Partial upper sternotomy vs full median sternotomy in obese patients undergoing aortic valve replacement: A meta-analysis

Abstract

BACKGROUND

Obese patients (body mass index ≥ 30 kg/m²) undergoing isolated aortic valve replacement (AVR) face increased surgical risks due to comorbidities. Partial upper sternotomy (PUS), a minimally invasive approach, may reduce complications compared to full median sternotomy (FMS). We hypothesize that PUS improves outcomes over FMS in obese patients undergoing AVR.

AIM

To compare the efficacy and safety of PUS vs FMS in obese patients undergoing isolated AVR.

METHODS

This systematic review and meta-analysis followed PRISMA guidelines, searching PubMed, EMBASE, and Cochrane databases for observational studies comparing PUS vs FMS in obese patients undergoing AVR. Outcomes were analyzed using odds ratios (OR), mean differences (MD), 95% confidence intervals (CI), I² statistic, and Newcastle-Ottawa Scale was used for quality assessment.

RESULTS

Four observational studies involving 677 patients were analyzed. PUS reduced intensive care unit stay (MD -2.67 days, 95%CI: -4.43 to -0.90, P = 0.003, I² = 78%) but increased cardiopulmonary bypass time (MD 5.62 minutes, 95%CI: -0.36 to 11.59, I² = 55%). No differences were observed in renal failure (OR 1.13, 95%CI: 0.63-2.94, I² = 0%), atrial fibrillation (OR 0.81, 95%CI: 0.43-1.54, I² = 30%), reexploration (OR 1.09, 95%CI: 0.48-2.47, I² = 0%), postoperative bleeding (OR 1.48, 95%CI: 0.53-4.15, I² = 60%), wound infection (OR 1.23, 95%CI: 0.70-2.14, I² = 0%), hospital stay (MD 0.51 days, 95%CI: -4.13 to 5.15, I² = 90%), or cross-clamp time (MD 4.03 minutes, 95%CI: -0.75 to 8.80, I² = 50%).

CONCLUSION

PUS is safe and effective for obese patients undergoing AVR, reducing intensive care unit stay and enhancing recovery, provided surgical expertise is available.

Key Words: Partial upper sternotomy; Full median sternotomy; Aortic valve replacement; Obesity; Intensive care unit stay; Cardiopulmonary bypass time; Minimally invasive surgery; Postoperative complications

Core Tip: This meta-analysis innovatively demonstrates that partial upper sternotomy (PUS) is a safe and effective alternative to full median sternotomy for obese patients (body mass index ≥ 30 kg/m²) undergoing aortic valve replacement, uniquely reducing intensive care unit stay by approximately 2.67 days. Despite a trend toward longer cardiopulmonary bypass times, PUS maintains equivalent safety profiles across major complications, offering a compelling minimally invasive option to enhance recovery in this high-risk population.

INTRODUCTION

Aortic valve disease, including aortic stenosis and regurgitation, significantly contributes to cardiovascular morbidity and mortality, often impairing left ventricular function and leading to heart failure[1]. Aortic valve replacement (AVR) is the definitive treatment for severe cases, with demand rising due to an aging population, where approximately 3% of individuals aged 75 or older develop aortic stenosis[2]. Obesity, defined as a body mass index (BMI) ≥ 30 kg/m², complicates AVR by increasing perioperative risks and postoperative complications[3]. The impact of obesity on AVR outcomes remains debated, with some studies suggesting an “obesity paradox” of reduced postoperative risks[4], while others report increased complications, such as sternal wound infections and prolonged hospital stays[5]. These conflicting findings underscore the need to identify optimal surgical approaches for obese patients undergoing AVR[6].

Full median sternotomy (FMS) is the standard approach for AVR, offering comprehensive cardiac access but posing challenges in obese patients, including higher risks of sternal wound infections and prolonged ventilation[7]. Since the 1990s, minimally invasive partial upper sternotomy (PUS) has emerged as an alternative, utilizing a smaller incision to reduce surgical trauma, blood loss, and recovery time, which may particularly benefit obese patients prone to complications[8-12]. Evidence suggests PUS provides comparable safety to FMS while reducing postoperative complications, making it a promising option for obese AVR patients[13,14].

This meta-analysis addresses the gap in understanding the optimal surgical strategy for obese patients undergoing AVR by comparing the efficacy, safety, and postoperative outcomes of PUS vs FMS. By synthesizing existing evidence, this study seeks to clarify the benefits and limitations of each approach to inform clinical decision-making and guide future research.

