Abstract
Purpose
Uninsured and underinsured patients face notable healthcare disparities in neurosurgery, but limited literature exists on the impact of insurance on non-functioning pituitary adenomas (NFPAs). We investigated how insurance affects outcomes of endoscopic transsphenoidal pituitary surgery (ETPS) for NFPAs.
Methods
We retrospectively reviewed NFPA patients who underwent ETPS at our institution from 2012 to 2023. Patients were grouped by insurance status, and insured patients were further subcategorized by insurance providers. Bivariate analyses used Fisher’s exact, chi-square, and t-tests. ANOVA or Kruskal-Wallis tests were applied for ≥ 2 groups. Logistic regression identified relationships between binomial variables and insurance.
Results
Our cohort (n = 651, 56.93 ± 15.53 years, 52.53% male) included 611 insured and 40 uninsured patients. Uninsured patients had lower preoperative KPS, higher rates of visual disturbances (VD), preoperative tumor volumes (TV), chiasm compression, and Knosp 4 grade, along with lower resection and longer hospital stays (LOS) (p < 0.05). Multivariate analysis showed lack of insurance was associated with increased VD (aOR 3.38), TV (aOR 2.63), Knosp 4 (aOR 3.44), subtotal resection (aOR 2.72), and prolonged LOS (aOR 7.03) (p < 0.05). When insured patients were grouped into Private (n = 361), Medicare (n = 223), and Medicaid (n = 23), Medicaid patients had larger preoperative TV, chiasm compression, Knosp 3 grade, and longer LOS (p < 0.05), with higher odds for Knosp 3 (aOR 3.00), subtotal resection (aOR 3.86), and prolonged LOS (aOR 8.38) (p < 0.05).
Conclusion
Our study highlights significant disparities in uninsured patients and those with Medicaid, underscoring the need for targeted interventions for these populations.
Keywords: Pituitary adenoma, Insurance, Outcome, Endoscopic transsphenoidal pituitary surgery
Introduction
It is well known that socioeconomic disparities influence patients’ access to health care and significantly impact outcomes following neurosurgical procedures [1–4]. Insurance status is an important socioeconomic factor and is considered a surrogate marker of social determinants of health [5]. Previous studies have demonstrated that uninsured patients and those with Medicaid often present with worse disease severity, experience increased length of hospital stay (LOS), and have a higher rate of postoperative complications following craniotomy and spine surgery [6, 7]. These factors translate to worse outcomes and higher mortality after neurosurgical procedures [6, 7].
Pituitary adenomas (PAs), although typically benign, pose a significant burden to patients and the healthcare system [8–10]. Several studies have highlighted the impact of socioeconomic factors on outcomes of endoscopic transsphenoidal surgery (ETPS) for PA [11, 12]. However, to our knowledge, there are only two studies that have specifically focused on the impact of insurance on patients with PAs, covering both functional PAs (FPAs) and non-functional PAs (NFPAs) in their cohorts [13, 14]. Given the differences in presentation, diagnosis, and treatments between FPAs and NFPAs, it is important to refine existing literature by elucidating the impact of insurance in each group specifically. Although the Affordable Care Act broadened insurance coverage [15], our experience at a large academic institution in an urban setting reveals ongoing challenges treating uninsured and underinsured patients.
There is little research focused on how insurance impacts outcomes of patients with NFPA. Therefore, in this study, we investigated the effect of insurance status and provider on outcomes of patients undergoing ETPS for NFPA. Elucidation of these disparities will aid in empowering neurosurgeons to initiate actionable changes to equilibrate future outcomes [1].
Methods
Patient
After Institutional Review Board Approval (IRB no. 20160437), a retrospective chart review was conducted of patients who underwent ETPS at our tertiary care institution from 2012 to 2023. Patients were evaluated preoperatively with magnetic resonance imaging (MRI) and hormonal workup. Surgical technique [16] and postoperative management [17] are described previously. Patients with age ≥ 18 years, NFPA as identified by presentation and preoperative hormonal workup, and histopathological confirmation of PA were included in this study. Patients with FPA were excluded. Informed consent was waived due to retrospective nature of study.
Data collection
Patient demographics and preoperative data, including age, gender, insurance status, comorbidities, preoperative Karnofsky performance score (KPS), visual disturbances (VD) as determined by neurological exam at admission as well as ophthalmological exams from prior follow-ups or ophthalmology consults, apoplexy, history of radiation to the pituitary, were collected. MRI scans were reviewed for preoperative tumor volume (TV), chiasm compression, macroadenoma (tumor diameter > 10 mm), and the probability of cavernous sinus invasion as determined by the Knosp-Steiner classification. TVs were calculated using the ellipsoid formula (length x width x height/2), based on the greatest dimensions in axial, coronal, and sagittal planes from MRI scans. Intraoperative data included CSF leak and lumbar drain use. Residual TV and extent of resection (EOR) were calculated through 24-hour postoperative MRI scans. Gross total resection (GTR) was defined as EOR equal to 100%. Data on outcomes included LOS, postoperative complications including CSF leak, diabetes insipidus (DI), hyponatremia, epistaxis, hyposmia, septal perforation, deep vein thrombosis, and vascular injury, 30-day readmission, and tumor recurrence.
Statistical analysis
Patients were categorized based on their insurance status (insured vs. uninsured) at the time of surgery. Insured patients were further categorized by type of insurance: Private, Medicare, Medicaid, and TRICARE. Categorical variables were analyzed using chi square or Fisher exact tests. For comparison of continuous variables between 2 groups, Student’s t-test or Welch’s t-tests were used depending on the equality of variance tested via Levene’s test. For comparison of continuous variables between ≥ 2 groups, ANOVA or Kruskal-Wallis tests were used, followed by Tukey or pairwise Wilcox tests for significant comparisons. Patients with TRICARE were excluded from insurance subgroup analysis due to low sample size (n = 4). Mean and standard deviation were reported for all continuous variables, except for preoperative KPS and LOS, where median and interquartile range (25-75th percentile) were used due to non-normal distribution.
Univariate analysis via logistic regression with cross validation was conducted to examine relationships between binomial outcome variables and insurance status. Linear variables were converted to binomial format for this analysis. Prolonged LOS (PLOS), defined as 90th percentile of LOS [18], was compared with standard LOS. TVs were divided into 2 groups based on the median TV and EOR was analyzed as gross total resection (GTR) or subtotal resection (STR). Variables were then assessed in a multivariate model controlling for age, gender, and preoperative KPS to obtain an adjusted odds ratio (aOR). Statistical analyses were performed using Python version 3.11.5 for Windows.
