Ambulatory surgery is defined as a procedure that does not require an overnight stay or inpatient service.1 The pace of ambulatory surgery has been on a steady rise in the US, largely due to the pressure to improve perioperative care, reduce length of stay, and ultimately uphold today’s healthcare values while improving quality.2 On the other hand, there is also an ongoing increase in ambulatory surgery centers (ASCs), which are intended to alleviate the volume and cost burden of tertiary care centers.2–4 These ASCs are aimed to provide a facility to perform services that require less than 24 hours as an alternative to performing them in a hospital outpatient (HO) setting. Currently, many spinal surgeries are gradually being transitioned from inpatient procedures to outpatient HO and ASC settings, as shown in several studies. One study reported that 84.2% of spine surgeons now perform surgeries in ASCs or outpatient settings.5
Lumbar decompression is one such procedure that has been increasingly performed in the outpatient setting, with some studies suggesting that up to 86% of all lumbar laminectomy procedures are now performed in the outpatient setting.6 With this rise in the use of outpatient settings, coupled with the goal of upholding healthcare values, there is a balance that needs to be found to alleviate healthcare costs that would otherwise result in inpatient settings. To date, most studies reporting outcomes and cost analysis for outpatient spinal procedures in the US have not distinguished the various outpatient settings from each other. In particular, freestanding ASCs, which have seen an exponential increase in number since the first one opened in 1970 in Arizona,7 are touted to provide more attractive profit margins for physicians due to lower administrative and overall costs.8
An increasing focus of efforts on curbing the high costs associated with back surgery has occurred in recent years.9 These efforts have been based around upholding the principles of value-based healthcare, in order to help mitigate low-value practices.10–12 Such low-value practices include unnecessarily prolonged length of stay and nonroutine discharge.13 Studies have suggested that unanticipated nonroutine discharge to a skilled nursing facility (SNF) or another acute care facility may be associated with significant cost increases. Given that a critical objective of outpatient surgery is to alleviate healthcare costs associated with inpatient services, it is imperative to investigate the differences in these low-value practices among various outpatient settings.
In the current study, the authors sought to compare index admission charges, as well as cumulative 7-, 30-, and 90-day charges of freestanding ASCs and HOs for patients undergoing outpatient lumbar decompression.
Methods
Data Source
The Florida State Ambulatory Surgery Database (SASD) was queried. This is a state-level database from the Agency for Healthcare Research and Quality (AHRQ) Healthcare Cost and Utilization Project (HCUP).14 The HCUP is a group of healthcare databases and related software tools which were developed by state-level and federal-level industry partnerships. The SASD includes encounter-level data for patients undergoing outpatient surgeries as well as other outpatient services such as observation stays, lithotripsy, radiation therapy, imaging, chemotherapy, and labor and delivery.15 The SASD includes encounter-level data synthesized from both HOs and independent ASCs.
Use of the SASD began in 1997, and at the time of this writing 5 states were enrolled. The information contained for each encounter includes a large set of clinical and nonclinical information, including but not limited to all listed diagnoses and procedures for the encounter, patient demographics (sex, age, race), expected payment source, and total charges.15
State Selection
We first limited our analysis to only those states that had available databases for the years 2013–2014, were high-volume states both in terms of population and in terms of popularity of outpatient surgery and ASCs, and were in two different census divisions. We initially selected Florida and New York databases for this project, since they are ranked second and seventh among states with the greatest number of ASCs, with 722 and 333 ambulatory centers, respectively. However, because we could only identify 41 HO lumbar decompression cases for New York, we limited analysis to Florida, which provided more HO cases.
Cohort Selection
Adult patients with lumbar degenerative disc disease undergoing lumbar decompression were included. The following Current Procedural Terminology (CPT) codes were used: 63012, posterior extradural laminotomy or laminectomy for exploration/decompression of neural elements or excision of herniated intervertebral disc; 63017, laminectomy with exploration and/or decompression of spinal cord and/or cauda equina, without facetectomy, foraminotomy; 63030, posterior extradural laminotomy or laminectomy for exploration/decompression of neural elements or excision of herniated intervertebral disc procedure; 63042, laminotomy, hemilaminectomy, with decompression of nerve root(s), including partial facetectomy, foraminotomy and/or excision of herniated intervertebral disc, reexploration, single interspace; 63047, posterior extradural laminotomy or laminectomy for exploration/decompression of neural elements or excision of herniated intervertebral disc procedures; or 63005, posterior extradural laminotomy or laminectomy for exploration/decompression of neural elements or excision of herniated intervertebral disc procedures. International Classification of Diseases, Ninth Revision (ICD-9) codes for lumbar disc degenerative disease (DDD) were used to select patients; these codes can be seen in Supplementary Table 1.
