Introduction

Dental caries is the progressive damage to the enamel caused by commensal bacteria in the mouth. External pathogens have not been shown to cause dental caries. However, a change in the homeostasis between normal commensals in the mouth and the surrounding tissues and structures has been shown to lead to dental caries causation [1]. Bacterial concentrations increase when there is inadequate removal of plaque and increased dietary sugar [2]. If untreated, dental caries can lead to pain and early loss of teeth, resulting in disfigurement and affect the oral health-related quality of life [3]. Early childhood caries (ECC) is defined as the presence of one or more cavitated or non-cavitated, decayed, missing (due to caries), or filled tooth in any primary (deciduous) tooth in a child aged under six years [4].

In 2017, the incidence of oral health conditions was ranked third-highest among all health problems and consisted of 3,6 billion cases, and approximately 530 million children suffer from deciduous caries globally [5]. The prevalence of ECC varies both across countries [6,7,8] and within the same country [7]. Similarly, the prevalence of ECC in South Africa (SA) differs between provinces. In 2004, the national prevalence of the disorder was 60.3% for 6 year olds [9]. No current national prevalence data for ECC in South Africa is available. However, Smit et al. [10] documented a significantly higher prevalence (84%) in the Western Cape than described by van Wyk et al. in 2004 [9].

The severity of the dental disease is expressed as the decayed, missing and filled tooth score and measured as the decayed, missing and filled tooth (dmft) index. International figures for the dmft indices vary: in Qatar it was reported as 7.6 [11], 3.65 in China [12], and 2.46 in Palestine [13]. In South Africa, the national dmft score was 2.4 [9], and 6.2 in the Western Cape [10].

South Africa is a densely populated developing country. It is listed as an upper- to middle-income country with 59.62 million inhabitants, of which children under the age of five constitute almost 10% or 5.7 million [14]. Historically, South Africa was immersed in political and racial division. Since the freedom charter was introduced in 2004, every South African is considered equal. Globally, South Africa has the highest income disparity within its constituents, with a Gini index of 63.0 in 2014 [15]. The Gini index determines the measure of inequality within a country. An index of 100 represents perfect inequality, and a measure of 0 means that the population is equal (all individuals have the same income) [16]. The country’s economic inequalities have resulted in an association between the prevalence of dental caries and unmet treatment needs [17].

Early Childhood Caries (ECC) has a significant burden in South Africa, particularly in the Western Cape Province. A few published studies report ECC’s prevalence in children under six years of age and under living in South Africa. To effectively prevent and manage ECC in South Africa, it is essential to know the disease prevalence and severity within this population. Therefore, the present study aimed to determine the prevalence and severity of ECC in South Africa.

Materials and methods

This study was conducted according to the Meta-Analysis of Observational Studies in Epidemiology (MOOSE) guidelines [18], Additional file 1: Table S1. The protocol of this systematic review was registered with PROSPERO, CRD42018112161, in November 2018. The protocol paper was published in JMIR Research Protocols in 2021 [19]. Ethics approval was not required as the present investigation was not a primary study involving participants.

A comprehensive search strategy was first developed by a research team comprising experts in paediatric dentistry, epidemiology, biostatistics, and librarian studies. There was no limit to the language of publication. Although all the studies were performed in South Africa, the studies were all published in English. The first and last authors (FKD and TR) independently conducted a pilot study to test the strategy, following which the authors confirmed the final search strategy. Peer-reviewed articles were searched in the following databases until the end of November 2020, MEDLINE; ERIC via EBSCOhost; Scopus; CINAHL via EBSCO (1900 to present); Dentistry and Oral Sciences Sources via EBSCOhost; Academic Search Complete via EBSCOhost; E-Journals via EBSCOhost; Health Source: Nursing Academic Edition via EBSCOhost and Cochrane Library. Using the key terms: (a) "early childhood caries" OR "caries" OR "decay" OR "dmft" OR "dental" OR "oral" OR "PUFA" (b) "prevalence" and (c) "children" OR "peri-natal" OR "paediatric" OR "pediatric" OR "neonatal" OR "infant" and (d) "South Africa". The keywords were used in the following combinations: a + b + c + d. Hand searching of included articles was performed. All eligible studies downloaded from the databases were uploaded into Rayyan [20], where duplicates were removed.

