Association of obstructive sleep apnea with postoperative outcomes after breast reconstruction

Abstract

Background

Obstructive sleep apnea (OSA) is prevalent among breast reconstruction patients, yet its impact on surgical outcomes remains inadequately characterized. This study assessed associations between OSA and postoperative complications, length of stay, and inpatient costs among patients undergoing breast reconstruction.

Methods

We performed a retrospective population-based study using the National Inpatient Sample from 2016 to 2022. Breast reconstruction hospitalizations were identified using ICD−10-PCS procedure codes, and OSA was identified using the ICD−10-CM diagnosis code G47.33. National estimates were generated using HCUP discharge weights, and the complex survey design of the NIS was accounted for in all analyses. Multivariable logistic regression was used to analyze binary postoperative complications. LOS and inpatient costs were summarized as median (interquartile range [IQR]) and compared using the Wilcoxon rank-sum test because of non-normal distributions. A prespecified two-sided P value < 0.001 was considered statistically significant.

Results

Based on weighted national estimates, 177,435 adult breast reconstruction hospitalizations were included, of which 7,865 (4.4%) involved patients with OSA. In multivariable-adjusted analyses, OSA was associated with increased odds of respiratory failure (adjusted odds ratio [aOR], 2.705; 95% confidence interval [CI], 2.089–3.504), heart failure (aOR, 2.282; 95% CI, 1.862–2.796), and thrombocytopenia (aOR, 1.552; 95% CI, 1.219–1.976) (all P < 0.001). OSA was also associated with lower odds of seroma (aOR, 0.555; 95% CI, 0.468–0.659). Compared with patients without OSA, those with OSA had a longer LOS (median, 3 [IQR, 2–4] vs. 2 [IQR, 1–3] days; P < 0.001) and higher inpatient costs (median, $99,066 [IQR, $64,118–$144,249] vs. $91,965 [IQR, $60,160–$139,225]; P < 0.001). In subgroup analyses, the associations with respiratory failure, heart failure, and seroma remained directionally consistent across autologous and implant-based reconstruction.

Conclusions

OSA is independently associated with increased cardiopulmonary complications, thrombocytopenia, prolonged hospitalization, and elevated costs in breast reconstruction patients. These findings support routine preoperative OSA screening and optimized perioperative management.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12905-026-04514-y.

Introduction

Globally, breast cancer represents the most commonly diagnosed cancer in women and ranks as the second leading cause of cancer-associated mortality among them [1–4]. Mastectomy—frequently employed as a primary treatment—requires complete excision of the breast tissue and frequently leads to substantial bodily disfigurement. This physical alteration can trigger severe psychological distress, impair health-related quality of life, and decrease compliance with postoperative adjuvant therapies. In response to these challenges, breast reconstruction has evolved into a cornerstone of holistic breast cancer management [5]. Accumulating clinical evidence demonstrates that reconstructive procedures significantly improve patients’ body image, sexual function, and psychosocial well-being, all while maintaining oncological safety. By concurrently addressing the physical aftermath of mastectomy and meeting essential emotional and social needs, this integrative strategy supports patients in resuming everyday activities with renewed self-assurance and satisfaction [6–8].

Although breast reconstruction markedly improves patients’ quality of life, optimizing surgical outcomes requires careful attention to perioperative risks. Among the factors influencing these risks, sleep-related disorders—particularly OSA—are increasingly recognized as critical determinants of patient safety and postoperative recovery. OSA is defined by repeated episodes of upper airway obstruction during sleep, resulting in chronic intermittent hypoxia and fragmented sleep architecture [9–12]. In North America, the condition affects an estimated 10–15% of women in the general population [13, 14]. Notably, its prevalence is substantially higher among individuals with breast cancer, especially postmenopausal women. Moreover, multiple prospective cohort studies have linked OSA to poorer breast cancer outcomes, and Mendelian randomization analyses provide evidence that genetic predisposition to OSA directly elevates breast cancer risk [15, 16]. Despite these well-established associations between OSA and both the pathophysiology and prognosis of breast cancer, the full scope of how untreated OSA influences perioperative morbidity and surgical complications throughout the entire continuum of breast cancer care remains poorly characterized.

