Safety and efficacy of Da Vinci robot-assisted atrial septal defect repair in patients with different body mass index levels: a single-center retrospective analysis

Abstract

Objective

This study aims to investigate the impacts of various body mass index (BMI) classifications on perioperative parameters and short-term outcomes in patients undergoing Da Vinci robot-assisted atrial septal defect (ASD) repair providing evidence for personalized perioperative management.

Methods

A retrospective analysis was conducted on 100 patients who underwent Da Vinci robot-assisted ASD repair at Qingdao University Affiliated Hospital between November 2014 and December 2025. Based on the Chinese adult BMI classification criteria, the patients were categorized into four groups: underweight group (BMI < 18.5 kg/m², n = 9), normal weight group (18.5 ≤ BMI < 24.9 kg/m², n = 52), overweight group (25.0 ≤ BMI < 29.9 kg/m², n = 26), and obese group (BMI ≥ 30.0 kg/m², n = 13). This study compared baseline characteristics, intraoperative indicators, postoperative recovery, and short-term prognosis across these groups. Furthermore, multivariate logistic regression analyzed the independent associations of BMI classification and age with postoperative surgical site infection and hospital readmission.

Results

No significant differences were observed among the four groups concerning baseline echocardiographic parameters, key intraoperative metrics (e.g., operation duration, cardiopulmonary bypass time), or most postoperative recovery measures (e.g., mechanical ventilation duration, length of intensive care unit stay), with all P > 0.05. No sternotomy conversion or early mortality occurred. Univariate analysis showed the obese group had significantly higher rates of surgical site infection (15.4%) and readmission (23.1%) (all P < 0.05), but multivariate regression did not identify BMI classification or age as independent predictors (all P > 0.05).

Conclusion

Da Vinci robot-assisted ASD repair is safe and effective across different BMI levels, with its minimally invasive nature appearing to mitigate weight-related surgical challenges within this cohort. Although obese patients face higher risks of postoperative infection and readmission, BMI was not found to be an independent influencing factor in this exploratory analysis; however, small subgroup sizes and sparse data limit the statistical reliability of this finding. Individualized enhanced perioperative management for obese populations is recommended, and weight status should not be a contraindication for this procedure.

Introduction

Atrial septal defect (ASD) represents a prevalent form of congenital heart disease, constituting approximately 13% of all such conditions [1]. In severe instances, prolonged hemodynamic abnormalities may lead to complications such as pulmonary hypertension and heart failure, necessitating surgical intervention to enhance patient prognosis [2]. The progression of minimally invasive cardiac surgery techniques has facilitated the adoption of Da Vinci robot-assisted surgery as a significant modality for ASD repair [3, 4]. This approach capitalizes on advantages such as three-dimensional visualization and the dexterity of robotic arm manipulation [5]. Its efficacy in minimizing surgical trauma, expediting recovery, and reducing postoperative complication rates has been extensively corroborated [6, 7].

Body mass index (BMI) serves as a fundamental indicator that elucidates the relationship between body weight and height, and it is intricately linked to metabolic status and anatomical structure [8]. Its influence on the efficacy and prognosis of cardiac surgery has been a longstanding subject of clinical investigation [9]. Prior research has suggested that both obesity and underweight conditions may impact the perioperative safety and recovery processes of cardiac surgery by modifying thoracic anatomical structures, increasing surgical complexity, and compromising postoperative wound healing and immune function [10, 11]. Nevertheless, comprehensive clinical research examining the effects of various BMI classifications on perioperative outcomes and short-term prognosis in Da Vinci robot-assisted ASD repair remains sparse, with inconsistent findings [12]. Some studies indicate that elevated BMI is associated with prolonged operation durations and heightened complication risks, whereas others do not identify a significant correlation between BMI and surgical outcomes [12, 13]. Furthermore, there is a pressing need for high-quality clinical evidence to inform individualized surgical strategies and perioperative management protocols for patients with atypical weight profiles, such as those who are underweight or obese.

