Spinal fusion surgeries are common procedures for treating a variety of spinal conditions, including degenerative spine disease, scoliosis, mechanical back pain, spinal stenosis, isthmic spondylolisthesis, cervical myelopathy, fractures, and tumors.1–3 With the advancement in fusion devices and the increase of aging population, the use of spinal fusion procedures have increased dramatically since 1990s in the United States (U.S.).3–5 During 1997 to 2003, the utilization of cervical, thoracolumbar, and lumbar fusions increased by 89%, 31%, and 134%, respectively.4 A study using the Nationwide Inpatient Sample estimated that between 1998 and 2008, 1,288,496 primary posterior lumbar fusions were performed.6 Among Medicare patients, the rate of complex spinal fusion procedures increased 15-fold between 2002 and 2007 resulting in increased life-threatening complications, higher readmission rate, and greater hospitalization costs.3 Despite the extensive efforts being made to reduce infections among patients undergoing spinal fusion surgeries, postsurgical infections including surgical site infections (SSIs) and blood stream infections (BSIs) still remain constant threats to these patients, especially to those receiving open posterior instrumented spinal fusion procedures.3,7–10 With increasing emphasis on quality of care, reducing readmission rates and related costs has become an important component of health care reform in the U.S. Through establishing the Hospital Readmissions Reduction Program, the Centers for Medicare & Medicaid Services has been required to reduce payments to hospitals with excess readmissions since 2012.11
The risk of SSIs among patients undergoing instrumented spinal fusion surgeries was estimated to be 3.8% (median, 4.2%; range, 0.4–20%) based on 39 cohorts with a total of 28,628 patients.12 SSIs following spinal fusion surgeries can be superficial wound infection or invasive infection occurring in deep wound or organ space.13 Among all the pathogens causing postsurgical infections, Staphylococcus aureus (S. aureus) accounts for nearly half of all cases.8,13,14
SSIs and BSIs post spinal fusion surgeries were reported to be associated with worse clinical outcomes and significant cost increase compared with those without such complications.15–18 However, prior studies had small sample size, were conducted in a single health care facility, or did not focus specifically on S. aureus, the most common pathogen. For example, Kuhns et al18 reported that deep wound infection among 22 patients undergoing dorsal cervical fusion was associated with an average of $12,619 higher cost than those without such infections. To better understand hospital resource utilization and costs related to postsurgical S. aureus infections among patients undergoing elective posterior instrumented spinal fusion surgeries in the U.S., use of data from large hospital discharge database would be beneficial.
This study aimed to assess the cost and hospital resource utilization related to postsurgical S. aureus SSIs and BSIs during 180-day follow-up period post elective, posterior, instrumented spinal fusion surgeries between 2010 and 2015 using data from 129 U.S. hospitals that consistently submitted microbiology data to the Premier Healthcare Database (PHD).
MATERIALS AND METHODS
Study Design, Data Source, and Participants
A retrospective cohort study using de-identified PHD data was conducted. The PHD is a complete census of inpatient and hospital-based outpatient visits from over 700 hospitals across all 50 states and contains 20% of all hospital discharges in the U.S. since 2000. PHD data are extracted from standard hospital discharge files and include patients’ demographics, disease status, and information on date-stamped billed services in patients’ daily service records. Patients can be tracked across visits within facilities with a unique identifier.19,20 About 25% of participating hospitals voluntarily contribute microbiology data to the PHD.
PHD data are HIPAA compliant according to 45 CRF 46.101(b)(4) and 45 CRF 164.506(d)(2)(ii)(B).
Patients who met all of the following inclusion criteria and none of the exclusion criteria were included. Inclusion criteria included (1) aged ≥ 18 years at time of index surgery; (2) had a principal/secondary ICD-9-CM procedure code for posterior instrumented spinal fusion surgeries including 81.01, 81.03, 81.05, 81.07, 81.08, 81.31, 81.33, 81.35, 81.37, or 81.38 during index hospitalization; (3) index hospitalization was elective; (4) was admitted between January 01, 2010, and June 30, 2015, to a qualifying hospital that submitted consistent microbiology data during the 12 months before and 6 months post index surgery. Exclusion criteria included (1) had a major surgery as defined by the National Healthcare Safety Network (NHSN) procedures done between index surgery and the positive culture;21 (2) had a positive S. aureus culture from a normally sterile site such as blood or cerebral spinal fluid (CSF) collected during the 12 months before or 2 days post index surgery; and (3) had the following infections present at admission of the index hospitalization: BSIs, SSI deep and organ/space infections, osteomyelitis, vertebral disc space infection, meningitis, and intra-abdominal infections defined by ICD-9-CM diagnosis codes.
