The selection of appropriate outcomes that matter to both patients and operators is increasingly appreciated, with core outcome sets in clinical trials gaining in popularity. The first step in core outcome set development is the generation of a list of possible important outcomes based on a scoping literature review. Moreover, outcome heterogeneity is known to detract from the findings of systematic reviews and meta-analyses. The aim of this study was to identify the range of outcome domains and specific outcome measures in contemporary orthodontic research.
Multiple electronic databases were searched from December 31, 2012, to December 31, 2016, to identify clinical trials of orthodontic interventions, with no language restrictions. Abstracts, eligible full texts, and reference lists were screened, and all reported primary and nonprimary outcomes and methods of measurement were recorded.
The search identified 1267 abstracts, of which 189 full-text articles were retrieved, and 164 studies were included in the analysis. A total of 54 outcomes were identified and categorized into 14 outcome domains. The most frequently measured outcomes were patient-reported pain, periodontal health, tooth angulation/inclination changes, and treatment duration, followed by rate of tooth movement and skeletal changes. Outcomes that followed the overall course of treatment were assessed in only 14 studies.
Patient perspectives are increasingly being accounted for in orthodontic trials; however, there is little consistency in outcome selection among them. The identified list of outcomes will be used to inform a ranking exercise with service users and providers to establish an agreed core outcome set for future orthodontic clinical trials.
The aim was to identify the range of outcomes and measurement tools used in contemporary orthodontic research.
Fifty four outcomes were identified and categorized into 14 outcome domains.
Pain; periodontal health; and tooth angulation/inclination were the most frequently measured outcomes.
Outcomes that followed the overall course of treatment were assessed in only 14 studies.
The identified list of outcomes will help to establish a core outcome set for future clinical trials.
Outcomes from clinical trials may be used to assess the relative merits and demerits of an intervention. These outcomes are measured by using tools to determine changes in the health state of a patient resulting from a health care intervention. These may be applied to a variety of contexts, from measuring outcomes relating to physiological change, disease status and delivery of care, to symptoms or self-perceptions. Outcomes and outcome measures should be clearly defined and relevant to key stakeholders, including consumers and providers of care, if they are to have meaning and relevance. When at least 1 outcome is used to reflect changes within a broader concept, which may not be directly measurable, the latter is known as an outcome domain. Different outcomes can thus be grouped together under the same umbrella outcome domain. For instance, in an orthodontic study evaluating the duration of treatment or the number of different archwires used to reach a desired state, both outcomes might be categorized to the same overall outcome domain of cost-effectiveness or health care utilization.
There is a wealth of evidence that outcome heterogeneity is pervasive across health care research. In orthodontic research, frequent conclusions of systematic reviews are lack of quality evidence; inability to synthesize disparate studies; and need for further research. This inability to perform meaningful syntheses is one of many issues relating to the use of inconsistent outcomes (termed “outcome heterogeneity”) in clinical research studies. This outcome heterogeneity was, for example, shown in a Cochrane review evaluating orthodontic interventions to distalize maxillary first molars, where differences in outcomes and incomplete reporting of data precluded meta-analysis of the 4 included studies assessing the effectiveness of a distalizing appliance compared to an untreated control.
Similar problems may be encountered in studies evaluating orthodontic treatment outcomes and occlusal stability. Numerous indices exist, each measuring slightly different outcomes, thus making comparisons between trials difficult. For example, the Index of Complexity Outcome and Need may be used to assess final occlusion, while the Peer Assessment Rating, American Board of Orthodontics system, or even a simple irregularity index, that assesses alignment of the anterior mandibular segment, may be applied in the evaluation of treatment outcome and stability. The correlation between such indices is varied, and the heterogeneity in measured outcomes renders comparisons problematic. This inconsistency among orthodontic studies considering the effectiveness of interventions may render evidence synthesis and meta-analysis impossible and, consequently, hinder interpretation of their results. This was evident in an analysis of 157 orthodontic systematic reviews in 5 leading orthodontics journals and the Cochrane Database, with meta-analysis present in only 43 of the reviews (27.4%) and a median of only 4 trials per meta-analysis. Similarly, in a recently published systematic review assessing oral health-related quality of life after orthodontic treatment, only 3 studies of a potential 13 were included in the meta-analysis, because the oral health-related quality of life outcome measure used in these studies was the Child Perception Questionnaire 11-14, whereas the remaining studies used alternatives including the Oral Health Impact Profile-14 or the Oral Impacts on Daily Performance instrument.
