A commentary on ‘Metabolic syndrome and surgical complications: a systematic review and meta-analysis of 13 million individuals’

Dear Editor,

We read a recently conducted systematic review that explored the impact of metabolic syndrome (MetS) on surgical complications among patients¹. The results found that patients with MetS had a significant risk of 30-day mortality, cardiovascular complications, surgical site infection, hospital readmission, and extended length of hospital stay compared to those without that condition. This study indicated that early identification and management of MetS among surgical patients were very important². The study had multiple advantages, such as a large-scale sample of 1, 919, 347, patients, overall surgical complications, broad subgroup analysis, and compliance with PRISMA, AMSTAR, and pre-registered PROSPERO. We commend the authors for conducting such a great study that strengthens the important impact of MetS on adverse surgical complications. However, with humility, we believe that some questions are worth discussing to improve the quality of this systematic review and meta-analysis.

In the statistical analysis section, the authors initially declared that they extracted the effect size of the association between MetS and surgical complications from the adjusted model, controlling for all confounding factors. However, the results presented in the figures showed the univariate association between MetS and surgical complications. For instance, in Angeloni’s study, the odds ratio (OR) of the association between MetS and 30-day mortality was 1.91 (95% CI: 1.05–3.46), reflecting the crude association. Conversely, the original Angeloni’s study, using a fully adjusted Cox hazards regression model, reported a hazard ratio (HR) for the independent association between MetS and 30-day mortality as 2.54 (95% CI: 1.51–5.26). As a result, the pooled association between MetS and 30-day mortality, as shown in Figure 2 of Norris’s meta-analysis, was actually the crude effect size, leading to an overestimation of the effect size. The same methodology was also applied to 30-day cardiovascular complications, surgical site infections, and readmission. We suggest that the authors explicitly remind readers that the pooled effect size represents the crude association, not the independent association.

Secondly, based on the original studies, we can address another important question: What is the prevalence of MetS among surgical patients? We pooled all the data from the included studies and found that the overall prevalence of MetS among surgical patients was 19.2% (15.07–24.1%). Subgroup analysis based on surgery type revealed that the pooled prevalence of MetS was highest in vascular surgeries at 44.4% (95% CI: 35.3–53.8%), followed by vascular (37.1%, 95% CI: 26.1–49.6%), colorectal (18.3%, 95% CI: 5.8–44.9%), bariatric (19.9%, 95% CI: 13.3–28.7%), gastrointestinal (19.8%, 95% CI: 8.7–39.0%), urology (17.7%, 95% CI: 7.4–36.5%), hepatobiliary (13.4%, 95% CI: 6.5–25.6%), endocrine (8.7%, 95% CI: 6.1–12.4%), orthopedic (7.8%, 95% CI: 6.1–9.9%), and plastic surgery (7.4%, 95% CI: 4.5–11.9%). Furthermore, the results of subgroup analysis based on region showed that the pooled prevalence of MetS was 40.1% (34.0–46.5%) in Europe, followed by 33.2% (21.3–47.7%) in Asia, and 12.0% (14.4–23.1%) in the USA, with a statistically significant difference of P-value <0.01.

Lastly, although the authors indicated that they followed the PRISMA guideline, we did not find some other important analyses, such as sensitivity analysis and publication bias, in this systematic review. Regarding 30-day mortality and 30-day readmission, the result of publication bias by Egger’s test yielded P-values of 0.709 and 0.718, respectively, indicating no potential publication bias among the included studies. Similar things were observed regarding cardiovascular complications (P=0.604) and surgical site infections (P=0.389).

In addition, sensitivity analysis for 30-day mortality by omitting each original study and combining the remaining studies found that the pooled OR for the association between MetS and 30-day mortality ranged from 1.60 (95% CI: 1.31–1.95) to 1.82 (95% CI: 1.44–2.30), as shown in Figure 1. The sensitivity analysis for 30-day readmission showed that the pooled OR ranged from 1.46 (95% CI: 1.34–1.59) to 1.58 (95% CI: 1.43–1.75). Similar sensitivity analyses were used for 30-day surgical site infections and 30-day cardiovascular complications, and the results found that the pooled OR for 30-day surgical site infections ranged from 1.57 (95% CI: 1.43–1.73) to 1.62 (95% CI: 1.48–1.77), and for 30-day cardiovascular complications ranged from 1.50 (95% CI: 1.37–1.64) to 1.58 (95% CI: 1.42–1.75), respectively. Therefore, all the results of sensitivity analysis for outcomes showed that the results were robust and reliable.

Figure 1.

Sensitivity analysis for the association between metabolic syndrome and 30-day mortality.

Ethical approval

Not applicable.

Consent

Not applicable.

Sources of funding

Not applicable.

Author contribution

Xiao-Ming Zhang is responsible for study design and writing.

Conflicts of interest disclosure

There are no conflicts of interest.

Research registration unique identifying number (UIN)

Not applicable.

Guarantor

Xiao-Ming Zhang is the guarantor.

Data availability statement

The data in this study can be obtained from the public internet.

Provenance and peer review

Commentary, internally reviewed.