to be applicable to the general younger populations, as found for instance in Great Britain, Tanzania, Uganda, Brazil, and Peru [
13‐
18]. The OHIP-14, one of the most commonly used generic OHRQoL measures, has proved reliable and valid among both young and middle-aged people in Sweden [
19], Brazil [
20], Scotland [
21], New Zealand [
22] and Japan [
23]. There have been few reports comparing OHIP-14 and OIDP. In a cross-sectional study involving adolescents in Myanmar, both OIDP and OHIP-14 showed reasonably satisfactory psychometric properties [
24]. However, OHIP-14 emerged as the superior measure with respect to construct validity in that it discriminated better than the OIDP between groups with impacts and those without [
24]. Robinson et al. [
25] found similar results comparing OIDP and OHIP-14 among dental attendees in the UK. Baker et al. [
26] compared OHIP-14 and OIDP in UK dental patients with xerostomia and found that the OHIP-14 inventory performed better overall than the OIDP. Bernabe et al. [
18], reported a moderate level of agreement between the OIDP and OHIP-14 in a sample of Brazilian adolescents, probably due to differences in the scoring systems and content of oral impacts between the two inventories.
Few attempts have been made to evaluate OHRQoL measures among young people from the general population in developing countries [
14‐
17]. Comparisons between OHRQoL measures are even more seldom in non-occidental contexts [
24]. The epidemiology of oral health-related quality of life among young people in China, the world’s most populous and second largest country by area, remains unclear. This is notable because young people are a major focus of dental public health care, globally. Moreover, where resources are scarce, patient-based oral health outcomes can help ensure that services are directed at the conditions most likely to negatively impact OHRQoL.
The present study assessed the reliability and validity of two OHRQoL measures in a sample of Chinese high school students: the abbreviated 14-item OHIP (OHIP-14) and the OIDP. These measures were chosen since they are both derived from theory and relatively short and are thus suitable for questionnaire-based population surveys [
13]. The study compared the internal consistency reliability, test-retest reliability, and construct validity of the OIDP and OHIP-14 inventories. Construct validity was assessed by evaluating the relationship of each inventory with global oral health ratings, socio-demographic factors, and oral health-related behaviors as well as the association and level of agreement between OIDP and OHIP-14.
Discussion
This study compared for the first time a Chinese version of the OIDP and OHIP-14 inventory in a population of adolescent students in Xi’an province, China. Whereas the OIDP has not previously been evaluated in this socio-cultural context, the applicability of the original OHIP-49 and two abbreviated OHIP-14 instruments have been tested in face-to-face interviews with the elderly population and recently among middle aged stroke patients in Hong Kong [
31,
32]. Since health is a dynamic state, application of the OHIP-14 among school-going adolescents in China required reestablishment of its psychometric properties when using self-administered questionnaires in accordance with the data collection of the original English OHIP-14 version [
33].
The Chinese versions of the OIDP and OHIP-14 appeared to be valid and reliable with psychometric properties similar to their original English versions [
6‐
8]. Moreover, the Chinese version of the OHIP-14 had psychometric properties similar to the abbreviated OHIP versions that were derived for use among elderly in Hong Kong (OHIP-14 original and OHIP-14 Chinese version) [
31]. Consistent with a previous study of Brazilian adolescents, this study revealed a moderate level of agreement between the OHIP-14 and OIDP (kappa value 0.45), reflecting differences in content validity [
17]. Such moderate agreement may reflect variation in scope, with OHIP-14 assessing oral impacts on all levels, in accordance with the model by Locker [
10], while OIDP emphasizes the most severe impacts only: the levels of handicap and disability. The strength of the correlation coefficient between the two inventories (Spearman’s rho 0.65) provided support for their common theoretical origin [
9,
10]. The correlation coefficient observed in this study is stronger than that of 0.40 reported by Soe et al. [
24] among Myanmar adolescents, but agrees with what was reported by Robinson et al. [
25] investigating dental attendees in UK. In general, OIDP and OHIP-14 performed almost equally well among the Chinese school-going adolescents investigated in this study. This indicates that the total burden on participants (additive scores) was as important in this young population as was the number of areas affected (simple count scores).
All Cronbach’s alpha values observed met Nunnally’s standard of 0.70 for appropriate internal consistency in studies involving group comparisons [
30]. These figures compare favorably with those obtained in other studies involving young people from high and low income countries [
13‐
19]. In the present study, agreement in terms of kappa values was 0.52 for the presence of an impact as detected by OHIP-14 and 0.66 for OIDP. This corresponds to the test-retest results reported in the study of Myanmar adolescents [
24]. Our finding indicates acceptable test-retest reliability, although the kappa value for the presence of an impact denoted moderate agreement for OHIP-14 and good agreement for OIDP [
30].
Cultural issues, in particular language, might cause problems with validity [
33]. Although no approach guarantees cross-cultural equivalence, the Chinese version of the OIDP and OHIP-14 seem to preserve the overall concepts of their corresponding English versions and do not differ in the sequence of questions, Likert scale, or recall memory period (six months) used. Previous experience regarding the usability of the OHIP inventory in its original- and abbreviated versions in personal interviews among the elderly, as well as recent self-administered questionnaires among school going adolescents, support the cross-cultural equivalence of this inventory. Completion rates for the OIDP (missing items varied from 0.2-1%) and the OHIP-14 (missing items varied from 0.2-0.6%) were acceptable, adding support to the face validity of both measures. There was no indication from the reference groups of academics, or from the pilot study among adolescents, that the relevance of any of the items was low in the context of Chinese school-going adolescents. This suggests that the Chinese high school students were capable of fully understanding the translated version without altering the meaning of the questions and that the Chinese and English versions of the OHIP-14 and OIDP inventories are comparable.
