- Email: [email protected]
Medicaid payer status and other factors associated with hospital length of stay in patients undergoing primary lumbar spine surgery
Clinical Neurology and Neurosurgery 188 (2020) 105570
Contents lists available at ScienceDirect
Clinical Neurology and Neurosurgery journal homepage: www.elsevier.com/locate/clineuro
Medicaid payer status and other factors associated with hospital length of stay in patients undergoing primary lumbar spine surgery
T
Jose F. Domingueza, Piyush Kalakotib, Xintong Chenc, Kaisen Yaod, Nam K. Leed, Smit Shahe, Meic Schmidta, Chad Colea, Chirag Gandhia, Fawaz Al-Muftia, Christian A. Bowersa,* a
Department of Neurosurgery, Westchester Medical Center, New York Medical College, Valhalla, NY 10595, USA Neurosurgery, Louisiana State University Health Sciences Center, Shreveport, LA 71103, USA c Icahn School of Medicine at Mount Sinai, New York, NY 10029, USA d New York Medical College, Valhalla, NY 10595, USA e Rutgers, Robert Wood Johnson Medical School, NJ, 08854, USA b
A R T I C LE I N FO
A B S T R A C T
Keywords: Medicaid Insurance Spine surgery Spinal fusion Hospitalization Length of stay Lumbar degenerative disease Charlson Comorbidity Index
Objective: The Medicaid patient population and health care costs for spine surgeries among these patients have increased since 2010. Hospital length of stay (LOS) contributes appreciably to hospital costs for patients undergoing primary lumbar spine surgery (PLSS). The aim of this study was to identify independent risk factors for increased LOS in patients undergoing PLSS. Patients and methods: In a single-center retrospective study, we reviewed demographic and clinical data from electronic medical records for 181 consecutive adult patients who underwent PLSS involving 1–3 levels from July 2014 to July 2017. We performed regression analyses to identify independent risk factors for increased LOS and to quantify their effects as percent changes in LOS. Results: Among 181 patients who underwent PLSS, the mean LOS was 3.57 days. Based on the Charlson Comorbidity Index (CCI) and American Society of Anesthesiologist (ASA) classification, patients with Medicaid insurance were healthier than non-Medicaid patients (mean CCI: 0.34 versus 0.65; p = 0.041, ASA: 1.71 versus 1.91; p = 0.046) yet Medicaid patients had a longer LOS compared with non-Medicaid patients (mean LOS: 4.03 versus 3.30 days; p = 0.047). There was no significant difference in discharge disposition between Medicaid and non-Medicaid patients (Home = 82.35 % versus 79.65 %; p = 0.855). Medicaid patients also had significantly less spinal levels involved in their surgery (1.44 versus 1.67; p = 0.027). Multivariable regression modeling identified independent risk factors positively associated with increased LOS as age (+1.0 % per year; p = 0.007), Medicaid insurance status (+28.7 %; p = 0.007), and CCI (10.1 % per increment in CCI; p = 0.030). Fusion surgery also was an independent risk factor for increased LOS when compared with laminectomy (−54.1 %; p < 0.001) or discectomy (−51.3 %; p < 0.001). Conclusions: Increasing age, Medicaid insurance status, higher CCI, and fusion surgery were independently associated with increased LOS after PLSS. This information is useful for preoperative patient counseling, shared decision-making, and risk stratification and may help to further ongoing discussion regarding contributors to rising health care costs. Findings of increased LOS among Medicaid patients will help direct efforts to identify factors that contribute to this health care expense.
1. Introduction Given the increased number of patients insured through Medicaid due to the Affordable Care Act, insurance status is a key component of discussions regarding health care costs. Since the Act was passed in 2010, as many as 17 million Medicaid patients have been added to the insurance pool. As a result, the number of spine surgeries has increased
⁎
dramatically, with a 150 % increase in Medicaid payments for spinal surgeries from 2010 to 2016 [1]. Medicaid patients have been shown to have worse surgical outcomes and higher costs of care than non-Medicaid patients [2]. Patient satisfaction scores can be used to measure provider and treatment effectiveness; Medicaid patients overall have lower levels of satisfaction than non-Medicaid patients, which has been shown to affect reimbursements in some models [3]. Since Medicaid
Corresponding author at: Department of Neurosurgery, Westchester Medical Center, New York Medical College, 100 Woods Road, Valhalla, NY, 10595, USA. E-mail address: [email protected] (C.A. Bowers).
https://doi.org/10.1016/j.clineuro.2019.105570 Received 13 September 2019; Received in revised form 17 October 2019; Accepted 21 October 2019 Available online 24 October 2019 0303-8467/ © 2019 Elsevier B.V. All rights reserved.
Clinical Neurology and Neurosurgery 188 (2020) 105570
J.F. Dominguez, et al.
were reported as means ± standard deviation (SD) and/or medians, as appropriate based on testing of assumptions for variable distribution. Differences in metric values between Medicaid versus non-Medicaid patients were assessed using the independent samples t-test or MannWhitney U-test, as applicable. To identify factors that independently impact LOS in patients undergoing PLSS, we constructed an ordinary least squares (OLS) regression model. Initial regression analysis demonstrated heteroskedasticity due to the non-Gaussian distribution of LOS. Therefore, we performed several arithmetic transformations, including cubic, square, square root, and logarithmic (natural and to the base 10) and found that the natural logarithmic (ln) transformation of LOS provided the best fit to normality. Using the ln transformed LOS, we constructed a multivariable OLS regression model to identify potential risk factors for increased LOS while adjusting for each independent variable. Regression diagnostics were tested using variance (coefficient of determination) and residual analyses. To interpret coefficient estimates from ln transformed OLS modeling, we back-transformed coefficients; the effects of individual variables were reported in terms of per increment percent change on LOS. All statistical analyses were performed using IBM SPSS v25.0 [22] and R v3.3.3 [23]. A p value < 0.05 was considered statistically significant.
patients have significantly lower surgeon reimbursement rates and worse surgical outcomes, many providers refuse to see Medicaid patients and primary care physicians can encounter difficulty finding appropriate specialty care for their Medicaid patients [4]. Lumbar degenerative disc disease (LDDD) is a common pathological spinal condition with associated lower back pain that can negatively impact patients’ quality of life [5,6]. Surgical intervention to alleviate back pain for patients refractory to conservative treatment measures frequently consists of laminectomy, discectomy, and/or fusion [7]. During the past decade, health care spending for lumbar spinal stenosis (LSS) surgery has increased from $1.8 to $8.2 billion, as hospitals have experienced a doubling of spine surgeries performed during that period [8,9]. Factors contributing to increased costs of hospitalization are multifactorial. Age, sex, American Society of Anesthesiologists (ASA) score [10], Oswestry Disability Index, Medicare/Medicaid, ambulation status, patient-reported allergies, fusion surgery, increased operating time, and medical comorbidities [often categorized by the Charlson Comorbidity Index (CCI)], have been associated with increased LOS and increased complication rates and readmission rates for spinal surgery patients [11–19]. Increased LOS is a significant contributor to hospital costs and patient health care costs [20]. In this retrospective study of patients undergoing primary lumbar spine surgery (PLSS) for LDDD, we analyzed multiple variables related to LOS, including age, sex, body mass index (BMI), insurance type, and CCI score [21].