Patients with any diagnosis ICD-9 code for tumor (192.2, 225.3) or infection (722.93, 730.00, and 730.10) or fracture (805.x) were excluded. We also excluded patients who had a concurrent CPT code for lumbar fusion, excision for lumbar spine tumor, cervical spine procedure, or thoracic spine procedure. These codes can be seen in Supplementary Tables 2–4.
Covariates
The following covariates were included: age, sex, race, insurance status, zip code median income quartile, number of Elixhauser comorbidities present, and number of operative levels. Age was grouped into distinct categories, which included 18–40, 41–65, and above 65 years; the following race categories are reported in the HCUP databases: African American, Asian or Pacific Islander, Hispanic, Native American, White, or other; the following insurances are documented: Medicaid, Medicare, private insurance, self-pay, no charge, or other; number of comorbidities were grouped as 0, 1, or 2+; number of operative levels were documented using CPT codes and were grouped into 1 level, 2 levels, or 3–5 levels.
Outcomes
The primary outcomes of interest included rate of overnight stays, nonroutine discharge, index admission charge, and cumulative follow-up charge at 7, 30, and 90 days. For overnight stay, the variable “STATE_OS” was used.16 For discharge destination, the variable “DISPUNIFORM” offered by the SASD was used; nonroutine discharge was described as a discharge destination other than “home routine” or “home healthcare.” For the state of Florida, the reported charge also contains the “professional fee,” which includes the fee charged by the surgeon.17 These charges include those that are billed by the hospitals but are not among the charges that are reimbursed.
The cumulative charge was calculated by adding together any hospital, ASC, or emergency department charge for subsequent admissions within the specific time point (7, 30, or 90 days) after the initial surgical encounter.
Statistical Analysis
Continuous variables were summarized using means and standard deviations and compared between the two groups using the t-test or Wilcoxon rank-sum test. Categorical variables were summarized using frequencies and proportions and compared between the two groups using the chi-square or Fisher’s exact test. Multivariable linear regression was used to assess the impact of outpatient type on index admission charge, adjusting for age, sex, insurance type, number of comorbidities, and number of operative levels. Within the multivariable regression, an interaction term between outpatient type and insurance type was added to assess the relationship to admission cost. Moreover, marginal effect analysis was used to assess the percent change in cost for ASC versus HO admission within each insurance type.18,19 All statistical analyses were performed using Stata 16 (StataCorp LP) and R version 4.0.0 (R Foundation for Statistical Computing).
Results
Demographics
We identified a total of 25,486 patients; of these, 7067 patients (27.7%) underwent lumbar decompression in a freestanding ASC and 18,419 (72.3%) in HO. Patients in the ASC cohort were found to be significantly younger than those in the HO group (50.9 ± 13.6 vs 58.1 ± 15.7 years; p < 0.001). Patients in the ASC group were also somewhat more likely to be male (59.2%, n = 4185, vs 56.4%, n = 10,390; p < 0.001). A much higher proportion of patients in the ASC group had private insurance (76.5%, n = 5405, vs 45.7%, n = 8418); in contrast, these patients were less likely to be on Medicaid or Medicare compared to HO patients (7.4%, n = 518, vs 45.4%, n = 8349; p < 0.001). None of the patients in the ASC group were found to have “no charge”; for HO, 0.6% (n = 119) of patients had no charge. We also found a higher proportion of “self-pay” patients in the ASC group (8.2%, n = 576, vs 1.1%, n = 192). For patients in the ASC group who had “other” insurance (8.1%, n = 568), most (54.2%, n = 308) had workers’ compensation. In addition, 9.7% (n = 55) had TRICARE or other federal government insurance, and 0.9% (n < 11) had other state/local government or Veterans Affairs (VA) insurance. The remaining 35.2% of these patients (n = 200) had “other unknown” insurance. For patients in the HO group who had “other” insurance (7.3%, n = 1341), similarly to the ASC group, most patients (48.7%, n = 653) had workers’ compensation. In addition, 35.8% (n = 480) had TRICARE or other federal government insurance, 4.5% (n = 61) had VA insurance, and 7% (n = 94) had other state/local government insurance, while the remaining 51 patients (3.8%) had “other unknown” insurance. These results are summarized in Table 1.