MATERIALS AND METHODS

This systematic review and meta-analysis adhered to the PRISMA statement and Cochrane Collaboration guidelines[15,16]. Institutional Review Board approval was not required, as data was publicly available from electronic databases.

Eligibility criteria

Inclusion criteria: This review included observational studies (e.g., cohort or case-control studies) of adult patients (aged ≥ 18 years) with obesity (defined as BMI ≥ 30 kg/m²) undergoing isolated AVR via PUS compared to FMS. No randomized controlled trials (RCTs) were identified. Eligible studies reported at least one of the following outcomes of interest: Operative time, cardiopulmonary bypass time, cross-clamp time, intraoperative blood loss, intensive care unit (ICU) stay duration, hospital stay duration, complication rates (e.g., renal failure, atrial fibrillation, reexploration for bleeding, postoperative bleeding, wound infection), and mortality. Studies were limited to English-language, full-text publications from database inception to April 2025, to reflect contemporary surgical techniques.

Exclusion criteria: Excluded studies included case reports, reviews, editorials, letters, and those involving pediatric populations (< 18 years), non-clinical outcomes, or procedures other than isolated AVR (e.g., coronary artery bypass grafting, combined procedures, or non-sternotomy approaches like thoracotomy). Studies lacking comparative data between PUS and FMS, sufficient details on clinical outcomes, or outcomes specific to obese patients were also excluded.

Search strategy and information sources

A comprehensive literature search was conducted across PubMed, EMBASE, and Cochrane databases, covering studies published from database inception to April 2025. The key terms included "aortic valve replacement", "AVR", "partial upper sternotomy", "PUS", "full median sternotomy", "FMS", "minimally invasive", "obesity", "obese", and "BMI". These terms were combined with outcome-specific keywords, such as "renal failure", "atrial fibrillation", "reexploration", "postoperative bleeding", "wound infection", "hospital stay", "ICU stay", "cardiopulmonary bypass time", and "cross-clamp time". Search terms were carefully selected to maximize study retrieval and ensure comprehensive capture of relevant literature on PUS vs FMS in obese AVR patients. Limitations of the search strategy included restriction to English-language publications and exclusion of gray literature (e.g., conference abstracts, theses), which may introduce language bias and miss unpublished data.

Study selection

A comprehensive literature search was conducted across PubMed, EMBASE, and Scopus to identify studies comparing PUS with FMS in obese adults. References were managed using Zotero for duplicate removal. Two independent reviewers screened titles and abstracts, followed by full-text assessments, based on predefined criteria. Discrepancies were resolved through discussion or consultation with a third reviewer. Studies included if they reported perioperative or clinical outcomes in obese patients undergoing AVR via PUS or FMS.

Data extraction and collection

Data were extracted independently by two reviewers (Gupta A and Chikhradze T) using a standardized, pre-designed Google Sheet to ensure consistency. Prior to extraction, reviewers underwent training to align on variable definitions and data entry protocols. Extracted data included study characteristics, patient demographics, surgical approaches, and clinical outcomes (e.g., operative time, intraoperative blood loss, ICU and hospital stay duration, complication rates, mortality). Discrepancies were resolved through consensus discussions, with a third reviewer consulted if agreement could not be reached. Authors were contacted for clarification when needed. For outcomes reported as medians with interquartile ranges, validated methods were used to convert these to means and standard deviations. A random sample of 20% of extracted data was cross-checked by an independent reviewer to verify accuracy and enhance reproducibility.

Quality assessment

The Newcastle-Ottawa Scale (NOS) was used to evaluate study quality[17]. Two independent authors (Sakrani RA and Khan Z) assessed each study across three domains: Selection of study groups (up to 4 points, evaluating representativeness of exposed cohort, selection of non-exposed cohort, ascertainment of exposure, and demonstration that outcome was not present at start); comparability of groups (up to 2 points, based on control for key confounders such as age, sex, and comorbidities); and outcome assessment (up to 3 points, evaluating adequacy of follow-up, completeness of follow-up, and independent blind assessment or record linkage). Total scores ranged from 0 (worst) to 9 (best), with classifications as good (≥ 7, indicating low risk of bias with strong representativeness, comparability, and outcome assessment), moderate (≥ 5 to < 7, indicating some limitations in one or more domains but overall acceptable quality), or poor (≤ 5, indicating high risk of bias due to major flaws in selection, comparability, or outcome). Summary scores for included studies are detailed in the Results section, with justifications provided for each classification based on domain-specific performance.

