Drivers of Variability in 90-Day Cost for Elective Anterior Cervical Discectomy and Fusion for Cervical Degenerative Disease

Drivers of Variability in 90-Day Cost for Elective Anterior Cervical Discectomy and Fusion for... Abstract BACKGROUND Value-based episode of care reimbursement models is being investigated to curb unsustainable health care costs. Any variation in the cost of index spine surgery can affect the payment bundling during the 90-d global period. OBJECTIVE To determine the drivers of variability in cost for patients undergoing elective anterior cervical discectomy and fusion (ACDF) for degenerative cervical spine disease. METHODS Four hundred forty-five patients undergoing elective ACDF for cervical spine degenerative diagnoses were included in the study. The direct 90-d cost was derived as sum of cost of surgery, cost associated with postdischarge utilization. Multiple variable linear regression models were built for total 90-d cost. RESULTS The mean 90-d direct cost was $17685 ± $5731. In a multiple variable linear regression model, the length of surgery, number of levels involved, length of hospital stay, preoperative history of anticoagulation medication, health-care resource utilization including number of imaging, any complications and readmission encounter were the significant contributor to the 90-d cost. The model performance as measured by R2 was 0.616. CONCLUSION There was considerable variation in total 90-d cost for elective ACDF surgery. Our model can explain about 62% of these variations in 90-d cost. The episode of care reimbursement models needs to take into account these variations and be inclusive of the factors that drive the variation in cost to develop a sustainable payment model. The generalized applicability should take in to account the differences in patient population, surgeons’ and institution-specific differences. Spine, ACDF, Cost, Bundled payment, Readmission, Postdischarge, DRG ABBREVIATIONS ABBREVIATIONS ACDF anterior cervical discectomy and fusion AP arm pain CAD coronary artery disease CHF chronic heart failure CPT current procedural terminology DRG diagnosis related group ER current procedural terminology EMG emergency room EQ-5D EuroQol-5D MI myocardial infraction, NDI neck disability index NP neck pain PROs patient reported outcome NRS numeric rating scale pain scores MRI magnetic resonance imaging. The current trajectory of healthcare expenditure in the United States is clearly unsustainable. Cervical spine surgeries are one of the commonly performed surgical procedures in the United States. There is an uptrend in the number elective cervical spine surgery performed over last decade with up to 206% increase in the number of cervical spine fusions among Medicare beneficiaries from 1992 to 2005.1-6 Therefore, it is important to investigate the costs associated with most commonly performed cervical spine surgery, anterior cervical discectomy and fusion (ACDF), and to determine the real-world clinical and cost benefit.7-9 To curb escalating healthcare spending, the hospital and physician reimbursement models are transitioning from fee-for-serve to pay-per-performance payment models. These models entail that the individual physician, hospital, and other providers will be held accountable for the quality and costs of care from the time of surgery to prespecified episode of care dur-ation.10-12 The goal is to provide a single comprehensive targeted price for the specific set of health care services delivered to the patient during that episode of care by hospital system or different providers over a defined set period of time. The individual physicians will submit a no-pay claim to Medicare and they will be paid by the hospital system under the predetermined payment amount. In addition, the cost associated with surgery, the inpatient facility cost, physical and occupation therapy costs, and any costs associated with postoperative complications or hospital readmission during the defined period will be covered by the bundled payment for that episode of care. Therefore, it is vital that all the stakeholders collaborate and investigate optimal strategies to propose a target price for a sustainable reimbursement model.11,13-17 Given the wide variation in patients, disease process, operative technique, and costs associated with different spine procedures, the strategy to lump different spine procedures into a single bundle might not be sustainable in a long term. As such, it is vital to investigate a relatively comparable group of patients to define a procedure-based risk-adjusted payment model. The variation in the cost of index surgery and postdischarge resource utilization affects the reimbursement for episode of care. In this analysis, we set forth to analyze the drivers of variability in 90-d cost for elective primary ACDF surgery for degenerative cervical spine pathology. METHODS All the patients undergoing elective spine surgery at a single comprehensive spine center over a period of 4 yr were enrolled in a prospective longitudinal registry. Of those undergoing elective ACDF surgery for degenerative spine pathology were included in this study. The study approval and waiver for informed consent was obtained from the institutional review board. Patients age >18 yr, those presenting with neck pain (NP) and arm pain (AP) after failed 3-mo of multimodal nonoperative care and those who have correlative imaging findings for the diagnosis of stenosis, disc herniation, and cervical instability, and patients in whom the number of vertebral level operated on were ≤4 (1-3 motion segment fused) were included in the study. Patients with any pathological spinal disease including tumor, trauma or infection, those with any extraspinal cause of NP or AP and patients that were unwilling or were unable to participate in the follow-up questionnaires were excluded. Patient reported outcome (PROs) for disability, pain, and quality of life were recorded at baseline and 3-mo after surgery. These outcomes were recorded via phone interviews, which were conducted by independent data coordinators not involved with clinical care. PRO instruments included (1) neck-related disability: neck disability index (NDI), (2) Quality of life—EuroQol-5D (EQ-5D),18 and (3) numeric rating scale pain scores (NRS) for NP and AP.19 Cost Data Total 90-d costs include costs associated with in-patient hospital stay, surgeons’ fee, and postdischarge resource utilization. The in-patient hospital stay costs were derived depending on the procedure being performed, the severity of the individual case, and whether complications occur, which collectively determined the diagnosis related group (DRG). Surgeon fee were derived from current procedural terminology (CPT) codes based on Medicare allowable amounts using the resource-based relative-value scale. The patient-reported postdischarge resource utilization costs were calculated for visits to health-care providers’ visits, diagnostic costs, and medication-associated costs. NP and AP-related outpatient visits to providers including surgeons, chiropractors, other physicians, physical therapists, and acupuncturists were captured. Diagnostic tests including radiographs, computed tomography (CT) scans, magnetic resonance imaging (MRI), and electromyography were recorded. The postoperative devices utilization (braces, canes, and walkers), epidural steroid injections, emergency department visits, back-specific medications (nonsteroidal anti-inflammatory drugs, Cox-2 inhibitors, oral steroids, narcotics, muscle relaxants, and antidepressants), and inpatient or outpatient rehabilitation (physical therapy and occupational therapy) days were assessed. The cost associated with 90-d readmissions and complications were also recorded. Such cost calculations are reported previously.20, 21 Statistical Analysis Descriptive statistics including mean, standard deviation, median, and range for continuous variables, and frequency for categorical variables were calculated. For continuous data, Student's t-test and Mann–Whitney U-test were used. Nominal data were compared via Chi-squared test. A P-value <.05 was considered statistically significant. Multivariable linear regression model for total direct cost and hospital cost was built. Variables: age at the time of surgery, American Society of Anesthesiologists (ASA) grades, BMI, number of comorbidities, comorbidity including diabetes, hypertension, coronary artery disease (CAD), history of arthritis, and preoperative anticoagulation, length of surgery, length of hospital stay, complication, number of days on pain medication (muscle relaxant, narcotic pain medication, nonsteroidal anti-inflammatory drug (NSAIDs)), number of ASA visits, number of times diagnostics required (X-ray, MRI scans, and CT scan), number of postoperative surgeons or physician visits associated with back or leg pain, spine-related readmission within 90-d were included. All the analyses were performed using SPSS version 22 (IBM, Armonk, New York). RESULTS A total of 445 consecutive patients undergoing primary ACDF were analyzed. The mean age of 241 male and 204 female patients was 52 ± 10 yr. Twenty-four percent of patients (n = 106) had diabetes, the mean body mass index (BMI) was 31 ± 7; and 26% (n = 116) were obese with BMI ≥ 35. Mean baseline NDI was 42.1 ± 17.9 percentage points, EQ-5D was 0.65 ± 0.48, NRS-AP was 5.6 ± 3.2 points, and NP was 6.0 ± 2.8 points. Four percent (n = 18) of patients developed 90-d complications after surgery and 8 (1.7%) had readmission for management of UTI (n = 1), instrumentation failure (n = 3), dysphagia (n = 1), deep venous thrombosis (DVT) (n = 1), syncope (n = 1), and wound infection (n = 1). Table 1 summarizes the patient variables. TABLE 1. Characteristics of the Study Cohort Variables  N (%)  Age (>65)  56 (13%)  Gender: Male  241(54%)  Obese (BMI > 35)  116 (26%)  Smoker  120 (27%)  Duration of symptoms > 12 mo  192 (43%)  Radiculopathy  212 (48%)  Myelopathy  233 (52%)  History of Diabetes  106 (24%)  History of Arthritis  263 (59%)  History of preoperative anticoagulation  5 (1.1%)  History of MI  18 (4%)  Number of comorbidities (Mean ± SD)  2.5 ± 1.7  ASA grade: ¾  268 (60%)  Number of levels (Mean ± SD)  1.7 ± 0.8  Length of surgery (minutes; Mean ± SD)  158.09 ± 56.5  Length of hospital stay (Mean ± SD)  1.3 ± 1.1  EBL (Mean ± SD)  115.5 ± 136.0  Complications  18 (4%)  90-d readmission  8 (1.7%)  Variables  N (%)  Age (>65)  56 (13%)  Gender: Male  241(54%)  Obese (BMI > 35)  116 (26%)  Smoker  120 (27%)  Duration of symptoms > 12 mo  192 (43%)  Radiculopathy  212 (48%)  Myelopathy  233 (52%)  History of Diabetes  106 (24%)  History of Arthritis  263 (59%)  History of preoperative anticoagulation  5 (1.1%)  History of MI  18 (4%)  Number of comorbidities (Mean ± SD)  2.5 ± 1.7  ASA grade: ¾  268 (60%)  Number of levels (Mean ± SD)  1.7 ± 0.8  Length of surgery (minutes; Mean ± SD)  158.09 ± 56.5  Length of hospital stay (Mean ± SD)  1.3 ± 1.1  EBL (Mean ± SD)  115.5 ± 136.0  Complications  18 (4%)  90-d readmission  8 (1.7%)  View Large TABLE 1. Characteristics of the Study Cohort Variables  N (%)  Age (>65)  56 (13%)  Gender: Male  241(54%)  Obese (BMI > 35)  116 (26%)  Smoker  120 (27%)  Duration of symptoms > 12 mo  192 (43%)  Radiculopathy  212 (48%)  Myelopathy  233 (52%)  History of Diabetes  106 (24%)  History of Arthritis  263 (59%)  History of preoperative anticoagulation  5 (1.1%)  History of MI  18 (4%)  Number of comorbidities (Mean ± SD)  2.5 ± 1.7  ASA grade: ¾  268 (60%)  Number of levels (Mean ± SD)  1.7 ± 0.8  Length of surgery (minutes; Mean ± SD)  158.09 ± 56.5  Length of hospital stay (Mean ± SD)  1.3 ± 1.1  EBL (Mean ± SD)  115.5 ± 136.0  Complications  18 (4%)  90-d readmission  8 (1.7%)  Variables  N (%)  Age (>65)  56 (13%)  Gender: Male  241(54%)  Obese (BMI > 35)  116 (26%)  Smoker  120 (27%)  Duration of symptoms > 12 mo  192 (43%)  Radiculopathy  212 (48%)  Myelopathy  233 (52%)  History of Diabetes  106 (24%)  History of Arthritis  263 (59%)  History of preoperative anticoagulation  5 (1.1%)  History of MI  18 (4%)  Number of comorbidities (Mean ± SD)  2.5 ± 1.7  ASA grade: ¾  268 (60%)  Number of levels (Mean ± SD)  1.7 ± 0.8  Length of surgery (minutes; Mean ± SD)  158.09 ± 56.5  Length of hospital stay (Mean ± SD)  1.3 ± 1.1  EBL (Mean ± SD)  115.5 ± 136.0  Complications  18 (4%)  90-d readmission  8 (1.7%)  View Large The mean 90-d direct cost was $17 685 ± $5731 (median $16, 394 and range $9494 to $62, 590). The mean hospital costs were $14, 152 ± $5200 (median $12 959 and range $7026 to $55, 656). Table 2 summarizes the cost breakdown for 90-d direct costs. Table 3 summarizes the 90-d postoperative resource utilization. The median cost for primary ACDF was not significantly different among surgeons (Figure). Mean percent contribution of health care resource utilization is 5% (0%-38%) including diagnostic 0.92% (0%-25%), medication cost 2.4% (0%-12%), and health care visits 2.8% (0%-42%). CPT-based surgeon reimbursement contributed 16% (4%-30%) and the DRG-based hospital cost contributed 79% (30%-93%) to the total 90-d cost at 90-d. For the patients readmitted within 90-d following surgery, the percent contribution of readmission episode was as high as 35% (11%-63%). FIGURE. View largeDownload slide Box plot demonstrating variability in total 90-d cost among the participating surgeons. The “o” and “*” represent the outliers above third quartile or below first quartile. FIGURE. View largeDownload slide Box plot demonstrating variability in total 90-d cost among the participating surgeons. The “o” and “*” represent the outliers above third quartile or below first quartile. TABLE 2. Cost Breakdown for 90-d Direct Costs   Mean (range)  Hospital cost  $14 153 ($7026-$55 565)  Surgeons fee  $2664 ($2571-$2711)  Postoperative health-care resource utilization  $1132 ($145-$1691)  Readmit cost  $13 615 ($2550-$23 628)    Mean (range)  Hospital cost  $14 153 ($7026-$55 565)  Surgeons fee  $2664 ($2571-$2711)  Postoperative health-care resource utilization  $1132 ($145-$1691)  Readmit cost  $13 615 ($2550-$23 628)  View Large TABLE 2. Cost Breakdown for 90-d Direct Costs   Mean (range)  Hospital cost  $14 153 ($7026-$55 565)  Surgeons fee  $2664 ($2571-$2711)  Postoperative health-care resource utilization  $1132 ($145-$1691)  Readmit cost  $13 615 ($2550-$23 628)    Mean (range)  Hospital cost  $14 153 ($7026-$55 565)  Surgeons fee  $2664 ($2571-$2711)  Postoperative health-care resource utilization  $1132 ($145-$1691)  Readmit cost  $13 615 ($2550-$23 628)  View Large TABLE 3. 90-d Postoperative Resource Utilization Mean (range)  Frequency  Cost  Health care visits     $359 (0-$1077)   Surgeon visits  2 (1-3)     Physician visits  0 (0-3)     Chiropractor/Acupuncture visit  0 (0-5)     Outpatient PT/OT visits  4 (0-12)     Inpatient PT/OT visits  0 (0-21)    Pain Medication    $134 ($35-$303)   Narcotic pain medication  12 (7-21)     Muscle relaxant  21 (0-60)     NSAIDs  5 (0 -14)    Imaging and ER visits    $639 ($110-$1260)   X-rays  1.7 (0-3)     MRI  1 (0-2)     CT-scan  0 (0)     Spine injection  0 (0)     ER visits  0.7 (0-1)    Mean (range)  Frequency  Cost  Health care visits     $359 (0-$1077)   Surgeon visits  2 (1-3)     Physician visits  0 (0-3)     Chiropractor/Acupuncture visit  0 (0-5)     Outpatient PT/OT visits  4 (0-12)     Inpatient PT/OT visits  0 (0-21)    Pain Medication    $134 ($35-$303)   Narcotic pain medication  12 (7-21)     Muscle relaxant  21 (0-60)     NSAIDs  5 (0 -14)    Imaging and ER visits    $639 ($110-$1260)   X-rays  1.7 (0-3)     MRI  1 (0-2)     CT-scan  0 (0)     Spine injection  0 (0)     ER visits  0.7 (0-1)    View Large TABLE 3. 90-d Postoperative Resource Utilization Mean (range)  Frequency  Cost  Health care visits     $359 (0-$1077)   Surgeon visits  2 (1-3)     Physician visits  0 (0-3)     Chiropractor/Acupuncture visit  0 (0-5)     Outpatient PT/OT visits  4 (0-12)     Inpatient PT/OT visits  0 (0-21)    Pain Medication    $134 ($35-$303)   Narcotic pain medication  12 (7-21)     Muscle relaxant  21 (0-60)     NSAIDs  5 (0 -14)    Imaging and ER visits    $639 ($110-$1260)   X-rays  1.7 (0-3)     MRI  1 (0-2)     CT-scan  0 (0)     Spine injection  0 (0)     ER visits  0.7 (0-1)    Mean (range)  Frequency  Cost  Health care visits     $359 (0-$1077)   Surgeon visits  2 (1-3)     Physician visits  0 (0-3)     Chiropractor/Acupuncture visit  0 (0-5)     Outpatient PT/OT visits  4 (0-12)     Inpatient PT/OT visits  0 (0-21)    Pain Medication    $134 ($35-$303)   Narcotic pain medication  12 (7-21)     Muscle relaxant  21 (0-60)     NSAIDs  5 (0 -14)    Imaging and ER visits    $639 ($110-$1260)   X-rays  1.7 (0-3)     MRI  1 (0-2)     CT-scan  0 (0)     Spine injection  0 (0)     ER visits  0.7 (0-1)    View Large Table 4 summarizes the drivers of 90-d cost. The baseline 90-d cost computed based on multiple variable linear regression model was $6368, which increases $32 with each increase in the length of surgery (minutes), $1457 with each increase in the number of levels involved, $2216 with each additional day spent in hospital postoperative, $6389 in patients with the history of preoperative anticoagulation, $4890 for any 90-d complications, and $9344 for any 90-d readmission, $758 for the number of diagnostic imaging performed within 90-d. The model performance as measured by R2 was 0.616. TABLE 4. Multivariable Linear Regression Model for 90-d Total Cost Summarizing the Drivers of Total 90-d Cost   Beta coefficients  P-value  95.0% Confidence interval for B        Lower bound  Upper bound  (Constant)  6368.49  <0.0001  2931.414  9805.565  Age (at time of surgery)  6.484  0.781  −39.333  52.301  Gender  −565.895  0.174  −1382.947  251.157  BMI  −12.147  0.755  −88.771  64.478  Smoker  480.185  0.317  −461.364  1421.734  History of diabetes  305.278  0.647  −1006.475  1617.031  Arthritis  5.179  0.993  −1189.904  1200.261  Atrial fibrillation  −88.806  0.943  −2535.493  2357.881  Hypertension  −218.873  0.717  −1403.999  966.253  Coronary artery disease  −1052.076  0.162  −2529.941  425.789  Chronic obstructive pulmonary Disease  −1185.964  0.252  −3217.98  846.053  Myocardial infarction  −1575.176  0.202  −4000.926  850.574  Congestive heart failure  −505.572  0.729  −3372.089  2360.945  Number of comorbidity  301.439  0.421  −434.459  1037.337  Preoperative anticoagulation  6389.587  0.001  2616.268  10162.905  Duration of preoperative narcotic use (days)  −0.487  0.062  −0.997  0.024  Duration of symptoms  −390.774  0.342  −1197.932  416.384  Length of surgery in minutes  32.708  <0.0001  23.547  41.869  Length of hospital stay in d  2216.524  <0.0001  1787.8  2645.248  Number of levels  1457.841  <0.0001  784.679  2131.002  Postdischargeresource utilization           Pain meds (NSAID + Muscle relaxant)  −16.764  0.094  −36.404  2.877   Health care visits  −11.511  0.935  −290.124  267.102   Imaging  20.278  0.018  3.458  37.097   Narcotic (pain) medication  7.478  0.282  −6.163  21.119   Inpatient physical therapy  119.79  0.292  −103.494  343.075   Out-patient PT/OT  64.141  0.055  −1.484  129.767  ER visits  −443.49  0.306  −1294.74  407.76  Spine injections  1288.019  0.121  −340.301  2916.34  EMGs  1831.462  0.249  −1291.159  4954.083  90-d complication  3653.398  0.001  1478.81  5827.985  90-d readmission  7375.999  <0.0001  4127.889  10624.108    Beta coefficients  P-value  95.0% Confidence interval for B        Lower bound  Upper bound  (Constant)  6368.49  <0.0001  2931.414  9805.565  Age (at time of surgery)  6.484  0.781  −39.333  52.301  Gender  −565.895  0.174  −1382.947  251.157  BMI  −12.147  0.755  −88.771  64.478  Smoker  480.185  0.317  −461.364  1421.734  History of diabetes  305.278  0.647  −1006.475  1617.031  Arthritis  5.179  0.993  −1189.904  1200.261  Atrial fibrillation  −88.806  0.943  −2535.493  2357.881  Hypertension  −218.873  0.717  −1403.999  966.253  Coronary artery disease  −1052.076  0.162  −2529.941  425.789  Chronic obstructive pulmonary Disease  −1185.964  0.252  −3217.98  846.053  Myocardial infarction  −1575.176  0.202  −4000.926  850.574  Congestive heart failure  −505.572  0.729  −3372.089  2360.945  Number of comorbidity  301.439  0.421  −434.459  1037.337  Preoperative anticoagulation  6389.587  0.001  2616.268  10162.905  Duration of preoperative narcotic use (days)  −0.487  0.062  −0.997  0.024  Duration of symptoms  −390.774  0.342  −1197.932  416.384  Length of surgery in minutes  32.708  <0.0001  23.547  41.869  Length of hospital stay in d  2216.524  <0.0001  1787.8  2645.248  Number of levels  1457.841  <0.0001  784.679  2131.002  Postdischargeresource utilization           Pain meds (NSAID + Muscle relaxant)  −16.764  0.094  −36.404  2.877   Health care visits  −11.511  0.935  −290.124  267.102   Imaging  20.278  0.018  3.458  37.097   Narcotic (pain) medication  7.478  0.282  −6.163  21.119   Inpatient physical therapy  119.79  0.292  −103.494  343.075   Out-patient PT/OT  64.141  0.055  −1.484  129.767  ER visits  −443.49  0.306  −1294.74  407.76  Spine injections  1288.019  0.121  −340.301  2916.34  EMGs  1831.462  0.249  −1291.159  4954.083  90-d complication  3653.398  0.001  1478.81  5827.985  90-d readmission  7375.999  <0.0001  4127.889  10624.108  View Large TABLE 4. Multivariable Linear Regression Model for 90-d Total Cost Summarizing the Drivers of Total 90-d Cost   Beta coefficients  P-value  95.0% Confidence interval for B        Lower bound  Upper bound  (Constant)  6368.49  <0.0001  2931.414  9805.565  Age (at time of surgery)  6.484  0.781  −39.333  52.301  Gender  −565.895  0.174  −1382.947  251.157  BMI  −12.147  0.755  −88.771  64.478  Smoker  480.185  0.317  −461.364  1421.734  History of diabetes  305.278  0.647  −1006.475  1617.031  Arthritis  5.179  0.993  −1189.904  