Comparison of demographic characteristics between patients treated in the ASC group and patients treated in the HO group
| ASC (n = 7067) | HO (n = 18,419) | p Value | |
|---|---|---|---|
| Age, yrs | 50.926 ± 13.588 | 58.081 ± 15.690 | <0.001 |
| Age group, yrs | <0.001 | ||
| 18–40 | 1604 (22.7%) | 2919 (15.8%) | |
| 41–65 | 4608 (65.2%) | 8404 (45.6%) | |
| >65 | 854 (12.1%) | 7096 (38.5%) | |
| Female | 2882 (40.8%) | 8029 (43.6%) | <0.001 |
| Race | <0.001 | ||
| No. missing | 224 | 136 | |
| African American | 493 (7.2%) | 1011 (5.5%) | |
| Asian or Pacific Islander | 54 (0.8%) | 92 (0.5%) | |
| Hispanic | 500 (7.3%) | 1351 (7.4%) | |
| Native American | 28 (0.4%) | 14 (0.1%) | |
| Other | 73 (1.1%) | 303 (1.7%) | |
| White | 5695 (83.2%) | 15,512 (84.8%) | |
| Insurance | <0.001 | ||
| Medicaid | <11 | 454 (2.5%) | |
| Medicare | 514 (7.3%) | 7895 (42.9%) | |
| No charge | 0 (0.0%) | 119 (0.6%) | |
| Private insurance | 5405 (76.5%) | 8418 (45.7%) | |
| Self-pay | 576 (8.2%) | 192 (1.0%) | |
| Other | 568 (8.0%) | 1341 (7.3%) | |
| KidCare | 0 (0.0%) | <11 | |
| Other state/local government | <11 | 94 (7.0%) | |
| TRICARE or other federal government | 55 (9.7%) | 480 (35.8%) | |
| VA | <11 | 61 (4.5%) | |
| Workers’ compensation | 308 (54.2%) | 653 (48.7%) | |
| Other/unknown | 200 (35.2%) | 51 (3.8%) | |
| Zip code income quartile | <0.001 | ||
| 1 | 2048 (29.8%) | 5325 (29.4%) | |
| 2 | 2338 (34.0%) | 6305 (34.8%) | |
| 3 | 1662 (24.2%) | 4658 (25.7%) | |
| 4 | 822 (12.0%) | 1833 (10.1%) |
Values are presented as number of patients (%) or mean ± SD unless otherwise indicated. Boldface type indicates statistical significance.
Clinical Characteristics
Patients in the ASC group were healthier on average, as evidenced by the substantially lower proportion of patients who had more than one comorbidity, compared to those in the HO group (1.7%, n = 120, vs 63.8%, n = 11,748; p < 0.001). Using ICD-9 codes, we also reported the particular degenerative diagnoses for each patient and found that patients in the ASC group were less likely to have a diagnosis code for sciatica (0.4%, n = 25, vs 2.3%, n = 431), more likely to have a diagnosis code for spondylosis or spondylolisthesis (56.1%, n = 3966, vs 10.%, n = 1928), more likely to have a diagnosis code for disc herniation (91.3%, n = 6449, vs 69.3%, n = 12,764), more likely to have a diagnosis code for spinal stenosis (70.1%, n = 4955, vs 37.3%, n = 6878), and more likely to have a diagnosis code for backache (43.6%, n = 3080, vs 14.4%, n = 2654, p < 0.001). We did not observe a substantial difference in terms of operative complexity as indicated by number of levels. The mean number of operative levels for patients in the ASC group was 1.04 ± 0.22 compared to 1.08 ± 0.31 in the HO group; patients in the ASC group were slightly more likely to undergo a single-level procedure compared to those in the HO group (95.9%, n = 6775, vs 92.9%, n = 17,109; p < 0.001). These results have been summarized in Table 2.
Comparison of clinical characteristics between patients in the ASC group and patients in the HO group
| ASC (n = 7067) | HO (n = 18,419) | p Value | |
|---|---|---|---|
| No. of comorbidities | <0.001 | ||
| >2 | <11 | 8319 (45.2%) | |
| 0 | 426 (6.0%) | 983 (5.3%) | |
| 1 | 6521 (92.3%) | 5688 (30.9%) | |
| 2 | 117 (1.7%) | 3429 (18.6%) | |
| Diagnosis | |||
| Spinal stenosis | 4955 (70.1%) | 6878 (37.3%) | <0.001 |
| Disc herniation | 6449 (91.3%) | 12,764 (69.3%) | <0.001 |
| Backache | 3080 (43.6%) | 2654 (14.4%) | <0.001 |
| Spondylosis/spondylolisthesis | 3966 (56.1%) | 1928 (10.5%) | <0.001 |
| Radiculopathy | 2743 (38.8%) | 3551 (19.3%) | <0.001 |
| Other degenerative disease | 1179 (16.7%) | 819 (4.4%) | <0.001 |
| Sciatica | 25 (0.4%) | 431 (2.3%) | <0.001 |
| Postlaminectomy syndrome | 107 (1.5%) | 171 (0.9%) | <0.001 |
| No. of levels | <0.001 | ||
| Mean | 1.044 ± 0.222 | 1.080 ± 0.310 | |
| Range | 1.000–5.000 | 1.000–5.000 | |
| No. of levels | |||
| 1 | 6775 (95.9%) | 17,109 (92.9%) | |
| 2 | 276 (3.9%) | 1173 (6.4%) | |
| 3–5 | 16 (0.2%) | 137 (0.7%) |
Values are presented as number of patient (%) or mean ± SD unless otherwise indicated. Boldface type indicates statistical significance.