While postoperative complications following breast reconstruction have been extensively documented, evidence specifically evaluating outcomes in individuals with OSA remains scarce. This study seeks to compare rates of postoperative complications, duration of hospitalization, and associated healthcare expenditures between patients undergoing breast reconstruction who have OSA and those who do not. We hypothesize that OSA is associated with an elevated risk of postoperative complications, longer hospital stays, and increased overall costs.

Methods

Data collection

This population-based, retrospective observational study leveraged data from the U.S. National Inpatient Sample (NIS)—the nation’s largest publicly accessible, all-payer inpatient database—curated by the Healthcare Cost and Utilization Project (HCUP) under the Agency for Healthcare Research and Quality (https://ahrq.gov/data/hcup/index.html). By incorporating cost-to-charge ratio files, the NIS enables nationally and regionally representative estimates of inpatient utilization, access patterns, healthcare expenditures, quality metrics, and clinical outcomes for hospitalizations occurring between January 1, 2016, and December 31, 2022. The database draws from a stratified 20% sample of discharges from approximately 1,050acute-care hospitals across44states, encompassing over 8 million annual hospital stays and representing more than 95% of theU.S resident population.

The NIS contains extensive patient-level data, including principal and secondary diagnoses, procedures performed during admission, admission and discharge dispositions, demographic variables, primary expected payer, and duration of hospitalization. It also includes hospital-level characteristics such as bed capacity, geographic region, teaching status, and urban/rural classification. Nationally representative estimates were generated using HCUP-provided discharge weights that account for the database’s complex survey design. Because the dataset is de-identified and publicly available, this study was exempt from institutional review board (IRB) approval [14].

Patient population

Patients who underwent breast cancer reconstruction were identified using procedure codes from the International Classification of Diseases, Tenth Revision (ICD-10). We identified adult patients (aged ≥ 18 years) who underwent breast reconstruction between January 1, 2016 and December 31, 2022 using ICD-10-PCS procedure codes. Reconstruction type was classified as autologous breast reconstruction (ABR) or implant-based breast reconstruction (IBR) according to predefined ICD-10-PCS code sets. OSA was identified using the ICD-10-CM diagnosis code G47.33. The full ICD-10 code lists used to define breast reconstruction type, OSA, and postoperative outcomes are provided in Supplementary Table S1. Adults undergoing breast reconstruction were included. Exclusion criteria were age < 18 years, in-hospital death, male sex or other sex-based ineligible records, and missing data in variables required for analysis (including age, payer, and total hospital charges).

Covariates

Covariates were selected a priori on the basis of clinical relevance and prior literature. The same prespecified covariate set was applied across all multivariable models for binary postoperative outcomes. Covariates included patient demographics (age; race/ethnicity), hospital characteristics (primary payer, hospital bed size, teaching status, and hospital region), type of admission (elective), location of hospital (urban) and comorbidities identified using the HCUP Elixhauser Comorbidity Software for ICD-10-CM. The comorbidities included obesity, diabetes without chronic complications, paralysis, hypothyroidism, diabetes with chronic complications, smoking, chronic pulmonary disease, hypertension, depression, peripheral vascular disease, renal failure, anemia, drug abuse, alcohol abuse, and acquired immunodeficiency syndrome (Table 1).

Table 1.

