Orthodontics aims to improve oral health-related quality of life (OHRQoL). In this systematic review, we examined the evidence for changes in OHRQoL after orthodontic treatment for patients treated before they were 18 years old.
The participants were patients aged less than 18 years. The interventions were nonorthognathic and cleft orthodontic treatment. The comparisons were before and after orthodontic treatment, or nonorthodontic control. The outcomes were validated measures of OHRQoL. The study designs were randomized controlled trials, controlled clinical trials, prospective cohort studies, and cross-sectional or case-control studies. Multiple electronic databases were searched, with no language restrictions; authors were contacted, and reference lists screened. The Newcastle-Ottawa scale was used for quality assessments. Screening, data extraction, and quality assessments were performed by 2 investigators independently.
We found 1590 articles and included 13 studies (9 cohort, 3 cross sectional, and 1 case control), with 6 in the meta-analyses. All were judged of low or moderate quality. A moderate improvement in OHRQoL was observed before and after orthodontic treatment (n = 243 participants; standardized mean difference, −0.75; 95% CI, −1.15 to −0.36) particularly in the dimensions of emotional well-being (n = 213 participants; standardized mean difference, −0.61; 95% CI, −0.80 to −0.41) and social well-being (n = 213 participants; standardized mean difference, −0.62; 95% CI, −0.82 to −0.43).
Orthodontic treatment during childhood or adolescence leads to moderate improvements in the emotional and social well-being dimensions of OHRQoL, although the evidence is of low and moderate quality. More high quality, longitudinal, prospective studies are needed.
We assessed changes in OHRQoL of young people after orthodontic treatment.
We included 13 studies in the review and 6 in the meta-analysis.
There was moderate improvement in overall OHRQoL after orthodontic treatment.
Domains of emotional and social well-being showed the most significant improvements.
The impact of malocclusion on both patients and populations has been explored extensively. Many traditionally held beliefs regarding the dental health implications of malocclusion, such as an increase in caries, periodontal disease, or temporomandibular disorders, are now considered ambiguous and are largely unsupported by the literature. However, it is now recognized that the impact of malocclusion on health must be explored beyond the mere influence that it may have on dental health. The World Health Organisation describes health as a “state of complete physical, mental and social well-being and not merely the absence of disease or infirmity.” It is therefore unsurprising that over the past decade the use of patient-reported outcomes measures, including measures of oral health-related quality of life (OHRQoL), have been recognized as crucial in identifying the functional, emotional, and social impacts of malocclusion.
As an outcome measure, a fundamental objective of OHRQoL is to provide a subjective evaluation of oral-health status. As a reflection of this, a universally accepted definition of OHRQoL describes it as a measure that focuses on “the impact of oral diseases and disorders on everyday life that a patient or person values, that are of sufficient magnitude, in terms of frequency, severity or duration to affect their experience and perceptions of their life overall.”
Recent systematic reviews have found evidence to suggest that malocclusion impacts negatively on OHRQoL. When the individual dimensions of OHRQoL are explored, malocclusion has been found to have no significant impact on functional limitations and oral symptoms. The dimensions of emotional and social well-being have been found to be significantly influenced. Not only has the impact of malocclusion been explored, but recently published literature has also sought to establish the effect of wearing orthodontic appliances on OHRQoL. To date, studies have determined that such appliances have a negative impact, particularly on the oral symptoms and functional limitations dimensions. Perhaps these results are unsurprising; one would expect most forms of dental intervention to have a negative impact on OHRQoL during treatment. Moreover, it is logical to assume that it is the subsequent improvement in at least 1 dimension after treatment that drives people to seek and undergo such care.
To date, there is evidence to suggest that malocclusion and subsequent treatment with orthodontic appliances have a negative impact on OHRQoL. It is only appropriate now to question whether completion of orthodontic treatment to correct a malocclusion will lead to an improvement in this multidimensional concept. Identifying whether orthodontic treatment has such a benefit is crucial if we are to safeguard against interventions that have little value and to prevent the wastage of limited health care resources in countries where treatment is funded by the state. To date, this question has not been addressed in the context of a systematic review.
Our aim was to systematically review the current literature to identify changes in OHRQoL before and after orthodontic treatment in children and adolescents.
Material and methods
Protocol registration, conflict of interest, and funding
The protocol for this systematic review was registered on the National Institute of Health Research Database (registration number CRD42014014825; www.crd.york.ac.uk/PROSPERO ). The source of funding was a National Institute for Health Research Academic Clinical Fellowship for the first author.
The following selection criteria were applied for the review.
Participants. Patients aged 17 years or under at the start of their orthodontic treatment. Exclusions were patients with craniofacial syndromes and cleft lip or palate and those who had undergone previous orthodontic treatment (if it was possible to identify them).
