Developing a new self-reported scale of oral health outcomes for 5-year-old children (SOHO-5)

Despite general improvements in oral health, dental caries is a major public health problem affecting 60%–90% of children globally [1, 2]. Consequently, children are the focus for oral health policy in many countries [3, 4] and measurement of oral health and quality of life should be key outcomes for evaluating oral health programs. Dental conditions, untreated dental caries in particular, have been linked with delayed growth and cognitive development of the child [5‐9]. Furthermore, they can negatively impact on the daily life of children and affect their families psychologically and economically [8, 10‐13].

Many “oral health related quality of life” (OHRQoL) measures have been developed and tested in various populations to assess the impacts of oral conditions on the daily life of people [14]. Such measures have also been developed specifically for child and adolescent populations, in line with the respective general health measures [15‐18]. However, they have predominantly focused on children aged 8 years and older [19‐22], with the exception of a parental report measure for children aged 6 years and older [12]. While a self-reported set of OHRQoL questions has been used on very young children [23], those were treated as independent questions, rather than a composite measure, without presenting concrete evidence on psychometric properties. Recently, an OHRQoL measure that included both parental and children reports was developed and validated for preschool and school-aged children, though the child self-report version was only used on children aged 8 years or older [24].

The relatively limited research in this area is a direct consequence of methodological and conceptual challenges for developing OHRQoL measures for very young children. Children’s perceptions about the impact of oral conditions on their life are based on their experience of oral diseases and are influenced by their immediate family environments and the wider social context including friends, schools and neighbourhoods [16, 25, 26]. Furthermore, their understanding of illness and health is age dependent due to social, language, emotional, and cognitive development [16, 26]; these developmental stages should be carefully considered within the appropriate social contexts when constructing subjective measures for young children [27]. Abstract thinking is not initiated before the age of 6 years and understanding of even basic health concepts may be problematic in younger children [28]. It is even later in childhood that they can evaluate their feelings and thoughts and compare them with those of their peers [29]. All these challenges have led to the lack of appropriate measures or the use of parental reports as proxies for young children’s perceptions of OHRQoL. On the other hand, there is evidence that 4–6 year-old children can report reliably on more concrete domains of their own general health and quality of life, including pain and dysfunction, though not on abstract domains such as emotional well-being [30]. However, using general health related quality of life measures would be useful but insufficient for measurement of quality of life aspects related specifically to oral health, as has also been shown in adult populations [31, 32].

Hence, it is important to develop and validate appropriate self-reported OHRQoL measures for very young children. This study aimed to develop and assess the reliability and validity of a new self-reported OHRQoL measure, the Scale of Oral Health Outcomes for 5-year-old children (SOHO-5), in the UK.

This cross-sectional study comprised of two phases. The first phase included consultation focus groups (CFGs) with parents of 5-year-olds. This informed the development of the questionnaires used in the second phase on 5-year-old children. Both phases were carried out in the Greater Glasgow and Clyde National Health Service (NHSGGC) Board area in the west of Scotland. The study was undertaken as part of the Childsmile program evaluation strategy. Childsmile is a program in Scotland designed to improve the children’s oral health and reduce inequalities in dental health and access to dental services [33]. Ethical approval was obtained from the West Glasgow Ethics Committee. All parents in the CFGs and parents of all children that participated in the second phase provided written consent.

To our knowledge, this is the first study to develop and test a self-reported OHRQoL measure among 5 year-old children. Previous studies on children of this age have mostly used composite OHRQoL measures based on parental reports [12, 37]. Despite being commonly used as proxies to assess impacts of chronic conditions on younger children, parents do not always accurately perceive their children’s quality of life, thereby parental proxy reports bring a different perception but do not substitute for children self-reports [38]. This study demonstrated that 5-year-old children are capable of providing their own perceptions of oral impacts and highlighted important challenges in this process. The SOHO-5 was developed in two phases; an initial phase routed on literature review and CFGs informed the generation and selection of the items, while a subsequent quantitative phase provided initial evidence for the reliability and validity of the new measure.

The initial phase ensured that the selected items were relevant and covered the physical, psychological and social aspects of oral impacts among very young children. Furthermore, the CFGs were pivotal in addressing the different conceptual and practical challenges for the development of SOHO-5. Challenges in developing OHRQoL measures for young children relate to their cognitive and emotional development [26, 39] as well as the social contexts they live in [25]. Potential problems include the appropriateness and relevance of the themes covered and content of the measures, relevance of the language, readability, comprehension and ability to respond to questions appropriately, as well as the (lack of) experience of oral conditions. Furthermore, respondent burden and limited attention span are key issues when considering quality of life questionnaires at this age.

Based on the CFG data, we attempted to address some of these challenges. We used thematic analyses of the CFGs to inform the content of the SOHO-5 and included questions on all the themes that emerged from these discussions with parents, thereby ensuring that all relevant aspects of oral health related quality of life for that age were explored. Furthermore, we developed a short interviewer-administered questionnaire so that children do not lose interest or focus [26]. The questions included simple words, sentences were kept short to aid comprehension, and language accommodated the local dialect and culture. In line with evidence that 6-year-olds tend to use only the extreme and middle ratings of a 5-point scale [40], we employed 3-point answering scales, a finding confirmed by all CFGs. Moreover, following evidence from pain studies [41], we used prompt cards (with faces) to facilitate rating of oral impacts and the study identified no respondent burden issues.