Overnight Stay and Discharge Destination
We found that no patient in the ASC group required an overnight stay: all were discharged on the same day. Among patients in the HO group, 9.2% (n = 1691) had an overnight stay (p < 0.001). Similarly, less than 11 patients had a nonroutine discharge (transfer to short-term hospital, SNF, or intermediate care facility (ICF), compared to 0.8% (n = 134) patients in the HO group (7 patients left against medical advice [LAMA], and 127 patients discharged to short-term hospital, SNF, or ICF). These results have been summarized in Table 3.
Analysis of length of stay and discharge destination between ASC and HO
| ASC (n = 7067) | HO (n = 18,419) | p Value | |
|---|---|---|---|
| LOS group | <0.001 | ||
| >1 day | 0 (0.0%) | 1691 (9.2%) | |
| Same day or day 1 discharge | 7067 (100.0%) | 16,728 (90.8%) | |
| Discharge destination | <0.001 | ||
| Home healthcare | <11 | 950 (5.2%) | |
| Home routine | 7061 (99.9%) | 17,334 (94.1%) | |
| LAMA | 0 (0.0%) | <11 | |
| Transfer to short-term hospital, SNF, or ICF | <11 | 127 (0.7%) |
LOS = length of stay.
Values are presented as number of patients (%) or mean ± SD unless otherwise indicated. Boldface type indicates statistical significance.
Rate and Reasons for Readmissions
While rates and reasons for readmissions and reoperations have been covered in a different manuscript (our unpublished data, 2020), we summarize here briefly. A total of 520 (7.36%) patients in the ASC group had a 7-day readmission. The reasons for readmissions were documented from the first two related ICD-9 diagnosis codes at the repeat visit. The most common reason for readmission was the same diagnosis as that at the time of the index procedure (93.7%, n = 485/520 patients; 6.9% overall), postoperative pain (3.6%, n = 19; 0.3% overall), operative complication (9.2%, n = 48; 0.7% overall), and surgical site infection (SSI) or other wound complication (< 11). In the HO group, 166 patients (0.9%) had a 7-day readmission; the most common reason was reemergence of symptom (56.6%, n = 94; 0.5% overall), followed by postoperative pain (26.5%, n = 44; 0.24% overall), SSI or wound complication (22.3%, n = 37; 0.21% overall), operative complication (10.8%, n = 18; 0.1% overall), and deep vein thrombosis (DVT) or pulmonary embolism (PE) (< 11).
A total of 670 patients (9.5%) had a 30-day readmission in the ASC group. Of these, 89.7% (n = 601/670 patients; 8.5% overall) had the same diagnosis as that at the time of the index procedure. These patients had the same or similar diagnosis codes as those at index admission. Other reasons for readmission in the ASC group included operative complication (14.9%, n = 100; 1.4% overall), postoperative pain (4.5%, n = 30; 0.42% overall), SSI or other wound complication (3.9%, n = 26; 0.36% overall), and DVT/PE (< 11). Of the 480 (2.6%) patients in the HO group who had a 30-day readmission, the most common reason was reemergence of symptoms (56.25%, n = 270; 1.5% overall), followed by SSI or other wound complications (32.1%, n = 154; 0.83% overall).
The rate of 90-day readmission in the ASC group was found to be 11.72%. The most common reason for readmission was the same diagnosis as the index operation (91.9%, n = 556/607 patients; 10.8% overall) with readmission due to reemergence of preoperative symptoms. Other reasons for readmission in the ASC group included operative complications (17.8%, n = 108; 2.1% overall), SSI or other wound complication (4.5%, n = 27; 0.5% overall), postoperative pain (3.1%, n = 19; 0.4% overall), and DVT/PE (< 11). The rate of 90-day readmission was found to be 5.56% for HO; the most common reason was readmission due to index operation diagnosis (73.2%, n = 557; 4.1% overall), followed by SSI or other wound complications (21.7%, n = 165; 1.2% overall). These results have been summarized in the Supplementary Tables.
Index Admission and Subsequent Charge
The mean index admission charge for the ASC group was found to be $35,017.28 ± $14,335.60, which was higher than that for the HO group ($33,881.50 ± $15,023.70; p < 0.001). We also analyzed 7-, 30-, and 90-day follow-up charges for patients who were readmitted to any facility. Patients in the ASC who had a 7-day readmission (n = 546) had a mean readmission charge of $74,860.06 ± $84,445.60, which was significantly higher than the mean 7-day readmission charge for HO ($49,132.41 ± $69,767.20; p < 0.001). The mean cumulative 7-day charge after adding the index admission charge for patients who had a readmission was found to be $120,500.20 ± $91,384.2) for ASC and $83,783.60 ± $75,736.20 for HO; the difference was significant (p < 0.001). We further evaluated average 7-day charges for the entire cohort, including both patients with and without 7-day readmissions; the average charge for patients in the ASC group was found to be $44,314.10 ± $46,785.71 compared to $34,649.21 ± $18,797.00 (p < 0.001).