1200.261  Atrial fibrillation  −88.806  0.943  −2535.493  2357.881  Hypertension  −218.873  0.717  −1403.999  966.253  Coronary artery disease  −1052.076  0.162  −2529.941  425.789  Chronic obstructive pulmonary Disease  −1185.964  0.252  −3217.98  846.053  Myocardial infarction  −1575.176  0.202  −4000.926  850.574  Congestive heart failure  −505.572  0.729  −3372.089  2360.945  Number of comorbidity  301.439  0.421  −434.459  1037.337  Preoperative anticoagulation  6389.587  0.001  2616.268  10162.905  Duration of preoperative narcotic use (days)  −0.487  0.062  −0.997  0.024  Duration of symptoms  −390.774  0.342  −1197.932  416.384  Length of surgery in minutes  32.708  <0.0001  23.547  41.869  Length of hospital stay in d  2216.524  <0.0001  1787.8  2645.248  Number of levels  1457.841  <0.0001  784.679  2131.002  Postdischargeresource utilization           Pain meds (NSAID + Muscle relaxant)  −16.764  0.094  −36.404  2.877   Health care visits  −11.511  0.935  −290.124  267.102   Imaging  20.278  0.018  3.458  37.097   Narcotic (pain) medication  7.478  0.282  −6.163  21.119   Inpatient physical therapy  119.79  0.292  −103.494  343.075   Out-patient PT/OT  64.141  0.055  −1.484  129.767  ER visits  −443.49  0.306  −1294.74  407.76  Spine injections  1288.019  0.121  −340.301  2916.34  EMGs  1831.462  0.249  −1291.159  4954.083  90-d complication  3653.398  0.001  1478.81  5827.985  90-d readmission  7375.999  <0.0001  4127.889  10624.108    Beta coefficients  P-value  95.0% Confidence interval for B        Lower bound  Upper bound  (Constant)  6368.49  <0.0001  2931.414  9805.565  Age (at time of surgery)  6.484  0.781  −39.333  52.301  Gender  −565.895  0.174  −1382.947  251.157  BMI  −12.147  0.755  −88.771  64.478  Smoker  480.185  0.317  −461.364  1421.734  History of diabetes  305.278  0.647  −1006.475  1617.031  Arthritis  5.179  0.993  −1189.904  1200.261  Atrial fibrillation  −88.806  0.943  −2535.493  2357.881  Hypertension  −218.873  0.717  −1403.999  966.253  Coronary artery disease  −1052.076  0.162  −2529.941  425.789  Chronic obstructive pulmonary Disease  −1185.964  0.252  −3217.98  846.053  Myocardial infarction  −1575.176  0.202  −4000.926  850.574  Congestive heart failure  −505.572  0.729  −3372.089  2360.945  Number of comorbidity  301.439  0.421  −434.459  1037.337  Preoperative anticoagulation  6389.587  0.001  2616.268  10162.905  Duration of preoperative narcotic use (days)  −0.487  0.062  −0.997  0.024  Duration of symptoms  −390.774  0.342  −1197.932  416.384  Length of surgery in minutes  32.708  <0.0001  23.547  41.869  Length of hospital stay in d  2216.524  <0.0001  1787.8  2645.248  Number of levels  1457.841  <0.0001  784.679  2131.002  Postdischargeresource utilization           Pain meds (NSAID + Muscle relaxant)  −16.764  0.094  −36.404  2.877   Health care visits  −11.511  0.935  −290.124  267.102   Imaging  20.278  0.018  3.458  37.097   Narcotic (pain) medication  7.478  0.282  −6.163  21.119   Inpatient physical therapy  119.79  0.292  −103.494  343.075   Out-patient PT/OT  64.141  0.055  −1.484  129.767  ER visits  −443.49  0.306  −1294.74  407.76  Spine injections  1288.019  0.121  −340.301  2916.34  EMGs  1831.462  0.249  −1291.159  4954.083  90-d complication  3653.398  0.001  1478.81  5827.985  90-d readmission  7375.999  <0.0001  4127.889  10624.108  View Large Total direct cost = $6348 + $6.5 (age) + (−565.9) (Male) + (−12.4) (BMI) + (480) (smoker) + (305.3) (history of diabetes) + (5.2) (arthritis) + (−88.8) (atrial fibrillation) + (−218.9) (hypertension) + (−1052.1) (coronary artery disease) + (−1185.9) (chronic obstructive pulmonary disease) + (−1575) (myocardial infraction (MI)) + (−505) (chronic heart failure (CHF)) + (301) (number of comorbidity) + (6389) (preoperative anticoagulation) + (−0.49) (duration of preoperative opioid medication) + (−390.8) (duration of symptoms) + (32.7) (length of surgery) + (2216.5) (length of hospital stay) + (1457.8) (number of levels operated on) + (−16.8) (postdischarge nonopioid use) + (−11.5) (number of health care visits) + (20.3) (postdischarge imaging) + (7.4) (postdischarge opioid pain medication use) + (119.8) (number of days spent in inpatient rehabilitation) + (64.1) (number of days spent in outpatient PT/OT) + (−443.8) (number of postdischarge emergency room (ER) visits) + (1288.1) (number of postdischarge spine injections) + (1831.5) (Electromyography (EMG) postdischarge) + (3653.4) (number of 90-d complications) + (7375.9) (number of 90-d readmissions). DISCUSSION In this analysis, we demonstrate that the hospital costs had highest percent contributions to the total 90-d cost followed by readmission, and surgeons’ fee and health care resource utilization. The significant drivers of 90-d cost for primary ACDF were length of surgery, number of levels involved, length of hospital stay, preoperative history of anticoagulation medication, health-care resource utilization including number of imaging, any complications and readmission within 90-d after surgery. Our model accounted for 62% of variation in the total 90-d cost following surgery. The length of surgery was the one of significant contributors of the 90-d cost. The cost of first 30 min of operating room time was $1054 and increases with each 30 min increase in operating room time. The median length of surgery was 157 min; the predicted added cost of performing ACDF surgery will be $5270 for 50% of the patients. In our analysis, the length of surgery was calculated as time from skin incision to closure, it does not account for operating room time associated with patient positioning and time needed for induction of anesthesia. The length of hospital stay was a significant contributor to the total 90-d cost, which is not surprising. Previously reported factors associated with an increased length of hospital stay post-ACDF include older age, female gender, history of diabetes, anemia, and in-patient complications, and should be accounted for when computing the cost for ACDF.22-24 Furthermore, as demonstrated previously outpatient surgery for 1-to-2 level ACDF might save additional costs associated with hospitalization.25 Increasing number of levels operated on was another significant driver of cost. The additional cost associated with an increasing number of levels exposed can be attributed to increased length of surgery, increased use of graft and implant, and increased length of hospital stay. Patients on preoperative anticoagulation medication significantly contributed to total 90-d cost. Five patients in our study cohort were on anticoagulation medication preoperatively. For each of these patients, the total 90-d cost will increase by about $5796. In addition, the number of comorbidities and individual comorbidities including diabetes, history of arthritis, hypertension, MI, and obesity contributed significantly to the total 90-d cost; however, these factors were not the significant drivers of cost in multivariable model. The patients on preoperative anticoagulants have higher likelihood of having DVT and pulmonary embolism following surgery,26 which may result in higher 90-d complications leading to higher associated cost. Analogous to our findings, prior studies have reported that the postdischarge care accounts for a relatively small portion of total 90-d costs (range, 4%-8%).12,13,17,27 Of all the variables contributing to the postoperative health care resource utilization, the cost associated with imaging was the only significant driver of total 90-d cost. The medication use including opioid pain medication drives the cost following ACDF; however, the contribution was not statistically significant. Readmission episode and ER visits accounted for a significant increase in the 90-d cost for ACDF. The mean percent contribution of readmission cost was 35% to the 90-d cost. The readmission rate within 90-d following surgery was 2%, which suggests that for every 100 patients 2 patients can add up to 37% to the total 90-d cost. The inclusion of readmission in the targeted payment is essential to embrace the care of complex and sickest patient without being penalized for the event for the readmission encounter within 90-d postdischarge. Furthermore, as reported previously a better understanding of factors associated with readmission and ER visits following spine surgery are needed to facilitate efforts aimed at quality improvement and cost containment associated with these significant drivers of cost.28-31 Understanding and accurately predicting which patients may require readmission within a global period after surgery may help facilitate the creation and implementation of risk-adjusted bundled payment systems that would more fairly compensate surgeons and hospitals for advanced services. Prior studies have demonstrated significant variability in the episode of care cost within each individual spine related DRG.12-14, 27 However, none of the studies have determined the drivers of variability in 90-d cost for a spine care episode. Given the significant heterogeneity in patient-specific and disease-specific factors among the patients undergoing spine surgery, the variation in cost is inevitable. Therefore, implementation of the bundled payment models in real-world practice is complex and challenging.32-38 The drivers of variability in cost, as defined in this analysis, will guide a fuller accounting of costs incurred during the 90-d postdischarge time frame. Furthermore, it will form the basis of debate on defining the ranges within the target bundled price that would account for the highs and lows in the total 90-d cost. The optimization of the pre- and perioperative drivers will result in the containment of cost, and reduction in complications, which will ultimately result in increasing the overall value of spine care. Limitations There are several limitations to this study. This is a single institution study and even though the costs are reported from payer's perspective and adjusted for Medicare allowable national payments amount, costs may differ based on the geographical location. The healthcare resource utilization costs are patient self-reported, which might have some recall bias leading to higher or lower frequency of utilization. The indirect resource utilization from societal perspective including patient workday losses, family work day loss, caregiver cost were not included in the total cost calculation, as the goal of this study was to define variability in direct cost associated with healthcare resource utilization. Utilizing comprehensive list of variables captured in our prospective registry, we present a risk-adjusted analysis for drivers of cost for primary ACDF. The differences in patient population, surgeons’ perioperative practice patterns and institution-specific difference such as operating room staffing, reimbursement models must be considered while applying the results of this