Building upon this premise, the current study conducted a retrospective analysis of clinical data from 100 patients who underwent Da Vinci robot-assisted ASD repair. The objective was to investigate the impact of varying BMI classifications (underweight, normal weight, overweight, and obese) on patients’ baseline characteristics, intraoperative indicators, postoperative recovery, and short-term outcomes. The study seeks to elucidate the relationship between BMI and the safety and efficacy of surgical procedures, thereby offering reliable insights for the development of individualized perioperative management strategies and the optimization of surgical decision-making.

Patients and methods

This study was conducted in compliance with the principles outlined in the Declaration of Helsinki (2013 revision) and received approval from the Institutional Ethics Committee of Qingdao University Affiliated Hospital. Due to the retrospective design of the study, the requirement for individual informed consent was waived.

Study population

A retrospective study was conducted involving 100 patients who underwent Da Vinci robot-assisted ASD repair at our hospital between November 2014 and December 2025, with comprehensive clinical data available for all patients. The inclusion criteria were as follows: (1) ASD diagnosis confirmed via echocardiography, with indications for surgical repair; (2) no history of prior cardiac surgery; and (3) availability of complete clinical data, encompassing baseline characteristics, intraoperative records, and postoperative follow-up information. The exclusion criteria were (1) presence of other complex congenital heart diseases, severe valvular heart disease, or coronary artery disease; (2) preoperative severe complications, including severe infection, liver or kidney failure, or coagulation disorders; and (3) extreme BMI values, defined as less than 15 kg/m² or greater than 50 kg/m².

The patients were stratified into four categories based on the Chinese adult BMI classification criteria and clinical practice guidelines: underweight group (BMI < 18.5 kg/m²), normal weight group (18.5 ≤ BMI < 24.9 kg/m²), overweight group (25.0 ≤ BMI < 29.9 kg/m²), and obese group (BMI ≥ 30.0 kg/m²).

This four-group stratification was employed for exploratory comparative analysis in line with clinical guidelines. Given the total sample size of 100 patients, subgroup numbers were relatively small (especially underweight and obese groups), which may limit statistical power and stability.

Surgical methods

All surgical procedures were conducted by a consistent team of experienced cardiac surgeons utilizing the Da Vinci Si or Xi robotic surgical systems. Following the successful induction of general anesthesia, a central venous catheter was inserted via percutaneous puncture of the right internal jugular vein, and a 16-Fr venous catheter was placed as the superior vena cava drainage tube. Cardiopulmonary bypass (CPB) was initiated through femoral arteriovenous cannulation. A primary operating port and surgical lens entry site were established via a 3–5 cm incision at the midclavicular line of the 4th intercostal space. Additional 5 mm incisions were made at the anterior axillary line of the 2nd and 6th intercostal spaces, as well as the parasternal region of the 5th intercostal space, to accommodate the insertion of the left and right robotic arms and the atrial retractor. Following pericardiotomy, tapes were placed around the superior and inferior vena cava for standby use.

A purse-string suture was initially placed at the aortic root, followed by the insertion of a catheter for the delivery of cardioplegic solution. Subsequently, a cross-clamp was applied to the ascending aorta to obstruct blood flow, and cardioplegic solution was administered to induce cardiac arrest. Thereafter, the superior and inferior vena cava were clamped, and an incision was made in the right atrium. Closure of the ASD was performed either through direct suturing or by using a pericardial patch, contingent upon the size and condition of the defect’s edges. Following the completion of suturing, a thorough evacuation of air was conducted, the cross-clamp on the ascending aorta was released, and the incision in the right atrium was sutured after the resumption of cardiac activity. In certain cases, ASD repair was conducted under beating heart conditions, depending on the patient’s clinical status and anatomical features.

After removing CPB catheters, thorough hemostasis was performed, followed by layered chest closure.