The mean 30-day readmission charge for patients in the ASC group (n = 700) was found to be $74,117.91 ± $82,167.90, which was significantly higher compared to patients in the HO group with a 30-day readmission (n = 488; $55,976.68 ± $67,015.90; p < 0.001). Cumulative 30-day charges for patients in the ASC group were also higher compared to those for patients in the HO group ($118,861.50 ± $89,000.50 vs $91,400.10 ± $74,338.24; p < 0.001). Mean 90-day readmission charges for patients in the ASC group (n = 605) were significantly higher compared to patients in the HO group (n = 761) ($71,337.60 ± $78,409.40 vs $59,574.20 ± $70,485.60; p = 0.001). Cumulative 90-day readmission charges were also significantly higher for patients in the ASC group compared to the HO group ($114,306 ± $85,769.43 vs $94,034.60 ± $76,654.50; p < 0.001). These results have been summarized in Table 4.
Analysis of index admission charge and 7-, 30-, and 90-day charges for ASC and HO
| ASC | HO | p Value | |
|---|---|---|---|
| Index admission | n = 7067 | n = 18,419 | |
| Mean | 35,017.28 ± 14,335.60 | 33,881.50 ± 15,023.70 | <0.001 |
| Range | 750.00–304,854.00 | 277.00–232,463.00 | |
| 7-day readmission | n = 546 | n = 168 | |
| Mean | 74,860.06 ± 84,445.63 | 49,132.44 ± 69,767.24 | <0.001 |
| Range | 1838.00–529,171.00 | 343.00–657,813.00 | |
| Cumulative 7-day mean | 120,500.20 ± 91,384.20 | 83,783.60 ± 75,736.20 | <0.001 |
| Range | 12,698.00–613,750.00 | 15,601.00–739,204.00 | |
| 30-day readmission | n = 700 | n = 488 | |
| Mean | 74,117.92 ± 82,167.90 | 55,976.61 ± 67,015.90 | <0.001 |
| Range | 1838.00–54,813.80 | 343.00–657,813.00 | |
| Cumulative 30-day mean | 118,861.47 ± 89,000.54 | 91,400.07 ± 74,338.18 | |
| Range | 11,200.00–632,717.00 | 15,601.00–739,204.00 | |
| 90-day readmission | n = 864 | n = 929 | |
| Mean | 71,137.56 ± 78,409.36 | 59,574.15 ± 70,485.56 | 0.001 |
| Range | 556.00–548,138.00 | 343.00–657,813.00 | |
| Cumulative 90-day readmission mean | 114,306.00 ± 85,769.37 | 94,034.57 ± 76,654.60 | <0.001 |
| Range | 4566.00–632,717.00 | 15,601.00–739,204.00 |
Values are presented as US dollars unless otherwise indicated. Mean values are presented ± SD. Boldface type indicates statistical significance.
Multivariable Analysis for Log of Index Admission Cost
With multivariable linear regression for log of index admission charges, we found HO to be significantly associated with lower index admission cost (B −0.05, 95% CI −0.06 to −0.03, p < 0.001). Other significant factors included age group, where patients aged 41–65 years had significantly higher admission costs compared to 18–40 years (B 0.07, 95% CI 0.05 to 0.08) and those aged above 65 years had significantly higher admission costs compared to those aged 18–40 years (B 0.08, 95% CI 0.07 to 0.12, p < 0.001); and insurance status, where private insurance compared to Medicare or Medicaid had higher admission costs (B 0.03, 95% CI 0.01 to 0.05, p < 0.001), those on self-pay compared to Medicare or Medicaid had significantly lower admission costs (B −0.11, 95% CI −0.15 to −0.07, p < 0.001), and those on other insurance compared to Medicare or Medicare also had lower significantly lower admission costs (B −0.15, 95% CI −0.18 to −0.13, p < 0.001). For operative levels, patients with a single-level surgery compared to 3–5 levels had significantly lower costs (B −0.34, 95% CI −0.41 to −0.26, p < 0.001), and patients who had 2-level surgery compared to 3 levels had significantly lower admission costs (B −0.20, 95% CI −0.28 to −0.13, p < 0.001). For presence of comorbidities, patients with no comorbidity compared to those with more than 2 comorbidities had significantly lower admission costs (B −0.19, 95% CI −0.22 to −0.17, p < 0.001), those with 1 comorbidity compared to those with more than 2 comorbidities had significantly lower admission costs (B −0.05, 95% CI −0.07 to −0.003, p < 0.001), and those with 2 comorbidities compared to those with more than 2 comorbidities also had significantly lower admission costs (B −0.03, 95% CI −0.047 to −0.011, p < 0.001). Multivariable regression results can be seen in Table 5.