study to other institution. The costs are adjusted for the Medicare allowable national payment amounts. Utilizing the actual payer status for each patient would make the costs greater given the higher private pay percentage. The rate of complication and readmissions is comparable to the published literature;25,39-41 however given the relatively smaller number of patients from a single-center, it is possible that less common readmissions are likely not accounted for in the cost analyses if they weren’t encountered. Despite of the model being robust, the drivers of cost presented here explain about 60% variability in the cost for ACDF. The granular data on different types of instrumentation or type of biologics used are not captured as part of registry, which can certainly differ from based on surgeons’ and institutional practice patterns. The included confounding variables are not exhaustive and adding more pre- and perioperative variables might account for other 40% variation in the cost following index surgery. The generalized applicability of these results needs to take these limitations into account. Further analysis using multicenter prospective registries such as the Quality Outcomes Database is imperative.42 CONCLUSION There was considerable variation in total 90-d cost for elective ACDF surgery. Our model can explain about 62% of these variations in 90-d cost. The episode of care reimbursement models needs to take into account these variations and be inclusive of the factors that drive the variation in cost to develop a sustainable payment model. The generalized applicability should take in to account the differences in patient population, surgeons’ and institution-specific differences. Disclosure The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article. Notes The abstract was presented on December 4, 2015 at the Cervical Spine Research Society 2015 Annual Meeting, San Diego, California. REFERENCES 1. Angevine PD, Arons RR, McCormick PC. National and regional rates and variation of cervical discectomy with and without anterior fusion, 1990–1999. Spine . 2003; 28( 9): 931- 939; discussion 940. Google Scholar PubMed  2. Deyo RA, Gray DT, Kreuter W, Mirza S, Martin BI. United States trends in lumbar fusion surgery for degenerative conditions. Spine . 2005; 30( 12): 1441- 1445; discussion 1446-1447. Google Scholar CrossRef Search ADS PubMed  3. Marquez-Lara A, Nandyala SV, Fineberg SJ, Singh K. Current trends in demographics, practice, and in-hospital outcomes in cervical spine surgery. Spine . 2014; 39( 6): 476- 481. Google Scholar CrossRef Search ADS PubMed  4. Oglesby M, Fineberg SJ, Patel AA, Pelton MA, Singh K. Epidemiological trends in cervical spine surgery for degenerative diseases between 2002 and 2009. Spine . 2013; 38( 14): 1226- 1232. Google Scholar CrossRef Search ADS PubMed  5. Patil PG, Turner DA, Pietrobon R. National trends in surgical procedures for degenerative cervical spine disease: 1990–2000. Neurosurgery . 2005; 57( 4): 753- 758; discussion 753-758. Google Scholar CrossRef Search ADS PubMed  6. Wang MC, Kreuter W, Wolfla CE, Maiman DJ, Deyo RA. Trends and variations in cervical spine surgery in the United States. Spine . 2009; 34( 9): 955- 961; discussion 962-953. Google Scholar CrossRef Search ADS PubMed  7. Angevine PD, Zivin JG, McCormick PC. Cost-effectiveness of single-level anterior cervical discectomy and fusion for cervical spondylosis. Spine . 2005; 30( 17): 1989- 1997. Google Scholar CrossRef Search ADS PubMed  8. McAnany SJ, Overley S, Baird EO et al.   The 5-year cost-effectiveness of anterior cervical discectomy and fusion and cervical disc replacement. Spine . 2014; 39( 23): 1924- 1933. Google Scholar CrossRef Search ADS PubMed  9. Qureshi SA, McAnany S, Goz V, Koehler SM, Hecht AC. Cost-effectiveness analysis: comparing single-level cervical disc replacement and single-level anterior cervical discectomy and fusion. J Neurosurg Spine  2013; 19( 5): 546- 554. Google Scholar CrossRef Search ADS PubMed  10. Lubelski D, Senol N, Silverstein MP et al.   Quality of life outcomes after revision lumbar discectomy. J Neurosurg Spine  2015; 22( 2): 173- 178. Google Scholar CrossRef Search ADS PubMed  11. Bozic KJ, Ward L, Vail TP, Maze M. Bundled payments in total joint arthroplasty: targeting opportunities for quality improvement and cost reduction. Clin Orthop Relat Res . 2014; 472( 1): 188- 193. Google Scholar CrossRef Search ADS PubMed  12. Ugiliweneza B, Kong M, Nosova K et al.   Spinal surgery. Spine . 2014; 39( 15): 1235- 1242. Google Scholar CrossRef Search ADS PubMed  13. Birkmeyer JD, Gust C, Baser O, Dimick JB, Sutherland JM, Skinner JS. Medicare payments for common inpatient procedures: implications for episode-based payment bundling. Health Serv Res . 2010; 45( 6p1): 1783- 1795. Google Scholar CrossRef Search ADS PubMed  14. Miller DC, Gust C, Dimick JB, Birkmeyer N, Skinner J, Birkmeyer JD. Large variations in medicare payments for surgery highlight savings potential from bundled payment programs. Health Aff . 2011; 30( 11): 2107- 2115. Google Scholar CrossRef Search ADS   15. Cromwell J, Dayhoff DA, Thoumaian AH. Cost savings and physician responses to global bundled payments for medicare heart bypass surgery. Health Care Financ Rev  1997; 19( 1): 41- 57. Google Scholar PubMed  16. Edmonds C, Hallman GL. Cardio Vascular Care Providers. A pioneer in bundled services, shared risk, and single payment. Tex Heart Inst J . 1995; 22( 1): 72- 76. Google Scholar PubMed  17. Grenda TR, Pradarelli JC, Thumma JR, Dimick JB. Variation in hospital episode costs with bariatric surgery. JAMA Surg . 2015; 50( 12): 1109- 1115. Google Scholar CrossRef Search ADS   18. EuroQol. EuroQol—a new facility for the measurement of health-related quality of life. Health Policy . 1990; 16( 3): 199- 208. CrossRef Search ADS PubMed  19. Langley GB, Sheppeard H. The visual analogue scale: its use in pain measurement. Rheumatol Int . 1985; 5( 4): 145- 148. Google Scholar CrossRef Search ADS PubMed  20. Devin CJ, Chotai S, Parker SL, Tetreault L, Fehlings MG, McGirt MJ. A cost-utility analysis of lumbar decompression with and without fusion for degenerative spine disease in the elderly. Neurosurgery . 2015; 77( suppl 4): S116- S124. Google Scholar CrossRef Search ADS PubMed  21. Parker SL, Fulchiero EC, Davis BJ et al.   Cost-effectiveness of multilevel hemilaminectomy for lumbar stenosis–associated radiculopathy. Spine J . 2011; 11( 8): 705- 711. Google Scholar CrossRef Search ADS PubMed  22. Arnold PM, Rice LR, Anderson KK, McMahon JK, Connelly LM, Norvell DC. Factors affecting hospital length of stay following anterior cervical discectomy and fusion. Evid-Based Spine Care J . 2011; 2( 03): 11- 18. Google Scholar CrossRef Search ADS PubMed  23. Basques BA, Bohl DD, Golinvaux NS, Gruskay JA, Grauer JN. Preoperative factors affecting length of stay after elective anterior cervical discectomy and fusion with and without corpectomy. Spine . 2014; 39( 12): 939- 946. Google Scholar CrossRef Search ADS PubMed  24. Gruskay JA, Fu M, Basques B et al.   Factors affecting length of stay and complications following elective anterior cervical discectomy and fusion. Clin Spine Surg . 2017; 29( 1): E34- 42. Google Scholar CrossRef Search ADS   25. McGirt MJ, Godil SS, Asher AL, Parker SL, Devin CJ. Quality analysis of anterior cervical discectomy and fusion in the outpatient versus inpatient setting: analysis of 7288 patients from the NSQIP database. Neurosurg Focus . 2015; 39( 6): E9. Google Scholar CrossRef Search ADS PubMed  26. Kearon C, Hirsh J. Management of anticoagulation before and after elective surgery. N Engl J Med . 1997; 336( 21): 1506- 1511. Google Scholar CrossRef Search ADS PubMed  27. Schoenfeld AJ, Harris MB, Liu H, Birkmeyer JD. Variations in medicare payments for episodes of spine surgery. Spine J . 2014; 14( 12): 2793- 2798. Google Scholar CrossRef Search ADS PubMed  28. Akamnonu C, Cheriyan T, Goldstein JA, Lafage V, Errico TJ, Bendo JA. Unplanned hospital readmission after surgical treatment of common lumbar pathologies. Spine . 2015; 40( 6): 423- 428. Google Scholar CrossRef Search ADS PubMed  29. Bernatz JT, Anderson PA. Thirty-day readmission rates in spine surgery: systematic review and meta-analysis. Neurosurg Focus . 2015; 39( 4): E7. Google Scholar CrossRef Search ADS PubMed  30. Bosco JA 3rd, Karkenny AJ, Hutzler LH, Slover JD, Iorio R. Cost burden of 30-day readmissions following medicare total hip and knee arthroplasty. J Arthroplasty . 2014; 29( 5): 903- 905. Google Scholar CrossRef Search ADS PubMed  31. Taylor BE, Youngerman BE, Goldstein H et al.   Causes and timing of unplanned early readmission after neurosurgery. Neurosurgery . 2016; 79( 3): 356- 369. Google Scholar CrossRef Search ADS PubMed  32. Cutler DM, Ghosh K. The potential for cost savings through bundled episode payments. N Engl J Med . 2012; 366( 12): 1075- 1077. Google Scholar CrossRef Search ADS PubMed  33. Delisle DR. Big things come in bundled packages. Am J Med Qual . 2013; 28( 4): 339- 344. Google Scholar CrossRef Search ADS PubMed  34. Iorio R. Strategies and tactics for successful implementation of bundled payments: bundled payment for care improvement at a large, urban, academic medical center. J Arthroplasty . 2015; 30( 3): 349- 350. Google Scholar CrossRef Search ADS PubMed  35. Kazberouk A, McGuire K, Landon BE. A survey of innovative reimbursement models in spine care. Spine (Phila Pa 1976) . 2016; 41( 4): 344- 352. Google Scholar CrossRef Search ADS PubMed  36. Mechanic RE. Opportunities and challenges for episode-based payment. N Engl J Med . 2011; 365( 9): 777- 779. Google Scholar CrossRef Search ADS PubMed  37. Rossi VJ, Ahn J, Bohl DD, Tabaraee E, Singh K. Economic factors in the future delivery of spinal healthcare. World J Orthop . 2015; 6( 5): 409- 412. Google Scholar CrossRef Search ADS PubMed  38. Sood N, Huckfeldt PJ, Escarce JJ, Grabowski DC, Newhouse JP. Medicare's bundled payment pilot for acute and postacute care: analysis and recommendations on where to begin. Health Aff . 2011; 30( 9): 1708- 1717. Google Scholar CrossRef Search ADS   39. Bhashyam N, De la Garza Ramos R, Nakhla J et al.   Thirty-day readmission and reoperation rates after single-level anterior cervical discectomy and fusion versus those after cervical disc replacement. Neurosurg Focus . 2017; 42( 2): E6. Google Scholar CrossRef Search ADS PubMed  40. Veeravagu A, Cole T, Jiang B, Ratliff JK. Revision rates and complication incidence in single- and multilevel anterior cervical discectomy and fusion procedures: an administrative database study. Spine J . 2014; 14( 7): 1125- 1131. Google Scholar CrossRef Search ADS PubMed  41. Adamson T, Godil SS, Mehrlich M, Mendenhall S, Asher AL, McGirt MJ. Anterior cervical discectomy and fusion in the outpatient ambulatory surgery setting compared with the inpatient hospital setting: analysis of 1000 consecutive cases. J Neurosurg Spine  2016; 24( 6): 878- 884. Google Scholar CrossRef Search ADS PubMed  42. Asher AL, Speroff T, Dittus RS et al.   