This population-based, retrospective observational study extracted all data from the US Nationwide Inpatient Sample (NIS) database, which is the largest all-payer, continuous inpatient care database in the United States, including about 8 million hospital stays each year (11). The database is administered by the Healthcare Cost and Utilization Project (HCUP) of the US National Institutes of Health (NIH) (https://hcup-us.ahrq.gov/db/nation/nis/nisdbdocumentation.jsp). Patient data include primary and secondary diagnoses, primary and secondary procedures, admission and discharge status, patient demographics, expected payment source, duration of hospital stay, and hospital characteristics (i.e., bed size/location/teaching status/hospital region). All admitted patients are initially considered for inclusion. The continuous, annually updated NIS database derives patient data from about 1,050 hospitals from 44 States in the US, representing a 20% stratified sample of US community hospitals as defined by the American Hospital Association (12).

All data were obtained through request to the Online HCUP Central Distributor. This study conforms to the NIS data-use agreement. Because this study analyzed pre-collected, de-identified data from the NIS and the public were not involved directly, our hospital exempted further Institutional Review Board (IRB) approval after the study protocol was submitted.

The data of adults aged 20 to 79 years who had been diagnosed with pituitary adenoma and had undergone transsphenoidal pituitary surgery between 2005 and 2018 were extracted from the NIS database. Diagnoses and surgeries were confirmed using the International Classification of Diseases, Ninth Revision (ICD-9) and Tenth (ICD-10) diagnostic codes for pituitary adenoma (ICD-9: 227.3; ICD-10: D35.2) and transsphenoidal pituitary surgery (ICD-9-PCS: 07.62, 07.65; ICD-10-PCS: 0GB03, 0GB04, or 0GT04). Patients lacking complete data of in-hospital mortality, discharge destination; hospital costs, sex, household income, primary payer, admission type and hospital bed size were excluded.

The main exposure variable of MetS was defined according to the diagnostic codes for MetS (ICD-9: 277.7; ICD-10: E88.81) or having at least three of the five components of MetS, including BMI ≥ 30 kg/m² (ICD-9: V85.3-V85.4; ICD-10: E66.0-E66.2, E66.8, E66.9, Z68.3-Z68.4), blood pressure ≥ 130/85 mmHg (ICD-9 codes: 401-405; ICD-10: I10-I16), fasting glucose ≥ 100 mg/dL (5.6 mmol/L) (ICD-9 codes: 250.x0, 250.x2; ICD-10: E11), serum HDL-C < 40 mg/dL in men and < 50 mg/dL in women (ICD-9: 272.5-272.6; ICD-10: E78.6, E88.1) and serum triglycerides ≥ 150 mg/dL (ICD-9 codes: 272.1-272.4; ICD-10: E78.1-E78.5). The definition we utilized is derived from the International Diabetes Federation (13).

Study outcomes were incidence of: 1) poor outcomes, a composite outcome defined by in-hospital mortality or unfavorable discharge (i.e., transfer to long-term care facilities); 2) pituitary-related complications defined by ICD diagnostic codes, including pituitary apoplexy, panhypopituitarism, corticoadrenal insufficiency, acquired hydrocephalus, meningitis, cerebrospinal fluid (CSF) rhinorrhea, diabetes insipidus, hypernatremia, and syndrome of inappropriate antidiuretic hormone secretion (SIADH); 3) presence of any patient safety indicators (PSIs); 4) prolonged length of stay (LOS), defined as LOS >= 75^th percentile; and 5) hospital costs. Specifically, the PSIs used in the present study were developed by the Agency for Healthcare Research and Quality’s (AHRQ) of the US to simplify the common inpatient complications or adverse events that indicate compromised patient safety. The PSIs include as follow: anesthetic complications – misplaced endotracheal tube, adverse effects in therapeutic use, other CNS depressants and anesthetics, poisoning by other CNS depressants and anesthetics, pressure ulcer, foreign body retained after procedure/surgery, iatrogenic pneumothorax, central venous line infection, postoperative hip fracture, postoperative physiologic & metabolic derangement – secondary diabetes with ketoacidosis, postoperative hemorrhage, postoperative respiratory failure, deep vein thrombosis & pulmonary embolism, sepsis, postoperative wound dehiscence, accidental puncture or laceration and transfusion reaction. The details and validation of the PSI in pituitary tumor patients was previously described elsewhere (14, 15).

Patients’ baseline characteristics included age, sex, household income, insurance status (primary payer), smoking status, and severity of comorbidities assessed by Charlson comorbidity index (CCI). Weekend admission, emergent admission and hospital-related characteristics (bed number, location/teaching status, hospital region) were also extracted from the database as part of the comprehensive data available for all patients.

Weighted samples (TRENDWT and DISCWT used before 2011 and after 2012, respectively), stratum (NIS_STRATUM), and cluster (HOSPID) were used to estimate the nationwide analyses. Descriptive statistics are expressed as unweighted counts (n) and weighted percentages (%) for categorical data with the SAS procedure of PROC SURVEYFREQ and mean ± standard error (SE) with PROC SURVEYMEANS for continuous data. Comparison between groups were conducted using the same SAS procedures. Logistic regressions were used to calculate the odds ratios (ORs) and 95% confidence intervals (CIs) to determine associations between MetS and outcomes using SAS procedure of PROC SURVEYLOGISTIC. Linear regressions were performed to analyze beta estimates and 95% CI of each factor associated with total hospital costs using PROC SURVEYREG. Multivariate models were adjusted for variables that were significant (p < 0.05) in univariate analyses. Stratified analysis to identify the associations between pituitary-related complication and MetS according to type of pituitary adenoma (secreting or non-secreting) was also performed. The statistical software package SAS software version 9.4 (SAS Institute Inc., Cary, NC, USA) was used for all statistical analyses.