3. Results 3.1. Patient demographics and clinical characteristics
2. Materials and methods Overall, 181 patients undergoing PLSS for HNP, lumbar stenosis, and/or spondylolisthesis were included. The mean age of the cohort was 52.29 ± 13.42 years, and 87 (48.07 %) were women; the average BMI was 30.83 ± 9.17 kg/m2. Of the 181 total patients, 105 (58.01 %) had fusion, 54 (29.83 %) had discectomy, and 22 (12.15 %) had laminectomy. The type of surgery performed did not differ significantly based on insurance status (Medicaid versus non-Medicaid). Prevalent comorbidities identified among the cohort were diabetes mellitus (DM) (28 patients, 15.47 %) and psychiatric diagnoses (32 patients, 17.68 %). Medicaid patients had fewer comorbidities than non-Medicaid patients, as indicated by the CCI score and ASA classification (mean 0.34 vs. 0.65; p = 0.041, ASA: 1.71 versus 1.91; p = 0.046). Discharge disposition was not significantly different between Medicaid and nonMedicaid patients (Home = 82.35 % versus 79.65 %; p = 0.855). Medicaid patients also had significantly less spinal levels involved in their surgery (1.44 versus 1.67; p = 0.027) (Table 1). Primary insurance providers for patients in the cohort included Medicaid (68 patients, 37.57 %), Medicare (38 patients, 20.99 %), private payers (36 patients, 19.89 %), workers’ compensation (24 patients, 13.26 %), and other sources (14 patients,7.73 %); one (0.55 %) patient was uninsured (Table 2). The mean LOS was 3.57 ± 2.81 days (median: 3.00 days). Despite having a lower CCI score, Medicaid patients had a longer mean LOS compared with non-Medicaid patients (4.03 vs. 3.30 days; p = 0.047) (Table 3).
2.1. Study design and patient characteristics In this retrospective single-center study, we reviewed the electronic medical records for 181 consecutive adult patients 18 years of age or older who had PLSS (1–3 levels of involvement) from July 1, 2014, to July 1, 2017. Patients who underwent laminectomy, discectomy, or fusion for herniated nucleus pulposus (HNP), lumbar stenosis, and/or spondylolisthesis were included. Patients with spinal neoplasia or congenital spinal disease and any revision operations were excluded. The study was approved by the Institutional Review Board of the Westchester Medical Center. 2.2. Variables Factors previously demonstrated to impact LOS or thought by the investigators to potentially impact LOS were defined as independent variables. These were patient demographics (age, sex, and insurance status), BMI, patient-reported allergies, discharge disposition, and comorbidities (ASA, CCI, diabetes mellitus [DM] and psychiatric comorbidities such as anxiety, depression, and/or substance abuse). The surgeon’s years of experience also was included as an independent variable in addition to whether there were two co-surgeons for the surgical procedure. Patient insurance status categories included Medicare, Medicaid, workers’ compensation, and various private commercial insurances. For purposes of statistical analysis, patients with private insurance as a supplement to their Medicare coverage were categorized as Medicare patients. Data on patient-reported allergies were also recorded. Medical comorbidities were used to calculate a CCI score.
3.2. Univariate analysis In univariate OLS regression analyses, factors that demonstrated borderline statistical significance (defined as p < 0.10) in their association with LOS included: advancing age (+0.9 % per yearly increment; p = 0.033), Medicaid insurance status (+22.0 %; p = 0.085), female sex (+21.0 %; p = 0.090); BMI (+1.1 %; p = 0.074); CCI (15.4 % per increment in CCI score; p = 0.003); DM (+45.4 %; 0.016); psychiatric disorder (+35.5 %; p = 0.039), and surgery type [laminectomy (−39.7 %; p = 0.003); discectomy (−52.6 %; p < 0.001); fusion (+137.1 %; p < 0.001)]. Patient-reported allergies, the surgeon’s length of experience, and the presence of a co-surgeon for the procedure were not associated with increased LOS (Table 4). Variables that demonstrated borderline statistical significance in univariate
2.3. Statistical analysis Our primary outcome was to identify independent predictors for LOS in patients undergoing PLSS. Descriptive and inferential statistical techniques were used for data analysis. Categorical variables were described as counts or percentages. Percentage differences between Medicaid and non-Medicaid patients were analyzed using the Pearson Chi-square test or Fisher exact test as appropriate. Continuous variables 2
Clinical Neurology and Neurosurgery 188 (2020) 105570
J.F. Dominguez, et al.
Table 1 Characteristics of patients undergoing lumbar spine surgery between 2014–2017 based upon Medicaid status at our institute. Characteristics
Medicaid
Non-Medicaid
Overall
P value
Age, in years Mean ± SD Median
47.96 ± 10.25 48.00
54.90 ± 14.43 55.00
52.29 ± 13.42 52.00
< 0.001 < 0.001∗
Gender, n (%) Female Male
37 (54.41) 31 (45.59)
50 (44.25) 63 (55.75)
87 (48.07) 94 (51.93)
0.185
Body mass index (BMI), kgm−2 Mean ± SD Median
32.18 ± 10.40 29.00
29.96 ± 8.22 29.50
30.83 ± 9.17 29.00
0.143 0.420∗
Diabetes mellitus, n (%)‖ Psychiatric comorbidities, n (%)‖
9 (13.24) 12 (17.65)
19 (16.81) 20 (17.70)
28 (15.47) 32 (17.68)
0.519 0.993
Charlson comorbidity index (CCI) CCI 0-1, n (%) CCI ≥ 2, n (%) Mean CCI ± SD
62 (91.18) 6 (8.82) 0.34 ± 0.68
97 (85.84) 16 (14.16) 0.65 ± 1.37
159 (87.85) 22 (12.15) 0.54 ± 1.17
0.352† 0.041
Surgery type, n (%)‖ Laminectomy Discectomies Fusion/arthrodesis
6 (8.82) 21 (30.88) 41 (60.29)
16 (14.16) 33 (29.20) 64 (56.64)
22 (12.15) 54 (29.83) 105 (58.01)
0.287 0.811 0.629
Two surgeons, n (%) Co-scrubs, n (%)
15 (22.06) 12 (17.65)
32 (28.32) 22 (19.47)
47 (25.97) 34 (18.78)
0.352 0.761
Surgeon experience, in years Mean ± SD Median
16.59 ± 6.22 19.00
17.98 ± 6.70 18.00
17.46 ± 6.54 19.00
0.158 0.435∗
Patient-reported allergies (PRA) Overall Medications Food Environment
0.99 0.69 0.15 0.15
1.08 0.74 0.22 0.12
1.04 0.72 0.19 0.13
0.725 0.776 0.563 0.643
ASA score Mean ± SD Median
1.71 ± 0.75 2.00
1.91 ± 0.69 2.00
1.83 ± 0.72 2.00
0.069 0.046*
Number of levels operated Mean ± SD Median
1.44 ± 0.63 1.00
1.67 ± 0.74 2.00
1.59 ± 0.71 1.00
0.027 0.032*
Discharge, n (%) Home Acute Rehab Expired
56 (82.35) 11 (16.18) 1 (1.47)
90 (79.65) 22 (19.47) 1 (0.88)
146 (80.66) 33 (18.23) 2 (1.10)
0.855†
± ± ± ±
1.58 1.27 0.40 0.50
± ± ± ±
2.00 1.32 1.26 0.44
± ± ± ±
1.85 1.30 1.02 0.46
†p-values derived from Fisher’s exact test. ∗ p-values derived from non-parametric Mann-Whitney U test. ‖ Estimates based upon missing data for DM (9.9 %), psychiatric comorbidities (9.9 %), surgery type (8.8 %), co-surgeons (8.8 %), Charlson comorbidity (8.8 %), and allergies (8.8 %). Bold p-values indicate statistical significance. Table 2 Patient characteristics based upon insurance type. Primary insurance provider
N (%)
Medicaid Medicare Private payer Uninsured Prisoners Workman compensation Military
68 (37.57) 38 (20.99) 36 (19.89) 1 (0.55) 8 (4.42) 24 (13.26) 6 (3.31)
Table 3 Length of hospital stay (primary outcome) in patients undergoing lumbar spine surgery. Characteristics
Medicaid
Non-Medicaid
Overall
P value
LOS, days Mean ± SD Median
4.03 ± 2.84 3.50
3.30 ± 2.76 3.00
3.57 ± 2.81 3.00
0.047 0.034*
* p-values derived from non-parametric Mann-Whitney U test.