Statistical methods

Data analysis was conducted using SPSS version 26.0. For normally distributed continuous variables, data were represented as mean ± standard deviation, and intergroup comparisons were made using one-way analysis of variance (ANOVA). For non-normally distributed continuous variables, data were expressed as median (interquartile range) [M (Q1, Q3)], and intergroup comparisons were performed using the Kruskal-Wallis H test. Categorical variables were presented as frequencies and percentages (n [%]), with intergroup comparisons conducted using either the χ² test or Fisher’s exact test. To investigate the independent associations of BMI classification and age with postoperative surgical site infection and hospital readmission rates, multivariate logistic regression analysis was employed. Variables with a P-value less than 0.05 in the univariate analysis were included in the regression model. A P-value of less than 0.05 was considered indicative of statistical significance. Potential confounding factors such as diabetes mellitus and operation duration were initially evaluated but excluded due to non-significance in univariate analysis (all P > 0.05).

Results

Baseline characteristics of patients

A cohort of 100 patients diagnosed with ASD was recruited for the study, with BMI values ranging from 15.40 to 46.88 kg/m². Notable differences were identified in the age distribution across the four groups (P = 0.034), with the obese group exhibiting the highest mean age of 47 years (range: 33–59 years), whereas the underweight group demonstrated the lowest mean age of 19 years (range: 14–31 years).

No statistically significant differences were observed in terms of gender distribution, New York Heart Association (NYHA) functional classification, or the presence of comorbidities, which included first-degree atrioventricular block, right bundle-branch block, atrial fibrillation, coronary heart disease, hypertension, diabetes mellitus, and cerebrovascular disease, among the four groups (all P > 0.05). Furthermore, echocardiographic assessments revealed no significant differences in ASD size, left ventricular ejection fraction (LVEF), or pulmonary artery pressure among the groups (all P > 0.05), as detailed in Table 1.

Table 1.

Baseline characteristics of the patients

	Underweight group (n = 9)	Normal weight group (n = 52)	Overweight group (n = 26)	Obese group (n = 13)	p-value
Age (years)	19 (14–31)	38 (27–53)	42 (30–58)	47 (33–59)	0.034
Male gender, n (%)	7 (77.8)	36 (69.2)	15 (57.7)	9 (69.2)	0.652
NYHA class, n (%)					0.788
I	5 (55.6)	21 (40.4)	8 (30.8)	3 (23.1)
II	2 (22.2)	23 (44.2)	14 (53.8)	9 (69.2)
III	2 (22.2)	8 (15.4)	4 (15.4)	1 (7.7)
IV	0 (0)	0 (0)	0 (0)	0 (0)
Comorbidities, n (%)
First - degree atrioventricular block or right bundle - branch block	2 (22.2)	11 (21.2)	6 (23.1)	4 (30.8)	0.915
Atrial fibrillation	1 (11.1)	4 (7.7)	3 (11.5)	0 (0)	0.523
Coronary heart disease	0 (0)	1 (1.9)	1 (3.8)	1 (7.7)	0.228
Hypertension	1 (11.1)	3 (5.8)	7 (26.9)	2 (15.4)	0.086
Diabetes mellitus	0 (0)	0 (0)	2 (7.7)	0 (0)	0.328
Cerebrovascular disease	0 (0)	2 (3.8)	2 (7.7)	0 (0)	0.864
Echocardiographic data
ASD size (cm)	2.42 ± 0.80	2.68 ± 0.76	2.67 ± 0.95	2.57 ± 0.97	0.836
LVEF (%)	62 (60–62)	62 (60–64)	62 (60–63)	61 (60–65)	0.705
Pulmonary artery pressure(mmHg)	52.22 ± 12.59	49.54 ± 13.19	44.73 ± 12.69	45.00 ± 9.74	0.243

BMI: body mass index; NYHA: New York Heart Association; ASD: atrial septal defect; LVEF: left ventricular ejection fraction

Intraoperative indicators

No statistically significant differences were identified among the four groups concerning operation duration, CPB time, the proportion of cardioplegic arrest surgeries, aortic cross-clamp time, the proportion of patch closures, or red blood cell (RBC) transfusion requirements (all P > 0.05). Additionally, there were no instances of sternotomy conversion in any of the groups, as detailed in Table 2.

Table 2.