Screening and selection criteria

Studies were included if they were conducted in South Africa; they were based on children six years and under from the general population, if they reported sufficient information on the prevalence of ECC (sample size, prevalence of disease, mean of dmft, standard deviation of dmft) (Additional file 2: Table S2). Articles were excluded if they were abstracts, commentaries, review articles, or intervention studies. Dissertations, conference proceedings, commentaries/letters and other grey literature were also excluded from this review. Cross-sectional and cohort studies were eligible for inclusion. The inclusion and exclusion of articles were performed in Rayyan [21]. Any disagreements in the screening of articles were clarified with all the authors (Additional file 2: Table S2).

Data extraction

Two authors independently screened (FKD and TR) and extracted data from the included articles into Excel. If there was any disagreement between the authors, a consensus was reached through discussion with all the authors. If possible and required, the corresponding authors were contacted to provide the additional or missing information. In instances where articles failed to reflect the data collection date and were too old, or the authors could not be contacted, a consensus was reached among the present study’s authors to impute a suitable missing year of data collection; usually, 3–5 years before the study was published.

The following information was extracted from each eligible study: author, year of publication, study design, location and period, sampling technique, sample size, number of cases, diagnostic criteria, type of examiners, number decayed, missing, and filled teeth (dmft). Where possible, each category was sub-grouped according to the year of publication, age, urban/rural area, and Province. If it was unclear whether the area was urban or rural, the information was designated to an “urban/rural” category.

Critical appraisal

The studies’ quality was assessed by two independent authors (FK and TR) using the Joanna Briggs Institute (JBI) Critical Appraisal Checklist for Studies Reporting Prevalence Data [22, 23]. The specific JBI Critical Appraisal tool contained nine explicit criteria, and a maximum score of nine indicated the lowest risk of bias, Table 1. The process was repeated twice by the same authors. Any inconsistencies which arose between the two authors were resolved by consulting with the remaining authors. Two independent authors (FK and TR) judged the scoring, and a final decision was reached by consensus with all the authors.

Table 1 Critical Appraisal of included studies
Full size table

Data synthesis

Meta-analyses were conducted using StataCorp. 2019. STATA Statistical Software: Release 17, College Station, TX: StataCorp LLC. The pooled estimates and 95% confidence intervals for each indicator were calculated by combining each study’s data. Q-test and I2-statistical analysis were used to determine statistical heterogeneity. A random-effects model was adopted because of significant heterogeneity (I2 > 50%); Subgroup analysis was conducted to explore possible factors, including urban and rural status, age distribution and Province.

To reflect ECC’s spatial distribution, pooled prevalence estimates for ECC in all children under six years in each Province during 1975–2014 were entered into the QGIS software 3.8.3 (2019) to form a prevalence map.

Patient and public involvement statement

Neither patients nor the public was involved in the design, conduct, reporting, or dissemination of research plans.

Results

Search and selection results

A total of 2441 publications were identified in the search strategy, and a further seven were identified through other sources. After 194 articles were removed due to duplication, 2254 articles were analysed.

After reading the titles and abstracts of the remaining articles, 2201 were excluded, and two authors independently evaluated the remaining 53 full-text articles for eligibility. After the full-text analysis, 24 were excluded because they did not meet the inclusion criteria, and 29 were included in the meta-analysis (Fig. 1, Table 2). Thus, the total sample size was 29,477 individuals. The characteristics of the 29 studies are summarised in Table 1. Sixteen articles had information on the dmft score, and 26 had information on prevalence. The overall prevalence was 44.94% (Table 3), and the overall dmft score was 2.422 (Table 4). Of these, 28 studies used diagnostic criteria established by the World Health Organization (WHO) Oral Basic Surveys Methods 2,3 and 4. In 12 studies, dentists examined the participants (Table 5). Characteristics of included studies can be found in Table 5.

Fig. 1
figure 1

Flow chart of literature search and selection

Full size image
Table 2 Table of excluded studies
Full size table
Table 3 Pooled prevalence per age group, year period and urban or rural status in South Africa from 1978–2019
Full size table
Table 4 Pooled dmft score by age, year period and urban and rural status in South Africa from 1978—2019
Full size table
Table 5 Table of included studies
Full size table

Prevalence of ECC in South Africa

The pooled overall prevalence of ECC was 44.94% (95% C.I. 39.73–50.15%). The prevalence for 4- and 5- year-olds were reported in 14 studies. ECC’s overall prevalence ranged from 57.37% between 1975 and 1979 to 61.75% between 2010 and 2014, indicating a U-shaped trend over time. The rural prevalence was higher than the urban prevalence of ECC.

The overall prevalence of dental caries by Province is illustrated in Fig. 2. The prevalence of dental caries decreased from 1975 to 1994 but showed an increase from 1995 to the present, Table 1.

