Adverse events in neonates with congenital heart disease after cardiovascular surgery: a systematic review and meta-analysis

Abstract

Introduction

Neonates with congenital heart disease (CHD) undergoing cardiovascular surgery are exposed to complex postoperative care and may experience adverse events.

Methods

A systematic review and meta-analysis was conducted. Searches were performed in the Cochrane Central Register of Controlled Trials, LILACS, CINAHL, EMBASE, PubMed, Web of Science Core Collection, and Scopus. Observational cohort studies evaluating adverse events in neonates were eligible for inclusion. Data extraction was performed independently by two reviewers; in the event of disagreement, a third reviewer was consulted. Search results were exported to the online reference management software EndNote Basic for duplicate removal and subsequently imported into the Rayyan platform.

Results

Of the 60 articles identified, 20 studies comprising a total of 25,422 neonates with congenital heart disease in the postoperative period were included. The overall prevalence of adverse events was 14%. The most prevalent events were central venous catheter (CVC)–related incidents (14%), healthcare-associated infections (12%), and extubation failure (11%).

Conclusions

The prevalence of adverse events in neonates with CHD undergoing cardiovascular surgery is substantial and likely underreported. Further studies are needed to improve the detection, mitigation, and reduction of these events.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12887-026-07164-z.

Introduction

Congenital heart disease (CHD) is a cardiac defect that affects approximately 1% of live births (LB) [1]. CHD comprises structural malformations that occur in the heart during embryonic development and accounts for nearly 30% of all congenital anomalies. Each year, approximately 1.35 million neonates with CHD are born worldwide. The global prevalence is estimated at 9.4 cases per 1,000 live births; in Asia, it is 9.3 per 1,000 LB; in Europe, approximately 8.2 per 1,000 LB; and in North America, 6.9 per 1,000 LB [2–4].

In Brazil, approximately 29.8 thousand neonates are born annually with CHD [5]. It is estimated that about 80% of CHD cases will require cardiovascular surgical intervention for treatment. Regarding neonatal mortality, CHD represents the third leading cause of death in this population [6–8].

The postoperative period following correction of congenital heart disease is a critical phase for neonates, as they are more vulnerable to numerous complications intrinsic to the disease itself, the surgical procedures, and adverse events (AEs). AEs are a type of incident that result in unnecessary harm to the patient and are considered preventable [9].

In the general neonatal population, there are no official data on the prevalence of AEs. However, studies have shown that such events are common in this population. One study, based on medical record review, identified an incidence of 0.74 AEs per patient, with a higher proportion observed among preterm neonates [10] in other similar studies, a prevalence ranging from 70% to 80% of adverse events and an incidence of 2.6 AEs per patient were reported [11, 12]. The most frequently reported AEs included healthcare-associated infections, medication errors, skin injuries, extubation failure, unplanned extubation—defined as the unintentional removal of the endotracheal tube, commonly occurring accidentally during patient care—and central venous catheter loss [12, 13].

Based on findings in non-cardiac neonates, adverse events represent a significant problem. Therefore, this systematic review and meta-analysis aims to identify the prevalence and the main adverse events in postoperative neonates with congenital heart disease.

Methods

This review was conducted in accordance with the criteria for systematic reviews and meta-analyses established by the PRISMA guidelines [14, 15]. The protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO) under registration number CRD42022354692. As this is a prevalence study based on observational research, the Meta-analysis Of Observational Studies in Epidemiology (MOOSE) guidelines were also incorporated into the methodological framework of this review [16].

The databases searched included the Cochrane Central Register of Controlled Trials (CENTRAL), accessed via the Cochrane Library; Latin American and Caribbean Health Sciences Literature (LILACS), accessed via the Virtual Health Library (BVS); the Cumulative Index to Nursing and Allied Health Literature (CINAHL), accessed via EBSCOhost; Excerpta Medica Database (EMBASE), accessed via Elsevier; PubMed/MEDLINE, accessed via the U.S. National Library of Medicine; Web of Science, accessed via Clarivate; and Scopus, accessed via Elsevier. Electronic searches were conducted on December 11, 2024, with no restrictions on publication date or language.

The search strategy was developed and conducted independently by Reviewer 01 and Reviewer 02 using combinations of controlled descriptors, keywords, and variants of the following terms: “Neonate,” “Congenital heart disease,” “Surgery,” “Postoperative,” and “Adverse events.” No searches of the grey literature were performed.