Statistical analysis

This meta-analysis followed PRISMA guidelines, analyzing dichotomous and continuous outcomes from cohort studies. Dichotomous outcomes were reported as odds ratios (OR) with 95% confidence intervals (CI), and continuous outcomes as mean differences (MDs) with 95%CI, based on measurement scale consistency. Heterogeneity was assessed using the I² statistic. The choice of model was based on heterogeneity levels: A fixed-effects model was applied for low heterogeneity (I² ≤ 50%), assuming studies estimate the same underlying effect; a random-effects model was used for moderate to high heterogeneity (I² > 50%) to account for between-study variability, providing more conservative estimates in the presence of differences in study populations or methods. Analyses were performed using RevMan version 5.4, with significance defined as P < 0.05. For each outcome, a forest plot was constructed to visually analyze the data and funnel plots were generated to check the publication bias. Sensitivity analyses, including leave-one-out approaches, were conducted to assess the robustness of findings by evaluating the impact of individual studies on overall heterogeneity, pooled estimates, and model choice; this helped identify influential studies and potential sources of bias. Following the statistical analysis, the quality of evidence for each outcome was evaluated using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach, assessing domains such as risk of bias, inconsistency, imprecision, indirectness, and effect size to determine the certainty of evidence.

RESULTS

Included studies

The PRISMA flow chart (Figure 1) outlines the literature screening, study selection, and exclusion criteria. The initial search yielded 430 articles, with 36 full-text articles assessed for eligibility. Four observational studies[12,18-20] met the criteria and were included in both qualitative and quantitative meta-analyses.

Figure 1  PRISMA flowchart outlining the literature screening process, study selection, and exclusion criteria.

Study characteristics

The four studies conducted between 2014 and 2023, compared PUS and FMS in obese adults undergoing isolated AVR. Mean ages ranged from 54.0-69.8 years for PUS and 55.9-70.0 years for FMS. Male patients comprised 46.0%-78.6% of PUS and 54.8%-79.5% of FMS groups. Mean BMI ranged from 30.56-38.3 kg/m² for PUS and 30.11-35.5 kg/m² for FMS. Diabetes mellitus prevalence was 6.0%-28.6% for PUS and 8.0%-38.5% for FMS. Hypertension affected 52.0%-90.0% of PUS and 62.0%-95.6% of FMS patients. Smoking, reported in three studies, ranged from 28.6%-29.4% for PUS and 26.4%-35.2% for FMS. Chronic obstructive pulmonary disease prevalence was 4.0%-9.5% for PUS and 5.0%-8.8% for FMS. Dyslipidemia, reported in two studies, affected 46.2%-90.0% of PUS and 48.4%-89.8% of FMS patients. Previous stroke, reported in two studies, occurred in 2.2%-6.3% of PUS and 4.4%-14.3% of FMS patients. Mean left ventricular ejection fraction ranged from 55.16%-62.5% for PUS and 58.33%-61.7% for FMS. Hemoglobin levels were 12.69-13.71 g/dL for PUS and 12.22-13.5 g/dL for FMS. Creatinine levels ranged from 1.12-1.68 mg/dL for PUS and 1.22-1.85 mg/dL for FMS. Hematocrit, reported in two studies, was 40.50%-41.57% for PUS and 38.32%-41.20% for FMS. Baseline characteristics were comparable, with minor variations in sex, diabetes, and stroke prevalence, supporting valid meta-analysis comparisons (Table 1).

Table 1 Detailed characteristics of each included studies and baseline characteristics.