Results
Insurance status: descriptive statistics
During the study period, out of 1070 patients who underwent ETPS at our institution, 651 patients met the inclusion criteria. Of these, 611 (mean age 57.37 ± 15.56 years, 52.70% male) were insured and 40 (mean age 49.12 ± 13.80 years, 50.00% male) were uninsured. There was no significant difference in gender, comorbidities, history of radiation, and apoplexy between groups (Table 1). Insured patients were significantly older (p ≤ 0.001). Uninsured patients had significantly lower preoperative KPS compared to insured patients (85 [80–90] vs. 90 [80–90], p = 0.004). Uninsured patients had higher rates of VD (82.50% vs. 55.81%, p = 0.002) and chiasm compression (87.18% vs. 75.41%, p = 0.009), higher preoperative TV (13.68 ± 16.89 vs. 6.62 ± 10.27 cm3, p = 0.005), and Knosp 4 grade (22.50% vs. 8.35%, p = 0.007).
Table 1.
Patient demographics, clinical, and perioperative characteristics for insured vs. uninsured group
| Variable | Insured (n = 611) | Uninsured (n = 40) | p value |
|---|---|---|---|
| Patient demographics: | |||
| Age (years), mean, SD | 57.37 ± 15.56 | 49.12 ± 13.80 | < 0.001 |
| Gender, male | 322 (52.70%) | 20 (50.00%) | 0.867 |
| Preop KPS, median, IQR | 90 (80–90) | 85 (80–90) | 0.004 |
| BMI, mean, SD | 29.48 ± 6.14 | 30.91 ± 10.00 | 0.447 |
| Diabetes mellitus | 173 (28.31%) | 11 (27.50%) | 1.000 |
| Hypertension | 326 (53.35%) | 24 (60.00%) | 0.514 |
| Active smoker at time of surgery | 46 (7.53%) | 2 (5.00%) | 0.760 |
| Hx of radiation | 6 (0.98%) | 0 (0.00%) | 1.000 |
| VD | 341 (55.81%) | 33 (82.50%) | 0.002 |
| Apoplexy | 38 (6.22%) | 4 (10.00%) | 0.317 |
| Adenoma characteristics: | |||
| Chiasm compression | 460 (75.41%) | 34 (87.18%) | 0.009 |
| Macroadenoma | 576 (94.27%) | 37 (92.50%) | 0.502 |
| Knosp grades: | |||
| Grade 0 | 98 (16.20%) | 10 (25.00%) | 0.221 |
| Grade 1 | 137 (22.42%) | 6 (15.00%) | 0.367 |
| Grade 2 | 110 (18.00%) | 4 (10.00%) | 0.281 |
| Grade 3 | 214 (35.02%) | 11 (27.50%) | 0.425 |
| Grade 4 | 51 (8.35%) | 9 (22.50%) | 0.007 |
| Preop TV (cm3), mean, SD | 6.62 ± 10.27 | 13.68 ± 16.89 | 0.005 |
| Operative characteristics: | |||
| Intraop CSF leak | 333 (54.50%) | 27 (70.00%) | 0.081 |
| Lumbar drain | 13 (2.13%) | 1 (2.50%) | 0.592 |
| Residual TV (cm3), mean, SD | 0.35 ± 1.43 | 2.85 ± 7.18 | 0.011 |
| EOR (%), mean, SD | 96.46 ± 9.60 | 85.44 ± 26.03 | 0.017 |
| Treatment outcomes: | |||
| LOS (days), median, IQR | 3 (3–4) | 5 (3–7) | < 0.001 |
| Overall complications | 354 (57.93%) | 27 (67.50%) | 0.306 |
| Postop CSF leak | 20 (3.27%) | 2 (5.00%) | 0.639 |
| Transient DI | 280 (45.83%) | 20 (50.00%) | 0.727 |
| Permanent DI | 67 (10.96%) | 6 (15.00%) | 0.576 |
| Hyponatremia | 70 (11.46%) | 7 (17.50%) | 0.371 |
| Epistaxis | 47 (7.69%) | 3 (7.50%) | 1.000 |
| Hyposmia | 30 (4.91%) | 0 (0.00%) | 0.247 |
| Septal perforation | 23 (3.76%) | 1 (2.50%) | 1.000 |
| DVT | 4 (0.65%) | 0 (0.00%) | 1.000 |
| Vascular injury | 1 (0.16%) | 1 (2.50%) | 0.119 |
| 30-day neurosurgical readmission | 7 (1.15%) | 2 (5.00%) | 0.101 |
| Postop radiation | 27 (4.42%) | 3 (7.50%) | 0.421 |
| Tumor recurrence | 29 (4.75%) | 4 (10.00%) | 0.137 |
| Follow up (days), median, IQR | 372 [57.5–716] | 365 [31–910] | 0.872 |
Bold entries signify statistical significance, p < 0.05
Preop, Preoperative; BMI, Body mass index; intraop, Intraoperative; KPS, Karnofsky performance score; hx, History; VD, Visual disturbances; TV, tumor volume; EOR, extent of resection; LOS, length of stay; postop, postoperative; DI, diabetes insipidus; DVT, deep vein thrombosis
Regarding operative characteristics, there were no significant difference in intraoperative CSF leaks and use of lumbar drains between groups. Uninsured patients experienced significantly lower EOR (85.44 ± 26.03% vs. 96.46 ± 9.60%, p = 0.017). Following surgery, uninsured patients experienced longer LOS (5 [3–7] vs. 3 [3–4] days, p ≤ 0.001). While there was an increased rate of overall complications (67.50% vs. 57.93%, p = 0.306), postoperative CSF leak (5.00% vs. 3.27%, p = 0.639) and hyponatremia (17.50% vs. 11.46%, p = 0.371) in uninsured group, the difference was not statistically significant. There were no differences in other postoperative complications, 30-day neurosurgical readmission, or follow up between groups (Table 1). While there was a trend towards higher rate of tumor recurrences in the uninsured group (10.00% vs. 4.75%), the difference was not statistically significant (p = 0.137).