Multivariable linear regression analysis for log of index admission cost
| Coefficient | SE | p Value | 95% CI | |
|---|---|---|---|---|
| Age category, yrs | ||||
| 18–40 | 0 | Ref | ||
| 41–65 | 0.06 | 0.007 | <0.001 | 0.046 to 0.076 |
| >65 | 0.08 | 0.011 | <0.001 | 0.062 to 0.111 |
| Insurance | ||||
| Medicare or Medicaid | 0 | Ref | ||
| Other | −0.32 | 0.101 | <0.001 | −0.52 to −0.12 |
| Private insurance | −0.15 | 0.09 | 0.117 | −0.34 to 0.03 |
| Self-pay | −0.23 | 0.027 | <0.001 | −0.283 to −0.174 |
| Outpatient type | ||||
| ASC | 0 | Ref | ||
| HO | −0.13 | 0.02 | <0.001 | −0.175 to −0.094 |
| No. of levels | ||||
| 3–5 | 0 | Ref | ||
| 1 | −0.36 | 0.036 | <0.001 | −0.431 to −0.289 |
| 2 | −0.21 | 0.037 | <0.001 | −0.29 to −0.142 |
| No. of comorbidities | ||||
| >2 | 0 | Ref | ||
| 0 | −0.18 | 0.013 | <0.001 | −0.214 to −0.163 |
| 1 | −0.06 | 0.007 | <0.001 | −0.076 to −0.04 |
| 2 | −0.034 | 0.009 | <0.001 | −0.052 to −0.017 |
Boldface type indicates statistical significance.
Marginal Effect Analysis
We also performed marginal effect analyses to further explore the differences in index admission costs between ASC and HO for each insurance type. We further explored the margins for each insurance type and outpatient type. For patients on Medicare or Medicaid, those undergoing the procedure in ASC were found to have adjusted mean charges of $37,457.75 (95% CI $36,003.65–$38,911.93), which was higher than the cost for those undergoing the procedure in an HO ($33,408.00, 95% CI $32,990.91–$33,826.04; p value for difference in means < 0.001). For patients on private insurance, adjusted mean charges for patients undergoing the procedure in ASC were $38,083.81 (95% CI $37,555.49–$38,612.12), which was higher than for those undergoing the procedure in an HO ($33,308.00, 95% CI $32,962.00–$33,654.16; p value for difference in means < 0.001). However, for patients on other insurance, those undergoing the procedure in an ASC were found to have lower adjusted charges than those undergoing the procedure in an HO (other: margin $23,172.28, 95% CI $22,330.36–24,014.19 vs $34,167.66, 95% CI $33,392.23–34,943.09; p value < 0.001 for difference in means). For patients who are on self-pay, adjusted charges were statistically similar between those undergoing the procedure in an ASC compared to those in an HO (self-pay: margin $32,340.7, 95% CI $31,084.63–$33,396.77, vs $31,987.52, 95% CI $30,029.06–$33,945.99; p = 0.827 for difference in means). Marginal effect analysis for each group can be seen in Table 6 and Fig. 1. Table 7 displays the difference in admission costs between ASC and HO within each insurance type.
Marginal effect analysis for each insurance type and outpatient type for index admission cost
| Insurance Type/Outpatient Type | Margin | SE | 95% CI |
|---|---|---|---|
| Medicare or Medicaid/ASC | 37,457.75 | 741.94 | 36,003.58–38,911.92 |
| Medicare or Medicaid/HO | 33,408.00 | 213.04 | 32,990.92–33,826.04 |
| Other/ASC | 23,172.28 | 429.56 | 22,330.36–24,014.19 |
| Other/HO | 34,167.66 | 395.63 | 33,392.23–34,943.09 |
| Private insurance/ASC | 38,083.81 | 269.55 | 37,555.49–38,612.12 |
| Private insurance/HO | 33,308.00 | 176.57 | 32,962.00–33,654.16 |
| Self-pay/ASC | 32,340.70 | 58,984.00 | 31,084.63–33,396.77 |
| Self-pay/HO | 31,987.52 | 999.23 | 30,029.06–33,945.99 |
Values are presented as US dollars.
Marginal effect analysis for interaction between insurance type and outpatient type. Figure is available in color online only.
Magnitude of cost difference for ASC in comparison with HO for each insurance type
| Delta | SE | p Value | 95% CI | |
|---|---|---|---|---|
| Medicare or Medicaid | 049.27 | 750.73 | <0.001 | 2577.87 to 5520.67 |
| Other | −10,995.40 | 576.22 | <0.001 | −12,124.80 to −9866.01 |
| Private insurance | 4775.72 | 308.51 | <0.001 | 4171.06 to 5380.38 |
| Self-pay | 253.18 | 1158.09 | 0.827 | −2016.65 to 2523.00 |
Values are presented as US dollars unless otherwise indicated.