Ref.	Year	Age (mean ± SD)		Male (%)		BMI (kg/m2)		Diabetes mellitus (%)		Hypertension (%)		Smoking (%)		COPD (%)		Peripheral vascular disease (%)		LVEF		Hemoglobin (mean ± SD)		Creatinine (mean ± SD)		Hematocrit (mean ± SD)		Dyslipidemia (%)		Previous stroke (%)
		PUS	FS	PUS	FS	PUS	FS	PUS	FS	PUS	FS	PUS	FS	PUS	FS	PUS	FS	PUS	FS	PUS	FS	PUS	FS	PUS	FS	PUS	FS	PUS	FS
Cammertoni et al[18]	2024	68.6 ± 14.5	69.2 ± 13.1	60.4	59.3	32.9 ± 1.3	32.6 ± 1.2	19.8	24.2	83.5	82.4	28.6	26.4	4.4	7.7	5.5	7.7	61.3 ± 6.2	60.1 ± 5.5	13.2 ± 1.4	13.5 ± 1.3	NA	NA	NA	NA	46.2	48.4	2.2	4.4
Luo et al[12]	2022	54.0 ± 11.5	55.9 ± 10.8	78.6	79.5	30.56 ± 3.09	30.11 ± 3.04	17.1	18.2	90	89.8	N/A	N/A	8.6	8	N/A	N/A	62.5 ± 10.2	61.7 ± 8.9	12.69 ± 2.01	12.66 ± 2.88	120.5 ± 88.5	119.6 ± 98.6	41.57 ± 3.88	40.94 ± 3.87	90	89.8	NA	NA
Welp et al[19]	2018	69.8 ± 10.4	70.0 ± 10.5	56	54.8	32.56 ± 3.09	33.11 ± 3.04	28.6	38.5	84.1	95.6	29.4	35.2	9.5	8.8	7.1	7.7	55.16 ± 14.90	58.33 ± 10.82	13.71 ± 1.48	13.39 ± 1.85	1.12 ± 0.77	1.22 ± 0.81	40.50 ± 4.81	41.20 ± 3.82	59.5	67	6.3	14.3
Xie et al[20]	2022	62.9 ± 10.7	57.8 ± 12.3	46	58	38.3 ± 5.3	35.5 ± 4.2	6	8	52	62	29	34	4	5	NA	NA	57 ± 11.31	58.33 ± 11.29	13.68 ± 1.68	12.22 ± 1.35	NA	NA	41.32 ± 3.68	38.32 ± 4.68	NA	NA	NA	NA

PUS: Partial upper sternotomy; FS: Full sternotomy; BMI: Body mass index; COPD: Chronic obstructive pulmonary disease; LVEF: Left ventricular ejection fraction; N/A: Not applicable; NA: Not available.

Quality assessment

Study quality was assessed using the NOS. Three studies were rated as good quality (scores 8/9), and one was rated moderate (score 7/9). All studies met the threshold for good quality (≥ 7), with consistent patient selection, comparability, and outcome ascertainment. Discrepancies between assessors were resolved by consensus (Supplementary Table 1).

Clinical outcomes

Outcomes are presented below under separate subheadings for clarity, with effect sizes (ORs or MDs) and 95%CI reported for each. Forest plots (Figures 1, 2 and 3) visually depict results, and Supplementary Table 2 summarizes GRADE assessments.

Figure 2 Forest plots comparing outcomes between partial upper sternotomy and full sternotomy in obese patients undergoing aortic valve replacement. A: Renal failure; B: Atrial fibrillation; C: Reexploration. PUS: Partial upper sternotomy; FS: Full sternotomy.

Figure 3 Forest plots comparing outcomes between partial upper sternotomy and full sternotomy in obese patients undergoing aortic valve replacement. A: Postoperative bleeding; B: Wound infection. PUS: Partial upper sternotomy; FS: Full sternotomy.

Renal failure: Four studies reported renal failure outcomes. No significant difference was observed between the PUS and full sternotomy (FS) groups. The pooled OR was 1.13 (95%CI: 0.63-2.94; P = 0.69), with no heterogeneity (I² = 0%) (Figure 2A). The lack of difference suggests comparable renal safety between PUS and FS, though the wide CI indicates limited precision, reducing clinical confidence in this finding.

Atrial fibrillation: Two studies assessed atrial fibrillation. No significant difference was found between the PUS and FS groups. The pooled OR was 0.81 (95%CI: 0.43-1.54; P = 0.84), with no heterogeneity (I² = 0%) (Figure 2B). This finding indicates similar arrhythmic risk, but the small number of studies and wide CI limit its clinical applicability.

Reexploration: Four studies reported reexploration rates. No significant difference was observed between the PUS and FS groups. The pooled OR was 1.09 (95%CI: 0.48-2.47; P = 0.84), with no heterogeneity (I² = 0%) (Figure 2C). The comparable reexploration rates suggest equivalent surgical precision, though the wide CI warrants cautious interpretation.