Insurance status: regression analysis
Univariate logistic regression revealed that patients with greater age were less likely to be uninsured (OR 0.25 [0.10, 0.58]; p = 0.001). Patients with lower KPS, VD, Knosp 4 grade, higher preoperative TV, STR, and PLOS were more likely to be uninsured (p < 0.05). Upon multivariate regression controlling for age, gender, and preoperative KPS, VD (aOR 3.38 [1.33, 8.59]; p = 0.010), higher preoperative TV (aOR 2.63 [1.15, 6.01]; p = 0.022), Knosp 4 grade (aOR 3.44 [1.42, 8.36]; p = 0.006), STR (aOR 2.72 [1.27, 5.84]; p = 0.010), and PLOS (aOR 7.03 [2.97, 16.63]; p ≤ 0.001) were significantly associated with uninsured status (Table 2; Fig. 1A).
Table 2.
Univariate and multivariate logistic regression of factors associated with lack of insurance
| Variable (ref: Insured) | Uninsured | |||
|---|---|---|---|---|
| OR (95% CI) | P value | aOR (95% CI) | P value | |
| Patient demographics: | ||||
| Age (> median) | 0.25 (0.10, 0.58) | 0.001 | ||
| Gender, male | 0.86 (0.42,1.76) | 0.678 | ||
| Preop KPS (< median) | 2.18 (1.07,4.46) | 0.034 | ||
| VD | 3.32 (1.34,8.20) | 0.009 | 3.38 (1.33, 8.59) | 0.010 |
| Apoplexy | 1.77 (0.51, 6.18) | 0.371 | 0.94 (0.23, 3.84) | 0.925 |
| Adenoma characteristics: | ||||
| Chiasm compression | 2.09 (0.72, 6.07) | 0.178 | 2.19 (0.73, 6.58) | 0.163 |
| Knosp grades: | ||||
| Grade 0 | 2.03 (0.91, 4.55) | 0.086 | 1.89 (0.81, 4.37) | 0.139 |
| Grade 1 | 0.37 (0.11, 1.22) | 0.102 | 0.41 (0.12, 1.40) | 0.155 |
| Grade 2 | 0.32 (0.07, 1.35) | 0.120 | 0.28 (0.07, 1.22) | 0.090 |
| Grade 3 | 0.85 (0.39, 1.83) | 0.669 | 0.80 (0.36, 1.76) | 0.579 |
| Grade 4 | 3.21 (1.37, 7.56) | 0.008 | 3.44 (1.42, 8.36) | 0.006 |
| Preop TV (cm3) (> median) | 2.65 (1.20, 5.85) | 0.016 | 2.63 (1.15, 6.01) | 0.022 |
| Operative characteristics: | ||||
| Intraop CSF leak | 1.73 (0.80, 3.72) | 0.164 | 1.66 (0.76, 3.62) | 0.205 |
| Lumbar drain | 1.28 (0.16, 10.18) | 0.814 | 1.58 (0.19, 13.04) | 0.671 |
| Postop TV (cm3) (> median) | 2.65 (1.29, 5.45) | 0.008 | 2.89 (1.37, 6.14) | 0.006 |
| STR | 2.59 (1.26, 5.32) | 0.010 | 2.72 (1.27, 5.84) | 0.010 |
| Treatment outcomes: | ||||
| PLOS | 8.31 (3.66, 18.87) | < 0.001 | 7.03 (2.97, 16.63) | < 0.001 |
| Overall complications | 1.47 (0.68, 3.18) | 0.325 | 1.33 (0.60, 2.94) | 0.480 |
| Postop CSF leak | 0.90 (0.12, 6.95) | 0.916 | 0.92 (0.12, 7.23) | 0.934 |
| Transient DI | 1.16 (0.60, 2.38) | 0.678 | 1.00 (0.48, 2.09) | 0.996 |
| Permanent DI | 1.62 (0.64, 4.09) | 0.308 | 1.76 (0.68, 4.54) | 0.242 |
| Hyponatremia | 0.73 (0.22, 2.46) | 0.606 | 0.70 (0.21, 2.41) | 0.576 |
| Epistaxis | 1.05 (0.31, 3.57) | 0.942 | 1.06 (0.31, 3.69) | 0.924 |
| Hyposmia | NA | NA | NA | NA |
| Septal perforation | 0.76 (0.10, 5.82) | 0.789 | 0.57 (0.07, 4.57) | 0.596 |
| DVT | NA | NA | NA | NA |
| Vascular injury | NA | NA | NA | NA |
| 30-day neurosurgical readmission | 4.59 (0.92, 23.07) | 0.064 | 5.13 (0.99, 26.47) | 0.051 |
| Postop radiation | 1.84 (0.53, 6.45) | 0.339 | 1.47 (0.41, 5.29) | 0.554 |
| Tumor recurrence | 2.54 (0.83, 7.80) | 0.102 | 2.23 (0.72, 6.92) | 0.168 |
Bold entries signify statistical significance, p < 0.05
Preop, Preoperative; KPS, Karnofsky performance score; VD, Visual disturbances; TV, Tumor volume; intraop, Intraoperative; postop, Postoperative; STR, Subtotal resection; PLOS, Prolonged length of stay; DI, Diabetes insipidus; DVT, Deep vein thrombosis
Fig. 1.
Forest plots with adjusted odds ratios of variables associated with A) uninsured, B) medicare, and C) medicaid group. *p < 0.05, **p < 0.01, and ***p < 0.001
Insurance subgroups: descriptive statistics
Insured patients were subcategorized based on their insurance providers into Private (n = 361, mean age 48.77 ± 12.01 years, 50.41% male), Medicare (n = 223, mean age 72.22 ± 7.48 years, 57.40% male), and Medicaid (n = 23, mean age 50.70 ± 15.52 years, 47.83% male). There was no significant difference in gender, BMI, smoking status, history of radiation, and apoplexy between groups (Table 3). Patients with private insurance had lower rates of diabetes mellitus and hypertension. While Medicaid group had lower preoperative KPS and higher rates of VD than Private and Medicare, the differences were not statistically significant. Moreover, Medicaid patients had higher preoperative TVs (p = 0.006), higher rates of chiasm compression (p = 0.007), and Knosp 3 grade (p = 0.014).
Table 3.