Interventions. Any form of orthodontic treatment provided in primary, secondary, or tertiary care settings were included. This included orthodontic treatment that involved extractions, surgical exposure of unerupted teeth, and surgical removal of teeth. Studies involving orthognathic surgery were excluded.
Comparator. Studies had to include either assessment of OHRQoL before and after orthodontic treatment or a comparison group of subjects who had not undergone orthodontic treatment. This could include subjects who were not due to undergo orthodontic treatment or patients who were on the waiting list but had not yet started treatment.
Outcome measures. The main outcome measure was OHRQoL at any time after orthodontic treatment. The OHRQoL must have been determined using a validated measure such as the Child Perception Questionnaire. Secondary outcome measures included the dimensions of OHRQoL comprising, but not limited to, functional limitations, oral symptoms, emotional well-being, and social well-being.
Study design. Randomized and controlled clinical trials, prospective cohort studies, and cross-sectional or case-control studies, with data collection or follow-up periods after orthodontic treatment were to be included.
Information sources, search strategy, and study selection
The following electronic databases were searched: MEDLINE via OVID (1946 to March week 3 in 2016) (see Appendix A for the search terms in the search strategy), the Cochrane Oral Health Group’s Trials Register (March 2016), the Cochrane Central Register of Controlled Trials (issue 3 of 12, March 2016), EMBASE (1974 to March 2016), PsychINFO (1806 to present), PubMed (inception to March 2016), Scopus (all years to March 25, 2016), and Web of Science (1900 to 2016). No language restrictions were applied. No search of the grey literature was undertaken.
Any systematic and narrative reviews on the topic were assessed, and any studies referenced therein that met the inclusion criteria for this systematic review were included; however, the reviews themselves were not included. The reference lists of eligible studies were also screened for additional relevant research. In addition, authors who are known to have an interest in this field of research were contacted to identify unpublished or ongoing trials.
Assessments of studies for inclusion in the review were performed independently and in duplicate. One author (H.J.) assessed all studies, and the other 2 authors (M.V. and P.E.B.) each assessed half of the retrieved studies. The investigators were not blinded to the authors or the results of the research, and any disagreements were resolved by discussion with a third author who was not involved with the original screening of that particular study (M.V. or P.E.B. as appropriate).
Data items and collection
Data were extracted independently and in duplicate in a similar method to that used for assessment of studies for inclusion. Prepiloted data extraction forms were used; where available, the following information was recorded: (1) year of publication, country, and study setting; (2) study design; (3) participants: sample size, ages before and after orthodontic treatment, sex, severity of malocclusion treated; (4) intervention: type of orthodontic treatment provided; (5) type of control or comparison group; and (6) outcomes: OHRQoL measure (including individual dimensions, where available) and the OHRQoL informant.
Authors were contacted to clarify data if required for further information regarding the OHRQoL outcome measure or additional data on the OHRQoL outcome.
Quality assessment and risk of bias in individual studies
The quality of the eligible trials was assessed independently and in duplicate, and disagreements were resolved, using the same methods outlined for study selection and data extraction. If appropriate randomized controlled trials were identified for the review, we intended to use the Cochrane Collaboration’s risk of bias tool to assess the risk and quality of these studies. An appropriately modified version of the Newcastle-Ottawa scale was used to assess the quality of nonrandomized studies. This included the scale designed for cohort and case-control studies, and an adapted version of the scale suitable for the assessment of cross-sectional studies. This tool evaluates studies based on 8 domains, which are divided into 3 broad criteria: patient selection, comparability of study groups, and outcome assessment. A star system was used, whereby high-quality studies at low risk of bias could receive a maximum of 9 stars. Studies achieving 8, 7, or 6 stars were considered to have moderate quality, and a rating of 5 stars or fewer signified low quality.
Summary measures and approach to statistical analysis
Studies were grouped based on their study designs (cohort, cross sectional, or case control). Clinical heterogeneity of the included studies was determined by assessing the study protocol and, in particular, the type of comparator used. Statistical heterogeneity was assessed using the I 2 test, and a threshold of less than 50% was assumed to demonstrate sufficient homogeneity. All studies that reported OHRQoL measures as scores (continuous outcomes) were combined to obtain the pooled mean values with 95% confidence intervals (95% CI), using the inverse variance method and the random effects model. Many instruments used to assess OHRQoL, including the Child Perception Questionnaire 11-14, are designed so that a higher score indicates a poorer level of OHRQoL. Based on this, a reduction in scores, demonstrated by a negative difference, was interpreted as an improvement in OHRQoL.