3
Clinical Neurology and Neurosurgery 188 (2020) 105570
J.F. Dominguez, et al.
Table 4 Regression models (univariate and multivariate) investigating factors associated with length of hospital stay. Univariate
Multivariate
Factors
% Change (95% CI)
P-value
% Change (95% CI)
P-value
Medicaid Age Female gender BMI DM Psychiatric CCI Surgery type Laminectomy Discectomies Fusion/arthrodesis Two-surgeons Co-scrub Surgeon experience Patient reported allergies Overall Medications Food Environment
+22.0% (-2.8 – +53.2) +0.9% (+0.1– +1.7) +21.0% (-3.0 – +50.8) +1.1% (-0.1 – +2.4) +45.4% (+7.5 – +96.6) +35.5% (+1.6 – +80.7) +15.4% (+5.2 – +26.6)
0.085^ 0.033^ 0.090^ 0.074^ 0.016^ 0.039^ 0.003^
+28.7% (+7.1 – +54.6) +1.0% (+0.3 – +1.8) +2.8% (-14.0 – +22.8) +0.8% (-0.2 – +1.8) +11.4% (-15.5 – +46.9) +20.6% (-4.1 – +51.7) +10.1% (1.0 – +20.1)
0.007 0.007 0.761 0.110 0.442 0.109 0.030
−39.7% (−56.7 – −16.0) −52.6% (−61.8 – −41.1) +137.1% (+97.0 – +185.3) −30.1% (−45.4 – −10.5) −40.4% (−54.7 – −21.7) +0.8% (−0.9 – 2.5)
0.003^ < 0.001^ < 0.001^ 0.005^ < 0.001^ 0.344
−54.1% (−65.1 – −39.6) −51.3% (−60.1 – −40.4) Reference +4.4% (−26.6 – +48.4) −30.7% (−53.5 – +3.2) –
< 0.001 < 0.001 Reference 0.811 0.071 –
0.687 0.406 0.544 0.603
– – – –
– – – –
+1.2% +3.7% −3.3% +6.8%
(−4.7 – +7.5) (−4.8 – +13.0) (−13.3 – +7.8) (−16.2 – +36.1)
^ indicates p-value < 0.1 in univariate analysis and will be included for multivariate analysis. Bold p-values indicate statistical significance.
increasingly important to identify and study factors that are associated with increased costs of PLSS in order to reduce health care expenditures. PLSS under general anesthesia costs approximately $13,206 [29], and an additional patient day of hospitalization raises hospital costs by $2102 [17]. A surgical patient’s LOS, which is a significant contributor to hospital and patient costs and a key hospital administrative priority [20], is potentially modifiable. Furthermore, since most PLSS operations are performed electively and the long-term efficacy of PLSS in alleviating pain cause by LDDD is controversial [30], preoperative patient counseling and risk stratification that incorporate factors relevant to both clinical outcome and health care expenditures are essential. As a single center retrospective study, our study has inherent limitations in its generalizability to other health care facilities in different geographic areas with different patient demographics and health care resources. Our analyses also focused on the short-term outcome of LOS. Additionally, the reluctance by health care providers to accept Medicaid patients may result in a diagnostic bias, whereby Medicaid patients' lower CCI score could be influenced by undiagnosed comorbidities. Additional research is indicated to replicate these findings in other populations and to evaluate the longer-term and broader implications of the several risk factors for increased LOS that we identified, including Medicaid insurance status.
analyses were included in subsequent multivariate analyses. 3.3. Multivariate analysis In a multivariable OLS regression model, independent risk factors associated with LOS were advancing age (+1.0 % increase in LOS per year; p = 0.007), Medicaid insurance status (+28.7 %; p = 0.007), CCI (10.1 % per increment in CCI score; p = 0.030), and undergoing laminectomy (−54.1 %; p < 0.001) or discectomy (−51.3 %; p < 0.001), compared with fusion (Table 4). 4. Discussion We present a retrospective cohort study of 181 adult patients who underwent PLSS for LDDD. Medicaid insurance status was an independent risk factor for increased LOS after PLSS. Despite being healthier than non-Medicaid patients (as indicated by the CCI and ASA classification), and more likely to be discharged home, Medicaid patients’ average LOS was 0.7 days longer than that of patients with any other type of insurance. The demonstration in this study of a positive association between Medicaid status and increased LOS in patients undergoing PLSS, which has not been reported previously among adult spine patients, is an important finding for several reasons. First, accompanying explosive growth in the number of Medicaid patients since enactment of the Affordable Care Act in 2010 has been an increasing reluctance among medical specialists to provide care for Medicaid patients. Providers’ hesitation to accept Medicaid patients is due to various factors that have been shown to be associated with Medicaid status, including a lower rate of provider reimbursement, poorer patient outcomes, and increased LOS across medical specialties [1,4,18,24–28]. Such concerns among providers are heightened in an era when pay for performance and episode-based bundles are common health care reforms, and patients’ increased LOS can lower a provider’s ranking or overall score [3,5]. Hospitals also may view Medicaid patients as less desirable due to their lower reimbursement rate [4]. These influences on providers’ decision making combined with the new finding that Medicaid patients undergoing PLSS have a longer mean LOS, may further increase the number of spine surgeons who are unwilling to treat Medicaid patients. Second, given the dramatic growth in the number of PLSS procedures being performed and the high costs of these surgeries, it is
5. Conclusions In this study, Medicaid insurance status, increasing age, higher CCI, and fusion surgery were independently associated with increased LOS among adult patients undergoing PLSS. Analysis of spinal levels involved, discharge disposition, and CCI demonstrated that the association between Medicaid insurance status and increased LOS was unlikely to be attributed to case complexity, discharge disposition or baseline health status. These findings may be useful for patient counseling, shared-decision making, risk stratification, and cost analyses. Identification of Medicaid status as an independent risk factor for increased LOS following PLSS merits further study, particularly given the increasing prevalence of spine surgery and the growth of the Medicaid patient population.
4
Clinical Neurology and Neurosurgery 188 (2020) 105570
J.F. Dominguez, et al.
Funding [14]
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
[15]
Declaration of Competing Interest The authors report no conflicts of interest.