Patient demographics, clinical, and perioperative characteristics for insured subgroup
| Variable | Private = 361 | Medicare = 223 | Medicaid = 23 | p value |
|---|---|---|---|---|
| Patient demographics: | ||||
| Age (years), mean, SD | 48.77 ± 12.01 a | 72.22 ± 7.48 a, b | 50.70 ± 15.52 b | < 0.001 |
| Gender, male | 182 (50.41%) | 128 (57.40%) | 11 (47.83%) | 0.230 |
| Preop KPS, median, IQR | 90 (80–90) | 90 (80–90) | 80 (80–90) | 0.097 |
| BMI, mean, SD | 29.98 ± 6.53 | 28.71 ± 5.34 | 29.43 ± 6.36 | 0.043 |
| Diabetes mellitus | 87 (24.10%) | 77 (34.53%) | 9 (39.13%) | 0.013 |
| Hypertension | 162 (44.87%) | 150 (67.26%) | 14 (60.87%) | < 0.001 |
| Active smoker at time of surgery | 23 (6.37%) | 20 (8.97%) | 3 (13.04%) | 0.227 |
| Hx of radiation | 4 (1.11%) | 2 (0.90%) | 0 (0.00%) | 1.000 |
| VD | 194 (53.74%) | 129 (57.85%) | 17 (73.91%) | 0.132 |
| Apoplexy | 29 (8.03%) | 9 (4.04%) | 0 (0.00%) | 0.085 |
| Adenoma characteristics: | ||||
| Chiasm compression | 256 (71.11%) | 182 (81.61%) | 20 (86.96%) | 0.007 |
| Macroadenoma | 337 (93.61%) | 213 (95.52%) | 22 (95.65%) | 0.658 |
| Knosp grades: | ||||
| Grade 0 | 66 (18.33%) | 30 (13.45%) | 2 (8.70%) | 0.210 |
| Grade 1 | 78 (21.67%) | 54 (24.22%) | 3 (13.04%) | 0.451 |
| Grade 2 | 62 (17.22%) | 46 (20.63%) | 2 (8.70%) | 0.323 |
| Grade 3 | 129 (35.83%) | 69 (30.94%) | 14 (60.87%) | 0.014 |
| Grade 4 | 25 (6.93%) | 24 (10.76%) | 2 (8.70%) | 0.238 |
| Preop TV (cm3), mean, SD | 6.25 ± 9.56 a | 6.43 ± 7.19 | 14.28 ± 28.82 a | 0.006 |
| Operative characteristics: | ||||
| Intraop CSF leak | 194 (53.89%) | 121 (54.26%) | 15 (65.22%) | 0.570 |
| Lumbar drain | 5 (1.39%) | 7 (3.14%) | 1 (4.35%) | 0.178 |
| Residual TV (cm3), mean, SD | 0.34 ± 1.53 a | 0.35 ± 1.19 b | 0.66 ± 1.32 a, b | 0.559 |
| EOR (%), mean, SD | 96.81 ± 9.58 a | 96.17 ± 9.55 b | 93.21 ± 10.88 a, b | 0.192 |
| Treatment outcomes: | ||||
| LOS (days), median, IQR | 3 (3–3) a, c | 3 (3–4) b, c | 5 (3–6) a, b | < 0.001 |
| Overall complications | 230 (63.91%) | 111 (49.78%) | 11 (47.83%) | 0.003 |
| Postop CSF leak | 13 (3.61%) | 7 (3.14%) | 0 (0.00%) | 0.918 |
| Transient DI | 195 (54.17%) | 74 (33.18%) | 9 (39.13%) | < 0.001 |
| Permanent DI | 39 (10.83%) | 25 (11.21%) | 2 (8.70%) | 0.971 |
| Hyponatremia | 41 (11.31%) | 23 (10.31%) | 5 (21.74%) | 0.259 |
| Epistaxis | 25 (6.94%) | 19 (8.52%) | 3 (13.04%) | 0.398 |
| Hyposmia | 14 (3.89%) | 14 (6.28%) | 2 (8.70%) | 0.215 |
| Septal perforation | 15 (4.16%) | 7 (3.14%) | 1 (4.35%) | 0.674 |
| DVT | 0 (0.00%) | 4 (1.79%) | 0 (0.00%) | 0.033 |
| Vascular injury | 0 (0.00%) | 1 (0.45%) | 0 (0.00%) | 0.406 |
| 30-day neurosurgical readmission | 2 (0.56%) | 5 (2.24%) | 0 (0.00%) | 0.200 |
| Postop radiation | 18 (5.00%) | 6 (2.69%) | 3 (13.04%) | 0.052 |
| Tumor recurrence | 19 (5.28%) | 9 (4.04%) | 1 (4.35%) | 0.137 |
| Follow up (days), median, IQR | 373 [81.75–727] | 366.5 [41–694] | 312 [35.5-526.5] | 0.171 |
a, b, c denote significance upon post-hoc analysis
Preop, Preoperative; BMI, Body mass index; intraop, Intraoperative; KPS, Karnofsky performance score; hx, History; VD, Visual disturbances; TV, Tumor volume; EOR, Extent of resection; LOS, Length of stay; postop, Postoperative; DI, Diabetes insipidus; DVT, Deep vein thrombosis
There were no differences in operative characteristics between insured groups (Table 3). Postoperatively, patients with Medicaid insurance experienced longer LOS compared to those with private insurance and Medicare (5 [3–6] vs. 3 [3–3] vs. 3 [3–4] days respectively, p ≤ 0.001). There was an increase in overall complications and transient DI in Private group and DVT in Medicare group (p < 0.05). However, there were no differences in other postoperative complications, 30-day neurosurgical readmissions or tumor recurrence between groups.
Insurance subgroups: regression analysis
On multivariate analysis, patients with Knosp 3 grade (aOR 3.00 [1.13, 7.95]; p = 0.027), STR (aOR 3.86 [1.44, 10.34]; p = 0.007), and PLOS (aOR 8.38 [2.76, 25.48]; p < 0.001) were more likely to have Medicaid insurance. Patients with Medicare had higher odds of PLOS as well (aOR 4.28 [1.22, 15.03]; p = 0.023) (Table 4; Fig. 1B and C).
Table 4.