Intraoperative data of the patients

	Underweight group (n = 9)	Normal weight group (n = 52)	Overweight group (n = 26)	Obese group (n = 13)	p-value
Operation duration (minutes)	240 (200–245)	200 (180–240)	228 (195–255)	240 (200–280)	0.150
CPB time (minutes)	74 (69–95)	85 (66–105)	97 (76–122)	90 (84–110)	0.267
Cardioplegic arrest surgery, n (%)	9 (100)	37 (71.2)	18 (69.2)	12 (92.3)	0.928
Aortic cross-clamp time (minutes)	0 (0–0)	0 (0–42)	0 (0–52)	0 (0–0)	0.138
Patch closure, n (%)	6 (66.7)	43 (82.7)	20 (76.9)	10 (76.9)	0.733
RBC transfusion (units)	0 (0–0)	0 (0–0)	0 (0–0)	0 (0–0)	0.787
Sternotomy conversion, n (%)	0 (0)	0 (0)	0 (0)	0 (0)	-

CPB: cardiopulmonary bypass; ASD: atrial septal defect; RBC: red blood cell

Postoperative indicators

The analysis revealed no statistically significant differences among the four groups concerning duration of mechanical ventilation, postoperative 24-hour drainage volume, length of intensive care unit (ICU) stay, length of postoperative stay, postoperative RBC transfusion, early mortality rates, 30 - day mortality, or LVEF and pulmonary artery pressure measurements both immediately after surgery and at one month postoperatively (all P > 0.05).

In terms of postoperative complications, the analysis revealed no statistically significant differences among the four groups concerning the incidence of residual ASD shunt, re-exploration for bleeding, new-onset stroke, new-onset atrial fibrillation, ventricular premature beats, pericardial effusion, second CPB, pulmonary infection, liver dysfunction, pneumothorax, or lower extremity vascular embolism (all P > 0.05). Conversely, significant differences were observed in the incidence of postoperative surgical site infection and hospital readmission rates (P = 0.009 and P = 0.001, respectively). Specifically, the obese group exhibited the highest incidence of postoperative surgical site infection at 15.4%, followed by the overweight group at 3.8%, with no cases reported in the underweight and normal weight groups. Similarly, the hospital readmission rate was highest in the obese group at 23.1%, followed by the overweight group at 7.7%, with no hospital readmission occurring in the underweight and normal weight groups, as detailed in Table 3.

Table 3.

Postoperative data of the patients

	Underweight group (n = 9)	Normal weight group (n = 52)	Overweight group (n = 26)	Obese group (n = 13)	p-value
Duration of mechanical ventilation (hours)	12 (10–13)	11 (9–15)	13(9–19)	13(10–18)	0.663
Postoperative 24-hour drainage volume (ml)	100 (55–130)	160 (80–230)	120 (70–170)	140 (90–220)	0.462
Length of ICU stay (hours)	67(49–72)	65(42–88)	71(60–96)	72 (43–87)	0.309
Length of postoperative stay (days)	10.78 ± 3.11	9.85 ± 3.23	10.19 ± 2.38	9.69 ± 2.96	0.803
RBC transfusion (units)	0 (0–0)	0 (0–0)	0 (0–0)	0 (0–0)	0.347
Early mortality, n (%)	0 (0)	0 (0)	0 (0)	0 (0)	-
Complications, n (%)
Residual ASD shunt	0 (0)	0 (0)	0 (0)	1 (7.7)	0.058
Re-exploration for bleeding	1 (11.1)	0 (0)	1 (3.8)	0 (0)	0.460
New-onset stroke	0 (0)	1 (1.9)	0 (0)	0 (0)	0.603
New-onset atrial fibrillation	1 (11.1)	2 (3.8)	2 (7.7)	2 (15.4)	0.526
Postoperative surgical site infection	0 (0)	0 (0)	1 (3.8)	2 (15.4)	0.009
Ventricular premature beats	0 (0)	2 (3.8)	0 (0)	1 (7.7)	0.617
Pericardial effusion	0 (0)	0 (0)	0 (0)	1 (7.7)	0.058
Second cardiopulmonary bypass	0 (0)	0 (0)	1 (3.8)	0 (0)	0.491
Pulmonary infection	0 (0)	3 (5.8)	2 (7.7)	0 (0)	0.934
Liver dysfunction	0 (0)	2 (3.8)	0 (0)	0 (0)	0.460
Pneumothorax	0 (0)	2 (3.8)	0 (0)	0 (0)	0.460
Lower extremity vascular embolism	0 (0)	0 (0)	1 (3.8)	0 (0)	0.491
Hospital readmission, n (%)	0 (0)	0 (0)	2 (7.7)	3 (23.1)	0.001
30 - day mortality, n (%)	0 (0)	0 (0)	0 (0)	0 (0)	-
Postoperative LVEF (%)	61 (60–64)	61 (60–63)	62 (60–63)	62 (60–63)	0.847
Postoperative pulmonary artery pressure (mmHg)	35.67 ± 12.47	31.25 ± 10.84	28.73 ± 4.50	29.62 ± 5.95	0.253
1-month postoperative LVEF (%)	61 (60–62)	62 (61–63)	62 (60–63)	62 (60–62)	0.338
1-month postoperative pulmonary artery pressure (mmHg)	28.33 ± 15.89	27.13 ± 9.72	27.19 ± 8.86	29.85 ± 7.81	0.834

