Abstract
Objectives
The current study aimed to evaluate the effectiveness of school-based dental screening versus no screening on improving oral health in children aged 3–18 years by a systematic review and meta-analysis of randomised controlled trials.
Sources and study selection
Three sets of independent reviewers searched MEDLINE, EMBASE, Web of Science and other sources through April 2016 to identify published and nonpublished studies without language restrictions and extracted data.
Data
Primary outcomes included prevalence and mean number of teeth with caries, incidence of dental attendance and harms of screening. Cochrane’s criteria for risk of bias assessment were used.
Results
A total of five cluster RCTs (of unclear or high risk of bias), including 28,442 children, were meta-analysed. For an intracluster correlation coefficient of 0.030, there was no statistically significant difference in dental attendance between children who received dental screening and those who did not receive dental screening (RR 1.11, 95% 0.97, 1.27). The Chi-square test for heterogeneity and the Higgin’s I 2 value indicated a substantial heterogeneity. Only one study reported the prevalence and mean number of deciduous and permanent teeth with dental caries and found no significant differences between the screening and no screening groups.
Conclusions
There is currently no evidence to support or refute the clinical benefits or harms of dental screening. Routine dental screening may not increase the dental attendance of school children, but there is a lot of uncertainty in this finding because of the quality of evidence.
Clinical significance
Evidence from the reviewed trials suggests no clinical benefit from school-based screening in improving children’s oral health. However, there is a lot of uncertainty in this finding because of the quality of evidence. There is a need to conduct a well-designed trial with an intensive follow-up arm and cost-effectiveness analysis.
Systematic review registration number
CRD42016038828 (PROSPERO database).
1
Introduction
Dental caries pose a major public health challenge in most countries in the world . In the Global Burden of Disease 2010 study, untreated caries in permanent teeth was found the most prevalent condition worldwide, affecting nearly 2.4 billion people, including children aged 5 years or older and adults . In the same study, untreated caries in deciduous teeth was the 10th most prevalent condition worldwide, affecting 621 million children. One of the three peaks in caries prevalence is at age 6 years . Furthermore, despite the overall decrease in the prevalence of untreated caries in industrialised countries, inequalities persist with the disadvantaged and vulnerable children bearing the greatest share of the untreated caries burden . In addition, untreated carious lesions may cause severe pain and mouth infection , which affect children’s school attendance and performance . Therefore, detecting such lesions, particularly at early stages, and providing the appropriate preventive and operative interventions are of paramount importance. Detecting and treating other oral diseases and conditions, such as pain, infection (oral sepsis), trauma, hard or soft tissues pathology, gross dental plaque and/or calculus, periodontal diseases, and malocclusion conditions at early stages have been considered important due to their impact on child’s wellbeing and quality of life .
School-based dental screening for oral health has been a popular and enduring public health intervention in many countries throughout the world . The World Health Organization has endorsed it stating that “Screening of teeth and mouth enables early detection, and timely interventions towards oral diseases and conditions, leading to substantial cost savings. It plays an important role in the planning and provision of school oral health services as well as health services” . There is a consensus on the importance and relevance of screening for untreated dental caries in children . Whilst screening for different oral diseases and conditions in children, such as periodontal diseases and orthodontic conditions, is controversial and of questionable value , professionals have included these diseases and conditions within the priority set of clinical criteria for school-based dental screening .
Despite the popularity of school-based dental screening in many countries and recommendations by the World Health Organization, there is currently no uniform public health policy in the UK. In the UK, school-based dental screening, known for a long time as school dental inspection, had been a statutory requirement, supported by a consecutive Acts of Parliament, for more than hundred years . In the mid-1980s and later in 2000, there have been governmental questioning and discussion on the aims and effectiveness, and therefore cost-wise justification, of such public health intervention . A number of small randomised controlled trials showed that school-based dental screening programmes were effective in stimulating dental attendance for children in need of treatment, particularly those from low socioeconomic position . However, later in 2006, the UK National Screening Committee recommended to the UK Chief Dental Officers , based on the findings of a large randomised controlled trial , that there was no evidence to support the effectiveness of school-based dental screening in increasing dental attendance rates or reducing caries levels for children, particularly those from low socioeconomic position. The decision to continue or cease the screening activity was left to the discretion of local authorities. This uncertainty in evidence, because of conflicting results in the studies, has substantial financial and social implications. It is very clear that the key to resolve the above mentioned uncertainty is to conduct a robust systematic review of available evidence on the effectiveness of school-based dental screening for oral health, as was previously called for by Baker . There have been few related reviews , however, none had systematically reviewed and assessed available evidence. Thus, the current study aimed to systematically review the randomised controlled trials (RCTs) that aimed to assess the effectiveness of school-based dental screening versus no screening on improving oral health in children aged 3–18 years.