When studies used different scales or instruments for the assessment of OHRQoL, the standardized mean difference for each study was used in the meta-analysis. Standardized mean differences were interpreted using thresholds described by Cohen, where 0.2 represents a small effect, 0.5 a moderate effect, and 0.8 a large effect. All analyses were carried out using RevMan (version 5.3.5; The Nordic Cochrane Centre, Copenhagen, Denmark), and a significance level of 5% was adopted for all analyses.
Risk of bias across studies
Publication bias could only be assessed when at least 10 studies were included in the meta-analysis. The 2 statistical measures that would be used for this purpose were the rank correlation of the Begg test and the Egger test.
Figure 1 is a flow diagram showing the retrieval, screening, and selection of articles for the review. The search returned 1590 studies after removal of duplicates. Contact with authors for the retrieval of possible unpublished studies returned 1 more study for the review. All titles and abstracts were reviewed, and 32 potentially relevant articles were retrieved in full. After a detailed assessment, which included contact with relevant authors for further clarification regarding the validation of 1 OHRQoL outcome measure (OASIS), as well as the translation of 1 article from Dutch to English, 18 studies were excluded (see Appendix B for screening details of full-text articles), and 14 studies remained. Of these 14 reported trials, 2 were duplicate reports of the same study. For this systematic review, these 2 were included, assessed, and analyzed as 1 study. This resulted in a total of 14 articles (13 studies) appropriate for inclusion. The Table summarizes the characteristics of the studies. No randomized controlled trials met the eligibility criteria. The study designs for the 13 studies were as follows: 8 studies were cohorts 1 study was a cohort but also included cross-sectional data, 3 studies were cross sectional, and 1 study was a case control. One cohort study included 3 distinct groups of patients: 1 group with cleft lip and palate, another group receiving orthognathic surgical treatment, and a third group receiving orthodontic treatment alone. These subjects were aged 17 years or under at the start of treatment. Since the third group provided data that met the inclusion criteria for this review, this study was deemed appropriate for inclusion.
|Study||Study design||Country||Setting (clinic/population based)||Participants||OHRQoL measure||OHRQoL informant||Type of orthodontic treatment||Control/comparison group|
|Agou et al, 2008||Cohort||Canada||University teaching hospital (clinic based)||Recruited (n = unreported); Follow-up at first recall appointment (45 children, 26 parents). Of those followed-up: 27 F, 18 M. Overall mean age, 12.6 years (SD, 1.4). Mean DAI score at baseline, 36.6; mean PAR score at baseline, 30.4.||CPQ 11-14||Self-reported and parent||Not stated||No control group|
|Agou et al, 2011||Cohort||Canada||University teaching hospital (clinic based)||Recruited (n = 199); follow-up at first retention check appointment (n = 118). Of those followed-up: 59 F, 59 M. Overall mean age, 12.9 years (SD, 0.98) initially. Based on DAI, 44.2% handicapping, 25.7% severe, 23.9% definite, and 6.2% had minor malocclusions.||CPQ 11-14||Self-reported||Routinely prescribed fixed appliance therapy||Control subjects were consecutively recruited from same clinics during their first orthodontic screening visit|
|Antoun et al, 2015||Cohort||New Zealand||Orthodontic unit of Christchurch Hospital (clinic)||Recruited (not reported); at follow-up within 3 months of treatment completion. Standard orthodontic treatment group (n = 30); CL/P group (n = 24); surgery group (n = 29), 37 F, 46 M. Pretreatment mean ages of standard group, 14.5 years (SD, 1.9); CL/P, 12.6 years (SD, 2.8); surgery group, 19.0 years (SD, 4.3). Pretreatment DAI of standard group, 45.5 (SD, 9.0), CL/P group, 45.4 (SD, 13.4); surgery group, 56.6 (SD, 12.8).||OHIP-14||Self-reported||Single or double arched fixed appliances||2 groups of comparison: CL/P group and orthoganthic Surgery group|
|Arrow et al, 2011||Cohort (follow-up) with cross-sectional elements||Australia||School dental service in South Australia (clinic based for treated group; population based for control group)||Orthodontically treated (n = 155); nonorthodontically treated (n = 286). Age of participants in both groups, approximately 30 years. Orthodontically treated group vs Nonorthodontically treated group, DAI ≤ 25, no need (n = 53; n = 144); DAI 26-30, elective (n = 32; n = 81); DAI 31-35, desirable (n = 27; n = 35); DAI ≥ 36, mandatory (n = 43; n = 26).||OHIP-14||Self-reported||Fixed orthodontic treatment||Random sample of adults the same age as the study cohort drawn from Adelaide’s electoral register|