Multivariate logistic regression of factors associated with Medicare and Medicaid insurance
| Variable (ref: Private) | Medicare | Medicaid | ||
|---|---|---|---|---|
| aOR (95% CI) | P value | aOR (95% CI) | P value | |
| Patient demographics: | ||||
| VD | 0.86 (0.42, 1.77) | 0.683 | 2.09 (0.71, 6.11) | 0.180 |
| Apoplexy | 2.97 (0.40, 21.98) | 0.287 | NA | NA |
| Adenoma characteristics: | ||||
| Chiasm compression | 2.15 (0.92, 5.02) | 0.076 | 3.09 (0.68, 14.00) | 0.144 |
| Knosp grades: | ||||
| Grade 0 | 0.79 (0.32, 1.98) | 0.617 | 0.57 (0.13, 2.61) | 0.468 |
| Grade 1 | 1.06 (0.50, 2.25) | 0.884 | 0.42 (0.09, 1.90) | 0.260 |
| Grade 2 | 1.04 (0.45, 2.40) | 0.928 | 0.28 (0.04, 2.19) | 0.227 |
| Grade 3 | 0.77 (0.38, 1.55) | 0.460 | 3.00 (1.13, 7.95) | 0.027 |
| Grade 4 | 2.31 (0.74, 7.19) | 0.150 | 1.34 (0.29, 6.23) | 0.712 |
| Preop TV (cm3) (> median) | 1.21 (0.63, 2.35) | 0.565 | 2.27 (0.79, 6.51) | 0.129 |
| Operative characteristics: | ||||
| Intraop CSF leak | 0.81 (0.42, 1.58) | 0.536 | 1.62 (0.59, 4.46) | 0.347 |
| Lumbar drain | 2.41 (0.35, 16.73) | 0.375 | 4.60 (0.47, 44.74) | 0.189 |
| Postop TV (cm3) (> median) | 0.74 (0.37, 1.50) | 0.406 | 4.21 (1.59, 11.17) | 0.004 |
| STR | 1.34 (0.69, 2.60) | 0.381 | 3.86 (1.44, 10.34) | 0.007 |
| Treatment outcomes: | ||||
| PLOS | 4.28 (1.22, 15.03) | 0.023 | 8.38 (2.76, 25.48) | < 0.001 |
| Overall complications | 1.13 (0.58, 2.22) | 0.720 | 0.53 (0.20, 1.35) | 0.180 |
| Postop CSF leak | 5.20 (0.95, 28.63) | 0.058 | NA | |
| Transient DI | 0.77 (0.40, 1.50) | 0.440 | 0.52 (0.20, 1.35) | 0.178 |
| Permanent DI | 0.85 (0.30, 2.37) | 0.754 | 0.82 (0.18, 3.73) | 0.797 |
| Hyponatremia | 0.73 (0.25, 2.16) | 0.574 | 2.07 (0.65, 6.66) | 0.221 |
| Epistaxis | 2.54 (0.66, 9.86) | 0.178 | 2.57 (0.68, 9.72) | 0.164 |
| Hyposmia | 0.48 (0.13, 1.81) | 0.281 | 2.44 (0.49, 12.07) | 0.276 |
| Septal perforation | 0.61 (0.08, 4.56) | 0.627 | 1.35 (0.16, 11.12) | 0.780 |
| DVT | NA | NA | NA | NA |
| Vascular injury | NA | NA | NA | NA |
| 30-day neurosurgical readmission | NA | NA | NA | NA |
| Postop radiation | 1.06 (0.26, 4.39) | 0.937 | 2.83 (0.74, 10.83) | 0.128 |
| Tumor recurrence | 1.06 (0.23, 4.97) | 0.943 | 0.80 (0.10, 6.39) | 0.836 |
Bold entries signify statistical significance, p < 0.05
Preop, Preoperative; KPS, Karnofsky performance score; VD, Visual disturbances; TV, Tumor volume; intraop, Intraoperative; postop, Postoperative; STR, Subtotal resection; PLOS, Prolonged length of stay; DI, Diabetes insipidus; DVT, Deep vein thrombosis
Discussion
Traditionally, employer-sponsored insurance has been the backbone of the US health coverage system, restricting healthcare access to those with the ability to work full-time [19]. As a result, older Americans and those with low incomes often did not have health insurance. In 1965, the Medicare and Medicaid programs were established in the US, extending health coverage to elderly, low-income, and disabled Americans [20]. Despite these legislative efforts, 7.9% of the US population remains uninsured [21]. These patients often have limited access to healthcare outside of emergency services [22], leading to more severe disease at the time of presentation as noted across both neurosurgical and non-neurosurgical fields [14, 23–25]. Moreover, individuals with Medicaid often experience issues with the consistency and adequacy of care due to variations in state-level Medicaid programs [26].
Study overview
In this study, we found that uninsured patients presented with more advanced disease, had higher preoperative TV leading to reduced EOR, and increased LOS compared to insured counterparts. Patients who experienced lower preoperative KPS, increased VD, preoperative TV, Knosp 4 grade, STR, and PLOS were more likely to be uninsured.
Upon subgroup analysis, Medicaid patients had increased preoperative TV, presented in advanced stage with associated comorbidities, and experienced longer LOS compared to non-Medicaid patients. Those who experienced Knosp 3 grade, STR, and PLOS were more likely to have Medicaid insurance compared to those with private insurance.
Patient presentation
Uninsured patients had a significantly lower preoperative KPS (85 [80–90] vs. 90 [80–90]; p = 0004). While Medicaid patients had lower preoperative KPS compared to Medicare and Private (80 [80–90] vs. 90 [80–90] vs. 90 [80–90]), this difference was not statistically significant. To our knowledge, there is no other literature analyzing differences in preoperative KPS by insurance status or type.
VD is one of the most prominent presentations of NFPA, with a recent meta-analysis on 35 case series reporting the incidence ranging from 28 to 100% [27]. In our study, uninsured patients were more likely to present with VD (82.50% vs. 55.81%; p = 0.002). On subgroup analysis of insured patients, Medicaid patients were more likely to present with VD than Private and Medicare patients (73.91% vs. 53.74% vs. 57.85%, respectively), although the difference did not reach statistical significance. This may be due to limited access to early care, leading to delayed presentation with more advanced symptoms. Similar to our findings, Osorio et al. noted a higher prevalence of VD among Medicaid patients compared to Private and Medicare (62.5% vs. 46.7% vs. 44.2%, respectively). However, they also failed to achieve statistical significance [10]. Contrary to Osorio et al. and our findings, Younus et al., including both NFPA and FPA in their cohorts, found no differences in rates of VD among Medicaid and non-Medicaid patients (44% vs. 43%, respectively) [13].