2
Materials and methods
The PRISMA guideline was followed to report this review, which is registered at PROSPERO platform (CRD42016038828) .
2.1
Inclusion and exclusion criteria
The present review included RCTs of school-based dental screening versus no screening for oral health, conducted on children aged 3–18 years, of both sexes, from different socio-demographic backgrounds, attending schools. There were no restrictions based on the country or year in which the trial was conducted, language of publication, and whether it was published as full journal article or only as a conference abstract. Although the plan was to translate non-English articles to English prior to data extraction, the translation was not required since there were no non-English articles that met the inclusion criteria.
2.2
Primary and secondary outcomes
As per protocol, information was sought on all the following primary and secondary outcomes, measured after a follow up period of two months or more.
The primary outcomes included:
1- Change in the prevalence and/or mean number of deciduous and/or permanent teeth with caries.
2- Incidence of dental attendance calculated as the number of children who attended a dentist at the follow-up out of the total number of children that were assigned to the trial’s arm.
3- Harms of screening (including adverse outcomes from false positive or false negative).
The secondary outcomes included:
1- Change in the prevalence of other oral diseases and conditions (infection/oral sepsis, pain, trauma, periodontal diseases, dental plaque, malocclusion, and pathological conditions of the hard or soft tissues of serious nature).
2- Oral health-related quality of life (OHRQOL).
3- School performance and attendance.
4- Costs.
2.3
Study selection
The following electronic bibliographic databases were searched: MEDLINE via Ovid, EMBASE via Ovid, The Cochrane Library (Cochrane Database of Systematic Reviews, Cochrane Central Register of Controlled Trials (CENTRAL), Cochrane Methodology Register), Web of Science (Science citation expanded), ClinicalTrials.gov, and the WHO International Clinical Trials Registry Platform until April 2016. Reference lists of eligible studies and review articles were searched for further eligible studies, and contact with experts to obtain grey literature was sought. The search keywords and medical subject headings (MeSH) terms related to school dental screening was combined with database-specific filters for controlled trials, where these were available. The search strategies used in the different databases have been presented in Appendix A . There were no language restrictions.
Titles and abstracts were screened independently by three sets of reviewers (EJ/EB, EJ/WS, EJ/KN). Full texts were sought when at least one of the authors considered the study as one that could potentially meet the inclusion criteria. The final decision was made on inclusion of the study based on full text and after discussion between the reviewers.
2.4
Data extraction
Data on demographical characteristics, risk of bias in the study, and the outcomes were extracted independently without blinding of the study authors, by two reviewers using a standardised data extraction form. Full details of the information sought is available in the published study protocol . Missing data were requested from study authors. Disagreements were resolved through discussion with a third author (the arbiter).
2.5
Risk of bias assessment
Cochrane’s criteria of risk of bias assessment were used . These included: sequence generation, allocation concealment, blinding of children and health care providers (screeners), blinding of outcome assessors, missing outcome data, selective outcome reporting, other sources of bias (including source of funding).