Covariates and outcome variables used in the regression analysis

Variables	Categories	Specific Variables
Covariates	Patient demographics	Age; Race and Ethnicity (Caucasian, African American, Hispanic, Asian, Native American, Other)
Covariates	Hospital characteristics	Type of insure (Medicare, Medicaid, Private insurance, Self-pay, No charge, Other); Bed size of hospital (Small, Medium, Large); Teaching status of hospital (non-teaching, teaching); Location of the hospital (Northeast, Midwest or North Central, South, West); Location of hospital (urban)
Covariates	Comorbidities	Acquired immune deficiency syndrome; Alcohol abuse; Depression; Diabetes without chronic complications; Diabetes with chronic complications; Drug abuse; Hypertension; Chronic pulmonary disease; Obesity; Paralysis; Peripheral vascular disease; Renal failure; Anemia; Smoking; Hypothyroidism
Outcome variables	Complications and economic outcomes	Thrombocytopenia; Heart failure; Respiratory failure; Seroma; Wound infection; Disruption of wound; Flap revision; Urinary tract infection; Sepsis; Pneumonia; Acute kidney injury; Blood transfusion; Hemorrhagic anemia; Capsular contracture; Excessive scarring; LOS; Inpatient costs

Primary outcome variables

The primary outcomes included individual in-hospital postoperative complications, which were interpreted in clinically meaningful categories as medical complications (heart failure, respiratory failure, urinary tract infection, sepsis, pneumonia, and acute kidney injury), wound-related or surgical complications (seroma, wound dehiscence, surgical site infection, flap revision, capsular contracture, and excessive scarring), and hematologic complications (thrombocytopenia). Secondary endpoints included hospital length of stay, requirement for blood transfusion, hemorrhagic anemia and total inpatient costs.

Statistical analysis

All analyses incorporated HCUP discharge weights to generate nationally representative estimates, and the complex survey design of the NIS, including stratification and clustering, was accounted for in all analyses, including regression models. Categorical variables were compared using weighted chi-square tests. Continuous variables, including LOS and inpatient costs, were non-normally distributed and were therefore compared using the Wilcoxon rank-sum test; these outcomes are presented as median with IQR. Multivariable logistic regression was used to estimate adjusted odds ratios and 95% confidence intervals for binary postoperative outcomes associated with OSA. To examine whether associations varied by reconstruction type, prespecified subgroup analyses were performed separately in the ABR and IBR subgroups. Because multiple outcomes were evaluated in a large national dataset, and consistent with prior studies using similar administrative data [17], a conservative two-sided P value threshold of < 0.001 was prespecified to define statistical significance. No formal adjustment for multiple comparisons was performed. Before model fitting, multicollinearity among covariates was assessed using tolerance statistics and variance inflation factors (VIFs), and no problematic multicollinearity was detected (all tolerance values > 0.85; all VIFs < 1.15) in Supplementary Table S6.

Results

Baseline characteristics

Based on weighted national estimates, 177,435 adult breast reconstruction hospitalizations met the inclusion criteria, including 169,570 without OSA and 7,865 with OSA; the weighted prevalence of OSA was 4.4%. Table 2 summarizes key demographic disparities between individuals diagnosed with OSA and those without the condition. Participants with OSA were significantly older than those without (median age, 57 [IQR, 51–63] vs. 51 [IQR, 44–59] years; P < 0.001). The racial distribution also varied between groups: a higher proportion of African American individuals were represented in the OSA cohort (16.8% vs. 11.7%), whereas the percentages of White, Asian, and Native American participants were comparatively lower. These patterns may align with known epidemiological differences in OSA susceptibility and the distribution of associated risk factors across population subgroups. Differences were also observed in payer and hospital characteristics. Compared with patients without OSA, those with OSA were more likely to have Medicare coverage (23.6% vs. 12.4%) and to undergo reconstruction at teaching hospitals (88.1% vs. 86.1%) (both P < 0.001).

Table 2.

Patient demographics, hospital characteristics, and comorbidities comparing patients with OSA versus those without OSA who underwent breast reconstruction (2016-2022)