[16]
References [17]
[1] J.E. Meyers, J. Wang, A. Khan, et al., Trends in physician reimbursement for spinal procedures since 2010, Spine (Phila Pa 1976) 43 (15) (2018) 1074–1079, https:// doi.org/10.1097/BRS.0000000000002516. [2] D.J. LaPar, C.M. Bhamidipati, C.M. Mery, et al., Primary payer status affects mortality for major surgical operations, Ann. Surg. 252 (3) (2010) 544–551, https:// doi.org/10.1097/SLA.0b013e3181e8fd75. [3] S. Chotai, A. Sivaganesan, S.L. Parker, et al., Patient-specific factors associated with dissatisfaction after elective surgery for degenerative spine diseases, Neurosurgery 77 (2) (2015) 157–163, https://doi.org/10.1227/NEU.0000000000000768 discussion 163. [4] S.K. Long, Physicians may need more than higher reimbursements to expand medicaid participation: findings from Washington State, Health Aff. (Millwood) 32 (9) (2013) 1560–1567, https://doi.org/10.1377/hlthaff.2012.1010. [5] D.M. Cutler, K. Ghosh, The potential for cost savings through bundled episode payments, N. Engl. J. Med. 366 (12) (2012) 1075–1077, https://doi.org/10.1056/ NEJMp1113361. [6] F.M. Kovacs, V. Abraira, J. Zamora, et al., Correlation between pain, disability, and quality of life in patients with common low back pain, Spine (Phila Pa 1976) 29 (2) (2004) 206–210, https://doi.org/10.1097/01.BRS.0000107235.47465.08. [7] X.L. Ma, X.W. Zhao, J.X. Ma, et al., Effectiveness of surgery versus conservative treatment for lumbar spinal stenosis: a system review and meta-analysis of randomized controlled trials, Int. J. Surg. 44 (2017) 329–338, https://doi.org/10.1016/j. ijsu.2017.07.032. [8] M. Raad, C.J. Donaldson, M.H. El Dafrawy, et al., Trends in isolated lumbar spinal stenosis surgery among working US adults aged 40-64 years, 2010–2014, J. Neurosurg. Spine 29 (2) (2018) 169–175, https://doi.org/10.3171/2018.1. SPINE17964. [9] N.K. Sharma, B. Olotu, A. Mathew, et al., Lumbar spine surgeries and medication usage during hospital stay: one-center perspective, Hosp. Pharm. 52 (11) (2017) 774–780, https://doi.org/10.1177/0018578717737431. [10] C.M. Kuza, K. Matsushima, W.J. Mack, et al., The role of the American Society of Anesthesiologists physical status classification in predicting trauma mortality and outcomes, Am. J. Surg. 10 (19) (2019) 30254–30255, https://doi.org/10.1016/j. amjsurg.2019.09.019. [11] J.A. Gruskay, M. Fu, D.D. Bohl, et al., Factors affecting length of stay after elective posterior lumbar spine surgery: a multivariate analysis, Spine J. 15 (6) (2015) 1188–1195, https://doi.org/10.1016/j.spinee.2013.10.022. [12] C. Lagman, B. Ugiliweneza, M. Boakye, et al., Spine surgery outcomes in elderly patients versus general adult patients in the United States: a market scan analysis, World Neurosurg. 103 (2017) 780–788, https://doi.org/10.1016/j.wneu.2017.04. 001. [13] M.W. Manoso, A.M. Cizik, R.J. Bransford, et al., Medicaid status is associated with
[18]
[19]
[20]
[21]
[22] [23] [24]
[25]
[26] [27]
[28]
[29]
[30]
5
higher surgical site infection rates after spine surgery, Spine (Phila Pa 1976) 39 (20) (2014) 1707–1713, https://doi.org/10.1097/BRS.0000000000000496. M.J. McGirt, S.L. Parker, S. Chotai, et al., Predictors of extended length of stay, discharge to inpatient rehab, and hospital readmission following elective lumbar spine surgery: introduction of the Carolina-Semmes Grading Scale, J. Neurosurg. Spine 27 (4) (2017) 382–390, https://doi.org/10.3171/2016.12.SPINE16928. P.G. Passias, C.M. Jalai, N. Worley, et al., Predictors of hospital length of stay and 30-day readmission in cervical spondylotic myelopathy patients: an analysis of 3057 patients using the ACS-NSQIP Database, World Neurosurg. 110 (2018) e450–e458, https://doi.org/10.1016/j.wneu.2017.11.009. M.J. Press, D.P. Scanlon, A.S. Navathe, et al., The importance of clinical severity in the measurement of hospital readmission rates for Medicare beneficiaries, 19972007, Med. Care Res. Rev. 70 (6) (2013) 653–665, https://doi.org/10.1177/ 1077558713496167. R.C. Puffer, R. Planchard, G.W. Mallory, et al., Patient-specific factors affecting hospital costs in lumbar spine surgery, J. Neurosurg. Spine 24 (1) (2016) 1–6, https://doi.org/10.3171/2015.3.SPINE141233. K. Rumalla, K.A. Smith, P.M. Arnold, Cervical total disc replacement and anterior cervical discectomy and fusion: reoperation rates, complications, and hospital resource utilization in 72 688 patients in the United States, Neurosurgery 82 (4) (2018) 441–453, https://doi.org/10.1093/neuros/nyx289. R.K. Wadhwa, J. Ohya, T.D. Vogel, et al., Risk factors for 30-day reoperation and 3month readmission: analysis from the Quality and Outcomes Database lumbar spine registry, J. Neurosurg. Spine 27 (2) (2017) 131–136, https://doi.org/10.3171/ 2016.12.SPINE16714. S. Missios, K. Bekelis, Hospitalization cost after spine surgery in the United States of America, J. Clin. Neurosci. 22 (10) (2015) 1632–1637, https://doi.org/10.1016/j. jocn.2015.05.005. W. D’Hoore, C. Sicotte, C. Tilquin, Risk adjustment in outcome assessment: the charlson comorbidity index, Methods Inf. Med. 32 (5) (1993) 382–387, https://doi. org/10.1055/s-0038-1634956. IBM Corp, IBM SPSS Statistics for Windows, Version 25.0. Armonk, NY, (2017). R Core Team, R: a Language and Environment for Statistical Computing, Version 3.3.3. R Foundation for Statistical Computing, Vienna, Austria, 2017. S.K. Cho, N.N. Egorova, The association between insurance status and complications, length of stay, and costs for pediatric idiopathic scoliosis, Spine (Phila Pa 1976) 40 (4) (2015) 247–256, https://doi.org/10.1097/BRS.0000000000000729. R.B. Hakim, M.B. Benedict, N.J. Merrick, Quality of care for women undergoing a hysterectomy: effects of insurance and race/ethnicity, Am. J. Public Health 94 (8) (2004) 1399–1405, https://doi.org/10.2105/ajph.94.8.1399. J. Kwok, The impact of health insurance status on the survival of patients with head and neck, Cancer 116 (2) (2010) 476–485, https://doi.org/10.1002/cncr.24774. J.J. Shen, E.L. Washington, Disparities in outcomes among patients with stroke associated with insurance status, Stroke 38 (3) (2007) 1010–1016, https://doi.org/ 10.1161/01.STR.0000257312.12989.af. D.B. Stulberg, J.X. Zhang, S.T. Lindau, Socioeconomic disparities in ectopic pregnancy: predictors of adverse outcomes from Illinois hospital-based care, 2000–2006, Matern. Child Health J. 15 (2) (2011) 234–241, https://doi.org/10. 1007/s10995-010-0579-6. P. Agarwal, J. Pierce, W.C. Welch, Cost Analysis of spinal versus general anesthesia for lumbar diskectomy and laminectomy spine surgery, World Neurosurg. 89 (2016) 266–271, https://doi.org/10.1016/j.wneu.2016.02.022. R.A. Deyo, S.K. Mirza, J.A. Turner, et al., Overtreating chronic back pain: Time to back off? J. Am. Board Fam. Med. 22 (1) (2009) 62–68, https://doi.org/10.3122/ jabfm.2009.01.080102.