We also noted an increasing trend of apoplexy in uninsured patients (10.00% vs. 6.20%; p = 0.317). Jahangiri et al. previously reported that lack of insurance was significantly associated with pituitary apoplexy [28]. In their study, 11.85% of patients were uninsured as compared to 6.13% in our study. We believe that since our study had a smaller proportion of uninsured patients, the difference in rates of apoplexy did not reach statistical significance. Moreover, upon subgroup analysis, our results were similar to Younus et al. who did not show a statistically significant difference in rates of apoplexy among Medicaid and non-Medicaid groups [13]. To our knowledge, there are no other studies on rates of apoplexy within patients with different types of insurance providers.
Adenoma characteristics
Our study used the ellipsoid formula to calculate tumor volume based on the maximum dimensions in axial, coronal, and sagittal planes. Although we recognize that volume segmentation software could yield more precise measurements, we lacked segmentation-based volume data in our dataset. However, as a natural sequela to delayed presentation, we noted that uninsured patients were more likely to have larger TV(13.68 ± 16.89 vs. 6.62 ± 10.27 cm³, p = 0.005) with significantly higher rates of chiasm compression and Knosp 4 grade. Similarly, on subgroup analysis, Medicaid patients had larger TV compared to Private and Medicare groups (14.28 ± 28.82 vs. 6.25 ± 9.56 vs. 6.43 ± 7.19 cm³, respectively), higher rates of chiasm compression, and Knosp 3 grade. There is a lack of literature regarding tumor characteristics among insured and uninsured patients. However, literature combining both FPAs and NFPAs noted that Medicaid patients had larger mean tumor diameters (26.1 mm vs. 23.1 mm) [13]. Regarding the severity of cavernous sinus invasion, in studying patients with NFPA, Osorio et al. noted higher rates of cavernous sinus invasion in Medicaid patients compared to Private and Medicare patients (62.5% vs. 33.7% vs. 43.0%, respectively; p = 0.009) [10]. Younus et al., with both FPAs and NFPAs in their study, failed to show a significant difference in cavernous sinus invasion between Medicaid and non-Medicaid groups [13]. This may be since 27% of patients in their study had FPAs which typically present as microadenomas. Keeping in mind a residual tumor doubling time of 3.4 years [29], it could be extrapolated that patients lacking insurance and Medicaid roughly present with a 3-4-year delay compared to their insured counterparts.
EOR
The lower EOR in uninsured patients (85.44 ± 26.03% vs. 96.46 ± 9.60%) is likely related to greater Knosp 4 grade in those patients. This aligns with previous studies associating higher grades of cavernous sinus extension with lower EOR [30, 31]. Upon subgroup analysis, our results correlate with other studies that do not find a significant difference between EOR among patients with different insurance providers [10, 13].
LOS
Our study uniquely demonstrates that uninsured patients have a median LOS that is 2 days longer compared to insured patients (p < 0.001). Similarly, the median was 2 days longer in Medicaid patients compared to Medicare and Private patients (p < 0.001). A previous study comparing NFPA noted that private insurance patients were discharged 1 day earlier than those with Medicare and Medicaid, although this did not reach statistical significance. With both FPAs and NFPAs in their cohort, Younus et al. noted that Medicaid patients had significantly longer LOS (9.4 vs. 3.55 days) compared to non-Medicaid patients [13]. Moreover, privately insured patients experienced shorter LOS [32]. The rate of overall complications and hyponatremia was higher in uninsured patients (67.50% vs. 57.93% and 17.50% vs. 11.46%, respectively), and this may be a factor responsible for PLOS. Among insured patients, the longer LOS in Medicaid patients could be attributable to a higher incidence of hyponatremia compared to Private and Medicare patients (21.74% vs. 11.31% vs. 10.31%). Similarly, a higher prevalence of diabetes mellitus in Medicaid patients could be a contributing factor as well [33].
Postoperative complications
While the rates of overall complications, postoperative CSF leak, transient DI and hyponatremia were higher in uninsured group, the differences were not statistically significant. Notably, despite being significantly younger than the insured group, the uninsured group showed a trend toward higher complication rates. Further analysis using propensity score matching in future studies may help uncover statistically significant differences in complication rates related to insurance status. Subgroup analysis revealed an increased rate of overall complications and transient DI in Private group. However, when transient DI was excluded from the overall complication rate, the difference between groups was not significant anymore. Moreover, there were no differences in other specific complications between patients with different insurance providers which aligns with a previous study by Osorio et al. [10]. Younus et al., with FPAs and NFPAs into their cohort, found a statistically significant increase in postoperative complications in Medicaid group compared to non-Medicaid (14% vs. 7%; p < 0.05) [13]. Performing propensity score matching to further evaluate the impact of insurance on complication rate may lead to emergence of statistically significance. On the other hand, high volume facilities and surgeons with high surgical caseload have been shown to have better short-term outcomes after ETPS [34, 35], potentially equalizing our complication rates across insurance groups despite advanced disease in uninsured and Medicaid patients.
Multivariate analysis
Lack of insurance was associated with reduced preoperative KPS, higher VD, preoperative TV, Knosp 4 grade, STR, and PLOS. Similarly, Medicaid was associated with Knosp 3 grade, STR, and PLOS (p < 0.05). These results are consistent with other studies in neurosurgical literature showing uninsured patients and those with Medicaid to have more severe disease upon presentation and worse postoperative outcomes [6, 12, 28, 36]. This delay in presentation may be due to reduced access to primary care physician or specialty consultation with lack of insurance and underinsurance [36, 37].
Limitations and strengths
Our study is inherently limited by its retrospective nature. To overcome these limitations, we included only patients with complete records available. Additionally, it is a single-institution study, and our results may be influenced by the social and geopolitical setting of our institution. Therefore, larger multi-institutional studies and creation of national pituitary adenoma dataset with granular data is warranted to further investigate our findings. In our study, the probability of cavernous sinus invasion was determined using the Knosp-Steiner classification, which does not distinguish between grade 3 A and grade 3B. Future studies may consider the modified Knosp classification to enable further subgroup analysis based on these distinctions. Another limitation of our study is the absence of data on postoperative hormonal status and long-term hormone replacement needs. Future studies could help clarify the impact of insurance status on postoperative pituitary function and endocrine recovery.