Characteristics	No OSA	OSA	P
Total (n=count)	169,570	7865
Total incidence (%)	4.4
Age (median/mean, years)	51 (44,59)	57 (51,63)	< 0.001
Race and Ethnicity (%)
Caucasian	111,355 (65.7%)	5225 (66.4%)	< 0.001
African American	19,865 (11.7%)	1320 (16.8%)
Hispanic	18,555 (10.9%)	710 (9.0%)
Asian	7855 (4.6%)	210 (2.7%)
Native American	440 (0.3%)	10 (0.1%)
Other	11,500 (6.8%)	390 (5.0%)
Type of insure (%)
Medicare	20,945 (12.4%)	1855 (23.6%)	< 0.001
Medicaid	18,455 (10.9%)	615 (7.8%)
Private insurance	123,665 (72.9%)	5100 (64.8%)
Self-pay	2230 (1.3%)	65 (0.8%)
No charge	175 (0.1%)	0 (0%)
Other	4100 (2.4%)	230 (2.9%)
Bed size of hospital (%)
Small	29,290 (17.3%)	1260 (16%)	0.004
Medium	38,770 (22.9%)	1765 (22.4%)
Large	101,510 (59.9%)	4840 (61.5%)
Elective admission (%)	156,190 (92.1%)	7360 (93.6%)	< 0.001
Type of hospital (teaching %)	145,975 (86.1%)	6930 (88.1%)	< 0.001
Region of hospital (%)
Northeast	48,075 (28.4%)	1935 (24.6%)	< 0.001
Midwest or North Central	28,415 (16.8%)	1575 (20%)
South	60,445 (35.6%)	2865 (36.4%)
West	32,635 (19.2%)	1490 (18.9%)
Location of hospital (urban, %)	167,745 (98.9%)	7815 (99.4%)	< 0.001
Comorbidities
Acquired immune deficiency syndrome	340 (0.2%)	25 (0.3%)	0.025
Alcohol abuse	615 (0.4%)	30 (0.4%)	0.787
Depression	18,935 (11.2%)	1920 (24.4%)	< 0.001
Diabetes without chronic complications	9275 (5.5%)	1310 (16.7%)	< 0.001
Diabetes with chronic complications	2685 (1.6%)	510 (6.5%)	< 0.001
Drug abuse	520 (0.3%)	65 (0.8%)	< 0.001
Hypertension	42,275 (24.9%)	4395 (55.9%)	< 0.001
Chronic pulmonary disease	15,795 (9.3%)	2055 (26.1%)	< 0.001
Obesity	21,575 (12.7%)	3165 (40.2%)	< 0.001
Paralysis	140 (0.1%)	5 (0.1%)	0.565
Peripheral vascular disease	2010 (1.2%)	190 (2.4%)	< 0.001
Renal failure	1530 (0.9%)	265 (3.4%)	< 0.001
Anemia	1990 (1.2%)	145 (1.8%)	< 0.001
Smoking	30,865 (18.2%)	2020 (25.7%)	< 0.001
Hypothyroidism	19,530 (11.5%)	1810 (23%)	< 0.001

The relationship between OSA and comorbidities

Patients undergoing breast reconstruction who had OSA exhibited a substantially greater comorbidity burden compared to those without OSA (Table 2). Specifically, OSA was associated with significantly higher prevalence rates of depression (24.4%vs.11.2%), both uncomplicated diabetes (16.7%vs.5.5%) and diabetes with chronic complications (6.5% vs. 1.6%), hypertension (55.9% vs. 24.9%), hypothyroidism (23.0%vs.11.5%), current smoking status (25.7%vs.18.2%), chronic pulmonary disease (26.1%vs.9.3%), obesity (40.2%vs.12.7%), peripheral vascular disease (2.4% vs.1.2%), and renal failure (3.4%vs.0.9%). A small but notable difference was also observed for drug abuse (0.8%vs.0.3%). Although rare, cases of acquired immunodeficiency syndrome (AIDS) were documented within the OSA cohort; given its known effects on immune competence and tissue repair, AIDS may further elevate perioperative risks by impairing wound healing and immune responses.

Patients with OSA often present with a complex array of comorbidities, which—when combined with perioperative variables such as intraoperative fluid administration and choice of anesthetic agents—can heighten surgical risk and drive up healthcare costs. These observations highlight the importance of thorough preoperative evaluation and targeted management of comorbid conditions in OSA patients to reduce both perioperative complications and associated expenditures.