ICU: intensive care unit; ASD: atrial septal defect; LVEF: left ventricular ejection fraction; RBC: red blood cell

Multivariate logistic regression analysis of hospital readmission and postoperative surgical site infection

Hospital readmission risk analysis

A binary logistic regression analysis was conducted, with hospital readmission serving as the dependent variable, while age and BMI classification (using the underweight group as the reference category) were included as independent variables. The analysis indicated that the model demonstrated a good overall fit, as evidenced by the Hosmer-Lemeshow test (χ²=2.781, P = 0.947), and accounted for 33.5% of the variance in hospital readmission risk (Nagelkerke R²=0.335). Furthermore, the model was found to be statistically significant (χ²=11.626, P = 0.020). However, no significant associations were identified between age or any BMI classification (normal weight, overweight, or obese groups) and the risk of hospital readmission, with all P-values exceeding 0.05 (refer to Tables 4 and 5, and Table 6). Notably, extremely large standard errors, odds ratios approaching 0.000, and incomplete 95% confidence intervals were observed in several BMI categories, indicating sparse-data bias and limited reliability of the regression model.

Table 4.

Model summary of hospital readmission

-2 Log likelihood	Cox & snell R square	Nagelkerke R square	Chi-square	Significance (p-value)
28.077a	0.110	0.335	11.626	0.020

Table 5.

Hosmer-Lemeshow test of hospital readmission

Chi-square	Significance (p-value)
2.781	0.947

Table 6.

Variable of hospital readmission

	B	Std. error	Wald	Significance (p-value)	Exp(B)	95% CI for Exp(B)
						Lower bound	Upper bound
Age (years)	-0.008	0.031	0.071	0.790	0.992	0.934	1.053
BMI group (Reference: underweight group)			1.733	0.630
Normal weight group	-20.162	13383.344	0.000	0.999	0.000	0.000	.
Overweight group	-20.058	5568.273	0.000	0.997	0.000	0.000	.
Obese group	-1.312	0.997	1.733	0.188	0.269	0.038	1.899
Constant	-0.834	1.524	0.300	0.584	0.434

BMI: body mass index

Postoperative surgical site infection risk analysis

A binary logistic regression analysis was performed with postoperative surgical site infection as the dependent variable, while age and BMI classification (using the underweight group as the reference category) served as independent variables. The findings demonstrated that the model exhibited a good fit, as evidenced by the Hosmer-Lemeshow test (χ²=2.915, P = 0.940), and accounted for 31.8% of the variance in the risk of postoperative surgical site infection, as indicated by the Nagelkerke R² value of 0.318. However, the overall statistical significance of the model was only marginally significant (χ²=7.811, P = 0.099). Furthermore, no significant associations were identified between age or any BMI classification and the risk of postoperative surgical site infection, with all P-values exceeding 0.05 (refer to Tables 7 and 8, and Table 9). Similarly, large standard errors, near-zero odds ratios, and non-informative confidence intervals suggested sparse-data issues and instability in parameter estimation, weakening the robustness of the model.