The National Neurosurgery Quality and Outcomes Database (N2QOD). Spine . 2014; 39( 22 Suppl 1): S106- S116. Google Scholar CrossRef Search ADS PubMed  COMMENTS Based on a desperate need for national healthcare cost containment and the fact that by 2050 healthcare cost will consume over 20% of the GDP, the investigators of this submission have undertaken a prospective study on 445 patients with ACDF in order to answer 1 question: What drives the cost of ACDF over a 90-day period? From among over 30 appropriate variables, regression analysis indicated that the major drivers of cost in 1 episode of care were hospital length of stay and readmission (mainly due to complications). Surgery related factors were length and levels of surgery and follow-up drivers of cost included imaging studies. The findings are simple to understand but how to make them achievable seems tough. With the baby boomers getting older, cervical spondylotic radiculomyelopathies due to degenerative disc-osteophyte complex has become abundant. Such patients have comorbidities and are usually on anticoagulants, making their hospital course rocky and problematic. It is clear from this submission and many similar publications that attempts are needed to shorten hospital length of stay, length of surgery, and factors predisposing to complications such as venous thromboembolism, infections, and unnecessary postoperative imaging studies to name a few. Further studies should shed light on additional drivers of cost and better means to contain them. Bizhan Aarabi Baltimore, Maryland The authors argue that a progressive movement toward bundled payments for given procedures is well underway, making understanding of the particular drivers of variability of costs important. They investigated the factors that resulted in variation of costs for patients undergoing elective ACDF surgery using a prospective registry. As would be expected, factors that increased total costs included length of stay, treatment of additional spinal levels, and the need for hospital readmission. Surgeon contribution to cost was modest, as the major factors in cost variation were hospital based. An interesting finding was the magnitude of impact on overall costs for patients on preoperative anticoagulation therapy in whom there is a higher incidence of postoperative deep vein thrombosis and pulmonary embolus. Overall, the cost drivers of the model that emerged explained 62% of costs. The authors acknowledge that the impact of the study is diminished by the lack of granular data on the specific expenses such as type of biologics or instrumentation employed for a given patient. Nevertheless, the investigation is a meaningful step toward understanding of costs in spinal surgery. John Kenneth Houten Bronx, New York The authors have offered a retrospective analysis of the major factors driving cost in elective (nontrauma, nontumor) 1–3 level ACDFs from a single center. Four hundred forty-five patients were analyzed and a multivariate analysis was provided which explained 60% of the cost variability. Not surprisingly, factors such as surgical time, days in hospital, levels operated, complications, and readmissions were the primary contributors to cost. It was also discovered that preoperative anticoagulation conveyed a significant increase to the total cost. While the manuscript is limited by being a single institutional study and by 40% of costs not ascertained, it nevertheless sets the foundation for understanding procedural costs as the healthcare system moves toward more bundled payments. These analyses are vital for setting fair reimbursements. There will undoubtedly need to be more granularity within future studies (types of complications and costs, biologics/instrumentation costs), but this manuscript sets the stage for more in-depth investigations. Nathan E. Simmons Lebanon, New Hampshire Copyright © 2018 by the Congress of Neurological Surgeons This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Neurosurgery Oxford University Press

Drivers of Variability in 90-Day Cost for Elective Anterior Cervical Discectomy and Fusion for Cervical Degenerative Disease

Loading next page...
 
/lp/ou_press/drivers-of-variability-in-90-day-cost-for-elective-anterior-cervical-IOlmqlic4G
Publisher
Oxford University Press
Copyright
Copyright © 2018 by the Congress of Neurological Surgeons
ISSN
0148-396X
eISSN
1524-4040
D.O.I.
10.1093/neuros/nyy140
Publisher site
See Article on Publisher Site

Abstract

Abstract BACKGROUND Value-based episode of care reimbursement models is being investigated to curb unsustainable health care costs. Any variation in the cost of index spine surgery can affect the payment bundling during the 90-d global period. OBJECTIVE To determine the drivers of variability in cost for patients undergoing elective anterior cervical discectomy and fusion (ACDF) for degenerative cervical spine disease. METHODS Four hundred forty-five patients undergoing elective ACDF for cervical spine degenerative diagnoses were included in the study. The direct 90-d cost was derived as sum of cost of surgery, cost associated with postdischarge utilization. Multiple variable linear regression models were built for total 90-d cost. RESULTS The mean 90-d direct cost was $17685 ± $5731. In a multiple variable linear regression model, the length of surgery, number of levels involved, length of hospital stay, preoperative history of anticoagulation medication, health-care resource utilization including number of imaging, any complications and readmission encounter were the significant contributor to the 90-d cost. The model performance as measured by R2 was 0.616. CONCLUSION There was considerable variation in total 90-d cost for elective ACDF surgery. Our model can explain about 62% of these variations in 90-d cost. The episode of care reimbursement models needs to take into account these variations and be inclusive of the factors that drive the variation in cost to develop a sustainable payment model. The generalized applicability should take in to account the differences in patient population, surgeons’ and institution-specific differences. Spine, ACDF, Cost, Bundled payment, Readmission, Postdischarge, DRG ABBREVIATIONS ABBREVIATIONS ACDF anterior cervical discectomy and fusion AP arm pain CAD coronary artery disease CHF chronic heart failure CPT current procedural terminology DRG diagnosis related group ER current procedural terminology EMG emergency room EQ-5D EuroQol-5D MI myocardial infraction, NDI neck disability index NP neck pain PROs patient reported outcome NRS numeric rating scale pain scores MRI magnetic resonance imaging. The current trajectory of healthcare expenditure in the United States is clearly unsustainable. Cervical spine surgeries are one of the commonly performed surgical procedures in the United States. There is an uptrend in the number elective cervical spine surgery performed over last decade with up to 206% increase in the number of cervical spine fusions among Medicare beneficiaries from 1992 to 2005.1-6 Therefore, it is important to investigate the costs associated with most commonly performed cervical spine surgery, anterior cervical discectomy and fusion (ACDF), and to determine the real-world clinical and cost benefit.7-9 To curb escalating healthcare spending, the hospital and physician reimbursement models are transitioning from fee-for-serve to pay-per-performance payment models. These models entail that the individual physician, hospital, and other providers will be held accountable for the quality and costs of care from the time of surgery to prespecified episode of care dur-ation.10-12 The goal is to provide a single comprehensive targeted price for the specific set of health care services delivered to the patient during that episode of care by hospital system or different providers over a defined set period of time. The individual physicians will submit a no-pay claim to Medicare and they will be paid by the hospital system under the predetermined payment amount. In addition, the cost associated with surgery, the inpatient facility cost, physical and occupation therapy costs, and any costs associated with postoperative complications or hospital readmission during the defined period will be covered by the bundled payment for that episode of care. Therefore, it is vital that all the stakeholders collaborate and investigate optimal strategies to propose a target price for a sustainable reimbursement model.11,13-17 Given the wide variation in patients, disease process, operative technique, and costs associated with different spine procedures, the strategy to lump different spine procedures into a single bundle might not be sustainable in a long term. As such, it is vital to investigate a relatively comparable group of patients to define a procedure-based risk-adjusted payment model. The variation in the cost of index surgery and postdischarge resource utilization affects the reimbursement for episode of care. In this analysis, we set forth to analyze the drivers of variability in 90-d cost for elective primary ACDF surgery for degenerative cervical spine pathology. METHODS All the patients undergoing elective spine surgery at a single comprehensive spine center over a period of 4 yr were enrolled in a prospective longitudinal registry. Of those undergoing elective ACDF surgery for degenerative spine pathology were included in this study. The study approval and waiver for informed consent was obtained from the institutional review board. Patients age >18 yr, those presenting with neck pain (NP) and arm pain (AP) after failed 3-mo of multimodal nonoperative care and those who have correlative imaging findings for the diagnosis of stenosis, disc herniation, and cervical instability, and patients in whom the number of vertebral level operated on were ≤4 (1-3 motion segment fused) were included in the study. Patients with any pathological spinal disease including tumor, trauma or infection, those with any extraspinal cause of NP or AP and patients that were unwilling or were unable to participate in the follow-up questionnaires were excluded. Patient reported outcome (PROs) for disability, pain, and quality of life were recorded at baseline and 3-mo after surgery. These outcomes were recorded via phone interviews, which were conducted by independent data coordinators not involved with clinical care. PRO instruments included (1) neck-related disability: neck disability index (NDI), (2) Quality of life—EuroQol-5D (EQ-5D),18 and (3) numeric rating scale pain scores (NRS) for NP and AP.19 Cost Data Total 90-d costs include costs associated with in-patient hospital stay, surgeons’ fee, and postdischarge resource utilization. The in-patient hospital stay costs were derived depending on the