Postoperative bleeding: Three studies evaluated postoperative bleeding. No significant difference was found between the PUS and FS groups. The pooled OR was 1.48 (95%CI: 0.53-4.15; P = 0.45), with no heterogeneity (I² = 0%) (Figure 3A). This suggests similar bleeding risks, but the wide CI indicates uncertainty in clinical decision-making.

Wound infection: Four studies reported wound infection outcomes. No significant difference was observed between the PUS and FS groups. The pooled OR was 1.23 (95%CI: 0.70-2.14; P = 0.47), with no heterogeneity (I² = 0%) (Figure 3B). The lack of difference is notable given the expected benefit of PUS in reducing infections in obese patients, though the CI suggests further research is needed.

Cardiopulmonary bypass time: Four studies reported cardiopulmonary bypass time. No significant difference was found between the PUS and FS groups. The pooled MD was 5.62 minutes (95%CI: -0.36 to 11.59; P = 0.07), with moderate heterogeneity (I² = 55%) that dropped significantly (I² = 22%) upon removal of Welp et al[19] (Figure 4A). Heterogeneity may stem from variations in surgical expertise or patient comorbidities (e.g., diabetes severity). Subgroup analyses by study design or patient BMI were considered but not feasible due to limited study numbers; meta-regression was not performed due to insufficient covariates. The small MD suggests minimal clinical impact on operative efficiency.

Figure 4 Forest plots comparing outcomes between partial upper sternotomy and full sternotomy in obese patients undergoing aortic valve replacement. A: Cardiopulmonary bypass time; B: Intensive care unit stay; C: Hospital stay; D: Cross-clamp time. PUS: Partial upper sternotomy; FS: Full sternotomy; ICU: Intensive care unit.

ICU stay: Six studies assessed ICU stay duration. The PUS was associated with a significantly shorter ICU stay compared to the FS group. The pooled MD was -2.67 days (95%CI: -4.43 to -0.90; P = 0.003), with considerable heterogeneity (I² = 78%) that reduced significantly (I² = 62%) upon removal of Welp et al[19] (Figure 4B). Heterogeneity may arise from differences in ICU protocols or patient characteristics (e.g., age, obesity severity). Subgroup analyses by study quality or hospital setting were planned but not conducted due to limited data; meta-regression was infeasible. The 2.67-day reduction is clinically significant, potentially reducing resource use and improving patient recovery in obese AVR patients.

Hospital stay: Four studies reported hospital stay duration. No significant difference was observed between the PUS and FS groups. The pooled MD was 0.51 days (95%CI: -4.13 to 5.15; P = 0.83), with considerable heterogeneity (I² = 90%) that reduced significantly (I² = 68%) upon removal of Welp et al[19] (Figure 4C). Heterogeneity likely reflects variations in discharge criteria or postoperative care across centers. Subgroup analyses or meta-regression were not performed due to insufficient study numbers and covariate data. The lack of difference suggests limited clinical impact on overall hospital resource use.

Cross-clamp time

Four studies evaluated cross-clamp time. No significant difference was found between the PUS and FS groups. The pooled MD was 4.03 minutes (95%CI: -0.75 to 8.80; P = 0.10), with considerable heterogeneity (I² = 73%) that reduced significantly (I² = 0%) upon removal of Cammertoni et al[18] (Figure 4D). Heterogeneity may be due to differences in surgical techniques or patient complexity. Subgroup analyses and meta-regression were considered but not feasible due to limited data. The small MD indicates negligible clinical impact on operative duration.

Publication bias

Publication bias was assessed using funnel plots for all outcomes. No evidence of publication bias was observed for renal failure, atrial fibrillation, reexploration, post-operative bleeding, and wound infection. Cardiopulmonary bypass time, ICU stay, hospital stay, and cross-clamp time showed evidence of publication bias, indicated by asymmetrical plots with gaps in the lower left quadrant, suggesting potential underrepresentation of smaller studies with non-significant or negative results (Supplementary Figures 1 and 2).