The search results retrieved from the databases were exported to the online version of the reference management software EndNote Basic for duplicate removal and subsequently imported into the Rayyan platform. Two authors independently screened all potentially eligible abstracts and discussed any disagreements to reach consensus; when necessary, a third reviewer with expertise in the subject area was consulted. In the second stage, the full texts of potentially eligible articles were independently reviewed, and consensus was reached regarding the studies included in this review.

The same authors independently extracted the outcomes of interest using a data extraction spreadsheet developed by the authors in Microsoft Excel, structured according to the eligibility criteria. Eligible studies included neonates with congenital heart disease in the postoperative period following cardiovascular surgery following cardiovascular surgery (including both corrective and palliative procedures, as defined by the primary studies). Information regarding prior percutaneous cardiac interventions was not used as an inclusion or exclusion criterion due to inconsistent reporting across studies. Exclusion criteria were: neonates with congenital heart disease managed clinically without the need for surgical intervention; adverse events occurring in the preoperative or intraoperative periods; and case–control studies.

The methodological quality of the included studies was assessed using the Newcastle–Ottawa Scale (NOS) for cohort studies. Based on this scale, each study was evaluated across three broad domains—selection, comparability, and outcome—using a star allocation system [17]. A maximum of nine stars can be awarded across all three domains; studies scoring fewer than five stars were considered low quality, those scoring five to six stars moderate quality, and those scoring seven to nine stars high quality.

For this study, adverse events were defined according to the International Classification for Patient Safety (ICPS) proposed by the World Health Organization, that is, a preventable incident that causes unnecessary harm to the patient [9]. The outcomes investigated in this systematic review and meta-analysis were: (1) prevalence of adverse events; and (2) types of adverse events.

The I² statistic was calculated to assess the degree of homogeneity among the studies included in the meta-analysis. The I² statistic ranges from 0% to 100%, with higher values indicating greater heterogeneity among the included studies. When substantial heterogeneity is present, a mixed-effects model is more appropriate for estimating the pooled meta-analytic measure. The pooled prevalence and corresponding 95% confidence intervals were estimated using a random-effects meta-analysis with the restricted maximum likelihood (REML) approach and the Freeman–Tukey double arcsine transformation [18].

Additionally, subgroup analyses were conducted according to the type of adverse event. Subgroup analyses based on type of surgical procedure or prior percutaneous interventions were not feasible due to lack of detailed and consistent reporting in the included studies.Publication bias was assessed using funnel plot asymmetry and Egger’s test [19], with a significance level of 10% to indicate the presence of publication bias. The synthesis of the meta-analysis results was performed using STATA software, version 18 (StataCorp LP, College Station, TX, USA).

Results

A total of 6,659 articles were identified through database searches. After removal of duplicate records (n = 2,693), 3,966 articles remained and were independently screened by two reviewers based on titles and abstracts. Following this screening process, 60 articles were deemed eligible for full-text review (Fig. 1).

Fig. 1.

Flowchart illustrating the selection criteria used for studies included in this systematic review [15]

The excluded articles are listed with their respective reasons in Fig. 1. A total of 20 studies met the inclusion criteria for this systematic review, comprising 1,701 adverse events. The general characteristics of these studies are presented in Table 1.

Table 1.

Characteristics of eligible studies in the systematic review

Year	Country	Author	Study Design	Períod	Type of event	Sample	Nº de AV
2000	Israel	Erez et al. [20]	RC*	1995–1997	Infection (HAI†)	108	3
2007	Índia	Bakshi et al. [21]	RC*	1999–2006	Infection (HAI†)	300	148
2007	USA	Tan et al. [22]	RC*	2002–2004	Catheter-Related Incident	115	13
2010	USA	Barker et al. [23]	RC*	2002–2006	Infection (HAI†)	6621	426
2011	USA	Das et al. [24]	RC*	2004–2008	Infection (HAI†)	110	24
2012	USA	Kansy et al. [25]	RC*	1995–2010	Infection (HAI†)	1037	95
2012	USA	Gupta at al. [26]	RC*	2005–2011	Extubation Failure	64	14
2013	Canada	Jacques et al. [27]	RC*	2001–2011	Medical error	191	126
2017	USA	Mastropietro et al. [28]	PO**	2015	Extubation Failure	283	35
2017	USA	Benneyworth et al. [29]	PO**	2013–2015	Extubation Failure	899	103
2017	Japan	Miura et al. [30]	RC*	2010–2016	Extubation Failure	156	25
2019	USA	Haughey et al. [31]	RC*	2013–2015	Infection (HAI†)	1836	28
2019	USA	Stein et al. [32]	RC*	2015–2016	Catheter-Related Incident	124	38
2020	China	Yu et al. [33]	RC*	2012–2018	Infection (HAI†)	85	28
2020	China	Lin et al. [34]	RC*	2010–2019	Infection (HAI†)	387	9
2022	USA	Beshish et al. [35]	RC*	2008–2019	Extubation Failure	1957	115
2023	USA	Wilson et al. [36]	RC*	2015–2018	Extubation Failure	240	40
2023	USA	Perry et al. [37]	RC*	2014–2020	Extubation Failure	10.337	386
2023	USA	Anton-Martin et al. [38]	RC*	2011–2018	Catheter-Related Incident	371	20
2024	Turkey	Ozalp et al. [39]	RC*	2020–2022	Infecção (HAI†)	201	25