Contents lists available at ScienceDirect
Clinical Neurology and Neurosurgery journal homepage: www.elsevier.com/locate/clineuro
Medicaid payer status and other factors associated with hospital length of stay in patients undergoing primary lumbar spine surgery
T
Jose F. Domingueza, Piyush Kalakotib, Xintong Chenc, Kaisen Yaod, Nam K. Leed, Smit Shahe, Meic Schmidta, Chad Colea, Chirag Gandhia, Fawaz Al-Muftia, Christian A. Bowersa,* a
Department of Neurosurgery, Westchester Medical Center, New York Medical College, Valhalla, NY 10595, USA Neurosurgery, Louisiana State University Health Sciences Center, Shreveport, LA 71103, USA c Icahn School of Medicine at Mount Sinai, New York, NY 10029, USA d New York Medical College, Valhalla, NY 10595, USA e Rutgers, Robert Wood Johnson Medical School, NJ, 08854, USA b
A R T I C LE I N FO
A B S T R A C T
Keywords: Medicaid Insurance Spine surgery Spinal fusion Hospitalization Length of stay Lumbar degenerative disease Charlson Comorbidity Index
Objective: The Medicaid patient population and health care costs for spine surgeries among these patients have increased since 2010. Hospital length of stay (LOS) contributes appreciably to hospital costs for patients undergoing primary lumbar spine surgery (PLSS). The aim of this study was to identify independent risk factors for increased LOS in patients undergoing PLSS. Patients and methods: In a single-center retrospective study, we reviewed demographic and clinical data from electronic medical records for 181 consecutive adult patients who underwent PLSS involving 1–3 levels from July 2014 to July 2017. We performed regression analyses to identify independent risk factors for increased LOS and to quantify their effects as percent changes in LOS. Results: Among 181 patients who underwent PLSS, the mean LOS was 3.57 days. Based on the Charlson Comorbidity Index (CCI) and American Society of Anesthesiologist (ASA) classification, patients with Medicaid insurance were healthier than non-Medicaid patients (mean CCI: 0.34 versus 0.65; p = 0.041, ASA: 1.71 versus 1.91; p = 0.046) yet Medicaid patients had a longer LOS compared with non-Medicaid patients (mean LOS: 4.03 versus 3.30 days; p = 0.047). There was no significant difference in discharge disposition between Medicaid and non-Medicaid patients (Home = 82.35 % versus 79.65 %; p = 0.855). Medicaid patients also had significantly less spinal levels involved in their surgery (1.44 versus 1.67; p = 0.027). Multivariable regression modeling identified independent risk factors positively associated with increased LOS as age (+1.0 % per year; p = 0.007), Medicaid insurance status (+28.7 %; p = 0.007), and CCI (10.1 % per increment in CCI; p = 0.030). Fusion surgery also was an independent risk factor for increased LOS when compared with laminectomy (−54.1 %; p < 0.001) or discectomy (−51.3 %; p < 0.001). Conclusions: Increasing age, Medicaid insurance status, higher CCI, and fusion surgery were independently associated with increased LOS after PLSS. This information is useful for preoperative patient counseling, shared decision-making, and risk stratification and may help to further ongoing discussion regarding contributors to rising health care costs. Findings of increased LOS among Medicaid patients will help direct efforts to identify factors that contribute to this health care expense.
1. Introduction Given the increased number of patients insured through Medicaid due to the Affordable Care Act, insurance status is a key component of discussions regarding health care costs. Since the Act was passed in 2010, as many as 17 million Medicaid patients have been added to the insurance pool. As a result, the number of spine surgeries has increased
⁎
dramatically, with a 150 % increase in Medicaid payments for spinal surgeries from 2010 to 2016 [1]. Medicaid patients have been shown to have worse surgical outcomes and higher costs of care than non-Medicaid patients [2]. Patient satisfaction scores can be used to measure provider and treatment effectiveness; Medicaid patients overall have lower levels of satisfaction than non-Medicaid patients, which has been shown to affect reimbursements in some models [3]. Since Medicaid
Corresponding author at: Department of Neurosurgery, Westchester Medical Center, New York Medical College, 100 Woods Road, Valhalla, NY, 10595, USA. E-mail address: [email protected] (C.A. Bowers).
https://doi.org/10.1016/j.clineuro.2019.105570 Received 13 September 2019; Received in revised form 17 October 2019; Accepted 21 October 2019 Available online 24 October 2019 0303-8467/ © 2019 Elsevier B.V. All rights reserved.
Clinical Neurology and Neurosurgery 188 (2020) 105570
J.F. Dominguez, et al.
were reported as means ± standard deviation (SD) and/or medians, as appropriate based on testing of assumptions for variable distribution. Differences in metric values between Medicaid versus non-Medicaid patients were assessed using the independent samples t-test or MannWhitney U-test, as applicable. To identify factors that independently impact LOS in patients undergoing PLSS, we constructed an ordinary least squares (OLS) regression model. Initial regression analysis demonstrated heteroskedasticity due to the non-Gaussian distribution of LOS. Therefore, we performed several arithmetic transformations, including cubic, square, square root, and logarithmic (natural and to the base 10) and found that the natural logarithmic (ln) transformation of LOS provided the best fit to normality. Using the ln transformed LOS, we constructed a multivariable OLS regression model to identify potential risk factors for increased LOS while adjusting for each independent variable. Regression diagnostics were tested using variance (coefficient of determination) and residual analyses. To interpret coefficient estimates from ln transformed OLS modeling, we back-transformed coefficients; the effects of individual variables were reported in terms of per increment percent change on LOS. All statistical analyses were performed using IBM SPSS v25.0 [22] and R v3.3.3 [23]. A p value < 0.05 was considered statistically significant.
patients have significantly lower surgeon reimbursement rates and worse surgical outcomes, many providers refuse to see Medicaid patients and primary care physicians can encounter difficulty finding appropriate specialty care for their Medicaid patients [4]. Lumbar degenerative disc disease (LDDD) is a common pathological spinal condition with associated lower back pain that can negatively impact patients’ quality of life [5,6]. Surgical intervention to alleviate back pain for patients refractory to conservative treatment measures frequently consists of laminectomy, discectomy, and/or fusion [7]. During the past decade, health care spending for lumbar spinal stenosis (LSS) surgery has increased from $1.8 to $8.2 billion, as hospitals have experienced a doubling of spine surgeries performed during that period [8,9]. Factors contributing to increased costs of hospitalization are multifactorial. Age, sex, American Society of Anesthesiologists (ASA) score [10], Oswestry Disability Index, Medicare/Medicaid, ambulation status, patient-reported allergies, fusion surgery, increased operating time, and medical comorbidities [often categorized by the Charlson Comorbidity Index (CCI)], have been associated with increased LOS and increased complication rates and readmission rates for spinal surgery patients [11–19]. Increased LOS is a significant contributor to hospital costs and patient health care costs [20]. In this retrospective study of patients undergoing primary lumbar spine surgery (PLSS) for LDDD, we analyzed multiple variables related to LOS, including age, sex, body mass index (BMI), insurance type, and CCI score [21].