GRADE assessment

The GRADE approach evaluated evidence certainty for clinical outcomes (Supplementary Table 2). Based on four observational studies (two for atrial fibrillation, three for postoperative bleeding), certainty started at low. Low risk of bias was noted (NOS scores 7-9/9). Renal failure, atrial fibrillation, reexploration, postoperative bleeding, and wound infection had no heterogeneity (I² = 0%), no publication bias, but serious imprecision (wide CIs, e.g., OR 1.13, 95%CI: 0.63-2.94 for renal failure), yielding low certainty. Cardiopulmonary bypass time (MD 5.62 minutes, I² = 55%), hospital stay (MD 0.51 days, I² = 90%), and cross-clamp time (MD 4.03 minutes, I² = 73%) showed serious inconsistency, imprecision, and publication bias, resulting in very low certainty. ICU stay (MD -2.67 days, I² = 78%) had serious inconsistency and publication bias but no imprecision, maintaining low certainty. No outcomes were downgraded for indirectness.

DISCUSSION

This meta-analysis of four observational studies compared PUS and FMS in obese adults undergoing isolated AVR. PUS showed comparable safety to FMS for renal failure, atrial fibrillation, reexploration, postoperative bleeding, and wound infection (all I² = 0%), as well as hospital stay (I² = 90%) and cross-clamp time (I² = 73%), with no significant differences. However, PUS significantly reduced ICU stay (MD -2.67 days, P = 0.003, I² = 78%) but showed a trend toward longer cardiopulmonary bypass time (MD 5.62 minutes, P = 0.07, I² = 55%).

The equivalence in most outcomes suggests PUS is as safe as FMS for obese AVR patients. Studies like Eqbal et al[21] and Khalid et al[22] report similar complication rates for minimally invasive and conventional sternotomy in AVR, even in non-obese populations. Santana et al[23] found no differences in major complications for obese patients undergoing minimally invasive AVR, supporting these findings. In the context of existing literature, these results align with broader evidence indicating that minimally invasive techniques like PUS offer benefits in reducing surgical trauma without compromising safety, particularly in high-risk groups such as obese patients where traditional FMS may exacerbate wound healing issues due to excess adipose tissue. Low heterogeneity for renal failure, atrial fibrillation, reexploration, postoperative bleeding, and wound infection indicates consistent results, likely due to standardized surgical protocols and comparable patient characteristics (e.g., hypertension prevalence: 52.0%-95.6%). Comorbidities (e.g., diabetes: 6.0%-38.5%) and postoperative care may outweigh surgical approach effects. For hospital stay, high heterogeneity (I² = 90%) may reflect non-clinical factors like bed availability, as noted in prior research[24]. Sensitivity analysis revealed that removing Welp et al[19] reduced heterogeneity to I² = 68%, suggesting that differences in hospital discharge protocols or patient management in that study may have contributed to variability. This heterogeneity complicates direct comparisons of hospital stay, indicating that local practices may influence this outcome more than the surgical approach itself.

The shorter ICU stay with PUS is a key advantage, particularly for obese patients prone to prolonged recovery. Wilbring et al[25] and Olds et al[26] attribute reduced ICU stays with minimally invasive approaches to less surgical trauma and faster mobilization. These align with our results, suggesting PUS may optimize ICU resource use. Considerable heterogeneity (I² = 78%) may arise from variations in ICU protocols, obesity severity (BMI: 30.56-38.3 kg/m²), or surgical expertise. Sensitivity analysis showed that removing Welp et al[19] reduced heterogeneity to I² = 62%, indicating that unique factors in that study, such as ICU management practices or patient comorbidities, likely contributed to variability. Differences in postoperative ventilation, pain management, or criteria for ICU discharge, as suggested by Snell et al[27], could also drive heterogeneity. This suggests that the ICU stay benefit of PUS may vary across centers, and results should be interpreted cautiously in settings with differing ICU protocols. Smaller incisions likely reduce incisional pain and improve respiratory function, benefiting obese patients with high comorbidity burdens.

The trend toward longer cardiopulmonary bypass time with PUS reflects technical challenges in obese patients. Ilcheva et al[28] note that minimally invasive approaches often extend bypass times due to limited exposure and complex anatomy. In obese patients, adipose tissue can displace the heart and aortic valve, requiring incision adjustments, as reported by D'Oria et al[29]. Moderate heterogeneity (I² = 55%) may stem from differences in surgeon experience or institutional protocols. Sensitivity analysis indicated that removing Welp et al[19] lowered heterogeneity to I² = 22%, suggesting that specific surgical techniques or equipment used in that study may have influenced bypass times. This heterogeneity implies that the extent of bypass time differences may depend on center-specific factors, and the modest increase observed may not universally apply. Despite concerns that prolonged bypass times increase risks like acute kidney injury or atrial fibrillation[30], our findings show no such complications, indicating the modest increase is clinically acceptable given the ICU stay reduction.