Economic outcomes

Patients with OSA had a longer inpatient stay than those without OSA (median LOS, 3 [IQR, 2–4] vs. 2 [IQR, 1–3] days; P < 0.001). Inpatient costs were also higher among hospitalizations involving OSA (median, $99,066 [IQR, $64,118–$144,249] vs. $91,965 [IQR, $60,160–$139,225]; P < 0.001).

Multivariable-adjusted perioperative complications

In multivariable logistic regression analyses using the same prespecified covariate set for all binary outcomes, the associations of OSA differed across complication categories. OSA was associated with higher odds of selected medical complications, including respiratory failure and heart failure, as well as thrombocytopenia, whereas lower odds were observed for certain seroma and wound infection (Table 3).

Table 3.

In-hospital outcomes and postoperative complications comparing patients with OSA versus those without OSA who underwent breast reconstruction

Complications	Univariate Analysis			Multivariate Logistic Regression
	No OSA	OSA	P	aOR	95% CI	P
Respiratory failure	300 (0.2%)	100 (1.3%)	< 0.001	2.705	2.089–3.504	< 0.001
Heart failure	575 (0.3%)	140 (1.8%)	< 0.001	2.282	1.862–2.796	< 0.001
Thrombocytopenia	1015 (0.6%)	80 (1.0%)	< 0.001	1.552	1.219–1.976	< 0.001
Seroma	4860 (2.9%)	145 (1.8%)	< 0.001	0.555	0.468–0.659	< 0.001
Wound infection	1445 (0.9%)	40 (0.5%)	< 0.001	0.491	0.354–0.681	< 0.001
Disruption of wound	1460 (0.9%)	55 (0.7%)	0.128	0.750	0.566–0.993	0.044
Flap revision	535 (0.3%)	20 (0.3%)	0.342	0.600	0.379–0.949	0.029
Urinary tract infection	645 (0.4%)	45 (0.6%)	0.008	0.944	0.685–1.300	0.723
Sepsis	645 (0.4%)	35 (0.4%)	0.364	0.954	0.662–1.374	0.799
Pneumonia	355 (0.2%)	25 (0.3%)	0.042	0.716	0.466–1.099	0.127
Acute kidney injury	1065 (0.6%)	175 (2.2%)	< 0.001	1.195	0.997–1.433	0.054
Blood transfusion	4750 (2.8%)	290 (3.7%)	< 0.001	1.064	0.937–1.208	0.339
Hemorrhagic anemia	12,605 (7.4%)	805 (10.2%)	< 0.001	1.058	0.978–1.146	0.162
Capsular contracture	5810 (3.4%)	300 (3.8%)	0.065	1.132	1.001–1.280	0.048
Excessive scarring	1885 (1.1%)	110(1.4%)	0.018	1.195	0.976–1.462	0.084

Economic outcomes	Univariate Analysis
	No OSA		OSA		P
Died (%)	20 (0.01%)		0 (0.0%)		1.000
LOS (median, d)	2(1,3)		3(2,4)		<0.001
Inpatient costs (median, $)	91,965(60,160,139,225)		99,066(64,118,144,249)		<0.001

A two-sided P value < 0.001 was prespecified as statistically significant; results with P values between 0.001 and 0.05 are shown for completeness but were not considered statistically significant in this study

aOR Adjusted odds ratio, CI confidence interval, LOS length of stay, TOTCHG total hospital charges

Among medical complications, OSA was associated with respiratory failure (aOR, 2.705; 95% CI, 2.089–3.504; P < 0.001), heart failure (aOR, 2.282; 95% CI, 1.862–2.796; P < 0.001). In relative terms, the odds of developing respiratory failure and heart failure were elevated by approximately 170.5% and 128.2%, respectively.

OSA was significantly linked to an increased likelihood of thrombocytopenia (aOR,1.552;95% CI,1.219–1.976;P < 0.001).