HAIs† Healthcare-associated infections, RC^* Retrospective Cohort, PO^** Prospective Observational

Of the 20 included articles, the earliest was published in 2000 [20], and the most recent in 2024 [39]. Only one article was not published in English [34]. Information regarding the type of surgical procedures (e.g., corrective vs. palliative) and prior percutaneous interventions was inconsistently reported across studies and therefore could not be systematically analyzed. Only one study did not report the population in absolute numbers, presenting data as percentages, which required conversion for prevalence calculation [22]. Most studies included in this systematic review were conducted in high-income countries (n = 16), namely the United States of America (n = 13), Canada (n = 1), Japan (n = 1), and Israel (n = 1). The remaining studies (n = 5) were conducted in middle- and low-income countries, specifically China (n = 2), India (n = 1), and Turkey (n = 1). Regarding study design, all 20 studies were observational, including retrospective cohort studies (n = 17) and prospective cohort studies (n = 3).

The methodological quality assessment is presented in Table 2.

Table 2.

Methodological quality assessment

Estudy	Design	Type	*NOS	Outcome
Erez et al. [20]	O	R	7	Good
Bakshi et al. [21]	O	R	8	Good
Tan et al. [22]	O	R	6	Fair
Barker et al. [23]	O	R	8	Good
Das et al. [24]	O	R	7	Good
Kansy et al. [25]	O	R	8	Good
Gupta at al. [26]	O	R	7	Good
Jacques et al. [27]	O	R	7	Good
Mastropietro et al. [28]	O	P	7	Good
Benneyworth et al. [29]	O	P	7	Good
Miura et al. [30]	O	R	6	Fair
Haughey et al. [31]	O	R	8	Good
Stein et al. [32]	O	R	8	Good
Yu et al. [33]	O	R	8	Good
Lin et al. [34]	O	R	6	Fair
Beshish et al. [35]	O	R	6	Fair
Wilson et al. [36]	O	R	8	Good
Perry et al. [37]	O	R	8	Good
Anton-Martin et al. [38]	O	R	8	Good
Ozalp et al. [39]	O	P	8	Good

^*Newcastle–Ottawa Scale

Of the 20 articles, 16 (80%) were classified as having good quality and 4 (20%) as having fair quality. According to the Newcastle–Ottawa Scale (NOS) criteria [17], no study was classified as having poor quality. The number of stars awarded to the studies ranged from 6 to 8.

From the 20 articles included in this review, a total population of 25,422 neonates was identified, with the occurrence of 1,701 adverse events. Four meta-analyses were conducted: one addressing adverse events overall and three reporting the most prevalent adverse events, respectively. Nine studies reported healthcare-associated infections [20, 21, 23, 24, 31, 33, 34, 39], seven studies reported extubation failure [26, 28–30, 35–37], three studies reported central venous catheter (CVC)-related incidents [22, 32, 38], and one study reported medical error [27].

In the statistical analysis presented in the forest plot (Fig. 2), the overall prevalence of adverse events was 14% [0.14 (95% CI 0.08–0.22, p < 0.01)]. The prevalence of central venous catheter (CVC)-related incidents was 14% [0.14 (95% CI 0.03–0.31, p < 0.01)] (Fig. 3).

Fig. 2.

Forest plot of the meta-analysis of the prevalence of adverse events

Fig. 3.