Main exposure variable was infection status during the 180-day follow-up period. Three infection status groups were identified: any S. aureus infection, invasive S. aureus infection, and no S. aureus infection. Any S. aureus infection referred to having nonsurveillance culture positive for S. aureus (including both invasive and superficial infections). Invasive S. aureus infection was defined as having culture-confirmed BSI, deep SSI, or organ/space SSI. Specifically, BSI was defined as having a positive blood culture for S. aureus or having a positive nonblood nonsurveillance S. aureus culture and having one of the following ICD-9-CM codes: 038.11 (Methicillin-susceptible S. aureus septicemia), 038.12 (Methicillin-resistant S. aureus septicemia), 790.7 (bacteremia), and 038.10 (Staphylococcal septicemia, unspecified). Deep SSI was defined as (1) having a wound culture positive for S. aureus together with one of the following reoperation posterior spinal fusion specific ICD-9-CM procedure codes: 03.02, 03.09, 03.4, 77.19, 77.49, 77.69, 78.69, 80.09, 80.39, or one of the reoperation nonspecific ICD-9-CM procedure codes: 81.07, 81.62, 81.63, 81.64, 38.97, 81.35, 81.37, 81.38, 81.39, 83.02, 83.09, 83.14, 83.44, 83.45, 84.51, 86.04, 86.22; or (2) having a deep wound culture positive for S. aureus. Organ/Space SSI was defined as having a CSF culture positive for S. aureus or having a positive culture for S. aureus from a qualified specimen (e.g., body fluid culture, tissue culture) together with an ICD-9-CM diagnosis codes of 730.0x, 730.2x, 324.1, 324.9, 996.66, 996.67, 567.xx, and 320.3 during the same hospitalization. No S. aureus infection group included patients who did not have any culture-confirmed S. aureus infections. ICD code descriptions were listed in Appendix A, http://links.lww.com/BRS/B395.
Main outcomes included total and variable hospitalization costs, hospital length of stay, number and risk of readmission, and discharge status for the last hospital admission during follow-up. Variable cost was captured as directly reported by the hospitals, which typically includes cost that are deemed directly related to patient's clinical care (e.g., salaries for clinical staff, cost of supplies, and/or medications).
Patient and clinical characteristics that were assessed included sex, age, race/ethnicity, insurance type, level of fusion (single vs. multilevel), diabetic status, surgical history, year, type, and location of index surgery. Type of index surgery was categorized as fusion (Cervical: 81.01, 81.03; Thoracolumbar/Lumbar: 81.05, 81.07, 81.08), refusion (Cervical: 81.31, 81.33; Thoracolumbar/Lumbar: 81.35, 81.37, or 81.38), and both fusion and refusion (having codes from both of the above categories). Fusion categories were mutually exclusive. Charlson Comorbidity Index (CCI) was assessed using ICD-9-CM diagnosis codes at index hospitalization using Deyo algorithm22 with Premier modifications.23 Hospital characteristics included hospital size, teaching status, population served (rural vs. urban), and region.
Continuous data were expressed as mean, standard deviation, median, interquartile range, and range. Categorical data were expressed as counts and percentages of patients in each category. Chi-square tests were used for testing statistical differences between infection and no infection groups for categorical variables. T-test or Wilcoxon Rank Sum test was used for testing differences in continuous variables between comparison groups. Cost estimates were adjusted to 2015 U.S. dollars based on Consumer Price Index for urban consumers for hospital and related services.24
Crude means were reported for cost variables by infection status. Generalized linear regression modeling was used to compare differences in cost between infection and no infection groups adjusting for known confounders, including age, race, fusion type, fusion level, CCI, hospital size, teaching status, and hospital region. Each model was bootstrapped using 1000 replicates with replacement to account for skewness in the data. For each group, mean, 2.5% and 97.5% estimates were used to determine the average costs and 95% confidence intervals (95% CIs).
Negative binomial regression modeling was used to compare total length of stay and number of readmissions and Poisson regression modeling with robust error variance analysis was used to compare risk of readmission adjusting for known confounders.