Despite these limitations, our study provides valuable insights into the relationship between insurance and healthcare outcomes in patient undergoing ETPS for NFPA. Unlike existing generic database studies, we were able to collect detailed information on preoperative characteristics, tumor features, and postoperative outcomes, which are not always available in broader datasets. This granularity allows us to highlight significant healthcare disparities driven by insurance status, revealing how these factors contribute to delayed presentation, advanced disease, and prolonged length of stay.
Conclusion
Our findings reveal substantial disparities, with uninsured and Medicaid patients presenting with more advanced disease, higher rates of STR, and experiencing PLOS compared to their insured counterparts. These disparities highlight the critical role that socioeconomic factors play in healthcare access and quality, reaffirming the need for targeted interventions to address these inequities.
Author contributions
KHS: conceptualization, methodology, investigation, data curation, interpretation of data, writing - original draft, writing - review and editing, prepared Fig. 1. NS: methodology, investigation, data curation, formal analysis, interpretation of data, writing - original draft, writing - review and editing. AMK: supervision, interpretation of data, project administration, methodology, writing - review and editing. VML: supervision, interpretation of data, project administration, methodology, writing - review and editing. MEI: conceptualization, supervision, writing - review and editing. RJK: conceptualization, supervision, writing - review and editing. ZBS: conceptualization, supervision, writing - review and editing. AHS: conceptualization, interpretation of data, supervision, project administration, writing - review and editing. All authors reviewed the manuscript.
Funding
None.
Data availability
No datasets were generated or analysed during the current study.
Declarations
Disclosures
None.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Khushi H. Shah and Nikola Susic contributed equally to this work.
References
- 1.Glauser G, Detchou DK, Dimentberg R, Ramayya AG, Malhotra NR (2021) Social Determinants of Health and Neurosurgical outcomes: current state and future directions. Neurosurg Apr 15(5):E383–e390. 10.1093/neuros/nyab030 [DOI] [PubMed] [Google Scholar]
- 2.Hoh BL, Rabinov JD, Pryor JC, Carter BS, Barker FG 2 (2003) In-hospital morbidity and mortality after endovascular treatment of unruptured intracranial aneurysms in the United States, 1996–2000: effect of hospital and physician volume. AJNR Am J Neuroradiol 24(7):1409–1420 [PMC free article] [PubMed]
- 3.Hackett AM, Adereti CO, Walker AP et al (2024) Racial and socioeconomic status among a Patient Population presenting with Aneurysmal Subarachnoid Hemorrhage versus Unruptured Intracranial Aneurysm: a single-center study. Brain Sci Apr 18(4). 10.3390/brainsci14040394 [DOI] [PMC free article] [PubMed]
- 4.Gautam D, Findlay MC, Karsy M (2024) Socioeconomic and racial disparities affect Access to high-volume centers during Meningioma Treatment. World Neurosurg Jul 187:e289–e301. 10.1016/j.wneu.2024.04.076 [DOI] [PubMed] [Google Scholar]
- 5.Snyder RA, Chang GJ (2020) Insurance Status as a Surrogate for Social Determinants of Health in Cancer Clinical Trials. JAMA Netw Open 3(4):e203890. 10.1001/jamanetworkopen.2020.3890 [DOI] [PubMed]
- 6.Curry WT Jr., Carter BS, Barker FG 2 (2010) Racial, ethnic, and socioeconomic disparities in patient outcomes after craniotomy for tumor in adult patients in the United States, 1988–2004. Neurosurg Mar 66(3):427–437 discussion 437-8. 10.1227/01.Neu.0000365265.10141.8e [DOI] [PubMed] [Google Scholar]
- 7.El-Sayed AM, Ziewacz JE, Davis MC et al (2011) Insurance status and inequalities in outcomes after neurosurgery. World Neurosurg Nov 76(5):459–466. 10.1016/j.wneu.2011.03.051 [DOI] [PubMed] [Google Scholar]
- 8.Daly AF, Beckers A (2020) The epidemiology of Pituitary Adenomas. Endocrinol Metab Clin North Am Sep 49(3):347–355. 10.1016/j.ecl.2020.04.002 [DOI] [PubMed] [Google Scholar]
- 9.Olsson DS, Svensson M, Labori F, De Geer A, Johannsson G (2023) Healthcare cost and survival in patients with non-functioning pituitary adenoma. Eur J Endocrinol Jun 7(6):477–484. 10.1093/ejendo/lvad057 [DOI] [PubMed] [Google Scholar]
- 10.Osorio RC, Pereira MP, Joshi RS et al (2022) Socioeconomic predictors of case presentations and outcomes in 225 nonfunctional pituitary adenoma resections. J Neurosurg May 1(5):1325–1336. 10.3171/2021.4.Jns21907 [DOI] [PubMed] [Google Scholar]
- 11.Goljo E, Parasher AK, Iloreta AM, Shrivastava R, Govindaraj S (2016) Racial, ethnic, and socioeconomic disparities in pituitary surgery outcomes. Laryngoscope Apr 126(4):808–814. 10.1002/lary.25771 [DOI] [PubMed] [Google Scholar]
- 12.McKee S, Yang A, Kidwai S, Govindaraj S, Shrivastava R, Iloreta A (2018) The socioeconomic determinants for transsphenoidal pituitary surgery: a review of New York State from 1995 to 2015. Int Forum Allergy Rhinol Oct 8(10):1145–1156. 10.1002/alr.22148 [DOI] [PubMed] [Google Scholar]
- 13.Younus I, Gerges M, Schwartz TH, Ramakrishna R (2020) Impact of Medicaid insurance on outcomes following endoscopic transsphenoidal pituitary surgery. J Neurosurg Mar 20(3):801–806. 10.3171/2020.1.Jns192707 [DOI] [PubMed] [Google Scholar]
- 14.Jahangiri A, Lamborn KR, Blevins L, Kunwar S, Aghi MK (2012) Factors associated with delay to pituitary adenoma diagnosis in patients with visual loss. J Neurosurg Feb 116(2):283–289. 10.3171/2011.6.Jns101663 [DOI] [PubMed] [Google Scholar]