Discussion
Using an all-payer database of state ambulatory surgery from the state with the second highest numbers of ASCs,20 we found higher overall charges for ASC for both the index admission and subsequent charges. At the index admission, ASCs incurred approximately $2000 more in charges; this was despite a 0% rate of overnight stays in the ASC group. Moreover, in exploring differences in adjusted charges for each insurance type, we found the largest difference in charges for patients on private insurance, where patients in ASC had, on average, $5000 higher charges compared to those undergoing lumbar decompression in an HO. Similarly, for patients on Medicare or Medicaid, this difference was almost $4000 higher for ASC.21–24 Our findings are contrary to those by Malik et al.,25 who investigated outcomes and costs of patients on Medicare Advantage and on commercial insurance undergoing outpatient lumbar decompression between ASC and HO using an insurance database. They found that for patients on Medicare Advantage undergoing the procedure at an ASC, the average reimbursed cost was, on average, $2000 less than their counterparts undergoing the same procedure in an HO; for patients on commercial insurance, this difference was $3000.25 A major difference between our study and the study by Malik et al. was the difference in region; our cohort only consisted of ASCs and HOs in Florida, while the other study encompassed the entire US. Thus, whether these cost differences are regional due to differences in payment policies and state regulation of ASCs remains to be investigated.
Interestingly, for patients who had “other” insurance, this difference was higher in HO, where patients incurred an additional average of $11,000 in charges for index admission. In both ASC and HO groups, the majority of such patients had workers’ compensation. This may reflect the fact that ASCs have been reported to be offering bundled pricing for patients on workers’ compensation, thus making them more attractive for companies to have contracts with.26
Study Limitations
There are some limitations to this study that should be acknowledged. The retrospective nature of this study may subject the cohorts to selection bias. Given the use of an administrative database, there is potential for miscoding. This is particularly true for ASCs, which have been found to have some coding discrepancies. Furthermore, we only found one state’s data sufficient to be able to distinguish HO from ASC. Covariates and outcomes included in the analysis were also limited by the availability of variables in the database. Factors such as preoperative functional status, patient-reported outcomes, and longer-term outcomes could not be evaluated. Moreover, charges analyzed in this study reflect billing costs by the hospital, and not what was reimbursed. Differences in patient management between centers were also not recorded. Moreover, the differences in comorbidities between the two groups may partially be attributed to coding errors in ASCs. There is also a possibility that some patients, who were discharged to another facility, may have been admitted from the same facility, as in the case of an elderly patient being admitted from hospice or an elderly care center and being discharged back to that facility.
Conclusions
These “real-world” results from an all-payer statewide database indicate that for outpatient spine surgery, ASCs may be associated with higher index admission and subsequent 7-, 30-, and 90-day charges. Given that ASCs are touted to have lower overall costs for patients and better profit margins for physicians due to lower administrative costs, our results warrant further investigation into billing policies for ASCs. Furthermore, additional research is necessary to determine whether these results are generalizable to practice in other states.
Acknowledgments
This publication was made possible by CTSA Grant Number UL1 TR002377 from the National Center for Advancing Translational Sciences (NCATS), a component of the National Institutes of Health (NIH). Its contents are solely the responsibility of the authors and do not necessarily represent the official view of NIH.
Shyam Kurian was also awarded the Department of Neurologic Surgery—Uihlein Professorship Research Grants for this project.
Disclosures
The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.
Author Contributions
Acquisition of data: Alvi. Analysis and interpretation of data: Alvi, Jeffery. Drafting the article: Alvi, Kurian. Critically revising the article: Alvi, Kurian, Naessens, Spinner. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Bydon. Statistical analysis: Alvi, Wahood, Zreik, Jeffery. Administrative/technical/material support: Bydon, Jeffery, Naessens, Spinner. Study supervision: Bydon.
Supplemental Information
Online-Only Content
Supplemental material is available with the online version of the article.
Supplementary Tables 1–7. https://thejns.org/doi/suppl/10.3171/2021.2.SPINE201820.
Previous Presentations
This paper was presented at the 2020 Gattle Neurosurgery Research Symposium, which was organized by Mayo Clinic, Rochester, MN, and conducted online, on October 16, 2020.
References
- 1↑
Jarrett P. The International Association for Ambulatory Surgery (IAAS). Ambul Surg. 2003;10(3):113.
- 2↑
Hollenbeck BK, Dunn RL, Suskind AM, et al. Ambulatory surgery centers and outpatient procedure use among Medicare beneficiaries. Med Care. 2014;52(10):926–931.
- 3
Hollenbeck BK, Dunn RL, Suskind AM, et al. Ambulatory surgery centers and their intended effects on outpatient surgery. Health Serv Res. 2015;50(5):1491–1507.
- 4
Best MJ, Buller LT, Eismont FJ. National trends in ambulatory surgery for intervertebral disc disorders and spinal stenosis: a 12-year analysis of the national surveys of ambulatory surgery. Spine(Phila Pa 1976).2015;40(21):1703–1711.