Therefore, to improve data synthesis and reduce outcome heterogeneity and reporting bias, agreement is needed concerning which outcomes to collect and how to measure them. This will be achieved through the establishment of a core outcome set (COS) that will need to be measured as a minimum in all clinical trials for a specific condition. COS development is now established and supported through the Core Outcome Measures in Effectiveness Trials (COMET) initiative with successful development of outcome sets in childhood asthma and otitis media, for example.
An initial stage of COS development is to perform a scoping systematic review to ascertain the nature of outcomes in a specific research area. The identified list of outcomes is typically complemented by data obtained from patients and other stakeholders before being refined in a subsequent consensus process, leading to the development of an orthodontic COS. The aims of this scoping review were to update a previous review in relation to reported orthodontic outcomes and also to identify both outcome domains and specific measures used in contemporary orthodontic research.
Material and methods
The protocol for the overall study of COS development has been registered on the COMET website and published. A scoping review of recently published orthodontic clinical trials was carried out, and a previous review was updated.
The following inclusion criteria were used in this scoping review.
Study design: randomized and controlled clinical trials. All parallel-group trials, including those of crossover or cluster design, were considered eligible for inclusion.
Participants: children and young people undergoing orthodontic treatment, with no age restrictions.
Interventions: any orthodontic treatment intervention was to be included.
Control: any comparison group was to be included with no restrictions on control groups.
Outcomes: all reported outcomes (primary and secondary) were identified with separate demarcations of primary and secondary outcomes and related measurement tools.
Exclusion: retrospective studies and laboratory-only studies were excluded. Studies involving solely adults or patients undergoing orthognathic surgery and patients with cleft lip or palate, obstructive sleep apnea, syndromic conditions, or medical history complications were excluded.
The following electronic databases were searched: MEDLINE via Ovid, EMBASE via Ovid, the Cumulative Index to Nursing and Allied Health Literature (CINAHL) via EBSCO, psycINFO via EBSCO and the Cochrane Central Register of Controlled Trials (CENTRAL), via the Cochrane Library ( Appendix 1 ) to identify relevant studies from December 31, 2012, to December 31, 2016. No language restrictions were applied, and attempts were made to translate any non-English studies identified. In addition, the reference lists and trials identified in recently published Cochrane systematic reviews were cross-checked to ensure that no relevant studies were omitted.
The abstracts of all studies identified were assessed by 1 reviewer (A.T.) with a range of expertise including orthodontics, patient-reported outcome measures, and trial design. Full-text reports of studies that met the inclusion criteria and for which there was insufficient information in the title or abstract to make a clear decision were obtained. A second reviewer (P.S.F.) helped to resolve any uncertainty regarding final inclusion until consensus was reached.
All primary and any secondary outcomes were identified and recorded together with the specific outcome measures or tools used to measure each outcome based on the data presented. When delineation of primary or secondary outcomes was unclear, the primary outcome was inferred from the aim of the study, the sample size calculation, or the first reported outcome in the results section. Any subsequent outcomes reported in the results were also identified and recorded as secondary outcomes. When uncertainty persisted in relation to primary or secondary outcomes, all were recorded as primary outcomes, and a note was made in the prepiloted data extraction sheet.
The specific stage of treatment during which the trial was conducted was also recorded. Finally, all identified outcomes were grouped under broader outcome domains. The outcome domains were developed iteratively after inspection of the results and refined by 2 reviewers (A.T. and P.S.F.) until consensus was reached.
One thousand two hundred sixty-seven studies were identified through electronic searching and cross-referencing of sources. After removal of duplicate records, 675 abstracts were screened, of which 189 full texts were assessed for eligibility. Of those, 164 met the inclusion criteria and were included in the review ( Fig 1 ). Publications derived from the same trial but involving different outcomes or follow-up periods were considered as separate studies. The characteristics of all included trials and the outcomes they measured are shown in Appendix 2 .
A significant proportion (n = 59; 36%) of the trials related to the initial stages of treatment: eg, investigating the rate of initial orthodontic alignment or pain experience after separator or fixed appliance placement. Twenty-four studies (15%) investigated the effects of different brackets or archwires during initial stages and midstages of treatment (typically alignment and leveling occurring in the first 6-9 months of treatment or until passive engagement of working archwires), with just 14 studies (8.5%) encompassing active treatment in its entirety ( Table I ). Treatment stage was unclear in 9 studies (5%).