2.6
Strategy for data synthesis
Both narrative and quantitative syntheses of included studies’ findings were performed. The findings of studies that used the same outcome measure were pooled using random- and fixed-effects meta-analysis. Risk ratios were calculated for binary outcomes, whereas standardised mean differences were planned for continuous outcomes. Ninety five per cent confidence intervals (95% CI) and two sided P values were calculated for each outcome. In studies where the effects of clustering were present, the standard error of the effect estimates was adjusted using the intra-class correlation coefficient (ICC) to account for the cluster effect. Where adjusted effect estimates or ICC were not available, the ICC from the study with the lowest risk of bias was used and sensitivity analysis was performed for twice the ICC and half the ICC reported in the study with the lowest risk of bias. Heterogeneity between the studies in effect measures was assessed using both the Chi-square test and the I 2 statistic. I 2 values were interpreted in line with Cochrane’s Handbook i.e. 30%–60%: may represent moderate heterogeneity; 50%–90%: may represent substantial heterogeneity; 75%–100%: considerable heterogeneity, along with whether the heterogeneity was only in magnitude or whether it was in the direction of effects, chi-squared test of heterogeneity, and overlap of confidence intervals. Sensitivity analyses with different methods of imputation of data and low risk of bias trials, subgroup analyses (e.g. type of consent, referral and screeners, unit of randomisation) and publication bias assessment using funnel plots were planned , but could not be performed because of the paucity of the trials.
2
Materials and methods
The PRISMA guideline was followed to report this review, which is registered at PROSPERO platform (CRD42016038828) .
2.1
Inclusion and exclusion criteria
The present review included RCTs of school-based dental screening versus no screening for oral health, conducted on children aged 3–18 years, of both sexes, from different socio-demographic backgrounds, attending schools. There were no restrictions based on the country or year in which the trial was conducted, language of publication, and whether it was published as full journal article or only as a conference abstract. Although the plan was to translate non-English articles to English prior to data extraction, the translation was not required since there were no non-English articles that met the inclusion criteria.
2.2
Primary and secondary outcomes
As per protocol, information was sought on all the following primary and secondary outcomes, measured after a follow up period of two months or more.
The primary outcomes included:
1- Change in the prevalence and/or mean number of deciduous and/or permanent teeth with caries.
2- Incidence of dental attendance calculated as the number of children who attended a dentist at the follow-up out of the total number of children that were assigned to the trial’s arm.
3- Harms of screening (including adverse outcomes from false positive or false negative).
The secondary outcomes included:
1- Change in the prevalence of other oral diseases and conditions (infection/oral sepsis, pain, trauma, periodontal diseases, dental plaque, malocclusion, and pathological conditions of the hard or soft tissues of serious nature).
2- Oral health-related quality of life (OHRQOL).
3- School performance and attendance.
4- Costs.
2.3
Study selection
The following electronic bibliographic databases were searched: MEDLINE via Ovid, EMBASE via Ovid, The Cochrane Library (Cochrane Database of Systematic Reviews, Cochrane Central Register of Controlled Trials (CENTRAL), Cochrane Methodology Register), Web of Science (Science citation expanded), ClinicalTrials.gov, and the WHO International Clinical Trials Registry Platform until April 2016. Reference lists of eligible studies and review articles were searched for further eligible studies, and contact with experts to obtain grey literature was sought. The search keywords and medical subject headings (MeSH) terms related to school dental screening was combined with database-specific filters for controlled trials, where these were available. The search strategies used in the different databases have been presented in Appendix A . There were no language restrictions.
Titles and abstracts were screened independently by three sets of reviewers (EJ/EB, EJ/WS, EJ/KN). Full texts were sought when at least one of the authors considered the study as one that could potentially meet the inclusion criteria. The final decision was made on inclusion of the study based on full text and after discussion between the reviewers.
2.4
Data extraction
Data on demographical characteristics, risk of bias in the study, and the outcomes were extracted independently without blinding of the study authors, by two reviewers using a standardised data extraction form. Full details of the information sought is available in the published study protocol . Missing data were requested from study authors. Disagreements were resolved through discussion with a third author (the arbiter).
2.5
Risk of bias assessment
Cochrane’s criteria of risk of bias assessment were used . These included: sequence generation, allocation concealment, blinding of children and health care providers (screeners), blinding of outcome assessors, missing outcome data, selective outcome reporting, other sources of bias (including source of funding).