Forest plot of the meta-analysis of CVC-related incidents

The prevalence of healthcare-associated infections was 12% [0.12 (95% CI 0.04–0.23, p < 0.01)] (Fig. 4), and the prevalence of extubation failure was 11% [0.11 (95% CI 0.07–0.17, p < 0.01)] (Fig. 5).

Fig. 4.

Forest plot of the meta-analysis of healthcare-associated infections (HAIs)

Fig. 5.

Forest plot of the meta-analysis of extubation failure

In addition to the meta-analyses, Egger’s test was performed and is presented in the funnel plot (Fig. 6). Egger’s test yielded a value of 0.0236, suggesting the presence of possible publication bias, indicating that there are relatively few studies exploring the topic addressed in this review. This finding is important, as it highlights the need for further studies investigating adverse events in this population.

Fig. 6.

Egger’s test [40]

Discussion

This systematic review quantitatively evaluated the prevalence of adverse events in neonates with congenital heart disease in the postoperative period following cardiovascular surgery. Of these neonates, 99.1% were classified as high complexity and are particularly more exposed to invasive procedures intrinsic to surgical intervention. They require multiple devices, such as venous and arterial catheters, tubes, drains, and endotracheal tubes, and undergo more interventions during hospitalization. This scenario is conducive to the occurrence of adverse events. The first 72 h of the postoperative period represent the most critical time for the occurrence of adverse events and complications. One study identified the following events: cardiorespiratory arrest, surgical site infection, and the need for surgical reintervention [41].

Adverse events related to central venous catheters (CVCs)

CVC-related adverse events were the most prevalent, accounting for 14% [23, 33, 39]. This finding is significant, as these devices are essential for neonates under these conditions. In this review, the catheters included peripherally inserted central catheters (PICCs) [23, 33], umbilical catheters [33], and other CVCs [33, 39].

The main catheter-related adverse events reported were obstruction (4%), accidental dislodgement (2.4%), and a primary bloodstream infection rate of 3.6 per 1,000 catheter-days [23]. It is noteworthy that several of these adverse events are potentially preventable and may be mitigated through the implementation of evidence-based practices. Additional CVC-related complications included thrombosis (n = 4), catheter malfunction, cardiac tamponade requiring pericardiocentesis, and pleural effusion requiring drainage [33, 39].

Central venous catheter (CVC)–related adverse events were defined as complications associated with catheter insertion, maintenance, or function [23]. These included obstruction, accidental dislodgement, thrombosis, catheter malfunction, cardiac tamponade, and pleural effusion, as reported in the included studies. Importantly, catheter-related bloodstream infections were not included in this category and were analyzed under healthcare-associated infections (HAIs) to avoid overlap between outcomes.

Neonates with congenital heart disease constitute a special population; they are often critically ill and require prolonged intravenous access [23]. The current use of CVCs in patients admitted to pediatric cardiac intensive care units is not well elucidated. One study explored the prevalence and risk factors for thrombosis and primary bloodstream infection, finding that among 17,846 hospitalizations, 69% used at least one CVC, with higher use among younger patients (86% neonates). Significant risk factors for both thrombosis and bloodstream infection were younger age, greater surgical complexity, and the total number of catheter-days. The total number of CVC-days was the only modifiable risk factor identified [42].

Another study that evaluated adverse events related to vascular catheters in neonates showed that PICCs have a low prevalence of primary bloodstream infection; however, adverse events were associated with catheter obstruction and rupture. Regarding CVCs, the highest prevalence was clinical sepsis. The authors highlighted that, to reduce adverse events related to vascular access, it is important to monitor CVCs through a specialized team [43].

One of the actions to reduce CVC-related adverse events is the structuring of a Vascular Access Team (VAT), which can ensure standardization of processes for catheter selection and care, regardless of who manages the device. The implementation of policies addressing indications, insertion, maintenance, and the use of standardized kits is part of a multidisciplinary approach involving nurses, physicians, and the healthcare team. In addition, VATs are responsible for mitigating adverse events, improving patient experience through greater success in access attempts, and helping to reduce financial risks, such as increased length of hospital stay or legal penalties related to hospital-acquired conditions. In summary, the VAT can act as a “guardian,” assessing whether the indication for PICC placement is aligned with established criteria [44, 45].