procedure being performed, the severity of the individual case, and whether complications occur, which collectively determined the diagnosis related group (DRG). Surgeon fee were derived from current procedural terminology (CPT) codes based on Medicare allowable amounts using the resource-based relative-value scale. The patient-reported postdischarge resource utilization costs were calculated for visits to health-care providers’ visits, diagnostic costs, and medication-associated costs. NP and AP-related outpatient visits to providers including surgeons, chiropractors, other physicians, physical therapists, and acupuncturists were captured. Diagnostic tests including radiographs, computed tomography (CT) scans, magnetic resonance imaging (MRI), and electromyography were recorded. The postoperative devices utilization (braces, canes, and walkers), epidural steroid injections, emergency department visits, back-specific medications (nonsteroidal anti-inflammatory drugs, Cox-2 inhibitors, oral steroids, narcotics, muscle relaxants, and antidepressants), and inpatient or outpatient rehabilitation (physical therapy and occupational therapy) days were assessed. The cost associated with 90-d readmissions and complications were also recorded. Such cost calculations are reported previously.20, 21 Statistical Analysis Descriptive statistics including mean, standard deviation, median, and range for continuous variables, and frequency for categorical variables were calculated. For continuous data, Student's t-test and Mann–Whitney U-test were used. Nominal data were compared via Chi-squared test. A P-value <.05 was considered statistically significant. Multivariable linear regression model for total direct cost and hospital cost was built. Variables: age at the time of surgery, American Society of Anesthesiologists (ASA) grades, BMI, number of comorbidities, comorbidity including diabetes, hypertension, coronary artery disease (CAD), history of arthritis, and preoperative anticoagulation, length of surgery, length of hospital stay, complication, number of days on pain medication (muscle relaxant, narcotic pain medication, nonsteroidal anti-inflammatory drug (NSAIDs)), number of ASA visits, number of times diagnostics required (X-ray, MRI scans, and CT scan), number of postoperative surgeons or physician visits associated with back or leg pain, spine-related readmission within 90-d were included. All the analyses were performed using SPSS version 22 (IBM, Armonk, New York). RESULTS A total of 445 consecutive patients undergoing primary ACDF were analyzed. The mean age of 241 male and 204 female patients was 52 ± 10 yr. Twenty-four percent of patients (n = 106) had diabetes, the mean body mass index (BMI) was 31 ± 7; and 26% (n = 116) were obese with BMI ≥ 35. Mean baseline NDI was 42.1 ± 17.9 percentage points, EQ-5D was 0.65 ± 0.48, NRS-AP was 5.6 ± 3.2 points, and NP was 6.0 ± 2.8 points. Four percent (n = 18) of patients developed 90-d complications after surgery and 8 (1.7%) had readmission for management of UTI (n = 1), instrumentation failure (n = 3), dysphagia (n = 1), deep venous thrombosis (DVT) (n = 1), syncope (n = 1), and wound infection (n = 1). Table 1 summarizes the patient variables. TABLE 1. Characteristics of the Study Cohort Variables  N (%)  Age (>65)  56 (13%)  Gender: Male  241(54%)  Obese (BMI > 35)  116 (26%)  Smoker  120 (27%)  Duration of symptoms > 12 mo  192 (43%)  Radiculopathy  212 (48%)  Myelopathy  233 (52%)  History of Diabetes  106 (24%)  History of Arthritis  263 (59%)  History of preoperative anticoagulation  5 (1.1%)  History of MI  18 (4%)  Number of comorbidities (Mean ± SD)  2.5 ± 1.7  ASA grade: ¾  268 (60%)  Number of levels (Mean ± SD)  1.7 ± 0.8  Length of surgery (minutes; Mean ± SD)  158.09 ± 56.5  Length of hospital stay (Mean ± SD)  1.3 ± 1.1  EBL (Mean ± SD)  115.5 ± 136.0  Complications  18 (4%)  90-d readmission  8 (1.7%)  Variables  N (%)  Age (>65)  56 (13%)  Gender: Male  241(54%)  Obese (BMI > 35)  116 (26%)  Smoker  120 (27%)  Duration of symptoms > 12 mo  192 (43%)  Radiculopathy  212 (48%)  Myelopathy  233 (52%)  History of Diabetes  106 (24%)  History of Arthritis  263 (59%)  History of preoperative anticoagulation  5 (1.1%)  History of MI  18 (4%)  Number of comorbidities (Mean ± SD)  2.5 ± 1.7  ASA grade: ¾  268 (60%)  Number of levels (Mean ± SD)  1.7 ± 0.8  Length of surgery (minutes; Mean ± SD)  158.09 ± 56.5  Length of hospital stay (Mean ± SD)  1.3 ± 1.1  EBL (Mean ± SD)  115.5 ± 136.0  Complications  18 (4%)  90-d readmission  8 (1.7%)  View Large TABLE 1. Characteristics of the Study Cohort Variables  N (%)  Age (>65)  56 (13%)  Gender: Male  241(54%)  Obese (BMI > 35)  116 (26%)  Smoker  120 (27%)  Duration of symptoms > 12 mo  192 (43%)  Radiculopathy  212 (48%)  Myelopathy  233 (52%)  History of Diabetes  106 (24%)  History of Arthritis  263 (59%)  History of preoperative anticoagulation  5 (1.1%)  History of MI  18 (4%)  Number of comorbidities (Mean ± SD)  2.5 ± 1.7  ASA grade: ¾  268 (60%)  Number of levels (Mean ± SD)  1.7 ± 0.8  Length of surgery (minutes; Mean ± SD)  158.09 ± 56.5  Length of hospital stay (Mean ± SD)  1.3 ± 1.1  EBL (Mean ± SD)  115.5 ± 136.0  Complications  18 (4%)  90-d readmission  8 (1.7%)  Variables  N (%)  Age (>65)  56 (13%)  Gender: Male  241(54%)  Obese (BMI > 35)  116 (26%)  Smoker  120 (27%)  Duration of symptoms > 12 mo  192 (43%)  Radiculopathy  212 (48%)  Myelopathy  233 (52%)  History of Diabetes  106 (24%)  History of Arthritis  263 (59%)  History of preoperative anticoagulation  5 (1.1%)  History of MI  18 (4%)  Number of comorbidities (Mean ± SD)  2.5 ± 1.7  ASA grade: ¾  268 (60%)  Number of levels (Mean ± SD)  1.7 ± 0.8  Length of surgery (minutes; Mean ± SD)  158.09 ± 56.5  Length of hospital stay (Mean ± SD)  1.3 ± 1.1  EBL (Mean ± SD)  115.5 ± 136.0  Complications  18 (4%)  90-d readmission  8 (1.7%)  View Large The mean 90-d direct cost was $17 685 ± $5731 (median $16, 394 and range $9494 to $62, 590). The mean hospital costs were $14, 152 ± $5200 (median $12 959 and range $7026 to $55, 656). Table 2 summarizes the cost breakdown for 90-d direct costs. Table 3 summarizes the 90-d postoperative resource utilization. The median cost for primary ACDF was not significantly different among surgeons (Figure). Mean percent contribution of health care resource utilization is 5% (0%-38%) including diagnostic 0.92% (0%-25%), medication cost 2.4% (0%-12%), and health care visits 2.8% (0%-42%). CPT-based surgeon reimbursement contributed 16% (4%-30%) and the DRG-based hospital cost contributed 79% (30%-93%) to the total 90-d cost at 90-d. For the patients readmitted within 90-d following surgery, the percent contribution of readmission episode was as high as 35% (11%-63%). FIGURE. View largeDownload slide Box plot demonstrating variability in total 90-d cost among the participating surgeons. The “o” and “*” represent the outliers above third quartile or below first quartile. FIGURE. View largeDownload slide Box plot demonstrating variability in total 90-d cost among the participating surgeons. The “o” and “*” represent the outliers above third quartile or below first quartile. TABLE 2. Cost Breakdown for 90-d Direct Costs   Mean (range)  Hospital cost  $14 153 ($7026-$55 565)  Surgeons fee  $2664 ($2571-$2711)  Postoperative health-care resource utilization  $1132 ($145-$1691)  Readmit cost  $13 615 ($2550-$23 628)    Mean (range)  Hospital cost  $14 153 ($7026-$55 565)  Surgeons fee  $2664 ($2571-$2711)  Postoperative health-care resource utilization  $1132 ($145-$1691)  Readmit cost  $13 615 ($2550-$23 628)  View Large TABLE 2. Cost Breakdown for 90-d Direct Costs   Mean (range)  Hospital cost  $14 153 ($7026-$55 565)  Surgeons fee  $2664 ($2571-$2711)  Postoperative health-care resource utilization  $1132 ($145-$1691)  Readmit cost  $13 615 ($2550-$23 628)    Mean (range)  Hospital cost  $14 153 ($7026-$55 565)  Surgeons fee  $2664 ($2571-$2711)  Postoperative health-care resource utilization  $1132 ($145-$1691)  Readmit cost  $13 615 ($2550-$23 628)  View Large TABLE 3. 90-d Postoperative Resource Utilization Mean (range)  Frequency  Cost  Health care visits     $359 (0-$1077)   Surgeon visits  2 (1-3)     Physician visits  0 (0-3)     Chiropractor/Acupuncture visit  0 (0-5)     Outpatient PT/OT visits  4 (0-12)     Inpatient PT/OT visits  0 (0-21)    Pain Medication    $134 ($35-$303)   Narcotic pain medication  12 (7-21)     Muscle relaxant  21 (0-60)     NSAIDs  5 (0 -14)    Imaging and ER visits    $639 ($110-$1260)   X-rays  1.7 (0-3)     MRI  1 (0-2)     CT-scan  0 (0)     Spine injection  0 (0)     ER visits  0.7 (0-1)    Mean (range)  Frequency  Cost  Health care visits     $359 (0-$1077)   Surgeon visits  2 (1-3)     Physician visits  0 (0-3)     Chiropractor/Acupuncture visit  0 (0-5)     Outpatient PT/OT visits  4 (0-12)     Inpatient PT/OT visits  0 (0-21)    Pain Medication    $134 ($35-$303)   Narcotic pain medication  12 (7-21)     Muscle relaxant  21 (0-60)     NSAIDs  5 (0 -14)    Imaging and ER visits    $639 ($110-$1260)   X-rays  1.7 (0-3)     MRI  1 (0-2)     CT-scan  0 (0)     Spine injection  0 (0)     ER visits  0.7 (0-1)    View Large TABLE 3. 90-d Postoperative Resource Utilization Mean (range)  Frequency  Cost  Health care visits     $359 (0-$1077)   Surgeon visits  2 (1-3)     Physician visits  0 (0-3)     Chiropractor/Acupuncture visit  0 (0-5)     Outpatient PT/OT visits  4 (0-12)     Inpatient PT/OT visits  0 (0-21)    Pain Medication    $134 ($35-$303)   Narcotic pain medication  12 (7-21)     Muscle relaxant  21 (0-60)     NSAIDs  5 (0 -14)    Imaging and ER visits    $639 ($110-$1260)   X-rays  1.7 (0-3)     MRI  1 (0-2)     CT-scan  0 (0)     Spine injection  0 (0)     ER visits  0.7 (0-1)    Mean (range)  Frequency  Cost  Health care visits     $359 (0-$1077)   Surgeon visits  2 (1-3)     Physician visits  0 (0-3)     Chiropractor/Acupuncture visit  0 (0-5)     Outpatient PT/OT visits  4 (0-12)     Inpatient PT/OT visits  0 (0-21)    Pain Medication    $134 ($35-$303)   Narcotic pain medication  12 (7-21)     Muscle relaxant  21 (0-60)     NSAIDs  5 (0 -14)    Imaging and ER visits    $639 ($110-$1260)   X-rays  1.7 (0-3)     MRI  1 (0-2)     CT-scan  0 (0)     Spine injection  0 (0)     ER visits  0.7 (0-1)    View Large Table 4 summarizes the drivers of 90-d cost. The baseline 90-d cost computed based on multiple variable linear regression model was $6368, which increases $32 with each increase in the length of surgery (minutes), $1457 with each increase in the number of levels involved, $2216 with each additional day spent in hospital postoperative, $6389 in patients with the history of preoperative anticoagulation, $4890 for any 90-d complications, and $9344 for any 90-d readmission, $758 for the number of diagnostic imaging performed within 90-d. The model performance as measured by R2 was 0.616. TABLE 4. Multivariable Linear