- 15.Gaffney A, McCormick D (2017) The affordable Care Act: implications for health-care equity. Lancet Apr 8(10077):1442–1452. 10.1016/s0140-6736(17)30786-9 [DOI] [PubMed] [Google Scholar]
- 16.Chen SH, Sprau A, Chieng L et al (2019) Transsphenoidal Approach for Pituitary adenomas in Elderly patients. World Neurosurg Jan 121:e670–e674. 10.1016/j.wneu.2018.09.187 [DOI] [PubMed] [Google Scholar]
- 17.Ghiam MK, Ali IA, Dable CL et al (2022) Multidisciplinary postoperative care pathway to reduce readmissions following endoscopic transsphenoidal pituitary surgery: improving Quality of Patient Care. J Neurol Surg B Skull Base Dec 83(6):626–634. 10.1055/a-1920-0758 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Vimawala S, Chitguppi C, Reilly E et al (2020) Predicting prolonged length of stay after endoscopic transsphenoidal surgery for pituitary adenoma. Int Forum Allergy Rhinol Jun 10(6):785–790. 10.1002/alr.22540 [DOI] [PubMed] [Google Scholar]
- 19.Corlette S, Monahan CH (2022) U.S. Health Insurance Coverage and Financing. N Engl J Med Dec 22(25):2297–2300. 10.1056/NEJMp2206049 [DOI] [PubMed] [Google Scholar]
- 20.Berkowitz E (2008) Medicare and Medicaid: the past as prologue. Health Care Financ Rev Spring 29(3):81–93 [PMC free article] [PubMed] [Google Scholar]
- 21.Health Insurance Coverage in the United States (2022) (U.S. Government Publishing Office) (2023)
- 22.Pitts SR, Carrier ER, Rich EC, Kellermann AL (Sep 2010) Where Americans get acute care: increasingly, it’s not at their doctor’s office. Health Aff (Millwood) 29(9):1620–1629. 10.1377/hlthaff.2009.1026 [DOI] [PubMed]
- 23.He G, Li C, Wang S, Wang H, Ding J (2022) Association of insurance status with chronic kidney disease stage at diagnosis in children. Pediatr Nephrol Nov 37(11):2705–2714. 10.1007/s00467-022-05493-6 [DOI] [PubMed] [Google Scholar]
- 24.Markt SC, Lago-Hernandez CA, Miller RE et al (2016) Insurance status and disparities in disease presentation, treatment, and outcomes for men with germ cell tumors. Cancer Oct 15(20):3127–3135. 10.1002/cncr.30159 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Farley TA, Flannery JT (1989) Late-stage diagnosis of breast cancer in women of lower socioeconomic status: public health implications. Am J Public Health Nov 79(11):1508–1512. 10.2105/ajph.79.11.1508 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Ramírez de Arellano A (2007) A ranking of State Medicaid Programs. Public Citizen Health Research Group. https://www.citizen.org/article/unsettling-scores/
- 27.Muskens IS, Zamanipoor Najafabadi AH, Briceno V et al (2017) Visual outcomes after endoscopic endonasal pituitary adenoma resection: a systematic review and meta-analysis. Pituit Oct 20(5):539–552. 10.1007/s11102-017-0815-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Jahangiri A, Clark AJ, Han SJ, Kunwar S, Blevins LS Jr., Aghi MK (2013) Socioeconomic factors associated with pituitary apoplexy. J Neurosurg Dec 119(6):1432–1436. 10.3171/2013.6.Jns122323 [DOI] [PubMed] [Google Scholar]
- 29.Chen Y, Wang CD, Su ZP et al (2012) Natural history of postoperative nonfunctioning pituitary adenomas: a systematic review and meta-analysis. Neuroendocrinology 96(4):333–342. 10.1159/000339823 [DOI] [PubMed] [Google Scholar]
- 30.Serra C, Staartjes VE, Maldaner N et al (2018) Predicting extent of resection in transsphenoidal surgery for pituitary adenoma. Acta Neurochir (Wien) Nov 160(11):2255–2262. 10.1007/s00701-018-3690-x [DOI] [PubMed] [Google Scholar]
- 31.Singla R, Sharma R, Suri A (2023) Role of cavernous sinus extension and MRI T2 hypointensity in the extent of resection following trans-sphenoidal surgery for Giant Pituitary Adenomas. Neurol India 71(5):907–915. 10.4103/0028-3886.388120 [DOI] [PubMed] [Google Scholar]
- 32.Hamill CS, Villwock JA, Sykes KJ, Chamoun RB, Beahm DD (2018) Socioeconomic factors affecting discharge status of patients with uncomplicated Transsphenoidal Adenohypophysectomy. J Neurol Surg B Skull Base Oct 79(5):501–507. 10.1055/s-0038-1635095 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Sharma A, Muir R, Johnston R, Carter E, Bowden G, Wilson-MacDonald J (2013) Diabetes is predictive of longer hospital stay and increased rate of complications in spinal surgery in the UK. Ann R Coll Surg Engl May 95(4):275–279. 10.1308/003588413x13511609958299 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Chalif EJ, Couldwell WT, Aghi MK (2022) Effect of facility volume on giant pituitary adenoma neurosurgical outcomes. J Neurosurg Sep 1(3):658–667. 10.3171/2021.11.Jns211936 [DOI] [PubMed] [Google Scholar]
- 35.Barker FG 2nd, Klibanski A, Swearingen B (2003) Transsphenoidal surgery for pituitary tumors in the United States, 1996–2000: mortality, morbidity, and the effects of hospital and surgeon volume. J Clin Endocrinol Metab Oct 88(10):4709–4719. 10.1210/jc.2003-030461 [DOI] [PubMed] [Google Scholar]
- 36.Dasenbrock HH, Wolinsky JP, Sciubba DM, Witham TF, Gokaslan ZL, Bydon A (2012) The impact of insurance status on outcomes after surgery for spinal metastases. Cancer 118(19):4833-41. 10.1002/cncr.27388 [DOI] [PubMed]
- 37.DeSantis C, Jemal A, Ward E (2010) Disparities in breast cancer prognostic factors by race, insurance status, and education. Cancer Causes Control Sep 21(9):1445–1450. 10.1007/s10552-010-9572-z [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
No datasets were generated or analysed during the current study.