Fig. 2
figure 2

Spatial Distribution of ECC prevalence and dmft scores in South Africa

Full size image

The dmft per Province was 3.850, 3.000, 2.370, 2.442 and 0.330 in Eastern Cape, Kwazulu Natal, Western Cape, Gauteng, and Limpopo Provinces, respectively (Fig. 3).

Fig. 3
figure 3

Prevalence of ECC in South Africa by Year

Full size image

Publication bias

Funnel plots and Begg’s test assessed potential publication bias; the result was significant if p ≤ 0.05.

Duval and Tweedie’s "Trim and Fill" method was used to assess publication bias for the prevalence and dmft scores. Under the random-effects model, the point estimate for prevalence and 95% confidence interval for the pooled was 45.7% (44.8–46.7%). Using Trim and Fill, the imputed point estimate was the same. The method suggests a total of 0 studies missing from this review for the prevalence estimate.

In addition, there were no missing studies for the dmft score as the pooled point estimate was 2.307 (2.236–2.378) and using Trim and Fill, these values remain unchanged. Egger's test results were significant for both dmft, p < 0.001 (Fig. 4) and prevalence, p = 0.0031 (Fig. 5). These results suggest that there was publication bias.

Fig. 4
figure 4

Publication bias dmft score

Full size image
Fig. 5
figure 5

Publication bias prevalence estimate

Full size image

Discussion

Summary of main findings

The oral and dental health of individuals are essential to general health, and even more so in children. The current investigation is the first systematic review on the prevalence and severity of ECC in South Africa. The results summarise the last 30 years of prevalence studies among children under 71 months in South Africa.

Agreements and disagreements with previous studies

The meta-analyses of the observational data collected from the eligible studies in the current study have provided a summary estimate of ECC’s prevalence in South Africa. The overall pooled prevalence of ECC was 44.94% and 51.72% for 5-year-old children. The figures are much lower than other middle to upper-income countries, including Albania, 84.0% [73], American Samoa 87.0% [74], Argentina, 80.4% [75], and Turkey, 70.5% [76]. However, they are much higher than that of Namibia [77], which has a prevalence of 31.69%.

The prevalence of dental caries increased as age increased. This corroboratesthe findings from a systematic review of the prevalence of early childhood caries in China [78]. Caries prevalence seems to have decreased from the 1970’s till the early 1990’s, Table 1. Thereafter, the prevalence of caries in children under 6 appears to increase over time. After 1994, the South African government aimed to improve the living standards by providing the poor with housing rather than providing the poor with higher incomes [79]. There was thus an increase in housing and infrastructure but, no change in employment [79]. Furthermore, there has also been an increase in urbanisation. The majority of the population in South Africa is Black, and under the apartheid regime, they were restricted in their mobility [ability to move to urban areas for work]. After the apartheid laws were revoked in June 1991 there was an increase in mobility towards the cities. Post-Apartheid urbanisation has resulted in greater access to sugar and junk food compared to rural areas [80]. This may have resulted in the higher caries prevalence noted post-apartheid.

Of interest was the declining trend in the prevalence of caries prevalence as one moved to the north of the country. The National Children's Oral Health Survey, 2001–2002, indicated that the Limpopo Province had the lowest rate of dental caries in children: 31.30% of 4–5-year-olds and 30.80% in 6-year-old children [9]. The Limpopo Province is one of the poorest regions of South Africa, with a large disparity between poor and affluent residents, especially in the rural areas [81]. Limpopo Province is an arid land, and 75.00% of the population is dependent on groundwater. The Limpopo and North-West provinces have been identified as having high fluoride levels, up to 30 mg/l [82] and they also present with the lowest prevalence of dental caries at 37.36% and 41.02%, respectively.

A challenge or limitation of the current review is that most of the studies were conducted in the Western Cape and Gauteng Provinces. The authors strongly suggest that examiners are thoroughly trained and that test–retest validity is conducted in all future prevalence studies in South Africa. It would also be favourable that a single tool (standardised) be used to examine dental prevalence. The choice of dental disease tool should be based on the exact outcome of the study performed. While the dmft score is sufficient for a dental prevalence and severity study, pulpal involvement, ulceration, fistula and abscess (pufa) score is better suited to determine the severity of clinical outcomes related to the dental treatment needs of the study under investigation.

Caution should be exercised when evaluating the current study results as there is a high heterogeneity among the included studies; this is not uncommon when evaluating systematic reviews of this nature.