3. Results 3.1. Patient demographics and clinical characteristics
2. Materials and methods Overall, 181 patients undergoing PLSS for HNP, lumbar stenosis, and/or spondylolisthesis were included. The mean age of the cohort was 52.29 ± 13.42 years, and 87 (48.07 %) were women; the average BMI was 30.83 ± 9.17 kg/m2. Of the 181 total patients, 105 (58.01 %) had fusion, 54 (29.83 %) had discectomy, and 22 (12.15 %) had laminectomy. The type of surgery performed did not differ significantly based on insurance status (Medicaid versus non-Medicaid). Prevalent comorbidities identified among the cohort were diabetes mellitus (DM) (28 patients, 15.47 %) and psychiatric diagnoses (32 patients, 17.68 %). Medicaid patients had fewer comorbidities than non-Medicaid patients, as indicated by the CCI score and ASA classification (mean 0.34 vs. 0.65; p = 0.041, ASA: 1.71 versus 1.91; p = 0.046). Discharge disposition was not significantly different between Medicaid and nonMedicaid patients (Home = 82.35 % versus 79.65 %; p = 0.855). Medicaid patients also had significantly less spinal levels involved in their surgery (1.44 versus 1.67; p = 0.027) (Table 1). Primary insurance providers for patients in the cohort included Medicaid (68 patients, 37.57 %), Medicare (38 patients, 20.99 %), private payers (36 patients, 19.89 %), workers’ compensation (24 patients, 13.26 %), and other sources (14 patients,7.73 %); one (0.55 %) patient was uninsured (Table 2). The mean LOS was 3.57 ± 2.81 days (median: 3.00 days). Despite having a lower CCI score, Medicaid patients had a longer mean LOS compared with non-Medicaid patients (4.03 vs. 3.30 days; p = 0.047) (Table 3).
2.1. Study design and patient characteristics In this retrospective single-center study, we reviewed the electronic medical records for 181 consecutive adult patients 18 years of age or older who had PLSS (1–3 levels of involvement) from July 1, 2014, to July 1, 2017. Patients who underwent laminectomy, discectomy, or fusion for herniated nucleus pulposus (HNP), lumbar stenosis, and/or spondylolisthesis were included. Patients with spinal neoplasia or congenital spinal disease and any revision operations were excluded. The study was approved by the Institutional Review Board of the Westchester Medical Center. 2.2. Variables Factors previously demonstrated to impact LOS or thought by the investigators to potentially impact LOS were defined as independent variables. These were patient demographics (age, sex, and insurance status), BMI, patient-reported allergies, discharge disposition, and comorbidities (ASA, CCI, diabetes mellitus [DM] and psychiatric comorbidities such as anxiety, depression, and/or substance abuse). The surgeon’s years of experience also was included as an independent variable in addition to whether there were two co-surgeons for the surgical procedure. Patient insurance status categories included Medicare, Medicaid, workers’ compensation, and various private commercial insurances. For purposes of statistical analysis, patients with private insurance as a supplement to their Medicare coverage were categorized as Medicare patients. Data on patient-reported allergies were also recorded. Medical comorbidities were used to calculate a CCI score.
3.2. Univariate analysis In univariate OLS regression analyses, factors that demonstrated borderline statistical significance (defined as p < 0.10) in their association with LOS included: advancing age (+0.9 % per yearly increment; p = 0.033), Medicaid insurance status (+22.0 %; p = 0.085), female sex (+21.0 %; p = 0.090); BMI (+1.1 %; p = 0.074); CCI (15.4 % per increment in CCI score; p = 0.003); DM (+45.4 %; 0.016); psychiatric disorder (+35.5 %; p = 0.039), and surgery type [laminectomy (−39.7 %; p = 0.003); discectomy (−52.6 %; p < 0.001); fusion (+137.1 %; p < 0.001)]. Patient-reported allergies, the surgeon’s length of experience, and the presence of a co-surgeon for the procedure were not associated with increased LOS (Table 4). Variables that demonstrated borderline statistical significance in univariate
2.3. Statistical analysis Our primary outcome was to identify independent predictors for LOS in patients undergoing PLSS. Descriptive and inferential statistical techniques were used for data analysis. Categorical variables were described as counts or percentages. Percentage differences between Medicaid and non-Medicaid patients were analyzed using the Pearson Chi-square test or Fisher exact test as appropriate. Continuous variables 2
Clinical Neurology and Neurosurgery 188 (2020) 105570
J.F. Dominguez, et al.
Table 1 Characteristics of patients undergoing lumbar spine surgery between 2014–2017 based upon Medicaid status at our institute. Characteristics
Medicaid
Non-Medicaid
Overall
P value
Age, in years Mean ± SD Median
47.96 ± 10.25 48.00
54.90 ± 14.43 55.00
52.29 ± 13.42 52.00
< 0.001 < 0.001∗
Gender, n (%) Female Male
37 (54.41) 31 (45.59)
50 (44.25) 63 (55.75)
87 (48.07) 94 (51.93)
0.185
Body mass index (BMI), kgm−2 Mean ± SD Median
32.18 ± 10.40 29.00
29.96 ± 8.22 29.50
30.83 ± 9.17 29.00
0.143 0.420∗
Diabetes mellitus, n (%)‖ Psychiatric comorbidities, n (%)‖
9 (13.24) 12 (17.65)
19 (16.81) 20 (17.70)
28 (15.47) 32 (17.68)
0.519 0.993
Charlson comorbidity index (CCI) CCI 0-1, n (%) CCI ≥ 2, n (%) Mean CCI ± SD
62 (91.18) 6 (8.82) 0.34 ± 0.68
97 (85.84) 16 (14.16) 0.65 ± 1.37
159 (87.85) 22 (12.15) 0.54 ± 1.17
0.352† 0.041
Surgery type, n (%)‖ Laminectomy Discectomies Fusion/arthrodesis
6 (8.82) 21 (30.88) 41 (60.29)
16 (14.16) 33 (29.20) 64 (56.64)
22 (12.15) 54 (29.83) 105 (58.01)
0.287 0.811 0.629
Two surgeons, n (%) Co-scrubs, n (%)
15 (22.06) 12 (17.65)
32 (28.32) 22 (19.47)
47 (25.97) 34 (18.78)
0.352 0.761
Surgeon experience, in years Mean ± SD Median
16.59 ± 6.22 19.00
17.98 ± 6.70 18.00
17.46 ± 6.54 19.00
0.158 0.435∗
Patient-reported allergies (PRA) Overall Medications Food Environment
0.99 0.69 0.15 0.15
1.08 0.74 0.22 0.12
1.04 0.72 0.19 0.13
0.725 0.776 0.563 0.643
ASA score Mean ± SD Median
1.71 ± 0.75 2.00
1.91 ± 0.69 2.00
1.83 ± 0.72 2.00
0.069 0.046*
Number of levels operated Mean ± SD Median
1.44 ± 0.63 1.00
1.67 ± 0.74 2.00
1.59 ± 0.71 1.00
0.027 0.032*
Discharge, n (%) Home Acute Rehab Expired
56 (82.35) 11 (16.18) 1 (1.47)
90 (79.65) 22 (19.47) 1 (0.88)
146 (80.66) 33 (18.23) 2 (1.10)
0.855†
± ± ± ±
1.58 1.27 0.40 0.50
± ± ± ±
2.00 1.32 1.26 0.44
± ± ± ±
1.85 1.30 1.02 0.46
†p-values derived from Fisher’s exact test. ∗ p-values derived from non-parametric Mann-Whitney U test. ‖ Estimates based upon missing data for DM (9.9 %), psychiatric comorbidities (9.9 %), surgery type (8.8 %), co-surgeons (8.8 %), Charlson comorbidity (8.8 %), and allergies (8.8 %). Bold p-values indicate statistical significance. Table 2 Patient characteristics based upon insurance type. Primary insurance provider
N (%)
Medicaid Medicare Private payer Uninsured Prisoners Workman compensation Military
68 (37.57) 38 (20.99) 36 (19.89) 1 (0.55) 8 (4.42) 24 (13.26) 6 (3.31)
Table 3 Length of hospital stay (primary outcome) in patients undergoing lumbar spine surgery. Characteristics
Medicaid
Non-Medicaid
Overall
P value
LOS, days Mean ± SD Median
4.03 ± 2.84 3.50
3.30 ± 2.76 3.00
3.57 ± 2.81 3.00
0.047 0.034*
* p-values derived from non-parametric Mann-Whitney U test.