Paradoxically, after multivariable adjustment, patients with OSA exhibited lower odds of several local wound-related complications: seroma (aOR 0.555;95% CI 0.468–0.659;P < 0.001), wound infection (aOR,0.491;95% CI,0.354–0.681;P < 0.001). By contrast, no association meeting the prespecified threshold for statistical significance was observed for wound dehiscence, flap revision, urinary tract infection, sepsis, pneumonia, acute kidney injury, blood transfusion, hemorrhagic anemia, capsular contracture, or excessive scarring.

Subgroup analyses by reconstruction type

Detailed subgroup results are provided in Supplementary Tables S2–S5. In prespecified subgroup analyses, the association of OSA with major cardiopulmonary complications remained broadly consistent across reconstruction types. Among patients undergoing ABR, OSA was associated with higher odds of thrombocytopenia (aOR, 2.021; 95% CI, 1.525–2.679; P < 0.001), heart failure (aOR, 1.883; 95% CI, 1.376–2.577; P < 0.001), and respiratory failure (aOR, 2.565; 95% CI, 1.732–3.799; P < 0.001), and with lower odds of seroma (aOR, 0.593; 95% CI, 0.474–0.742; P < 0.001). Among patients undergoing IBR, OSA was associated with higher odds of heart failure (aOR, 2.263; 95% CI, 1.735–2.951; P < 0.001) and respiratory failure (aOR, 3.043; 95% CI, 2.149–4.310; P < 0.001), as well as lower odds of seroma (aOR, 0.496; 95% CI, 0.385–0.641; P < 0.001) and wound infection (aOR, 0.436; 95% CI, 0.288–0.659; P < 0.001). In the IBR subgroup, OSA was also associated with higher odds of capsular contracture (aOR, 1.471; 95% CI, 1.211–1.787; P < 0.001) and excessive scarring (aOR, 1.741; 95% CI, 1.297–2.336; P < 0.001).

After stratification by reconstruction type, median LOS was 3 days in both the OSA and non-OSA groups among patients undergoing ABR (IQR, 3–4 vs. 3–4; P < 0.001) and 2 days in both groups among those undergoing IBR (IQR, 1–2 vs. 1–2; P < 0.001). Inpatient costs were higher in the OSA group in the ABR subgroup (median, $115,373.5 [IQR, $78,069.6–$164,669.6] vs. $111,628.0 [IQR, $74,564.0–$163,940.0]; P < 0.001), whereas no significant difference in inpatient costs was observed between the OSA and non-OSA groups in the IBR subgroup (median, $76,713.5 [IQR, $52,067.6–$114,164.0] vs. $76,915.0 [IQR, $51,686.5–$114,334.8]; P = 0.790).

Discussion

Breast reconstruction following mastectomy plays a vital role in holistic breast cancer management, offering marked improvements in both aesthetic results and psychological health [18]. Within our patient cohort, individuals diagnosed with OSA exhibited higher rates of medical complications, particularly cardiopulmonary events, as well as longer hospital stays and increased healthcare expenditures, whereas wound-related outcomes warrant cautious interpretation. These heightened risks—including respiratory failure, heart failure, and thrombocytopenia—are likely attributable to a systemic susceptibility arising from chronic intermittent hypoxia, persistent low-grade inflammation, and reduced cardiopulmonary functional capacity.

The observed inverse association with selected local wound complications may reflect factors other than a direct effect. Given the greater comorbidity burden in the OSA group, alternative explanations should be considered, including residual confounding, surveillance bias, differential coding practices, differences in perioperative management, and procedure-related heterogeneity. In our subgroup analyses, the inverse association with seroma remained directionally consistent across ABR and IBR, whereas the inverse association with wound infection was observed only in the IBR subgroup, and the overall cost difference was attenuated after stratification, suggesting that procedure mix contributed to at least part of the observed associations. Prior comparative studies have shown that complication profiles differ meaningfully between autologous and implant-based reconstruction, with generally greater short-term systemic morbidity and longer length of stay reported for flap-based procedures, whereas implant-based reconstruction has been associated with higher risks of seroma and reconstructive failure in longer-term follow-up [19–21]. Timing of reconstruction may also influence complication profiles and length of stay [22]. In this context, the lower odds of seroma and wound infection observed in our cohort may reflect differences in procedure mix or perioperative pathways rather than a direct biologic benefit of OSA itself.