|Treatment stage||Number of studies (n = 164)|
|Initial (initial days or weeks of treatment)||59|
|Initial mid/mid (alignment and leveling up to placement of working archwires)||24|
|Final (space closure and finishing)||17|
Overall, 54 outcomes were identified from the 164 included trials. These were subsequently grouped into relevant outcome domains ( Table II ) with the frequency of their use as primary or secondary outcomes also calculated ( Table III ). The most frequently reported primary outcome was pain (n = 26; 16%), followed by rate of tooth movement (n = 19; 12%) and skeletal relationship (n = 17; 10%). Treatment duration was the most frequently reported secondary outcome (n = 18; 11%), followed by tooth angulation and inclination changes (n = 12; 7%) and periodontal condition (n = 9; 5%). When both primary and secondary outcomes were combined, pain was still the most frequently reported outcome (n = 30; 18%), followed by periodontal health (n = 25; 15%) and tooth angulation/inclination (n = 23; 14%; Fig 2 ).
|Harms (n = 60)||Function (n = 2)||Hard tissues/skeletal (n = 25)||Soft tissues (n = 14)||Occlusal/Alignment change (n = 38)||Dental development and morphology (n = 5)||Periodontal (n = 25)||Microbiologic/Physiological (n = 18)||Knowledge and satisfaction (n = 5)||Quality of life (n = 4)||Compliance (n = 13)||Appliance integrity (n = 18)||Efficiency/Cost effectiveness (n = 39)||Other (n = 5)|
|Pain (n = 30)||Speech (n = 1)||Skeletal relationship (n = 20)||Soft tissue profile changes (n = 13)||Tooth angulation/inclination (n = 23)||Eruptive changes (n = 6): Eruption/non-eruption (n = 3); Improvement in position (n = 3)||Periodontal health (n = 25)||Microbial composition/count (n = 13)||Satisfaction (n = 2)||Anxiety (n = 3)||Appliance breakage (n = 2)||Fixed appliance attachment and bond failure/strength (n = 9)||Treatment duration (n = 22): overall treatment time (n = 4); functional/interceptive (n = 8); chairside duration (n = 4); initial alignment (n = 3); space closure (n = 1)||Personality traits (n = 3)|
|Enamel demineralization (n = 13)||Mandibular excursions (n = 1)||Bone levels (n = 2)||Gingival margin esthetics (n = 1)||Arch dimensions/changes (n = 13)||Enamel reduction (n = 1)||Inflammatory markers/response (n = 4)||Information comprehension/recall (n = 2)||Self-esteem (n = 1)||Oral hygiene including tootbrushing duration and fluoride consumption (n = 6)||Fracture of functional appliance (n = 2)||Rate of tooth movement (n = 20): space closure (n = 13); initial alignment (n = 6); overall treatment (n = 1)||Orthognathic treatment need (n = 1)|
|Patient reported adverse effects (n = 5)||Bone density (n = 1)||Alignment relapse (n = 2)||Plaque and/or salivary pH (n = 3)||Acceptability (of appliance) (n = 1)||Malocclusion impact (n = 1)||Duration of removable appliance wear/day (n = 5)||Miniscrew force/stability (n = 2)||Treatment success (n = 3)||Airway volume (n = 1)|
|Root resorption (n = 5)||Condylar changes (n = 1)||Occlusal outcome (n = 3)||Attendance (n = 1)||Archwire coating (n = 1)||Direct and indirect costs of materials/appliances/societal (n = 2)|
|Enamel roughness post debond (n = 1)||Suture anatomy (n = 1)||Archwire strength (n = 1)|
|Caries (n = 3)||Adhesive retention (n = 4)|
|Halitosis and tongue coating (n = 2).|
|Nickel/chromium levels in saliva (n = 2).|
|Mucosal ulceration (n = 1).|
|Gingival irritation (n = 1).|
|Root contact (n = 1).|
|TAD fracture (n = 1).|
|Other adverse periodontal effects (n = 1).|
|Primary outcome (number of studies)||Secondary outcome (number of studies)|
|Appliance (miniscrew) stability||2||0|
|∗ Arch-form changes||12||1|
|∗ Bond failure||8||2|
|Direct (appliance/material) and indirect (societal) costs||1||1|
|∗ Enamel demineralisation||9||4|
|Gingival margin aesthetics||1||0|
|∗ Microbial composition/count||6||7|
|Mobility/failure (of TAD)||0||1|
|Oral hygiene compliance||4||2|
|Orthognathic treatment need||0||1|
|Other periodontal adverse effects||0||1|
|Patient-reported adverse effects||1||4|
|∗ Periodontal health/condition||16||9|
|Salivary metal ions||2||0|
|∗ Skeletal relationship||17||3|
|∗ Soft tissue changes||10||3|
|∗ Tooth angulation/inclination||11||12|
|∗ Tooth movement rate||19||1|
|∗ Treatment (stage) duration||4||18|