Both measures had significant validity in that they varied systematically, equally strongly, and in the expected direction with global oral health measures (Table
3). Thus, independent of the scoring system, the OIDP and OHIP-14 indicated (to the same degree) lower levels of oral impacts when self-perceived oral health was better. This similarity in performance agrees with the results reported by Baker et al. [
26]. It disagrees, however, with the results of Soe et al. [
24] and Robinson et al. [
25] who reported that the strength of the associations with oral health ratings were weaker for the OIDP than for the OHIP-14. Since the present study considered extent scores of OIDP only, it disagrees with recent findings of Kristapong et al. [
34] among Thai adolescents. In that study intensity OIDP scores associated with global oral health ratings whereas extent scores did not. The extent of oral impacts is calculated as a simple count scores (OIDPSC), whereas intensity scores calculate the percentage of respondents with impacts [
6]. As in this study, clinical measures have traditionally been excluded from previous validations of the OIDP instrument [
6]. The rationale behind the decision to omit clinical variables derives from the conceptual distinction between health and disease [
6,
9].
The OIDP and OHIP-14 scores were applicable across age and gender, with females and older students being most likely to report any impact (Table
4). Locker and Miller [
35] found younger Canadians to be as likely as their older counterparts to report impacts of oral disorders. Similar findings have been reported among students in Tanzania [
14]. Reports have shown that women perceive more negative impacts than do men, suggesting gender differences in their life-course influences [
22,
33]. The higher prevalence of impacts reported by disadvantaged students (rural residents from the poorest family wealth category and having mothers with lower education) is probably partly due to material and social deprivation [
36]. Greater frequency of smoking and sugar intake seem to imply less favorable students’ OIDP and OHIP-14 ratings. Moreover, the better the brushing and dental attendance ratings the more favorable the oral quality of life ratings. This is consistent with findings among Ugandan adolescents as reported by Åstrøm and Okullo [
16] as well as with studies conducted elsewhere in non-occidental cultural contexts [
14,
15].
Relatively high proportions of students reported being affected by an oral impact during the six months preceding the survey both for OIDP (45%) and OHIP 14 (62%). The higher number of impacts found with OHIP-14 compared to OIDP could be due to the greater number of items and to different content of the questions included (24, 25). Contrary to the OHIP-14, with its designed sensitivity to less severe impacts of oral condition, OIDP concentrates only on the most severe impacts and may thus be less sensitive in younger populations with lower levels of oral disease. Other studies have reported higher proportions of impacts using OHIP-14 than for OIDP [
17,
24,
25]. Eating was the most commonly reported OIDP impact, affecting about 19% of urban and 26% of rural residents using both OIDP and OHIP-14 [
13‐
19,
36]. In contrast, self-consciousness was the most frequently reported OHIP-14 impact (Table
2). Elderly Chinese have reported low frequencies of negative impacts in the psychological, social, and disability domains of OHIP [
31]. The present study of young Chinese people found psychological discomfort (self-consciousness) and functional aspects to have the greatest impact in both the OHIP-14 and OIDP. Thus, elderly Chinese may be more likely to accept their oral condition so that oral problems do not hinder their social life as much as for the younger part of the population [
31].
The OIDP estimates obtained in this study fell short when compared to the prevalence of OIDP reported among young people elsewhere such as in Uganda [
16]. On the other hand, the present OIDP prevalence was higher than that reported among young adults in Tanzania as well as in a nationwide Norwegian study (18%) [
14,
36]. The OHIP-14 prevalence and the mean additive OHIP-14 scores compared favorably with those obtained among subjects of the same age in Myanmar [
24]. It fell short, however, of the corresponding estimate for dental attendees in UK [
22]. The observed variety in prevalence of impacts using identical OHRQoL instruments might be due to differing perceptions of oral health in different populations or to reporting biases. Notably, comparison of prevalence estimates across surveys should be done with caution. Many previous analyses have selected the number of impacts at the frequency of “fairly often” or more with OHIP-14 and at “once a week” or more with the OIDP. Such a high threshold for prevalence was not suitable in the present study because of the skewed distribution of impacts. Thus, in accordance with McGrath and Bedi [
37] our study selected the criterion “having impact at any frequency” (all categories included regardless of their frequency) to capture those subjects experiencing only a single impact. Such differences in the use of the instruments may explain variation in levels of impacts across various populations.
Some additional limitations should be acknowledged when interpreting the results. Since regression analyses did not adjust for clinical measures of oral diseases, it is uncertain whether or not the social gradient observed is related to various levels of oral diseases. Moreover, without the possibility of confirming any causal relationship between socio-behavioral factors and OHRQoL with a cross sectional design, the present findings strongly suggest that perceived oral health status is shaped by lifestyles and prevailing social circumstances. The accuracy of reporting perceived impairments in population based studies may be limited. Another caveat may be the inventories using a 6 month recall period relying on self- reports which implies they can be prone to recall bias.
LH: PhD candidate, University of Bergen, Norway. TL: Professor Oral Facial pain and jaw function, Specialist center for oral rehabilitation, Norrkøping, Sweden, IMN: Clinical consultant, Specialist Center for Oral Rehabilitation, Norrkøping, Sweden. AJ: Professor, University of Bergen, Norway. ANÅ: professor, University of Bergen, Norway.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
LH : principle investigator, designed the study and collected data, performed statistical analyses and wrote the manuscript. TL: Provided valuable guidance in analyses of data and manuscript writing. IMN: provided guidance in manuscript writing. AJ: main supervisor participated in design of study, analyses and manuscript writing. ANÅ: co supervisor participated in analyses and manuscript writing. All authors read and approved the final manuscript.