Analyses were performed using SAS/STAT software, Version 9.4 of the SAS system [Copyright (2016) SAS Institute Inc., Cary, NC). Statistical significance level was assessed at 0.05 alpha level.
Patient and Hospital Characteristics
A total of 13,212 patients were included in the final analysis: 294 (2.22%) with any S. aureus infection, including 151 (1.14%) with invasive S. aureus infection and 12,918 with no S. aureus infection (Figure 1). Most patient characteristics were comparable between infection and no infection groups (Table 1). Compared with no S. aureus infection group, a higher percentage of patients in invasive group had multilevel fusion (81.46% vs. 69.25%, P = 0.0012). Any infection group had similar distribution in all but race variables with a higher percentage of white patients than in no S. aureus infection group. There was significant difference in geographical location and hospital size between invasive and no S. aureus infection groups (Table 1).
Baseline Charlson Comorbidities
Only prevalence of diabetes without chronic complications varied significantly between groups (Table 2). A higher percentage of patients in invasive (32.45%, P < 0.001) and any (26.53%, P = 0.0074) infection groups had diabetes without chronic complications during the index hospitalization than no S. aureus infection group (20.39%). Patients in invasive infection group had higher mean CCI score than those in no S. aureus infection group (1.09 vs. 0.83, P = 0.0006).
Unadjusted Analysis Results
All-cause 180-day readmission risk was 96.03% in invasive infection group, 54.76% in any infection group, and 19.18% in no S. aureus infection group (P < 0.001) (Table 3). Total length of hospital stay during index hospitalization and follow-up in invasive infection group was over three times that of no S. aureus infection group. Consistent with readmission risk and hospital length of stay observed, total and variable hospitalization costs were the highest in invasive group, followed by any and no S. aureus infection groups. Compared with no S. aureus infection group, a higher percentage of patients in invasive group were discharged to nursing homes, long-term care facilities, hospice, or other acute care facilities (49% vs. 32%) at their last hospital admission during follow-up (Table 3).
Adjusted Analysis Results
After adjusting for confounders and using bootstrapping method, the mean number of 180-day all-cause readmissions was highest among invasive infection group and lowest among no S. aureus infection group (adjusted mean: 1.65 vs. 0.25, P < 0.001) (Table 4). Mean length of hospital stay during index hospitalization and 180-day follow-up period in invasive group was three times of that in no S. aureus infection group (adjusted mean: 20.98 vs. 6.77 days, P < 0.001). The adjusted mean length of stay in the any infection group was two times of that in no S. aureus infection group (adjusted mean: 13.15 vs. 6.77 days, P < 0.001). The adjusted mean of total hospitalization cost remained highest in invasive infection group (Mean: $88,353, 95% CI: 78,907–100,893), followed by any infection group (Mean: $64,356, 95% CI: 58,106–70,893), and lowest in no S. aureus infection group (Mean: $47,366, 95% CI: 46,840–47,889) (Figure 2). The adjusted total variable hospitalization cost followed similar patterns and was highest among patients with invasive infection (Mean: $50,966, 95% CI: 45,802–57,975), followed by any infection group (Mean: $39,820, 95% CI: 36,199–43,673), and lowest in no S. aureus infection group (Mean: $30,243, 95% CI: 29,918–30,592).
Compared with patients under 50 years, patients ≥75 years were twice as likely to get readmitted. Nonwhite patients and those with multilevel fusion, having both fusion and refusion surgeries during index hospitalization, having higher CCI, from medium size hospitals or hospitals in Midwest were more likely to get readmitted than their peers (Table 5). After controlling confounders, the relative risk of 180-day all-cause readmission in invasive and any infection group was 2.15 times (95% CI: 2.06–2.25) and 1.70 times (95% CI: 1.61–1.80) that of no S. aureus infection group, respectively.
As one of the first studies using large national hospital discharge data to assess hospital resource utilization and cost related to culture-confirmed S. aureus infections post elective posterior instrumented spinal fusion surgeries, our study demonstrated that nearly all patients (96.03%) with invasive S. aureus infection had at least one readmission compared with 19.18% among no S. aureus infection group. After adjusting for confounders, invasive S. aureus infection was associated with a 115% increased risk of all-cause readmission during the 180-day follow-up period compared with no S. aureus infection group. Patients with invasive S. aureus infection usually need aggressive surgical debridement and prolonged antibiotic treatment, which often requires readmission.25 Using data from a single hospital, Schairer et al26 estimated that the reoperation rate for instrumented spinal fusion patients was 89.2%, which is consistent with the high readmission rate reported by our study.