2.6
Strategy for data synthesis
Both narrative and quantitative syntheses of included studies’ findings were performed. The findings of studies that used the same outcome measure were pooled using random- and fixed-effects meta-analysis. Risk ratios were calculated for binary outcomes, whereas standardised mean differences were planned for continuous outcomes. Ninety five per cent confidence intervals (95% CI) and two sided P values were calculated for each outcome. In studies where the effects of clustering were present, the standard error of the effect estimates was adjusted using the intra-class correlation coefficient (ICC) to account for the cluster effect. Where adjusted effect estimates or ICC were not available, the ICC from the study with the lowest risk of bias was used and sensitivity analysis was performed for twice the ICC and half the ICC reported in the study with the lowest risk of bias. Heterogeneity between the studies in effect measures was assessed using both the Chi-square test and the I 2 statistic. I 2 values were interpreted in line with Cochrane’s Handbook i.e. 30%–60%: may represent moderate heterogeneity; 50%–90%: may represent substantial heterogeneity; 75%–100%: considerable heterogeneity, along with whether the heterogeneity was only in magnitude or whether it was in the direction of effects, chi-squared test of heterogeneity, and overlap of confidence intervals. Sensitivity analyses with different methods of imputation of data and low risk of bias trials, subgroup analyses (e.g. type of consent, referral and screeners, unit of randomisation) and publication bias assessment using funnel plots were planned , but could not be performed because of the paucity of the trials.
3
Results
A reference flow describing the review search results is presented in Fig. 1 . The search yielded 1938 unique citations. After screening titles and abstracts, we excluded 1927 citations as clearly irrelevant to this review, leaving 11 for full-text review. Only five studies were included in the current review ( Table 1 ).
| Reference | Target population | Sample size (drop outs) | Number of subjects and details of dental screening intervention | Number of subjects and details of no dental screening | Duration of follow-up | Outcome(s) measured |
|---|---|---|---|---|---|---|
| Cunningham et al. (2009) UK, Scotland/Edinburgh | All children (aged 12–13 years) in state schools in Lothian and Fife, who are unregistered with a dentist, and without urgent treatment needs or evidence of recent treatment. | 3923 (0) | 3104 received dental screening against a checklist of treatment need criteria. Personalised letters for every child, tailored (or not tailored) to the child’s registration status (never registered or lapsed) were sent to home via the child with a list of local dentists accepting NHS child patients. | 819 did not receive dental screening until after the end of the study. | 3 months | Incidence of dentist registration from relevant databases. |
| Donaldson and Kinirons (2001) UK, Northern Ireland | All children (aged 5.5–7.5 years) in schools in the Causeway Health and Social Services Trust. | 2321 (316) | 1161 received dental screening for cavitated caries and treatment sub-components according to BASCD. Personalised referral letters for positively screened children were sent to home via the child. | 1160 did not receive dental screening until after the end of the study. | 2 months | Incidence of dental attendance as reported by parents/carers. |
| Hebbal and Nagarajappa (2005) India | All children (aged 6–15 years) in public schools in Davangere, which were almost equidistant from the dental college. | 4500 (0) | 2100 received dental screening for treatment needs according to the WHO criteria 1997. Personalised referral letters for positively screened children tailored to their required treatment were sent to home via the child. Oral health education was also provided. | 2400 did not receive dental screening until after the end of the study. | 3 months | Incidence of dental attendance at the dental college. |
| Milsom et al. (2006) UK, England | All children (aged 6–8 years) in state schools in St Helen and Knowsley. | 17098 (3528 only in relation to dental caries as an outcome) | 12872 received dental screening by dentists or parents. The former was done against a set of criteria that were based on either consensus view or the opinion of the screening dentist. Personalised referral letters for positively screened children were posted to home. For those who received screening by parents, a dental information leaflet, distributed via the schools was, sent to encourage parents to examine their child’s mouth and to take their child to a dentist if any problems were noted. | 4226 did not receive dental screening. | 4 months | 1- Incidence of dental attendance from relevant databases. 2- Change in the prevalence and mean number of deciduous and permanent teeth with caries (calculated as dt a > 0, dt, DT b > 0, and DT). |
| Praveen et al. (2014) India | All children (aged 6–13 years) in schools in Vikarabad town. | 600 (0) | 300 received dental screening against the American Dental Association specified type III clinical examination criteria. Personalised referral letters for positively screened children, tailored to their required treatment were sent to home via the child. | 300 did not receive dental screening until after the end of the study. | 3 months | Incidence of dental attendance at the dental college. |
Stay updated, free dental videos. Join our Telegram channel
VIDEdental - Online dental courses