Patients diagnosed with OSA experienced a hospital stay that was one day longer than that of patients without OSA. This difference should be interpreted with caution, as length of stay and inpatient costs may partly reflect reconstruction procedure mix and the unmeasured timing of reconstruction, in addition to OSA itself. The attenuation of the overall cost difference after stratification by reconstruction type further supports that procedural heterogeneity may have contributed to the unadjusted economic findings. Nevertheless, the associated increase in inpatient costs underscores the health-economic relevance of OSA in breast reconstruction and supports greater attention to perioperative risk stratification, resource planning, and postoperative surveillance in this population.

OSA was also associated with higher odds of thrombocytopenia. Although administrative data cannot establish the underlying mechanism, prior studies have linked OSA to altered platelet indices and suggested that CPAP may partially normalize these parameters [23–26]. In reconstructive surgery, this finding is clinically relevant because perioperative hematologic abnormalities may complicate bleeding risk assessment and postoperative monitoring. Given these findings, a thorough preoperative assessment for thrombocytopenia is warranted in patients with OSA. Furthermore, perioperative strategies—including optimization of CPAP adherence and vigilant hematologic monitoring—should be integrated into clinical management to reduce the risks of bleeding and compromised tissue repair.

Similarly, the observed association with heart failure is clinically plausible and consistent with prior literature showing substantial overlap between OSA and heart failure, as well as adverse cardiovascular interactions related to recurrent hypoxemia, sympathetic activation, and intrathoracic pressure swings [27–31]. In the reconstructive setting, these considerations support careful preoperative cardiovascular assessment and postoperative monitoring rather than implying a mechanism that can be confirmed from administrative data alone. In patients with OSA scheduled for breast reconstruction, subclinical cardiovascular impairment may be present, and the added hemodynamic stress of surgery could unmask or worsen underlying heart failure. Consequently, thorough preoperative cardiovascular assessment—and timely implementation of therapies such as continuous positive airway pressure (CPAP) [31]—is essential to mitigate perioperative HF risk.

The association between OSA and postoperative respiratory failure is also consistent with prior perioperative literature, particularly in patients with coexisting pulmonary disease and opioid exposure [32–35]. For patients undergoing breast reconstruction, these findings support attention to preoperative optimization, opioid-sparing analgesia, and appropriate postoperative respiratory monitoring.

Taken together, these findings suggest that OSA should be regarded as an important perioperative risk marker in breast reconstruction rather than merely a background comorbidity. Future studies with more detailed clinical data are needed to clarify the mechanisms underlying the inverse associations observed for selected wound-related outcomes and to determine whether targeted perioperative interventions can improve outcomes in this high-risk population.

Limitations

This large-scale retrospective analysis reveals that OSA exerts a significant adverse influence on outcomes following breast reconstruction surgery. Patients diagnosed with OSA exhibited elevated rates of thrombocytopenia, heart failure, and respiratory failure, required longer hospital stays, and accrued substantially higher healthcare expenditures. These findings underscore the importance of implementing routine preoperative OSA screening, optimizing perioperative care protocols, and enhancing postoperative monitoring to improve clinical outcomes and alleviate the associated healthcare burden in this vulnerable cohort. Nevertheless, several limitations must be acknowledged: there is a risk of underdiagnosis and undercoding (i.e., many patients may have OSA but lack corresponding ICD-10-CM coded documentation). the retrospective design precludes causal inference, and residual confounding may persist despite multivariable adjustment; the availability of granular clinical indicators was constrained; the study sample may lack full generalizability to the broader surgical population; and long-term postoperative outcomes were not evaluated. To confirm these associations, elucidate potential pathophysiological mechanisms, and assess whether targeted perioperative interventions can mitigate complications and costs, future research should prioritize prospective, multicenter investigations incorporating detailed clinical phenotyping and extended follow-up periods. In addition, the NIS does not provide information on OSA severity, CPAP use, treatment status, or treatment adherence, and information on immediate versus delayed reconstruction was not available for adjustment. Moreover, although hospitalizations with non-routine discharge dispositions were not excluded, these patients may represent a clinically more complex subgroup with a greater burden of comorbidities and higher baseline severity, such that some adverse outcomes may have been related, at least in part, to underlying illness severity rather than OSA itself. In addition, the analytic cohort excluded records with missing data in variables required for analysis as well as predefined ineligible hospitalizations (e.g., age < 18 years, in-hospital death, and sex-based exclusions), which may have introduced selection bias and limited the generalizability of the findings.