Cross-clamp time also showed considerable heterogeneity (I² = 73%), which sensitivity analysis traced to Cammertoni et al[18], as its removal reduced heterogeneity to I² = 0%. This suggests that study-specific factors, such as variations in surgical team expertise, aortic valve pathology, or intraoperative techniques, may have driven variability. This heterogeneity indicates that cross-clamp time differences between PUS and FMS may not be consistent across all settings, and results should be interpreted with caution in contexts with differing surgical practices.

The benefits of PUS include reduced ICU stay, potentially leading to faster recovery, lower healthcare costs, and decreased exposure to ICU-related risks like infections in obese patients. Limitations of PUS encompass technical challenges that may prolong operative times, requiring experienced surgeons to mitigate risks, and potential variability in outcomes due to institutional differences.

Clinical implications

The comparable safety of PUS and FMS supports their use in obese AVR patients. The shorter ICU stay with PUS suggests benefits in recovery and resource utilization, making it preferable where technical expertise exists. The reduced ICU stay (MD -2.67 days, P = 0.003) accelerates recovery for obese patients, who are at higher risk of prolonged hospitalization due to comorbidities. This can lower ICU-related complications and enhance mobility, while also optimizing hospital resources by freeing up critical care beds. Skilled surgical teams are essential to manage the modest increase in bypass time (MD 5.62 minutes, P = 0.07). High heterogeneity in ICU stay, hospital stay, bypass time, and cross-clamp time underscores the influence of center-specific factors like ICU protocols, discharge criteria, and surgical expertise. Standardized perioperative protocols (e.g., incision techniques, perfusion strategies, ICU discharge criteria) could reduce variability and strengthen the reliability of PUS benefits. Future research should prioritize larger, multicenter studies controlling for confounders (e.g., obesity severity, surgeon experience) and explore long-term outcomes like quality of life, valve durability, and cost-effectiveness.

Limitations

This meta-analysis has limitations. Only four observational studies were included, limiting statistical power and generalizability. Observational designs are prone to selection bias and confounding, despite NOS ratings of good to moderate quality. Considerable heterogeneity for ICU stay (I² = 78%), hospital stay (I² = 90%), bypass time (I² = 55%), and cross-clamp time (I² = 73%) likely stems from differences in surgical expertise, patient characteristics (e.g., BMI: 30.56-38.3 kg/m²), and perioperative protocols, as confirmed by sensitivity analyses identifying Welp et al[19] and Cammertoni et al[18] as key contributors. This heterogeneity limits the precision of pooled estimates for these outcomes and suggests that findings may not be uniformly applicable across all settings. Long-term outcomes (e.g., quality of life, valve durability) were not reported, restricting analysis to short-term results. Findings are specific to obese patients undergoing isolated AVR and may not apply to non-obese populations or combined procedures.

Future research

To address current limitations, future research should include larger sample sizes to enhance statistical power and reduce heterogeneity. RCTs are recommended to minimize selection bias and confounding inherent in observational designs. Additionally, studies with long-term follow-up (e.g., 5-10 years) are essential to evaluate outcomes such as valve durability, quality of life, and late complications. Multicenter collaborations could improve generalizability by incorporating diverse patient populations and surgical practices. Exploring subgroup analyses by obesity severity (e.g., class I vs class III) and surgeon experience, as well as cost-effectiveness evaluations, would further inform clinical decision-making.

CONCLUSION

PUS is a safe and effective alternative to FMS for obese patients undergoing isolated AVR, significantly reducing ICU stay (MD -2.67 days, P = 0.003) while maintaining comparable complication rates, despite a slight increase in cardiopulmonary bypass time. These key findings highlight the clinical significance of PUS in shortening critical care duration, which can mitigate risks associated with prolonged ICU exposure in obese patients, such as ventilator-associated pneumonia and deconditioning. This shorter ICU stay enhances patient recovery and optimizes hospital resource utilization, making PUS a valuable option in centers with surgical expertise. Standardized protocols are needed to address heterogeneity in outcomes and ensure consistent benefits, with future research focusing on long-term outcomes to support broader adoption in clinical practice.