Table 7.

Model summary of the postoperative surgical site infection

-2 Log likelihood	Cox & snell R square	Nagelkerke R square	Chi-square	Significance (p-value)
19.138a	0.075	0.318	7.811	0.099

Table 8.

Hosmer-Lemeshow test of the postoperative surgical site infection

Chi-square	Significance (p-value)
2.915	0.940

Table 9.

Variable of the postoperative surgical site infection

	B	Std. error	Wald	Significance (p-value)	Exp(B)	95% CI for Exp(B)
						Lower bound	Upper bound
Age (years)	0.030	0.044	0.453	0.501	1.030	0.945	1.122
BMI group (Reference: underweight group)			1.251	0.741
Normal weight group	-18.960	13177.067	0.000	0.999	0.000	0.000	.
Overweight group	-19.308	5508.820	0.000	0.997	0.000	0.000	.
Obese group	-1.440	1.287	1.251	0.263	0.237	0.019	2.955
Constant	-3.131	2.351	1.774	0.183	0.044

BMI: body mass index

Discussion

This study aimed to examine the influence of various BMI classifications (underweight, normal weight, overweight, and obese) on perioperative indicators and short-term prognosis in patients undergoing Da Vinci robot-assisted ASD repair. A retrospective analysis of clinical data from 100 enrolled patients indicated no significant differences in baseline characteristics, key intraoperative indicators (such as operation duration and CPB time), or most postoperative recovery indicators (including duration of mechanical ventilation and length of ICU stay) across different BMI groups (all P > 0.05). Nevertheless, the obese group exhibited significantly higher rates of postoperative surgical site infection (15.4%) and hospital readmission (23.1%) compared to other groups (P < 0.05). Multivariate logistic regression analysis did not identify BMI classification or age as independent predictors of these adverse outcomes (all P > 0.05). However, the multivariable models suffered from sparse‑data bias and unstable estimates, so this null finding should be interpreted with extreme caution. These findings suggest that obesity may be a potential factor influencing specific postoperative complications in ASD patients and that clinical outcomes may be modulated by the interaction of multiple factors. This provides valuable clinical evidence for optimizing the perioperative management of such patients.

In this study, baseline echocardiographic parameters, including ASD size, LVEF, and pulmonary artery pressure, as well as the distribution of comorbidities, were generally comparable across the four groups, with the exception of age. The obese group exhibited a higher mean age, a difference potentially attributable to the cumulative effects of obesity-related metabolic disorders [14]. Maria-Dolores et al.‘s decade-long cohort study demonstrated a significant increase in BMI with advancing age among adults aged 18–60 years (P < 0.05) [15]. The underlying mechanisms may involve age-associated reductions in basal metabolic rate, alterations in lifestyle, and hormonal changes, all contributing to weight gain and increased BMI [16]. Furthermore, older individuals are more susceptible to comorbidities such as metabolic syndrome, diabetes mellitus, and cardiovascular diseases, underscoring the importance of considering age as a critical factor in discussions of obesity and its related health implications [17].

