The protocol for this systematic review and meta-analysis [14] was registered with PROSPERO (CRD42023442266) and conducted in accordance with the MOOSE (Meta-Analysis of Observational Studies in Epidemiology) and PRISMA-P (Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols) guidelines [15,16].

RCTs and clinical studies (prospective and retrospective) evaluating crowns fabricated using stainless steel, zirconia, composite resin, compomer, and resin-modified glass ionomer cement (RMGIC) were included. The predefined inclusion criteria for this systematic review encompassed studies with an English abstract that evaluated crown restorations in pediatric patients aged 1-10 years, reporting crown-retention data, reasons for retentive loss, and technical and biological complications. Qualitative interviews, quasi-experimental studies, single-case studies, and series of single-case studies were excluded. Additionally, articles based on conference abstracts and dissertations were excluded.

Biological factors in this analysis included secondary caries and periodontal pathology (including the periodontal index, which assesses the severity of periodontal disease). Clinical parameters included the gingival index (evaluating gingival inflammation) and other periodontal health indicators. Technical factors included anatomic form, marginal adaptation, color match, surface texture, and wear on the opposing tooth. Success rates, if reported, were also included for the purposes of this systematic review and meta-analysis.

Using a database-appropriate search strategy, electronic databases, including Cochrane, Embase, and PubMed (MEDLINE), were systematically searched to identify RCTs and clinical studies (prospective and retrospective) based on the eligibility criteria described above. The PICOS (Population, Intervention, Comparison, Outcomes, and Study design) format for this systematic review and meta-analysis was as follows:

Multimedia Appendix 1 presents the search syntax used for searching the electronic databases. The search strategy was considered adequate to reduce the risk of selection and detection bias.

The search results were imported into Zotero, and duplicates were removed to create a virtual library. Study selection was conducted in 2 stages. In the first stage, 2 independent reviewers (SL and KJ) screened the titles and abstracts of identified studies to determine their eligibility. In the second stage, the same 2 reviewers independently assessed the full-text articles of potentially relevant studies to confirm whether they met the eligibility criteria. Quality assessment of included RCTs and prospective and retrospective clinical studies was also independently performed by SL and KJ. Any disagreements at any stage, including study selection, quality assessment, and data extraction, were resolved through discussion. If consensus could not be reached, a third expert (DG) was consulted, with DG making the final decision.

Data extraction was independently conducted by 2 reviewers (SL and KJ) using a standardized data extraction form. Extracted data included author details, year of publication, number of children, age, number of teeth, tooth type, crown material, dropout, follow-up duration, crown retention, and retention complications including biological and technical complications. Discrepancies in data extraction were resolved through discussion between the 2 reviewers. If consensus could not be reached, a third expert (DG) was consulted to make the final decision.

The Risk of Bias 2 tool developed by the Cochrane Collaboration was used to assess the quality of included RCTs [17]. Two independent reviewers evaluated various bias domains in RCTs, including randomization, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other sources of bias. After assessing each included RCT for risk-of-bias domains specified by the Risk of Bias 2 tool, an overall risk-of-bias judgment (low, some concerns, or high) was assigned to each trial [17]. A checklist proposed by Moga et al [18] was adapted for assessing the risk of bias in the included prospective and retrospective clinical studies. Only studies with a moderate or low risk of bias were included in this analysis. Furthermore, a funnel plot was generated to assess for any publication bias.

For quantitative analysis (meta-analysis), retention was defined as the number of crowns that were in situ, regardless of technical and biological complications. Total exposure time refers to the cumulative duration that all crowns remain in the mouth across all included studies. It is calculated by summing the duration each crown is in place for each study and then aggregating these times across all studies. Failure rates resulting from retentive complications were calculated by dividing the number of failures by the total exposure time. Exposure time for each included study was calculated by summing the exposure time for all restorations. A Poisson regression model was used to analyze the calculated rates. Proportions of crowns retained at 3 years and 5 years were estimated with an assumption of constant event rates. The Pearson goodness-of-fit test was performed to assess heterogeneity across studies and to determine whether a fixed-effects or random-effects model should be used. Since no statistically significant heterogeneity was detected in any of the individual groups and rates (P>.05), a fixed-effects Poisson regression model was used to estimate the parameters. A P value <.05 was considered significant. All analyses were performed using R statistical software (version 4.1.2; R Core Team).