Patients’ characteristics and outcomes following TSS for pituitary adenoma are summarized in Table 1 . Mean age of the study cohort was 52.1 ± 0.1 years, while 9,649 (50.6%) patients were females and 10,012 (52.6%) patients were White. Mean total hospital costs was 69.8 ± 1.2 thousand USD. Patients with MetS were older, and a greater percentage were female, Black, with lower household income, and insurance was covered by Medicare/Medicaid as compared to those without MetS. Higher percentages of smoking (26.1% vs. 20.5%, p < 0.001) and greater Charlson comorbidity index (CCI) (1: 20.8 vs. 14.3%; 2+: 14.8% vs. 6.1%, p < 0.001) were observed in patients with MetS. Non-secreting tumors were fewer in patients with MetS than in those without MetS (76.8% vs. 83.2%, p < 0.001). Higher frequencies of emergent admission (17.9% vs. 13.8%, p < 0.001), weekend admission (4.4% vs. 3.4%, p = 0.032), and staying in mid-Western or Southern hospitals rather than other regions (Midwest: 24.8% vs. 18.4%; South: 38.8% vs. 36.1%, p < 0.001) were also observed in patients with MetS. Distribution of the components of MetS in the entire study cohort is summarized in Supplementary Table S1 .

Postoperative outcomes of TSS are documented in Table 2 , including: 2.1% patients had pituitary-related complications, 3.1% had poor outcomes after surgery, 19.2% had prolonged LOS, and 4.2% had the presence of any PSI. Patients with MetS also had an overall greater proportion of poor outcomes (5.6% vs. 2.8%, p < 0.001), prolonged LOS (25.5% vs 18.4%, p < 0.001), and any PSI (6.7% vs 3.9%, p < 0.001). More specifically, among pituitary-related complications, patients with MetS had significantly more CSF rhinorrhea (6.5% vs 5.3%, p < 0.027) and hypernatremia (6.5% vs. 5.3%, p < 0.001).

Associations between MetS and pituitary-related complications, poor outcomes, prolonged LOS, PSI, and total hospital costs are summarized in Table 3 . After adjusting for all relevant confounders from univariate analysis in the multivariable models, MetS was not significantly associated with the presence of pituitary-related complications and poor outcomes. Although MetS was significantly associated with a slightly increased risk for prolonged LOS (adjusted OR (aOR) = 1.19; 95% CI: 1.05-1.34). In addition, subjects with MetS were more prone to have PSIs (aOR = 1.31; 95% CI: 1.07-1.59) and higher hospital costs (adjusted β = 8.63 thousand USD; 95% CI: 4.98- 12.29). The complete models of the associations are shown in Supplementary Tables S2 , S3 .

Sensitivity analyses conducted on patients who met at least two of five criteria for MetS revealed similar findings, as shown in Supplementary Table S4 .

Associations between MetS and the specific pituitary-related complications, including pituitary apoplexy, panhypopituitarism, corticoadrenal insufficiency, acquired hydrocephalus, meningitis, CSF rhinorrhea, diabetes insipidus, hypernatremia, and SIADH are documented in Table 4 and Figure 2 . After adjusting multivariate analysis for significant variables from univariate analysis, patients with MetS had a significantly higher risk of CSF rhinorrhea (aOR, 1.22; 95% CI: 1.01-1.47) and a lower risk of diabetes insipidus (aOR, 0.83; 95% CI: 0.71-0.97). Sensitivity analyses conducted on patients who met at least two of five criteria for MetS showed no significant association with CSF rhinorrhea. However, similar to the original results, there was a significantly reduced risk of developing SIADH, with an aOR of 0.77 (95% CI: 0.61-0.96), as detailed in Supplementary Table S5 .

We further carried out stratified analysis according to type of pituitary adenomas (secreting and non-secreting). The results are shown in Table 5 . In patients with non-secreting pituitary adenoma, those with MetS exhibit a notably increased risk of prolonged LOS (aOR, 1.15; 95% CI: 1.01-1.32) and PSI (aOR, 1.33; 95% CI: 1.07-1.65). Meanwhile, patients with secreting pituitary adenoma and MetS display a significantly elevated risk of poor outcome (aOR, 2.04; 95% CI: 1.22-3.42) and prolonged LOS (aOR, 1.49; 95% CI: 1.17-1.90).

First, the present study is inherently limited by its retrospective and observational nature, which limits the extent to which results can be generalized to other populations. Second, the possibility of coding errors exists as in other studies that used ICD code systems. Third, the intraoperative parameters such as duration of operation, blood loss, tumor size and biochemical data were not recorded by the NIS database thus could not be further investigated. Also, microscopic and endonasal endoscopic procedures could not be distinguished. Fourth, because data for each inpatient is confined to the time of a single hospitalization, this study also lacks follow-up data after discharge, precluding the analyses on the late complications after TSS. Fifth, conditions such as acromegaly and Cushing’s disease, which are linked to MetS components, could potentially be subject to underdiagnosis or divergent classification. Sixth, another important limitation of this study is the unavailability of data on unplanned admissions post-discharge. While studies have highlighted the significance of 30-day readmission rates as a quality indicator for transsphenoidal pituitary surgery, our study could not evaluate this due to data constraints (24, 25).