Healthcare-associated infections (HAIs)

HAIs represented the second most prevalent type of adverse event in this review [20, 21, 23–25, 31, 33, 34, 39]. All infections were considered late-onset, that is, occurring after 48 h of life and associated with care and interventions. Those reported in the studies included unspecified late-onset HAIs [20, 23, 24, 33], HAIs associated with delayed chest closure [25], surgical site infections [34], primary bloodstream infections [39], and septic embolism caused by vascular access [21]. Regardless of their origin, it should be emphasized that these infections are considered preventable [46].

A study investigating the risks of HAIs in neonates with congenital heart disease identified the main risk factors as the presence of a central venous catheter (CVC) and CVC use for more than 14 days; mechanical ventilation for more than 7 days; total parenteral nutrition (TPN) and TPN use for more than 14 days; chest drainage; delayed sternal closure for more than 3 days; cardiac surgical reintervention; and other surgical procedures [47]. HAIs after cardiac surgery in neonates with congenital heart disease result from multiple factors, and continuous surveillance together with the implementation of best practices is essential, as most HAIs are considered preventable. A HAI prevention program promoting early removal of CVCs and mechanical ventilation could reduce HAIs in this population by up to 30% [34, 46, 48].

After the implementation of a bundle for the prevention of primary bloodstream infection, a 59.5% reduction in this type of HAI was observed. The bundle was developed for a multidisciplinary team and included a team responsible for CVC insertion, provision of a bedside CVC kit, staff training and education, and the use of a checklist. In addition, audits and adherence calculations to the checklist were performed, as well as monthly feedback over 12 consecutive months post-intervention. Primary bloodstream infection rates per 1,000 catheter-days, CVC dwell time, and device utilization rates were calculated. A temporal trend toward reduced infection rates suggests that collaborative learning to disseminate information on how to implement changes in practice through processes and protocols can enable rapid improvements in the quality of care in resource-limited settings [49, 50].

Healthcare-associated infections (HAIs) were generally defined as infections occurring after 48 h of hospitalization and associated with healthcare interventions. However, the included studies did not consistently apply standardized definitions, which may impact comparability across studies.

The types of infections reported included primary bloodstream infections (frequently associated with central venous catheters), surgical site infections, infections related to delayed sternal closure, and septic embolism associated with vascular access. Despite heterogeneity in reporting, bloodstream infections appeared to be the most frequently described.

Extubation failure (EF)

Extubation failure was the third most prevalent adverse event in this review, accounting for 11%. It is defined as the need for reintubation within 48 h after extubation [26, 28–30, 35–37].

A retrospective cohort study identified 49 cases (8%) of reintubation, with extubation failure occurring more frequently within 6 h after extubation. Preoperative respiratory support, younger age, and more complex surgical procedures were associated with postoperative extubation failure [51].

Extubation failure after neonatal cardiac surgery has been associated with considerable postoperative morbidity, although data identifying its risk factors remain limited. In a study evaluating the risk of extubation failure in 120 neonates, 21 (17.5%) experienced this event. Factors associated with extubation failure included neonates with genetic anomalies (24% vs. 6%; p = 0.023), hypoplastic left heart syndrome (43% vs. 17%; p = 0.009), delayed sternal closure (38% vs. 12%; p = 0.004), postoperative infection before extubation (38% vs. 11%; p = 0.002), and longer duration of mechanical ventilation (median, 142 vs. 58 h; p = 0.009). Therefore, a more conservative approach may be justified in these patients, including longer duration of antibiotic therapy prior to extubation for postoperative infections [52]. Identifying the risks for extubation failure and promptly assessing post-extubation respiratory issues may help improve outcomes. It is important to emphasize that neonates at risk for extubation failure require a more cautious approach, making multidisciplinary discussion essential to determine the best strategy and optimal timing for extubation.

Medical error

Only one study addressed this type of event using the term “medical error” [27]. Therefore, its analysis was conducted qualitatively and it was not included in the meta-analysis. The objective of the study was to evaluate the frequency and impact of “medical errors” during palliation in patients with hypoplastic left heart syndrome (HLHS). In this context, the term “medical error” is synonymous with adverse events. 23% of adverse events occurred in the postoperative period and were determinants of death and the need for transplantation. Technique-related adverse events accounted for 34% and were associated with delayed recovery and prolonged length of hospital stay (approximately 24 days). Postoperative adverse events during stage I palliation led to a reduction in the overall success of the treatment strategy by approximately 30% of patients. Stage II postoperative adverse events were determinants of death and the need for transplantation.