3
Clinical Neurology and Neurosurgery 188 (2020) 105570
J.F. Dominguez, et al.
Table 4 Regression models (univariate and multivariate) investigating factors associated with length of hospital stay. Univariate
Multivariate
Factors
% Change (95% CI)
P-value
% Change (95% CI)
P-value
Medicaid Age Female gender BMI DM Psychiatric CCI Surgery type Laminectomy Discectomies Fusion/arthrodesis Two-surgeons Co-scrub Surgeon experience Patient reported allergies Overall Medications Food Environment
+22.0% (-2.8 – +53.2) +0.9% (+0.1– +1.7) +21.0% (-3.0 – +50.8) +1.1% (-0.1 – +2.4) +45.4% (+7.5 – +96.6) +35.5% (+1.6 – +80.7) +15.4% (+5.2 – +26.6)
0.085^ 0.033^ 0.090^ 0.074^ 0.016^ 0.039^ 0.003^
+28.7% (+7.1 – +54.6) +1.0% (+0.3 – +1.8) +2.8% (-14.0 – +22.8) +0.8% (-0.2 – +1.8) +11.4% (-15.5 – +46.9) +20.6% (-4.1 – +51.7) +10.1% (1.0 – +20.1)
0.007 0.007 0.761 0.110 0.442 0.109 0.030
−39.7% (−56.7 – −16.0) −52.6% (−61.8 – −41.1) +137.1% (+97.0 – +185.3) −30.1% (−45.4 – −10.5) −40.4% (−54.7 – −21.7) +0.8% (−0.9 – 2.5)
0.003^ < 0.001^ < 0.001^ 0.005^ < 0.001^ 0.344
−54.1% (−65.1 – −39.6) −51.3% (−60.1 – −40.4) Reference +4.4% (−26.6 – +48.4) −30.7% (−53.5 – +3.2) –
< 0.001 < 0.001 Reference 0.811 0.071 –
0.687 0.406 0.544 0.603
– – – –
– – – –
+1.2% +3.7% −3.3% +6.8%
(−4.7 – +7.5) (−4.8 – +13.0) (−13.3 – +7.8) (−16.2 – +36.1)
^ indicates p-value < 0.1 in univariate analysis and will be included for multivariate analysis. Bold p-values indicate statistical significance.
increasingly important to identify and study factors that are associated with increased costs of PLSS in order to reduce health care expenditures. PLSS under general anesthesia costs approximately $13,206 [29], and an additional patient day of hospitalization raises hospital costs by $2102 [17]. A surgical patient’s LOS, which is a significant contributor to hospital and patient costs and a key hospital administrative priority [20], is potentially modifiable. Furthermore, since most PLSS operations are performed electively and the long-term efficacy of PLSS in alleviating pain cause by LDDD is controversial [30], preoperative patient counseling and risk stratification that incorporate factors relevant to both clinical outcome and health care expenditures are essential. As a single center retrospective study, our study has inherent limitations in its generalizability to other health care facilities in different geographic areas with different patient demographics and health care resources. Our analyses also focused on the short-term outcome of LOS. Additionally, the reluctance by health care providers to accept Medicaid patients may result in a diagnostic bias, whereby Medicaid patients' lower CCI score could be influenced by undiagnosed comorbidities. Additional research is indicated to replicate these findings in other populations and to evaluate the longer-term and broader implications of the several risk factors for increased LOS that we identified, including Medicaid insurance status.
analyses were included in subsequent multivariate analyses. 3.3. Multivariate analysis In a multivariable OLS regression model, independent risk factors associated with LOS were advancing age (+1.0 % increase in LOS per year; p = 0.007), Medicaid insurance status (+28.7 %; p = 0.007), CCI (10.1 % per increment in CCI score; p = 0.030), and undergoing laminectomy (−54.1 %; p < 0.001) or discectomy (−51.3 %; p < 0.001), compared with fusion (Table 4). 4. Discussion We present a retrospective cohort study of 181 adult patients who underwent PLSS for LDDD. Medicaid insurance status was an independent risk factor for increased LOS after PLSS. Despite being healthier than non-Medicaid patients (as indicated by the CCI and ASA classification), and more likely to be discharged home, Medicaid patients’ average LOS was 0.7 days longer than that of patients with any other type of insurance. The demonstration in this study of a positive association between Medicaid status and increased LOS in patients undergoing PLSS, which has not been reported previously among adult spine patients, is an important finding for several reasons. First, accompanying explosive growth in the number of Medicaid patients since enactment of the Affordable Care Act in 2010 has been an increasing reluctance among medical specialists to provide care for Medicaid patients. Providers’ hesitation to accept Medicaid patients is due to various factors that have been shown to be associated with Medicaid status, including a lower rate of provider reimbursement, poorer patient outcomes, and increased LOS across medical specialties [1,4,18,24–28]. Such concerns among providers are heightened in an era when pay for performance and episode-based bundles are common health care reforms, and patients’ increased LOS can lower a provider’s ranking or overall score [3,5]. Hospitals also may view Medicaid patients as less desirable due to their lower reimbursement rate [4]. These influences on providers’ decision making combined with the new finding that Medicaid patients undergoing PLSS have a longer mean LOS, may further increase the number of spine surgeons who are unwilling to treat Medicaid patients. Second, given the dramatic growth in the number of PLSS procedures being performed and the high costs of these surgeries, it is
5. Conclusions In this study, Medicaid insurance status, increasing age, higher CCI, and fusion surgery were independently associated with increased LOS among adult patients undergoing PLSS. Analysis of spinal levels involved, discharge disposition, and CCI demonstrated that the association between Medicaid insurance status and increased LOS was unlikely to be attributed to case complexity, discharge disposition or baseline health status. These findings may be useful for patient counseling, shared-decision making, risk stratification, and cost analyses. Identification of Medicaid status as an independent risk factor for increased LOS following PLSS merits further study, particularly given the increasing prevalence of spine surgery and the growth of the Medicaid patient population.
4
Clinical Neurology and Neurosurgery 188 (2020) 105570
J.F. Dominguez, et al.
Funding [14]
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
[15]
Declaration of Competing Interest The authors report no conflicts of interest.
