Development of a core outcome set for orthodontic trials using a mixed-methods approach: protocol for a multicentre study

Orthodontic treatment is carried out, using braces, to correct abnormalities related to the teeth and/or jaw relationships, which are often referred to as a ‘malocclusion’. The most recent Children’s Dental Health Survey in the UK revealed that 35% of 12-year-olds and 28% of 15-year-olds were embarrassed to smile or laugh because of their teeth with 37% and 20% of 12- and 15-year-olds, respectively, having an unmet need for orthodontic treatment [1]. Malocclusion has been linked to bullying and teasing in 13% of 10- to 14-year-olds [2], as well as impaired oral health-related quality of life and social and emotional wellbeing [3]. If not corrected in adolescence, malocclusion may remain a long-term condition that could continue to adversely affect quality of life [4]. By improving appearance and function, orthodontic treatment improves the oral health-related quality of life of young people, particularly in the domains of emotional and social wellbeing [5] and might be considered integral for optimal oral health, where oral health is defined as ‘a standard of health of the oral and related tissues which enables an individual to eat, speak and socialise without active disease, discomfort or embarrassment and which contributes to general wellbeing’ [6].

Orthodontics is currently provided to approximately one third of young people in England and Wales at a cost to the National Health Service (NHS) of £273 million in 2013–2014, accounting for 11.2% of the dental budget [7]. Young people are considered eligible for NHS orthodontic treatment based on an assessment using the Index of Orthodontic Treatment Need (IOTN) [8]. Importantly, the IOTN does not take account of either patient- or parent-important outcomes or the impact of dentofacial anomalies on young peoples’ daily lives.

While the number of clinical trials in biomedical and dental research is increasing, it has been suggested that the reported outcome measures may resonate more with clinicians than with our patients [9]. This is important because the usefulness of a study relies on the outcomes it explores. Concentrating on measures that are solely important to clinicians may result in pertinent issues to patients being overlooked. This situation has been exposed in the field of respiratory medicine with Sinha et al. [10] bemoaning ‘wasted resources or misleading information that overestimates, underestimates or completely misses the potential benefits of an intervention’ secondary to the selection of inappropriate outcomes.

The issues stemming from the use and incomplete reporting of, often inappropriate, outcomes are, therefore, multifaceted. Specifically, the time and resources invested in the research may be wasted with approximately 40 to 89% of published trials being non-replicable due to inadequate descriptions of the interventions and outcomes [11]. Furthermore, accounting for patient values is central to evidence-based medicine, and the integration of these values with clinical research evidence is necessary to enable informed decision-making [12]. Valuable information regarding the effectiveness of an intervention may be overlooked by neglecting to consider outcome measures important to patients. An additional problem is the difficulty in combining the results from studies within systematic reviews. This outcome heterogeneity was exemplified in the Cochrane review of treatment of excessively prominent upper teeth, where the included studies all used allied, but distinct radiographic analyses to answer the same question, precluding meta-analysis [13].

Surprisingly, given the high prevalence of malocclusion and the extensive provision of orthodontic treatment, there is limited information derived from clinical trials to inform decision-making from a patient viewpoint. Specifically, when randomised controlled trials (RCTs) in orthodontics published between 2008 and 2012 were analysed, 42 general outcome measures were identified, with most (63%) focussing on technical or morphological changes that result from treatment, with little reference to values that can be discerned or appreciated by patients [9]. While the previous measures are both necessary and important in assessing the effectiveness of care from a technical perspective, other outcomes of comparable value, such as cost-benefit analyses, adverse effects of treatment, patient perceptions, and the impact of, and compliance with, treatment, have remained largely unexplored. The latter are relevant to patients and provide essential information when operators and patients make shared decisions about care. The findings are comparable to those of Sinha et al. [14], who conducted a systematic review of 159 RCTs concerning children with asthma. They found that short-term disease activity was the most frequently measured outcome domain, with quality of life, functional status and physical consequence of disease being underrepresented.

Recently, work has commenced on the development of an agreed standardized set of outcome measures for health care described as a core outcome set (COS). The Core Outcome Measurement in Effectiveness Trials (COMET) initiative is an international network experienced in the development and promotion of these universally agreed outcome sets across the biomedical field [15]. It is suggested that ‘these outcomes should be measured as a minimum in trials assessing effectiveness of interventions, and would help eliminate issues relating to outcome heterogeneity and reporting bias, while ensuring that wide-ranging perspectives are measured, thus enhancing the value of RCTs and systematic reviews’ [16]. A number of consensus approaches have been used to develop COS, including semistructured and unstructured group discussions, the Delphi approach, consensus development conferences, surveys and nominal group technique [17, 18]. While inclusion of key stakeholders including researchers, clinicians, patients, public, policy-makers and public health professionals is key to developing a meaningful consensus, a recent review has revealed that just 16% of these studies reported public involvement [17]. A search of the COMET database and current literature indicates that COS development has not been undertaken within orthodontics.

In light of the significant cost of orthodontics to the NHS, allied to the lack of agreed outcome measures identified within our previous systematic review [9], the need to develop COS for orthodontic treatment of children and young people is clear. The aim of this study is, therefore, to develop a COS for use in clinical trials involving orthodontic treatment of children and young people. The results will ultimately benefit decision-making by all stakeholders.

The research methodology follows advice set out by COMET on COS development [15] and is similar to that used in the project that developed a COS for the treatment of otitis media with effusion for children with cleft palate [18]. The main research stages are shown in Table 1.

The research process will follow a sequential exploratory design, whereby patient views will be explored with the intent of using this information to further build a quantitative survey instrument [29]. This mixed-methods research design, combining quantitative and qualitative approaches, was selected in order to provide a better understanding of patient experiences and expectations of orthodontic treatment, thus ensuring that the COS and future research outcomes are relevant to the former.

In theory, a simpler method, such as a questionnaire, could be used to gather information from patients, as was the case in the COS development for childhood asthma, which had the advantage of minimising the burden on participants and enabling the involvement of a larger participant sample [30]. However, for the present research, it was felt that it would be best if unprompted patient views and attitudes concerning orthodontic outcomes were obtained, rather than providing quantified answers to preconceived ideas as would be the case with a clinician-derived questionnaire [31]. Additionally, questionnaires do not offer opportunities to clarify ambiguous data and equally there is no means of knowing if participants fully understood or even if they misunderstood the questions posed. Finally, since this COS is directed at children and young people, it was also felt that their opinions would require in-depth analysis in order to be able to subsequently convert these opinions into outcomes. Consequently, it was felt that this research question would best be addressed using an integration of quantitative and qualitative data (i.e. mixed methods). A similar approach was also used in determining key health outcomes and development of a COS for children and young people with neurodisability [32].

At present a diverse range of outcomes is used in orthodontic research, with little or no consistency between studies. This research will lead to a COS for use in future clinical research of orthodontic treatment interventions. It will be challenging to identify the most important outcomes to measure in view of the breadth of this area. Patient involvement is, therefore, central to this study to ensure that patient views are represented, and multiple stakeholder groups will be involved to ensure that the COS is acceptable and adopted in future trials. Development of a COS will establish which outcomes to measure in clinical research, but not how to measure them. It would, therefore, be important for future research to be directed towards developing and validating tools to measure the outcomes included in the final COS, if none already exist.

The scoping review has been completed and we are currently recruiting patients for the qualitative interviews and focus groups.