When comparing hospital resource utilization between infection groups, our findings indicated that invasive and any S. aureus infection groups had an average of 14.21 and 6.38 more days of hospital stay during the index hospitalization and follow-up than no S. aureus infection group after adjusting for confounders. Pooled results from the study by Patel et al12 also indicated that patients with any SSI had longer length of stay (range, 7.1–19.3 days) than those without SSI (4.0–9.3 days).
Higher readmission risk and longer total hospital length of stay in invasive and any postsurgical S. aureus infection groups resulted in an average of $40,987 and $16,900 more total hospitalization cost compared with no S. aureus infection group over the 180-day follow-up period, respectively. The infection groups also had higher total variable hospitalization cost than the no infection group, which indicated that patients with postsurgical S. aureus infections consume more clinical resources in the hospital than those without such infections. Although the absolute amount of extra cost estimated by our study differs from estimates of previous studies due to different cost variable definitions and study populations, the general association between postsurgical infection and higher health care cost remains consistent across studies.18,27 Because this is a study using existing hospital discharge data, we were only able to assess the hospital burdens of S. aureus infections post target spinal fusion surgeries. The overall burden to the healthcare system and to the society would be even more substantial.
Compared with studies published before, this study has multiple strengths. First, we used data collected from 129 hospitals nationwide, which is more representative than studies conducted in one single institution. The large sample size provided sufficient power to test differences in key outcomes between comparison groups. Second, all cases were culture-confirmed, which is more accurate than cases identified by ICD diagnosis codes only. Third, we included not only invasive SSI but also BSI and noninvasive infections in the analysis, which provided a broad spectrum of S. aureus infections for analysis while further segmenting the infections by invasiveness (invasive and any S. aureus infection). Finally, on the basis of hospital reporting, variable hospitalization costs could be assessed, which may reflect attributable clinical resource utilization more accurately than total hospitalization cost.28
This study also has limitations. First, although we have high-quality cost and clinical data, we miss some clinical details that may help differentiate superficial from invasive S. aureus infections. This limited our ability to assess the impact of superficial infections separately. Second, because PHD only captures readmissions or outpatient visits to the same hospitals where the index surgeries occurred, the number of S. aureus infections, number of readmissions, and costs may be underestimated. This limitation is expected to be nondifferential across infection groups. Third, we only focused on S. aureus infections among patients with posterior instrumented spinal fusion surgeries in this study. More research is needed to study the impact of infections due to other pathogens and among patients with other types of spinal fusion surgeries.
In conclusion, our findings indicate that among patients undergoing elective, posterior, instrumented spinal fusion surgeries, S. aureus infections, especially invasive infections, are associated with significantly higher hospitalization cost, risk and number of all-cause readmissions, and total length of hospital stay during the 180-day follow-up period postindex surgery compared with patients experiencing no S. aureus infection. Reducing postsurgical S. aureus infection risk among these patients may reduce risk of readmission and economic burden. Although the effectiveness of many preoperative, intraoperative, and postoperative infection control measures on reducing SSIs have been studied, few have been proven effective with strong clinical evidence support based on guidelines from the World Health Organization and the Association for Professionals in Infection Control and Epidemiology.29,30 The review by Agarwal et al on implant contamination and septic methods concluded that more research is needed to explore effective methods to prevent postsurgical infections among patients undergoing elective, posterior, instrumented spinal fusion surgeries.31
The authors would like to acknowledge Drs. Beth Begier and Alejandra Gurtman for their scientific input.
1. Chaichana KL, Bydon M, Santiago-Dieppa DR, et al. Risk of infection following posterior instrumented lumbar fusion for degenerative spine disease in 817 consecutive cases. J Neurosurg Spine
2. Choy W, Lam SK, Smith ZA, Dahdaleh NS. Predictors of 30-Day Hospital Readmission After Posterior Cervical Fusion in 3401 Patients. Spine
3. Deyo RA, Mirza SK, Martin BI, et al. Trends, major medical complications, and charges associated with surgery for lumbar spinal stenosis in older adults. JAMA
4. Cowan JA Jr, Dimick JB, Wainess R, et al. Changes in the utilization of spinal fusion
in the United States. Neurosurgery
2006; 59:15–20. discussion 15–20.
