This meta-analysis evaluated all studies according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. All studies were reviewed independently by two authors. A systematic electronic literature search of Web of Science, Cochrane Library, PubMed, and Embase databases was conducted from inception to August 2024 using the following combined MeSH heading and keyword search terms. The search terms were combined from the following medical subject headings: “ cardiac surgical procedures, congenital heart disease, cardiac septal defect, atrial septal defect, mini-thoracotomy, thoracotomy, incision, cardiac surgery, coronary artery bypass grafting, CABG, cardiac surgery, valve replacement, or thoracic surgery, score, etc.” [AND] and [OR] were used as Boolean operators in the search strategy. The search was limited to clinical trials involving adult participants and published in English. The reference list of each full text used for the review and any existing reviews identified were manually searched to identify additional potentially eligible citations not found in the electronic database search.

To ensure selection of relevant studies, the following inclusion criteria were used. This included studies with scar assessment and measuring depression/anxiety scores as secondary outcomes in order to include more studies in areas that have been exhausted. Unpublished and ongoing studies were not identified and therefore excluded. The inclusion criteria for the current meta-analysis were as follows: participants: Patients who have undergone cardiac surgery; interventions: the full sternotomy and the limited sternotomy; comparators: scar formation and depression and anxiety symptoms.

The literature was managed using Endnote X9 (Clarivate, London, UK), and studies that were irrelevant, duplicates, or non-experimental (especially those based on animals) were excluded. Following this, the titles and abstracts were carefully reviewed to eliminate studies that did not meet the inclusion criteria. The remaining literature was examined in detail, ensuring a thorough and systematic process. The full texts of the identified studies were reviewed for a secondary screening, and any disagreements were settled through consensus discussions among three researchers. Two researchers assessed the quality of the literature independently using the Cochrane bias tool.

Meta-analysis was conducted using Review Manager version 5.4 (The Cochrane Collaboration). Effect sizes were calculated by converting measurement data into mean differences (MD), while count data were represented as odds ratios (OR). The heterogeneity among the included studies was assessed using Cochran’s Q statistic, and the Higgins I² test was employed to evaluate the variability of effect sizes. A statistically significant heterogeneity was indicated by an I² value exceeding 50%. To address any apparent heterogeneity, parallel subgroup and sensitivity analyses were performed. A fixed-effects model was applied for the meta-analyses when no significant heterogeneity was detected among the studies. A P-value of less than 0.05 was considered statistically significant.

The quality of the articles and our research was evaluated using the Risk of Bias Tool 2 (RoB2) and GRADE approaches. The RoB2 tool focused on five critical areas: (i) the process of randomization, (ii) deviations from the intended interventions, (iii) missing outcome data, (iv) the measurement of outcomes, and (v) the selection of reported results. In cases of disagreement, the reviewers collaborated until they achieved a consensus.

Figure 1 shows the flowchart of the study selection process. After the initial review of titles and abstracts, 402 studies remained after removing duplicates. Upon further examination of the abstracts, 21 studies met the preliminary screening criteria, and ultimately, 4 studies (16, 18, 28, 37) were selected for analysis after reviewing the full texts again.

Table 1 summarizes the characteristics of the studies included in this review.

The analysis compared the baseline status of patients in four studies. Figures 2 , 3 show tables outlining the “risk of bias in included studies.” A detailed description of the study features can be found in the section labeled “Features included in the study.”

All four studies compared the scores 6 months after cardiac surgery. There was no heterogeneity between the two groups of studies (I² = 0%). The scar after cardiac surgery in the full sternotomy group was significantly smaller than that in the limited sternotomy group (OR 0.94 [95% CI 0.28-1.61]; P = 0.005) (see Figure 4 ).

All four studies compared depression scores 6 months after cardiac surgery. There was no heterogeneity between the two groups (I² = 0%). The depression score after cardiac surgery in the Full sternotomy group was significantly higher than that in the Limited sternotomy group (OR1.61[95%CI:0.63-2.60]; P = 0.001) (see Figure 5 ).

All four studies compared anxiety scores 6 months after cardiac surgery. There was slight heterogeneity between the two studies (I² = 58%). Therefore, a random effects model was used. Compared with the Limited sternotomy group, the OR of post-cardiac surgery anxiety score in the Full sternotomy group was 0.70 [95% CI:1.40-2.80]; P = 0.51) (see Figure 6 ), which was not statistically significant difference.

Funnel plots were employed to evaluate publication bias for each outcome. As illustrated in Figures 7A through 7C , the funnel plots exhibit good symmetry, with scatter points evenly distributed within the funnel. This suggests that publication bias is unlikely to significantly impact the results of the meta-analysis (see Figure 7 ).

This meta-analysis has some limitations. Firstly, due to the limited number of studies on the effect of scar formation after cardiac surgery on postoperative anxiety and depression, this study only included 4 studies, and the sample size of each study was relatively small. Therefore, the relationship between scar formation after cardiac surgery and depression and anxiety needs to be further confirmed. Secondly, although the heterogeneity between the included studies was low, other potential factors may affect the summary results, such as gender, social support, and marital status. Notably, we did not account for Hypermobility Spectrum (defined by a Beighton score above 4), a known risk factor associated with both hypertrophic scars/keloids and significantly higher rates of anxiety (22-fold increased risk) and depression (over 3-fold increased risk). Given that approximately 15% of the general population meets this criterion, it is likely that some included patients had this comorbidity, which may have confounded the observed associations between scar formation and psychological symptoms. Future studies can consider multicenter prospective studies to obtain more diverse data and improve the external validity of the study. Thirdly, our inclusion strategy was relatively narrow, as we focused primarily on studies using the Patient and Observer Scar Assessment and Vancouver Scar Scale for scar evaluation, and specific instruments for anxiety and depression assessment. We excluded studies employing alternative scar metrics or other psychological scales. This may have limited the breadth of data and introduced potential selection bias, as different assessment tools may capture distinct aspects of scar characteristics and psychological states, affecting the generalizability of our findings. Future studies should adopt a broader inclusion strategy, incorporating diverse scar and psychological assessment instruments, and clearly justify any exclusions to ensure a more comprehensive synthesis of evidence. Fourthly, the follow-up time for scar formation in this study was 6 months after surgery, which is relatively short. Long-term follow-up studies can be carried out in the future to understand the long-term changes of scars on patients’ psychological state. At the same time, personalized psychological intervention programs can be designed in the future, which may be more effective in improving the mental health of patients after scar formation.