Regression Model for 90-d Total Cost Summarizing the Drivers of Total 90-d Cost   Beta coefficients  P-value  95.0% Confidence interval for B        Lower bound  Upper bound  (Constant)  6368.49  <0.0001  2931.414  9805.565  Age (at time of surgery)  6.484  0.781  −39.333  52.301  Gender  −565.895  0.174  −1382.947  251.157  BMI  −12.147  0.755  −88.771  64.478  Smoker  480.185  0.317  −461.364  1421.734  History of diabetes  305.278  0.647  −1006.475  1617.031  Arthritis  5.179  0.993  −1189.904  1200.261  Atrial fibrillation  −88.806  0.943  −2535.493  2357.881  Hypertension  −218.873  0.717  −1403.999  966.253  Coronary artery disease  −1052.076  0.162  −2529.941  425.789  Chronic obstructive pulmonary Disease  −1185.964  0.252  −3217.98  846.053  Myocardial infarction  −1575.176  0.202  −4000.926  850.574  Congestive heart failure  −505.572  0.729  −3372.089  2360.945  Number of comorbidity  301.439  0.421  −434.459  1037.337  Preoperative anticoagulation  6389.587  0.001  2616.268  10162.905  Duration of preoperative narcotic use (days)  −0.487  0.062  −0.997  0.024  Duration of symptoms  −390.774  0.342  −1197.932  416.384  Length of surgery in minutes  32.708  <0.0001  23.547  41.869  Length of hospital stay in d  2216.524  <0.0001  1787.8  2645.248  Number of levels  1457.841  <0.0001  784.679  2131.002  Postdischargeresource utilization           Pain meds (NSAID + Muscle relaxant)  −16.764  0.094  −36.404  2.877   Health care visits  −11.511  0.935  −290.124  267.102   Imaging  20.278  0.018  3.458  37.097   Narcotic (pain) medication  7.478  0.282  −6.163  21.119   Inpatient physical therapy  119.79  0.292  −103.494  343.075   Out-patient PT/OT  64.141  0.055  −1.484  129.767  ER visits  −443.49  0.306  −1294.74  407.76  Spine injections  1288.019  0.121  −340.301  2916.34  EMGs  1831.462  0.249  −1291.159  4954.083  90-d complication  3653.398  0.001  1478.81  5827.985  90-d readmission  7375.999  <0.0001  4127.889  10624.108    Beta coefficients  P-value  95.0% Confidence interval for B        Lower bound  Upper bound  (Constant)  6368.49  <0.0001  2931.414  9805.565  Age (at time of surgery)  6.484  0.781  −39.333  52.301  Gender  −565.895  0.174  −1382.947  251.157  BMI  −12.147  0.755  −88.771  64.478  Smoker  480.185  0.317  −461.364  1421.734  History of diabetes  305.278  0.647  −1006.475  1617.031  Arthritis  5.179  0.993  −1189.904  1200.261  Atrial fibrillation  −88.806  0.943  −2535.493  2357.881  Hypertension  −218.873  0.717  −1403.999  966.253  Coronary artery disease  −1052.076  0.162  −2529.941  425.789  Chronic obstructive pulmonary Disease  −1185.964  0.252  −3217.98  846.053  Myocardial infarction  −1575.176  0.202  −4000.926  850.574  Congestive heart failure  −505.572  0.729  −3372.089  2360.945  Number of comorbidity  301.439  0.421  −434.459  1037.337  Preoperative anticoagulation  6389.587  0.001  2616.268  10162.905  Duration of preoperative narcotic use (days)  −0.487  0.062  −0.997  0.024  Duration of symptoms  −390.774  0.342  −1197.932  416.384  Length of surgery in minutes  32.708  <0.0001  23.547  41.869  Length of hospital stay in d  2216.524  <0.0001  1787.8  2645.248  Number of levels  1457.841  <0.0001  784.679  2131.002  Postdischargeresource utilization           Pain meds (NSAID + Muscle relaxant)  −16.764  0.094  −36.404  2.877   Health care visits  −11.511  0.935  −290.124  267.102   Imaging  20.278  0.018  3.458  37.097   Narcotic (pain) medication  7.478  0.282  −6.163  21.119   Inpatient physical therapy  119.79  0.292  −103.494  343.075   Out-patient PT/OT  64.141  0.055  −1.484  129.767  ER visits  −443.49  0.306  −1294.74  407.76  Spine injections  1288.019  0.121  −340.301  2916.34  EMGs  1831.462  0.249  −1291.159  4954.083  90-d complication  3653.398  0.001  1478.81  5827.985  90-d readmission  7375.999  <0.0001  4127.889  10624.108  View Large TABLE 4. Multivariable Linear Regression Model for 90-d Total Cost Summarizing the Drivers of Total 90-d Cost   Beta coefficients  P-value  95.0% Confidence interval for B        Lower bound  Upper bound  (Constant)  6368.49  <0.0001  2931.414  9805.565  Age (at time of surgery)  6.484  0.781  −39.333  52.301  Gender  −565.895  0.174  −1382.947  251.157  BMI  −12.147  0.755  −88.771  64.478  Smoker  480.185  0.317  −461.364  1421.734  History of diabetes  305.278  0.647  −1006.475  1617.031  Arthritis  5.179  0.993  −1189.904  1200.261  Atrial fibrillation  −88.806  0.943  −2535.493  2357.881  Hypertension  −218.873  0.717  −1403.999  966.253  Coronary artery disease  −1052.076  0.162  −2529.941  425.789  Chronic obstructive pulmonary Disease  −1185.964  0.252  −3217.98  846.053  Myocardial infarction  −1575.176  0.202  −4000.926  850.574  Congestive heart failure  −505.572  0.729  −3372.089  2360.945  Number of comorbidity  301.439  0.421  −434.459  1037.337  Preoperative anticoagulation  6389.587  0.001  2616.268  10162.905  Duration of preoperative narcotic use (days)  −0.487  0.062  −0.997  0.024  Duration of symptoms  −390.774  0.342  −1197.932  416.384  Length of surgery in minutes  32.708  <0.0001  23.547  41.869  Length of hospital stay in d  2216.524  <0.0001  1787.8  2645.248  Number of levels  1457.841  <0.0001  784.679  2131.002  Postdischargeresource utilization           Pain meds (NSAID + Muscle relaxant)  −16.764  0.094  −36.404  2.877   Health care visits  −11.511  0.935  −290.124  267.102   Imaging  20.278  0.018  3.458  37.097   Narcotic (pain) medication  7.478  0.282  −6.163  21.119   Inpatient physical therapy  119.79  0.292  −103.494  343.075   Out-patient PT/OT  64.141  0.055  −1.484  129.767  ER visits  −443.49  0.306  −1294.74  407.76  Spine injections  1288.019  0.121  −340.301  2916.34  EMGs  1831.462  0.249  −1291.159  4954.083  90-d complication  3653.398  0.001  1478.81  5827.985  90-d readmission  7375.999  <0.0001  4127.889  10624.108    Beta coefficients  P-value  95.0% Confidence interval for B        Lower bound  Upper bound  (Constant)  6368.49  <0.0001  2931.414  9805.565  Age (at time of surgery)  6.484  0.781  −39.333  52.301  Gender  −565.895  0.174  −1382.947  251.157  BMI  −12.147  0.755  −88.771  64.478  Smoker  480.185  0.317  −461.364  1421.734  History of diabetes  305.278  0.647  −1006.475  1617.031  Arthritis  5.179  0.993  −1189.904  1200.261  Atrial fibrillation  −88.806  0.943  −2535.493  2357.881  Hypertension  −218.873  0.717  −1403.999  966.253  Coronary artery disease  −1052.076  0.162  −2529.941  425.789  Chronic obstructive pulmonary Disease  −1185.964  0.252  −3217.98  846.053  Myocardial infarction  −1575.176  0.202  −4000.926  850.574  Congestive heart failure  −505.572  0.729  −3372.089  2360.945  Number of comorbidity  301.439  0.421  −434.459  1037.337  Preoperative anticoagulation  6389.587  0.001  2616.268  10162.905  Duration of preoperative narcotic use (days)  −0.487  0.062  −0.997  0.024  Duration of symptoms  −390.774  0.342  −1197.932  416.384  Length of surgery in minutes  32.708  <0.0001  23.547  41.869  Length of hospital stay in d  2216.524  <0.0001  1787.8  2645.248  Number of levels  1457.841  <0.0001  784.679  2131.002  Postdischargeresource utilization           Pain meds (NSAID + Muscle relaxant)  −16.764  0.094  −36.404  2.877   Health care visits  −11.511  0.935  −290.124  267.102   Imaging  20.278  0.018  3.458  37.097   Narcotic (pain) medication  7.478  0.282  −6.163  21.119   Inpatient physical therapy  119.79  0.292  −103.494  343.075   Out-patient PT/OT  64.141  0.055  −1.484  129.767  ER visits  −443.49  0.306  −1294.74  407.76  Spine injections  1288.019  0.121  −340.301  2916.34  EMGs  1831.462  0.249  −1291.159  4954.083  90-d complication  3653.398  0.001  1478.81  5827.985  90-d readmission  7375.999  <0.0001  4127.889  10624.108  View Large Total direct cost = $6348 + $6.5 (age) + (−565.9) (Male) + (−12.4) (BMI) + (480) (smoker) + (305.3) (history of diabetes) + (5.2) (arthritis) + (−88.8) (atrial fibrillation) + (−218.9) (hypertension) + (−1052.1) (coronary artery disease) + (−1185.9) (chronic obstructive pulmonary disease) + (−1575) (myocardial infraction (MI)) + (−505) (chronic heart failure (CHF)) + (301) (number of comorbidity) + (6389) (preoperative anticoagulation) + (−0.49) (duration of preoperative opioid medication) + (−390.8) (duration of symptoms) + (32.7) (length of surgery) + (2216.5) (length of hospital stay) + (1457.8) (number of levels operated on) + (−16.8) (postdischarge nonopioid use) + (−11.5) (number of health care visits) + (20.3) (postdischarge imaging) + (7.4) (postdischarge opioid pain medication use) + (119.8) (number of days spent in inpatient rehabilitation) + (64.1) (number of days spent in outpatient PT/OT) + (−443.8) (number of postdischarge emergency room (ER) visits) + (1288.1) (number of postdischarge spine injections) + (1831.5) (Electromyography (EMG) postdischarge) + (3653.4) (number of 90-d complications) + (7375.9) (number of 90-d readmissions). DISCUSSION In this analysis, we demonstrate that the hospital costs had highest percent contributions to the total 90-d cost followed by readmission, and surgeons’ fee and health care resource utilization. The significant drivers of 90-d cost for primary ACDF were length of surgery, number of levels involved, length of hospital stay, preoperative history of anticoagulation medication, health-care resource utilization including number of imaging, any complications and readmission within 90-d after surgery. Our model accounted for 62% of variation in the total 90-d cost following surgery. The length of surgery was the one of significant contributors of the 90-d cost. The cost of first 30 min of operating room time was $1054 and increases with each 30 min increase in operating room time. The median length of surgery was 157 min; the predicted added cost of performing ACDF surgery will be $5270 for 50% of the patients. In our analysis, the length of surgery was calculated as time from skin incision to closure, it does not account for operating room time associated with patient positioning and time needed for induction of anesthesia. The length of hospital stay was a significant contributor to the total 90-d cost, which is not surprising. Previously reported factors associated with an increased length of hospital stay post-ACDF include older age, female gender, history of diabetes, anemia, and in-patient complications, and should be accounted for when computing the cost for ACDF.22-24 Furthermore, as demonstrated previously outpatient surgery for 1-to-2 level ACDF might save additional costs associated with hospitalization.25 Increasing number of levels operated on was another significant driver of cost. The additional cost associated with an increasing number of levels exposed can be attributed to increased length of surgery, increased use of graft and implant, and increased length of hospital stay. Patients on preoperative anticoagulation medication significantly contributed to total 