5. Deyo RA, Gray DT, Kreuter W, et al. United States trends in lumbar fusion surgery for degenerative conditions. Spine (Phila Pa 1976)
2005; 30:1441–1445. discussion 1446–1447.
6. Pumberger M, Chiu YL, Ma Y, et al. Perioperative mortality after lumbar spinal fusion
surgery: an analysis of epidemiology and risk factors. Eur Spine J
7. Gerometta A, Rodriguez Olaverri JC, Bitan F. Infections in spinal instrumentation. Int Orthop
8. Mackenzie WG, Matsumoto H, Williams BA, et al. Surgical site infection following spinal instrumentation for scoliosis: a multicenter analysis of rates, risk factors, and pathogens. J Bone Joint Surg Am
2013; 95:800–806. s801–802.
9. Su AW, Habermann EB, Thomsen KM, Milbrandt TA, Nassr A, Larson AN. Risk factors for 30-day unplanned readmission and major perioperative complications after spine fusion surgery in adults: a review of the national surgical quality improvement program database. Spine
10. Wang MC, Chan L, Maiman DJ, et al. Complications and mortality associated with cervical spine surgery for degenerative disease in the United States. Spine (Phila Pa 1976)
12. Patel H, Khoury H, Girgenti D, et al. Burden of surgical site infections associated with select spine operations and involvement of Staphylococcus aureus
. Surg Infect (Larchmt)
13. Chahoud J, Kanafani Z, Kanj SS. Surgical site infections following spine surgery: eliminating the controversies in the diagnosis. Front Med (Lausanne)
14. Takizawa T, Tsutsumimoto T, Yui M, et al. Surgical site infections caused by methicillin-resistant Staphylococcus epidermidis following spinal instrumentation surgery. Spine (Phila Pa 1976)
15. Calderone RR, Garland DE, Capen DA, et al. Cost
of medical care for postoperative spinal infections. Orthop Clin North Am
16. Chen SH, Lee CH, Huang KC, et al. Postoperative wound infection after posterior spinal instrumentation: analysis of long-term treatment outcomes. Eur Spine J
17. Drazin D, Nuno M, Shweikeh F, et al. Outcomes and national trends for the surgical treatment of lumbar spine trauma. Biomed Res Int
18. Kuhns BD, Lubelski D, Alvin MD, et al. Cost
and quality of life outcome analysis of postoperative infections after subaxial dorsal cervical fusions. J Neurosurg Spine
19. Magee G, Zaloga GP, Turpin RS, et al. A retrospective, observational study of patient outcomes for critically ill patients receiving parenteral nutrition. Value Health
20. Pontes-Arruda A, Liu FX, Turpin RS, et al. Bloodstream infections in patients receiving manufactured parenteral nutrition with vs without lipids: is the use of lipids really deleterious? JPEN J Parenter Enteral Nutr
22. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol
23. Rosenthal NA, Cao Z, Chung J, et al. Updated Coding Algorithm for Assessing Charlson Comorbidity Index Using Large Hospital Administrative Data. Presented at the 2017 ISPOR Annual Conference; May 21, 2017; Boston, MA.
24. US Department of Labor, Bureau of Labor Statistics. Inpatient Hospital Services Consumer Price Index. Available at: http://www.bls.gov/cpi/
. Accessed March 4, 2017.
25. Maruo K, Berven SH. Outcome and treatment of postoperative spine surgical site infections: predictors of treatment success and failure. J Orthop Sci
26. Schairer WW, Carrer A, Deviren V, et al. Hospital readmission after spine fusion for adult spinal deformity. Spine (Phila Pa 1976)
27. Parker SL, Shau DN, Mendenhall SK, et al. Factors influencing 2-year health care costs in patients undergoing revision lumbar fusion procedures. J Neurosurg Spine
28. Obeid T, Alshaikh H, Nejim B, et al. Fixed and variable cost
of carotid endarterectomy and stenting in the United States: a comparative study. J Vasc Surg
cost; hospital resource utilization; postsurgical infection; spinal fusion; Staphylococcus aureus
Supplemental Digital Content
Copyright © 2018 The Author(s). Published by Wolters Kluwer Health, Inc.