The issue of time recall was also a considerable challenge. In this young age, children’s notion of time and recall periods is a difficult concept to grasp [42‐44] and this was confirmed by parents in the CFG discussions. A very short time frame may have been more helpful in this respect [43], however this would not provide relevant information in terms of the impacts of oral conditions, as very few children would be expected to have experienced them, say, in the last 7 days. A more reasonable time frame would have been 3 months, but this was not feasible. Therefore, we avoided specific time recalls and opted to use the word “ever” and ask about lifetime experience of oral impacts. We do acknowledge, though, that this provides a broader picture of oral impacts and overestimates the prevalence at any given time period.

All children were able to understand and respond appropriately to the questionnaire and the vast majority did not have any difficulty comprehending the SOHO-5 questions. Actually, these questions required fewer repetitions compared to other questions, such as those about toothache experience and tooth eruption (9.1% and 9.8% respectively). The relatively more difficult concepts to comprehend were those tapping on psychosocial domains (avoiding smiling) while questions on functioning were hardly ever repeated. There may be a technical reason for this, as these were the only two questions with similar content. On a more conceptual note, this difference in comprehension between psychosocial and functional items is expected at this young age, as understanding and regulating emotions are more complex processes that require children becoming more aware of their internal world and are linked to their cognitive development later on in childhood [39, 45].

Compared to OHRQoL measures for older children, the SOHO-5 questions tapped on similar broader concepts, but employed much simpler wording while there were also some notable content differences such as the question on playing and the use of two different questions on smiling. The differentiation between avoiding smiling due to toothache and due to appearance was a direct and consistent suggestion of the CFGs and may be a reflection of the relative importance of smiling for very young children. In addition to its standard role as a meaningful indicator of aesthetics and perceptions, avoiding smiling may also present a way to react to toothache which is most often a “new” and therefore unfamiliar experience at this very young age.

In terms of the psychometric properties of SOHO-5, these initial results were very satisfactory. Internal consistency reliability was established through different statistics. All inter-item correlations were positive and none was very high, and all item-total correlation coefficients were above the recommended level of 0.2 [46]. Cronbach’s alpha was 0.74, above the arbitrary threshold of 0.70 [47, 48], and it was lower when any of the items was deleted. While the value of alpha tends to be higher on indices that have more questions, this study revealed very good internal consistency for the SOHO-5 despite the fact that it contains only a handful of questions.

For construct validity, we looked at the associations between SOHO-5 and different subjective oral health measures (current toothache, toothache lifetime experience, satisfaction with teeth, and presence of oral cavities), as well as with an aggregate oral health measure, based on both clinical (caries, pulp involvement) and subjective (current toothache, toothache experience) variables. All associations were significant and in the expected direction with higher SOHO-5 scores, indicating worse quality of life, for the groups reporting worse perceptions and having worse oral health. These consistent findings provide strong support for the validity of the new measure. Furthermore, clinically diagnosed active dental caries, pulp involvement and dental sepsis were associated with worse SOHO-5 scores. This indicates that dental caries may play an important role in shaping the overall perception of OHRQoL in very young children, in line with an earlier study [23]. More importantly, it practically demonstrates the ability of SOHO-5 to discriminate between different clinical groups.

Based on our results, oral diseases have considerable negative effects on children’s daily lives, as half of them reported an oral impact. This figure may actually underestimate the oral impacts on a population level, as our sample had lower prevalence of dental caries than the general child population in NHSGGC. The most prevalent impact related to difficulty eating, but other aspects of daily life, such as sleeping and smiling, were also affected. A previous study on a sample consisting also of young children has also shown that poor oral health affects smiling patterns [13]. It seems that oral diseases impact not only on the functional but also the psychosocial dimensions of life, even at a very young age.

Despite the positive initial results, the assessment of SOHO-5 should be an on-going process, by extending psychometric testing to properties not evaluated in this study, and assessing its applicability and performance in other populations. A future study, with repeated questionnaire administration, should provide information on test–retest reliability. This was not feasible in this study as we could not access the selected schools for a second visit because data collection took place very close to the end of the academic year and there was lack of available timetable spaces. Further studies should also complement the psychometric properties by looking into the responsiveness and interpretability of SOHO-5 scores. Despite being selected from the general population, our sample had considerable experience of dental caries; therefore future research should look into the discriminative properties of SOHO-5 in a healthier population sample. In the developmental phase, the study would have benefited if more parents participated in the CFGs; however, it is reassuring to note that discussions have been concordant in terms of the key themes, content, and technical characteristics of the new measure. Furthermore, involvement of children directly, rather than just their parents, in the qualitative work could also have provided further relevant information, though we acknowledge that conducting CFGs with very young children may also be open to criticism.

Developing a valid and reliable OHRQoL measure for very young children has important implications for research and practice. First, it can provide insight into the impacts of oral conditions on the daily life of young children. Dental caries is a chronic disease that can affect children from a very young age and it is important to measure its impacts on quality of life, as they may affect the psychological, social and educational development of the child [3, 5, 8, 12]. Future research should also explore the extent to which socioeconomic and clinical factors determine subjective measures of oral health and quality of life. In addition, OHRQoL measures could be used in clinical decision making to assess the effectiveness of dental treatment, thereby advancing the pediatric outcomes research agenda [49]. Finally, SOHO-5 could also potentially be a valuable outcome for evaluating oral health promotion programs and/or service initiatives for this age group [50].

This study has developed a new OHRQoL self-reported measure for 5-year-old children in the UK. The initial assessment of its reliability and validity showed very promising findings. Future studies should complement its psychometric testing and extend its application to patient samples as well as public health programs.