This population-based, retrospective observational study extracted data from the US NIS database, which is the largest continuous inpatient care database in the US and includes data from about 8 million hospital stays each year [16]. The database is administered by the Healthcare Cost and Utilization Project (HCUP) of the US National Institutes of Health (NIH). The patient data consist of primary and secondary diagnoses, primary and secondary procedures, admission and discharge status, patient demographics, projected payment source, hospital stay duration, and hospital characteristics (i.e., bed size, location, teaching status, and hospital area). We initially consider all hospitalized patients for inclusion in the study. The continuously updated, annual NIS database contains patient information from around 1050 hospitals in 44 states, representing a stratified sample of 20% of US community hospitals as defined by the American Hospital Association.

All data were obtained through a request to the Online HCUP Central Distributor (available at: https://www.distributor.hcup-us.ahrq.gov/), which administers the database (certificate HCUP-6CVV58M82). This study conforms to the NIS data-use agreement with HCUP. Because this study analyzed secondary data from the NIS database, patients and the public were not involved directly. The study protocol was submitted to the institutional review board (IRB) of our hospital, which exempted the study from IRB approval. Since all data in the NIS database are de-identified, the requirement for informed consent was also waived.

Data from patients hospitalized with an ACL injury who received reconstruction surgery between 2005 and 2018 were extracted. Patients with a concomitant diagnosis of posterior cruciate ligament (PCL) disruption or with missing study variables of interest were excluded. All diagnoses and procedures were identified through the International Classification of Diseases, Ninth Revision and Tenth Revision, Clinical Modification (ICD-9-CM, ICD-10-CM) and Procedure Coding System (ICD-9-PCS, ICD-10-PCS), listed in Supplementary Table S1. Patients aged < 20 years, those with a concomitant diagnosis of PCL disruption, and those with missing information were excluded. Patients were then divided into two groups based on their BMI: the non-obese group and the obese group (BMI ≥ 30 kg/m2), with obesity status confirmed through corresponding diagnostic codes.

Primary study outcomes were concomitant meniscus injury, length of hospital stay (LOS), post-procedural complications, and non-routine discharge. LOS was calculated by subtracting the admission date from the discharge date. Post-procedural complications, including venous thromboembolism (VTE), pneumonia, infection, bleeding complication, major blood loss, wound dehiscence, acute kidney injury (AKI), urinary tract infection (UTI), failure of reconstruction (defined as stiffness, effusion, instability, and post-procedural pain), hemarthrosis/joint fistula, post-traumatic osteoarthritis, and any other complication were identified in the patient records. Non-routine discharge was defined as discharge to a long-term care facility.

The patients’ demographic and clinical data were analyzed, including age, sex, insurance status/primary payer, household income, smoking, study year, weekend admission, and Elixhauser comorbidities. The comorbidities of interest were alcohol abuse, anemia, rheumatoid arthritis/collagen vascular diseases, congestive heart failure, chronic pulmonary disease, coagulopathy, depression, uncomplicated diabetes, complicated diabetes, drug abuse, hypertension, hypothyroidism, liver disease, fluid/electrolyte disorders, neurological disorders, paralysis, peripheral vascular disorders, psychoses, pulmonary circulation disorders, renal failure, valvular disease, and weight loss [17]. The codes used to identify the complications and comorbidities are also listed in Supplementary Table 1.

The NIS database covers 20% of the US annual inpatient admissions. Weighted samples (TRENDWT before 2011; DISCWT after 2012), strata (NIS_STRATUM), and clusters (HOSPID) were used to generate national estimates for all analyses. TRENDWT and DISCWT are weights to discharges in the universe, NIS_STRATUM is used to post-stratify hospitals for the calculation of universe and frame weights, and HOSPID is the HCUP hospital identification number. The SURVEY procedure in the SAS software was employed for analyzing sample survey data. Categorical data were presented as the number (n) and weighted percentage (%), and continuous data were presented as the mean and standard error (SE). PROC SURVEYFREQ was used for analyzing categorical data, while the PROC SURVEYREG procedure was used for analyzing continuous data. To further balance the baseline characteristics of the comparison groups, the study population was matched using the propensity-score-matching (PSM) method based on age, sex, and study year, with a 1:2 ratio of patients with and without obesity.

Associations between the study variables and the dichotomized outcomes were determined using logistic regression analysis with the PROC SURVEYLOGISTIC statement, and they are presented as odds ratios (ORs) and 95% confidence intervals (CIs). Linear regression analysis was employed to estimate the relation of LOS to the study variables using the PROC SURVEYREG statement, and the results are presented as beta and 95% CI. In cases where significant variables were identified for outcomes, these were included in multivariable regression models for adjustments when comparing differences between the obesity and non-obesity groups. All p values were two-sided, and the level of significance was set at 0.05. All statistical analyses were performed using SAS software version 9.4 (SAS Institute Inc., Cary, NC, USA).