Conclusion

A comprehensive retrospective investigation reveals that OSA profoundly influences patient outcomes following breast reconstruction surgery. Individuals diagnosed with OSA demonstrate an elevated occurrence of cardiovascular and respiratory complications, encompassing conditions such as thrombocytopenia, cardiac insufficiency, and pulmonary failure. Moreover, these patients necessitate extended hospitalizations and are associated with substantially higher healthcare expenditures. The lower odds of selected local wound complications should be interpreted cautiously and may reflect residual confounding or procedure-related differences rather than a true protective effect. Consequently, these observations emphasize the paramount importance of thorough preoperative OSA screening, refinement of perioperative management strategies, and intensified postoperative surveillance to enhance surgical results and mitigate the financial strain for high-risk patients undergoing breast reconstruction.

Authors’ contributions

Author Contributions StatementYinPing Li: Conceptualization, data curation, formal analysis, investigation, methodology, project administration, resources, supervision, validation, visualization, writing – original draft, writing – review & editing.Haoxi Feng: Data curation, formal analysis, investigation, methodology, software, validation, visualization, writing – original draft, writing – review & editing.LiBing He: Data curation, formal analysis, investigation, methodology, validation, visualization, writing – original draft, writing – review & editing.Wenhua Yang: Conceptualization, funding acquisition, project administration, resources, supervision, writing – review & editing.Hao Xie: Conceptualization, funding acquisition, project administration, resources, supervision, writing – review & editing.RuJie Mo: Conceptualization, funding acquisition, project administration, resources, supervision, writing – review & editing.All authors have read and approved the final manuscript. YinPing Li, Haoxi Feng, and LiBing He contributed equally as co-first authors; Wenhua Yang, Hao Xie, and RuJie Mo share corresponding authorship.

Funding

This study received no direct funding from any third-party donor or funding institution in the public, commercial, or non-profit sectors.

Data availability

This investigation utilized data obtained from the Nationwide Inpatient Sample (NIS), a comprehensive administrative database maintained under the Healthcare Cost and Utilization Project by the Agency for Healthcare Research and Quality (AHRQ). The NIS represents the largest publicly accessible all-payer inpatient healthcare database in the United States, encompassing substantial volumes of hospitalization records across diverse healthcare facilities.Access to the database can be obtained through the official AHRQ portal at https://www.ahrq.gov/data/hcup/index.html. Given that this research employed exclusively de-identified, publicly available datasets, the study was classified as exempt from institutional review board oversight and informed consent requirements under federal research regulations.In accordance with HCUP data use agreements and privacy protection protocols, the authors are unable to redistribute individual patient records or aggregated datasets. Researchers seeking to replicate this analysis should obtain independent access authorization through the aforementioned AHRQ data portal following standard application procedures and licensing agreements.

Declarations

Ethics approval and consent to participate

Ethical review and approval were not required for this investigation as it utilized exclusively de-identified, publicly available data from the Nationwide Inpatient Sample database, the largest all-payer inpatient healthcare database in the United States. This study did not involve direct human participant recruitment, intervention, or interaction by the authors. Given the use of anonymized administrative data already in the public domain, the research was classified as exempt from Institutional Review Board oversight under federal regulations governing human subjects research.

Consequently, informed consent procedures were not applicable, and no additional ethical permissions were necessary. All methods are carried out following relevant guidelines and regulations.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.