The adverse events reported in this study included: (1) judgment failure; (2) failures occurring in intensive care and ward settings; (3) extubation failure; (4) delayed response to clinical deterioration; (5) inappropriate discharge from intensive care; (6) delayed readmission to intensive care; (7) premature attempt at sternal closure; (8) catheter- and extracorporeal membrane oxygenation (ECMO)-related failures; and (9) anticoagulation-related failures.

The complexity of neonates with congenital heart disease in the postoperative period increases the risk of adverse events, particularly in intensive care settings. Nevertheless, adverse events remain significant predictors of poor outcomes, which are directly related to postoperative management. A specific concern in the postoperative period is the challenge of identifying clinical changes in such a critical and fragile condition. This is reflected in delays in recognition and management of postoperative clinical scenarios, which constituted a large proportion of the adverse events occurring in the intensive care unit [53, 54].

The challenges in identifying and reducing adverse events in neonates undergoing corrective cardiac surgery are substantial. However, certain strategies, such as fast-track care—which consists of prioritized practices including early postoperative extubation, early removal of vascular access, and initiation of enteral therapy—have been described. In addition, fast-track care has been associated with reduced hemodynamic or respiratory support, improved nutritional status, and shorter intensive care unit length of stay compared with standard care. Improving the quality of care and reducing adverse events require risk management as well as the analysis of indicators and outcomes [55–57].

Limitations

The main limitations of this systematic review and meta-analysis are related to publication bias on the investigated topic and the lack of a standardized method in the literature for identifying adverse events in this population. This limitation impacts the underreporting and classification of adverse events. Thus, the events identified in this review are those that are widely recognized; however, it is likely that many other events are being classified as complications or remain unclassified. Another important limitation is the lack of detailed reporting regarding the type of cardiovascular surgical procedures and the presence of prior percutaneous interventions in the included studies. This prevented subgroup analyses based on surgical complexity or prior interventions, which may influence the occurrence of adverse events.

Conclusions

The findings of this systematic review indicate that adverse events are frequent among neonates with congenital heart disease in the postoperative period following cardiovascular surgery. Most of the studies included were conducted in high-income countries, suggesting that the prevalence and types of adverse events observed may not fully represent the reality in low- and middle-income settings.

The overall prevalence of adverse events identified across the included studies was 14%. These events were grouped into three main subgroups: vascular access–related incidents (14%), healthcare-associated infections (12%), and extubation failure (11%).

Adverse events represent a relevant challenge in the postoperative care of neonates with congenital heart disease. Understanding the risks, prevalence, and potential contributing factors associated with these events may support the development of management strategies, including the implementation of protocols, standardization of practices, and risk-control measures aimed at improving patient safety.

This review highlights the need for more standardized approaches to the detection and reporting of adverse events in this population. Future studies should focus on improving surveillance methods, including the development of specific trigger tools for neonates undergoing cardiovascular surgery, as well as strategies to better analyze, mitigate, and evaluate the impact of these events.

Abbreviations

AE

Adverse Events

CHD

Congenital Heart Disease

CVC

Central Venous Catheter

HAI

Healthcare-Associated Infection

EF

Extubation Failure

ICPS

International Classification for Patient Safety

NOS

Newcastle–Ottawa Scale

PRISMA

Preferred Reporting Items for Systematic Reviews and Meta-Analyses

MOOSE

Meta-analysis Of Observational Studies in Epidemiology

REML

Restricted Maximum Likelihood

CI

Confidence Interval

Authors’ contributions

Conceptualization: E.N., F.R.E.G.Methodology: E.N.; R.A.P., R.C.C.P.S. F.R.E.GFormal analysis: E.N., D.G. R.A.P., F.R.E.G., R.C.C.P.S.Investigation: E.N., D.G. R.A.P., F.R.E.G., R.C.C.P.S.Data curation: E.N., D.G. R.A.P., F.R.E.G., R.C.C.P.S.Writing – original draft: R.A.P., E.N.Writing – review & editing: R.A.P., F.R.E.G., R.C.C.P.S.Supervision: F.R.E.G. R.C.C.P.S.All authors read and approved the final manuscript.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

All data generated or analyzed during this study are included in this published article and its supplementary information files. The supplementary materials include the complete search strategies for all databases and the extracted data used in the systematic review and meta-analysis.

Declarations

Ethics approval and consent to participate

This study is a systematic review and meta-analysis based on previously published data. Therefore, ethical approval and consent to participate were not required.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.