[16]
References [17]
[1] J.E. Meyers, J. Wang, A. Khan, et al., Trends in physician reimbursement for spinal procedures since 2010, Spine (Phila Pa 1976) 43 (15) (2018) 1074–1079, https:// doi.org/10.1097/BRS.0000000000002516. [2] D.J. LaPar, C.M. Bhamidipati, C.M. Mery, et al., Primary payer status affects mortality for major surgical operations, Ann. Surg. 252 (3) (2010) 544–551, https:// doi.org/10.1097/SLA.0b013e3181e8fd75. [3] S. Chotai, A. Sivaganesan, S.L. Parker, et al., Patient-specific factors associated with dissatisfaction after elective surgery for degenerative spine diseases, Neurosurgery 77 (2) (2015) 157–163, https://doi.org/10.1227/NEU.0000000000000768 discussion 163. [4] S.K. Long, Physicians may need more than higher reimbursements to expand medicaid participation: findings from Washington State, Health Aff. (Millwood) 32 (9) (2013) 1560–1567, https://doi.org/10.1377/hlthaff.2012.1010. [5] D.M. Cutler, K. Ghosh, The potential for cost savings through bundled episode payments, N. Engl. J. Med. 366 (12) (2012) 1075–1077, https://doi.org/10.1056/ NEJMp1113361. [6] F.M. Kovacs, V. Abraira, J. Zamora, et al., Correlation between pain, disability, and quality of life in patients with common low back pain, Spine (Phila Pa 1976) 29 (2) (2004) 206–210, https://doi.org/10.1097/01.BRS.0000107235.47465.08. [7] X.L. Ma, X.W. Zhao, J.X. Ma, et al., Effectiveness of surgery versus conservative treatment for lumbar spinal stenosis: a system review and meta-analysis of randomized controlled trials, Int. J. Surg. 44 (2017) 329–338, https://doi.org/10.1016/j. ijsu.2017.07.032. [8] M. Raad, C.J. Donaldson, M.H. El Dafrawy, et al., Trends in isolated lumbar spinal stenosis surgery among working US adults aged 40-64 years, 2010–2014, J. Neurosurg. Spine 29 (2) (2018) 169–175, https://doi.org/10.3171/2018.1. SPINE17964. [9] N.K. Sharma, B. Olotu, A. Mathew, et al., Lumbar spine surgeries and medication usage during hospital stay: one-center perspective, Hosp. Pharm. 52 (11) (2017) 774–780, https://doi.org/10.1177/0018578717737431. [10] C.M. Kuza, K. Matsushima, W.J. Mack, et al., The role of the American Society of Anesthesiologists physical status classification in predicting trauma mortality and outcomes, Am. J. Surg. 10 (19) (2019) 30254–30255, https://doi.org/10.1016/j. amjsurg.2019.09.019. [11] J.A. Gruskay, M. Fu, D.D. Bohl, et al., Factors affecting length of stay after elective posterior lumbar spine surgery: a multivariate analysis, Spine J. 15 (6) (2015) 1188–1195, https://doi.org/10.1016/j.spinee.2013.10.022. [12] C. Lagman, B. Ugiliweneza, M. Boakye, et al., Spine surgery outcomes in elderly patients versus general adult patients in the United States: a market scan analysis, World Neurosurg. 103 (2017) 780–788, https://doi.org/10.1016/j.wneu.2017.04. 001. [13] M.W. Manoso, A.M. Cizik, R.J. Bransford, et al., Medicaid status is associated with
[18]
[19]
[20]
[21]
[22] [23] [24]
[25]
[26] [27]
[28]
[29]
[30]
5
higher surgical site infection rates after spine surgery, Spine (Phila Pa 1976) 39 (20) (2014) 1707–1713, https://doi.org/10.1097/BRS.0000000000000496. M.J. McGirt, S.L. Parker, S. Chotai, et al., Predictors of extended length of stay, discharge to inpatient rehab, and hospital readmission following elective lumbar spine surgery: introduction of the Carolina-Semmes Grading Scale, J. Neurosurg. Spine 27 (4) (2017) 382–390, https://doi.org/10.3171/2016.12.SPINE16928. P.G. Passias, C.M. Jalai, N. Worley, et al., Predictors of hospital length of stay and 30-day readmission in cervical spondylotic myelopathy patients: an analysis of 3057 patients using the ACS-NSQIP Database, World Neurosurg. 110 (2018) e450–e458, https://doi.org/10.1016/j.wneu.2017.11.009. M.J. Press, D.P. Scanlon, A.S. Navathe, et al., The importance of clinical severity in the measurement of hospital readmission rates for Medicare beneficiaries, 19972007, Med. Care Res. Rev. 70 (6) (2013) 653–665, https://doi.org/10.1177/ 1077558713496167. R.C. Puffer, R. Planchard, G.W. Mallory, et al., Patient-specific factors affecting hospital costs in lumbar spine surgery, J. Neurosurg. Spine 24 (1) (2016) 1–6, https://doi.org/10.3171/2015.3.SPINE141233. K. Rumalla, K.A. Smith, P.M. Arnold, Cervical total disc replacement and anterior cervical discectomy and fusion: reoperation rates, complications, and hospital resource utilization in 72 688 patients in the United States, Neurosurgery 82 (4) (2018) 441–453, https://doi.org/10.1093/neuros/nyx289. R.K. Wadhwa, J. Ohya, T.D. Vogel, et al., Risk factors for 30-day reoperation and 3month readmission: analysis from the Quality and Outcomes Database lumbar spine registry, J. Neurosurg. Spine 27 (2) (2017) 131–136, https://doi.org/10.3171/ 2016.12.SPINE16714. S. Missios, K. Bekelis, Hospitalization cost after spine surgery in the United States of America, J. Clin. Neurosci. 22 (10) (2015) 1632–1637, https://doi.org/10.1016/j. jocn.2015.05.005. W. D’Hoore, C. Sicotte, C. Tilquin, Risk adjustment in outcome assessment: the charlson comorbidity index, Methods Inf. Med. 32 (5) (1993) 382–387, https://doi. org/10.1055/s-0038-1634956. IBM Corp, IBM SPSS Statistics for Windows, Version 25.0. Armonk, NY, (2017). R Core Team, R: a Language and Environment for Statistical Computing, Version 3.3.3. R Foundation for Statistical Computing, Vienna, Austria, 2017. S.K. Cho, N.N. Egorova, The association between insurance status and complications, length of stay, and costs for pediatric idiopathic scoliosis, Spine (Phila Pa 1976) 40 (4) (2015) 247–256, https://doi.org/10.1097/BRS.0000000000000729. R.B. Hakim, M.B. Benedict, N.J. Merrick, Quality of care for women undergoing a hysterectomy: effects of insurance and race/ethnicity, Am. J. Public Health 94 (8) (2004) 1399–1405, https://doi.org/10.2105/ajph.94.8.1399. J. Kwok, The impact of health insurance status on the survival of patients with head and neck, Cancer 116 (2) (2010) 476–485, https://doi.org/10.1002/cncr.24774. J.J. Shen, E.L. Washington, Disparities in outcomes among patients with stroke associated with insurance status, Stroke 38 (3) (2007) 1010–1016, https://doi.org/ 10.1161/01.STR.0000257312.12989.af. D.B. Stulberg, J.X. Zhang, S.T. Lindau, Socioeconomic disparities in ectopic pregnancy: predictors of adverse outcomes from Illinois hospital-based care, 2000–2006, Matern. Child Health J. 15 (2) (2011) 234–241, https://doi.org/10. 1007/s10995-010-0579-6. P. Agarwal, J. Pierce, W.C. Welch, Cost Analysis of spinal versus general anesthesia for lumbar diskectomy and laminectomy spine surgery, World Neurosurg. 89 (2016) 266–271, https://doi.org/10.1016/j.wneu.2016.02.022. R.A. Deyo, S.K. Mirza, J.A. Turner, et al., Overtreating chronic back pain: Time to back off? J. Am. Board Fam. Med. 22 (1) (2009) 62–68, https://doi.org/10.3122/ jabfm.2009.01.080102.