In terms of intraoperative indicators, no significant differences were observed in operation duration, CPB time, or aortic cross-clamp time across the four groups (all P > 0.05), thereby affirming the clinical applicability of Da Vinci robot-assisted surgery in patients with varying BMI levels. This finding contradicts the conclusion drawn by Sergio et al., which posits that “high BMI prolongs operation duration” [18]. The observed discrepancy may be attributed to the specific anatomical characteristics involved in ASD repair: the surgical field is primarily concentrated within the atrium, and the influence of thoracic fat accumulation on atrial exposure in obese patients is relatively minimal [19]. Furthermore, the three-dimensional visualization and the dexterous manipulation capabilities of the Da Vinci robotic system effectively mitigate the spatial constraints imposed by obesity, facilitating enhanced anatomical exposure, reducing the complexity of the procedure, and thereby preserving surgical efficiency [5]. Individualized technical adjustments and perioperative management strategies are essential for patients with varying BMI levels. For obese patients, it is recommended to conduct preoperative chest CT scans to evaluate chest wall fat thickness and to tailor the positioning of operating ports accordingly [6]. During surgery, the use of adaptive instruments, such as lengthened trocars, is advised, along with the prioritization of beating heart surgery to minimize the risk of aortic injury [20]. Postoperatively, enhanced blood glucose control and meticulous incision care are crucial to reduce the risk of infection [3]. Conversely, for young and underweight patients, who possess slender blood vessels with high wall elasticity and are susceptible to vasospasm, a detailed preoperative vascular assessment is imperative [21]. Intraoperatively, real-time ultrasound should be employed to localize vascular pathways, and local application of nitroglycerin prior to puncture is recommended to prevent spasm [22]. The use of small-caliber, thin-walled puncture needles and adaptive cannulas, along with gentle handling to avoid repeated puncture or aggressive dilation, is advised [23]. If vasospasm occurs, the operation should be paused and spasmolytic drugs administered; open intubation under direct vision may be necessary [24]. Additionally, CPB should be conducted using a low-flow and low-pressure perfusion mode [3].

In terms of short-term postoperative recovery, no significant differences were observed among the four groups in duration of mechanical ventilation, length of ICU stay, length of postoperative stay, or cardiac function recovery, including LVEF and pulmonary artery pressure immediately after surgery and one month postoperatively (all P > 0.05). Notably, there were no instances of sternotomy conversion, early mortality, or 30-day mortality within the entire cohort. These findings further substantiate the efficacy of Da Vinci robot-assisted surgery in mitigating the impact of BMI-related physiological variations on postoperative recovery. The study by Li et al. corroborates these results, demonstrating that, compared to traditional thoracotomy, Da Vinci robot-assisted surgery is associated with smaller incisions, reduced trauma to chest wall muscles and bones, and less postoperative pain, facilitating earlier mobilization and consequently reducing mechanical ventilation time and hospital stay [25]. Remarkably, both underweight and obese patients exhibited comparable short-term recovery outcomes to those of normal-weight patients, thereby challenging the conventional notion that “obese patients experience delayed postoperative recovery” following traditional thoracotomy [12].

While univariate analysis revealed statistically significant associations between BMI classification and both postoperative surgical site infection and hospital readmission rate (all P < 0.05), multivariate logistic regression analysis did not identify BMI classification or age as independent predictors of these adverse outcomes (all P > 0.05). However, the regression models showed clear signs of instability, including extremely large standard errors, odds ratios of effectively zero, and uninterpretable confidence intervals in several BMI groups, all indicative of sparse-data bias. Therefore, we cannot definitively conclude that BMI is not an independent risk factor based on the present analysis. Several factors may account for this discrepancy: firstly, the relatively small sample size of the study, particularly within the underweight (n = 9) and obese (n = 13) groups, may have resulted in insufficient statistical power to detect potential independent associations between BMI and the outcomes. Secondly, the regression model was limited to age and BMI classification as independent variables, whereas postoperative infection and hospital readmission are likely influenced by numerous confounding factors, such as perioperative antibiotic use, quality of incision care, postoperative activity level, and management of comorbid metabolic diseases [26]. The exclusion of these variables may obscure the independent effect of BMI. Thirdly, the Da Vinci robot-assisted surgery is characterized by an extensive learning curve, with increased surgical complexity observed in obese patients [27, 28]. These procedures are typically conducted once surgeons have surpassed the learning curve, and the implementation of standardized operating procedures may mitigate outcome disparities among patients with varying BMI levels. Consequently, it becomes challenging for BMI alone to exhibit an independent risk contribution. Fourth, the presence of exceedingly large standard errors in certain BMI categories (specifically, the normal weight and overweight groups) within the regression model indicates potential issues of multicollinearity or uneven sample distribution, which could compromise the reliability of parameter estimation.