- 5↑
Baird EO, Brietzke SC, Weinberg AD, et al. Ambulatory spine surgery: a survey study. Global Spine J. 2014;4(3):157–160.
- 6↑
Bu D, Gang CH, Kushwaha B, Qureshi SA. Ambulatory spine surgery: outcomes compared to inpatient surgery. Sem Spine Surg. 2018;30(3):160–168.
- 7↑
Ambulatory Surgery Center Association (ASCA). Accessed April 1, 2021. https://www.ascassociation.org/aboutus/whatisanasc/history
- 8↑
Witiw CD, Wilson JR, Fehlings MG, Traynelis VC. Ambulatory surgical centers: improving quality of operative spine care?. Global Spine J. 2020;10(1)(suppl):29S–35S.
- 9↑
Martin BI, Deyo RA, Mirza SK, et al. Expenditures and health status among adults with back and neck problems. JAMA. 2008;299(6):656–664.
- 10
Asher AL, Speroff T, Dittus RS, et al. The National Neurosurgery Quality and Outcomes Database (N2QOD): a collaborative North American outcomes registry to advance value-based spine care. Spine (Phila Pa 1976).2014;39(22)(suppl 1):S106–S116.
- 11
Gentry S, Badrinath P. Defining health in the era of value-based care: lessons from England of relevance to other health systems. Cureus. 2017;9(3):e1079.
- 12
Mummaneni PV, Bydon M, Knightly J, et al. Predictors of nonroutine discharge among patients undergoing surgery for grade I spondylolisthesis: insights from the Quality Outcomes Database. J Neurosurg Spine. 2020;32(4):523–532.
- 13↑
Chou R, Baisden J, Carragee EJ, et al. Surgery for low back pain: a review of the evidence for an American Pain Society Clinical Practice Guideline. Spine (Phila Pa 1976).2009;34(10):1094–1109.
- 14↑
Coffey RM, Farley D. HCUP-2 Project Overview. US Department of Health and Human Services, Public Health Service,. National Center for Health Services Research and Health Care Technology Assessment;1988.
- 15↑
HCUP-US SASD Overview. HCUP. October 21, 2020. Accessed April 1, 2021. https://www.hcup-us.ahrq.gov/sasdoverview.jsp
- 16↑
STATE_OS -. State indicator of observation stay record. Central Distributor SASD: Description of Data Elements. HCUP. April 11, 2008. Accessed April 1, 2021. https://www.hcup-us.ahrq.gov/db/vars/sasddistnote.jsp?var=state_os
- 17↑
TOTCHG –. Total charges, cleaned. Central Distributor SASD: Description of Data Elements. HCUP. April 11, 2008. Accessed April 1, 2021. https://www.hcup-us.ahrq.gov/db/vars/sasddistnote.jsp?var=totchg
- 18↑
Buis ML. Stata tip 87: Interpretation of interactions in nonlinear models. Stata J. 2010;10(2):305–308.
- 19↑
Long JS, Freese J. Regression Models for Categorical Dependent Variables Using Stata. 2nd ed. Stata Press;2006.
- 20↑
Moriarty A. How many ambulatory surgery centers are in the US? Definitive Healthcare . April 10, 2019. Accessed July 30, 2020. https://blog.definitivehc.com/how-many-ascs-are-in-the-us
- 21
Makanji HS, Bilolikar VK, Goyal DKC, Kurd MF. Ambulatory surgery center payment models: current trends and future directions. J Spine Surg. 2019;5(2)(suppl 2):S191–S194.
- 22
Manchikanti L, Singh V, Hirsch JA. Saga of payment systems of ambulatory surgery centers for interventional techniques: an update. Pain Physician. 2012;15(2):109–130.
- 23
Munnich EL, Parente ST. Procedures take less time at ambulatory surgery centers, keeping costs down and ability to meet demand up. Health Aff (Millwood). 2014;33(5):764–769.
- 24
Korol S. Orthopedic surgery payments lower at ASCs than HOPDs. Becker’s Spine Review. November 21, 2018.Accessed April 1, 2021.https://www.beckersspine.com/orthopedic-spine-practices-improving-profits/item/43629-orthopedic-surgery-payments-lower-at-ascs-than-hopds.html
- 25↑
Malik AT, Xie J, Retchin SM, et al. Primary single-level lumbar microdisectomy/decompression at a free-standing ambulatory surgical center vs a hospital-owned outpatient department-an analysis of 90-day outcomes and costs. Spine J. 2020;20(6):882–887.
- 26↑
Zasa R, Todorovich R. Trends and relevant strategies for ASC contracting. Becker’s Hospital Review. March 18, 2015. Accessed April 1, 2021.https://www.beckershospitalreview.com/finance/trends-and-relevant-strategies-for-asc-contracting.html