|Benson et al, 2014||Cohort||United Kingdom||Seven publicly funded schools (population based)||Recruited (n = 374); follow-up at 3 years (n = 258), 252 F, 122 M. At baseline, all aged 11-12 years. IOTN DHC, no need (n = 96); borderline need, (n = 138); definite need (n = 139).||CPQ 11-14 ISF-16||Self-reported||Not stated||Those who had no history of orthodontic treatment|
|Bernabé et al, 2008||Case control||Brazil||Secondary schools in Bauru (population based)||Cases n = (279); controls (n = 558), 485 F, 352 M. Aged 15 years, n = 552; aged 16 years, n = 285. Orthodontic treatment need of cases vs controls: no need (n = 116; n = 360), moderate need (n = 58; n = 122), definite need (n = 105; n = 76).||OIDP||Self-reported||Any history of orthodontic treatment irrespective of the type of appliance used.||Adolescents who had never received, or had not completed orthodontic treatment|
|Chen et al, 2010||Cohort||China||University teaching hospital (clinic based)||Recruited (n = 250); follow-up at posttreatment (n = 222), 148 F, 74 M. Overall mean age, 15.7 years.||OHIP-14||Self-reported||Fixed appliance treatment||No control group|
|de Oliveira and Sheiham, 2003 and 2004||Cross sectional||Brazil||Public and private schools (population based)||Orthodontically treated (n = 258); having orthodontic treatment (n = 357); untreated (n = 1060) 951 F, 724. Participants aged 15 years (n = 1110), aged 16 years (n = 565). IOTN DHC, no/slight need (n = 1031), moderate need (n = 351), need (n = 293)||OIDP and OHIP-14||Self-reported||Not stated||2 groups: adolescents undergoing orthodontic treatment and adolescents not having orthodontic treatment|
|Feu et al, 2013||Cohort||Brazil||Dental school for treatment group (TG) and waiting list group (WG) (clinic based). Public school for school group (SG) (population based)||Recruited TG (n = 92); WG (n = 124); SG (n = 102); follow-up at 2 years in TG (n = 87); WG (n = 101); SG (n = 96), 169 F, 149 M. Mean age in TG, 13.4 years (SD, 1.1); WG, 13.7 years (SD, 1.1); SG, 13.7 years (SD, 1.2). IOTN DHC mean TG, 3.5 (SD, 1.1); WG, 3.4 (1.2); SG 3.0, (0.9).||OHIP-14||Self-reported||Not stated||2 groups of comparison: orthodontic WG and SG included children from a public school who had never undergone or sought orthodontic treatment|
|Healey et al, 2016||Cohort||New Zealand||Nineteen Private specialist orthodontic practices (Clinic)||Recruited (n = 174); follow-up at end of orthodontic treatment (n = 152); follow-up at end of study period approximately 21 months after end of treatment (n = 104). 112 F, 62 M. Mean age at baseline, 13.5 years (SD, 1.3). Mean DAI at baseline, 35.8 (8.4).||CPQ 11-14||Self-reported||Maxillary and mandibular fixed orthodontic treatment||No control/comparison group|
|Olivieri et al, 2013||Cross sectional||Italy||Adolescents attending the last year of middle school (population based)||Undergone orthodontic treatment (n = 115); no orthodontic treatment (n = 444), 269 F, 292 M. All participants 14 years of age.||CPQ 11-14||Self-reported||Not stated||Those who had not already undergone orthodontic treatment|
|Seehra et al, 2013||Cohort (follow-up)||United Kingdom||Dental hospitals, Kent and Canterbury Hospitals, William Harvey Hospital, Guy’s campus of King’s College London Dental Institute (clinic based)||Recruited at follow-up (n = 27) 14 F, 13 M. Mean age of sample, 14.6 years (SD, 1.5). Pretreatment IOTN, grade 5 (n = 16), grade 4 (n = 9), grade 3 (n = 1), grade 2 (n = 1).||CPQ 11-14||Self-reported alone or with assistance from caregiver||Fixed appliances either alone or combined with functional appliances||No control/comparison group|
|Taylor et al, 2009||Cross sectional||United States||University of Washington School of Dentistry and Odessa Brown Children’s Clinic (clinic based)||Recruited precomprehensive group (PC) (n = 93); postinterceptive group (PI) (n = 44); comparison group (C) (n = 156): PC, 45 F, 48 M; PI, 21 F, 23 M; C, 76 F, 80 M. Mean age in PC, 12.5 (±1.1); PI, 12.5 years (±1.1); C, 12 years 9 months (±1.1). Pre-ICON total in PC, 69.0 (±21.5); PI, 79.0 (±20.1).||CPQ 11-14||Self-reported||Interceptive orthodontic treatment||2 groups of pediatric dental patients: precomprehensive group with no orthodontic treatment and comparison group with no plans for treatment|