From a clinical standpoint, the regression results do not dismiss the potential risks associated with obesity but rather highlight the complexity of clinical outcomes. The influence of obesity on postoperative infection and hospital readmission may not be a direct, independent effect; instead, it may occur through synergistic interactions with factors such as blood glucose control, immune function, and the quality of incision care [14, 16]. Consequently, it is imperative to optimize perioperative management strategies based on BMI characteristics. For obese patients, this includes preoperative regulation of metabolic indicators such as blood glucose and lipids to improve insulin resistance; precise intraoperative puncture positioning to avoid areas with significant fat accumulation; optimization of incision location; enhanced hemostasis and sterile techniques to reduce infection risks; and postoperative measures such as improved incision care (e.g., maintaining a dry and clean incision, regular dressing changes), early rehabilitation exercises (to promote local blood circulation and accelerate wound healing), and the judicious use of antibiotics, while developing a comprehensive long-term follow-up mechanism designed to promptly identify signs of infection and implement targeted interventions aimed at reducing the risk of hospital readmission [9, 12, 13].

This study is subject to several limitations. Firstly, the four-group BMI stratification among only 100 patients resulted in very small subgroups, which reduced statistical power and compromised the stability of between-group comparisons and regression analyses. Secondly, the multivariable logistic regression models were limited by sparse data, leading to unstable coefficients, inflated standard errors, and non-valid odds ratios in several categories; accordingly, no definitive conclusion regarding the independent effect of BMI can be drawn. Thirdly, although our results suggest that robotic ASD repair is feasible across BMI levels, the small sample size and low event rates prevent definitive conclusions about the ability of the robotic approach to overcome weight-related surgical challenges. Fourthly, instability in parameter estimation was observed in certain BMI groups within the regression model, potentially attributable to sample distribution characteristics or model specifications. Lastly, the absence of a traditional thoracotomy control group precludes direct comparison of the clinical advantages associated with different surgical approaches across varying BMI levels.

Future research should aim to increase the sample size and employ a multicenter prospective design. It is essential to include a control cohort comprising both Da Vinci robot-assisted surgery and traditional thoracotomy to rigorously assess the long-term clinical benefits of minimally invasive techniques in patients with ASD across various BMI categories. Additionally, refining observational indicators and meticulously documenting surgical procedures and postoperative rehabilitation interventions will aid in developing a more comprehensive prognostic prediction model. Extending the follow-up period to at least one year is recommended to thoroughly evaluate the impact of BMI on long-term cardiac function, metabolism-related complications, and quality of life. This approach will provide robust clinical evidence to support the personalized application of Da Vinci robot-assisted surgery.

Conclusion

In conclusion, Da Vinci robot‑assisted ASD repair appears safe and feasible across patients with different BMI levels in this single‑center cohort. The minimally invasive approach seemed to alleviate weight‑related surgical challenges in this sample, but this is an exploratory observation rather than a definitive conclusion. Although obese patients exhibited higher rates of postoperative surgical site infection and hospital readmission in univariate analysis, BMI was not identified as an independent predictor in multivariable models; however, small subgroup sizes and sparse‑data bias limit the reliability of this finding. We recommend intensified, individualized perioperative management for obese patients. Weight status should not be regarded as a contraindication for robotic ASD repair, but all conclusions regarding BMI’s independent effect should be considered exploratory due to limited sample size and statistical power.

Abbreviations

ASD

Atrial Septal Defect

BMI

Body Mass Index

CPB

Cardiopulmonary Bypass

ICU

Intensive Care Unit

LVEF

Left Ventricular Ejection Fraction

NYHA

New York Heart Association

RBC

Red Blood Cell

ANOVA

Analysis of Variance

Author contributions

Sumin Yang, Wei Wang, and Qingjiang Wang participated in the surgery. Haoyan Li, Xun Chi, and Ziang Sun participated in data collection. Qingjiang Wang and Rui Dai contributed to revising and drafting the manuscript. All authors read and approved the final manuscript.

Funding

This work received no specific grant from any funding agency in the

public, commercial, or not-for-profit sectors.

Data availability

The datasets generated and analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the institutional review board of The Affiliated Hospital of Qingdao University (Approval number QYFY WZLL 42136). Since the study was a retrospective study, informed consent was waived by the Ethics Committee. This study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.