A total of 13 RCTs and 5 clinical studies (prospective and retrospective) were included in this systematic review [19-31]. Figure 1 presents the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flow diagram illustrating the study selection process. Tables 1 and 2 provide the risk-of-bias assessments for the included RCTs and clinical studies, respectively.

Table 3 summarizes the 13 included RCTs, presenting data on crown retention and retentive complications [19-31]. These RCTs were conducted between 2014 and 2022 and recruited 454 children who received a total of 1172 crowns [19-31]. The follow-up duration ranged from 6 to 36 (median 12, IQR 9) months, with a cumulative dropout rate of 10.6% (124/1172 crowns). Table 4 outlines the 5 included clinical studies, reporting data on crown retention and retentive complications [32-36]. These clinical studies were conducted between 2008 and 2020 and included data from 810 children, with a follow-up duration ranging from 12 to 24 (median 20.8, IQR 5) months [32-36].

Figure 2 provides a descriptive analysis of retention rates for crowns fabricated with different materials, as reported in the included RCTs [19-31]. Retention rates varied from 77.8% to 100%, from 80.6% to 100%, and from 92.3% to 100% for compomer, composite resin, and stainless steel crowns, respectively. Additionally, retention rates for strip crowns and zirconia crowns ranged from 78% to 100% and from 86.4% to 100%, respectively. Table 5 illustrates various retentive complications reported in the RCTs for different materials. Regarding clinical studies, debonding and loss of crown were reported for zirconia crowns, while crown fractures were reported for composite resin crowns (Table 4).

Table 6 depicts the distribution of biological complications reported in the included RCTs [18-30]. In the prospective and retrospective clinical studies, 8.3% (6/67) of children receiving compomer crowns and 8.8% (11/121) receiving composite resin crowns reported secondary caries [31-35]. Additionally, 3.3% (10/289) and 21.8% (80/368) of children with stainless steel crowns and zirconia crowns, respectively, in these studies exhibited gingival inflammation [32-36].

Table 7 outlines the distribution of technical complications reported in the included RCTs [18-30]. In the prospective and retrospective clinical studies, 29% (19/67) of compomer crowns exhibited marginal adaptation complications, while stainless steel crowns did not demonstrate marginal adaptation issues [32-36]. No marginal discoloration or plaque retention complications were observed for compomer crowns. Among the 399 zirconia crowns evaluated, 4.8% (19/399) demonstrated shade mismatches, and 1% (4/399) exhibited marginal integrity complications [32-36].

Table 8 presents the estimated 3-year and 5-year retention rates of crowns fabricated using various materials, as reported in the included RCTs [19-31].

The Pearson goodness-of-fit test was performed to assess heterogeneity across studies and to determine whether a fixed-effects or random-effects model should be used. Since no statistically significant heterogeneity was detected in any of the individual groups and rates (P>.05), a fixed-effects Poisson regression model was used to estimate the parameters.

Meta-analysis of pooled 5-year retention rates for different crown materials revealed the following retention rates: 88.90% for compomer crowns (Figure 3), 92.18% for composite resin crowns (Figure 4), 90.30% for RMGIC crowns (Figure 5), and 97.88% for stainless steel crowns (Figure 6). Additionally, strip crowns exhibited a retention rate of 83.48% (Figure 7), while zirconia crowns had a retention rate of 97.09% (Figure 8). Figure 9 presents the retention rates, irrespective of the material, and found a retention rate of 92.20% (95% CI 92.14%‐92.26%).