90-d cost. Five patients in our study cohort were on anticoagulation medication preoperatively. For each of these patients, the total 90-d cost will increase by about $5796. In addition, the number of comorbidities and individual comorbidities including diabetes, history of arthritis, hypertension, MI, and obesity contributed significantly to the total 90-d cost; however, these factors were not the significant drivers of cost in multivariable model. The patients on preoperative anticoagulants have higher likelihood of having DVT and pulmonary embolism following surgery,26 which may result in higher 90-d complications leading to higher associated cost. Analogous to our findings, prior studies have reported that the postdischarge care accounts for a relatively small portion of total 90-d costs (range, 4%-8%).12,13,17,27 Of all the variables contributing to the postoperative health care resource utilization, the cost associated with imaging was the only significant driver of total 90-d cost. The medication use including opioid pain medication drives the cost following ACDF; however, the contribution was not statistically significant. Readmission episode and ER visits accounted for a significant increase in the 90-d cost for ACDF. The mean percent contribution of readmission cost was 35% to the 90-d cost. The readmission rate within 90-d following surgery was 2%, which suggests that for every 100 patients 2 patients can add up to 37% to the total 90-d cost. The inclusion of readmission in the targeted payment is essential to embrace the care of complex and sickest patient without being penalized for the event for the readmission encounter within 90-d postdischarge. Furthermore, as reported previously a better understanding of factors associated with readmission and ER visits following spine surgery are needed to facilitate efforts aimed at quality improvement and cost containment associated with these significant drivers of cost.28-31 Understanding and accurately predicting which patients may require readmission within a global period after surgery may help facilitate the creation and implementation of risk-adjusted bundled payment systems that would more fairly compensate surgeons and hospitals for advanced services. Prior studies have demonstrated significant variability in the episode of care cost within each individual spine related DRG.12-14, 27 However, none of the studies have determined the drivers of variability in 90-d cost for a spine care episode. Given the significant heterogeneity in patient-specific and disease-specific factors among the patients undergoing spine surgery, the variation in cost is inevitable. Therefore, implementation of the bundled payment models in real-world practice is complex and challenging.32-38 The drivers of variability in cost, as defined in this analysis, will guide a fuller accounting of costs incurred during the 90-d postdischarge time frame. Furthermore, it will form the basis of debate on defining the ranges within the target bundled price that would account for the highs and lows in the total 90-d cost. The optimization of the pre- and perioperative drivers will result in the containment of cost, and reduction in complications, which will ultimately result in increasing the overall value of spine care. Limitations There are several limitations to this study. This is a single institution study and even though the costs are reported from payer's perspective and adjusted for Medicare allowable national payments amount, costs may differ based on the geographical location. The healthcare resource utilization costs are patient self-reported, which might have some recall bias leading to higher or lower frequency of utilization. The indirect resource utilization from societal perspective including patient workday losses, family work day loss, caregiver cost were not included in the total cost calculation, as the goal of this study was to define variability in direct cost associated with healthcare resource utilization. Utilizing comprehensive list of variables captured in our prospective registry, we present a risk-adjusted analysis for drivers of cost for primary ACDF. The differences in patient population, surgeons’ perioperative practice patterns and institution-specific difference such as operating room staffing, reimbursement models must be considered while applying the results of this study to other institution. The costs are adjusted for the Medicare allowable national payment amounts. Utilizing the actual payer status for each patient would make the costs greater given the higher private pay percentage. The rate of complication and readmissions is comparable to the published literature;25,39-41 however given the relatively smaller number of patients from a single-center, it is possible that less common readmissions are likely not accounted for in the cost analyses if they weren’t encountered. Despite of the model being robust, the drivers of cost presented here explain about 60% variability in the cost for ACDF. The granular data on different types of instrumentation or type of biologics used are not captured as part of registry, which can certainly differ from based on surgeons’ and institutional practice patterns. The included confounding variables are not exhaustive and adding more pre- and perioperative variables might account for other 40% variation in the cost following index surgery. The generalized applicability of these results needs to take these limitations into account. Further analysis using multicenter prospective registries such as the Quality Outcomes Database is imperative.42 CONCLUSION There was considerable variation in total 90-d cost for elective ACDF surgery. Our model can explain about 62% of these variations in 90-d cost. The episode of care reimbursement models needs to take into account these variations and be inclusive of the factors that drive the variation in cost to develop a sustainable payment model. The generalized applicability should take in to account the differences in patient population, surgeons’ and institution-specific differences. Disclosure The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article. Notes The abstract was presented on December 4, 2015 at the Cervical Spine Research Society 2015 Annual Meeting, San Diego, California. REFERENCES 1. Angevine PD, Arons RR, McCormick PC. National and regional rates and variation of cervical discectomy with and without anterior fusion, 1990–1999. Spine . 2003; 28( 9): 931- 939; discussion 940. Google Scholar PubMed  2. Deyo RA, Gray DT, Kreuter W, Mirza S, Martin BI. United States trends in lumbar fusion surgery for degenerative conditions. Spine . 2005; 30( 12): 1441- 1445; discussion 1446-1447. Google Scholar CrossRef Search ADS PubMed  3. Marquez-Lara A, Nandyala SV, Fineberg SJ, Singh K. Current trends in demographics, practice, and in-hospital outcomes in cervical spine surgery. Spine . 2014; 39( 6): 476- 481. Google Scholar CrossRef Search ADS PubMed  4. Oglesby M, Fineberg SJ, Patel AA, Pelton MA, Singh K. Epidemiological trends in cervical spine surgery for degenerative diseases between 2002 and 2009. Spine . 2013; 38( 14): 1226- 1232. Google Scholar CrossRef Search ADS PubMed  5. Patil PG, Turner DA, Pietrobon R. National trends in surgical procedures for degenerative cervical spine disease: 1990–2000. Neurosurgery . 2005; 57( 4): 753- 758; discussion 753-758. Google Scholar CrossRef Search ADS PubMed  6. Wang MC, Kreuter W, Wolfla CE, Maiman DJ, Deyo RA. Trends and variations in cervical spine surgery in the United States. Spine . 2009; 34( 9): 955- 961; discussion 962-953. Google Scholar CrossRef Search ADS PubMed  7. Angevine PD, Zivin JG, McCormick PC. Cost-effectiveness of single-level anterior cervical discectomy and fusion for cervical spondylosis. Spine . 2005; 30( 17): 1989- 1997. Google Scholar CrossRef Search ADS PubMed  8. McAnany SJ, Overley S, Baird EO et al.   The 5-year cost-effectiveness of anterior cervical discectomy and fusion and cervical disc replacement. Spine . 2014; 39( 23): 1924- 1933. Google Scholar CrossRef Search ADS PubMed  9. Qureshi SA, McAnany S, Goz V, Koehler SM, Hecht AC. Cost-effectiveness analysis: comparing single-level cervical disc replacement and single-level anterior cervical discectomy and fusion. J Neurosurg Spine  2013; 19( 5): 546- 554. Google Scholar CrossRef Search ADS PubMed  10. Lubelski D, Senol N, Silverstein MP et al.   Quality of life outcomes after revision lumbar discectomy. J Neurosurg Spine  2015; 22( 2): 173- 178. Google Scholar CrossRef Search ADS PubMed  11. Bozic KJ, Ward L, Vail TP, Maze M. Bundled payments in total joint arthroplasty: targeting opportunities for quality improvement and cost reduction. Clin Orthop Relat Res . 2014; 472( 1): 188- 193. Google Scholar CrossRef Search ADS PubMed  12. Ugiliweneza B, Kong M, Nosova K et al.   Spinal surgery. Spine . 2014; 39( 15): 1235- 1242. Google Scholar CrossRef Search ADS PubMed  13. Birkmeyer JD, Gust C, Baser O, Dimick JB, Sutherland JM, Skinner JS. Medicare payments for common inpatient procedures: implications for episode-based payment bundling. Health Serv Res . 2010; 45( 6p1): 1783- 1795. Google Scholar CrossRef Search ADS PubMed  14. Miller DC, Gust C, Dimick JB, Birkmeyer N, Skinner J, Birkmeyer JD. Large variations in medicare payments for surgery highlight savings potential from bundled payment programs. Health Aff . 2011; 30( 11): 2107- 2115. Google Scholar CrossRef Search ADS   15. Cromwell J, Dayhoff DA, Thoumaian AH. Cost savings and physician responses to global bundled payments for medicare heart bypass surgery. Health Care Financ Rev  1997; 19( 1): 41- 57. Google Scholar PubMed  16. Edmonds C, Hallman GL. Cardio Vascular Care Providers. A pioneer in bundled services, shared risk, and single payment. Tex Heart Inst J . 1995; 22( 1): 72- 76. Google Scholar PubMed  17. Grenda TR, Pradarelli JC, Thumma JR, Dimick JB. Variation in hospital episode costs with bariatric surgery. JAMA Surg . 2015; 50( 12): 1109- 1115. Google Scholar CrossRef Search ADS   18. EuroQol. EuroQol—a new facility for the measurement of health-related quality of life. Health Policy . 1990; 16( 3): 199- 208. CrossRef Search ADS PubMed  19. Langley GB, Sheppeard H. The visual analogue scale: its use in pain measurement. Rheumatol Int . 1985; 5( 4): 145- 148. Google Scholar CrossRef Search ADS PubMed  20. Devin CJ, Chotai S, Parker SL, Tetreault L, Fehlings MG, McGirt MJ. A cost-utility analysis of lumbar decompression with and without fusion for degenerative spine disease in the elderly. Neurosurgery . 2015; 77( suppl 4): S116- S124. Google Scholar CrossRef Search ADS PubMed  21. Parker SL, Fulchiero EC, Davis BJ et al.   Cost-effectiveness of multilevel hemilaminectomy for lumbar stenosis–associated radiculopathy. Spine J . 2011; 11( 8): 705- 711. Google Scholar CrossRef Search ADS PubMed  22. Arnold PM, Rice LR, Anderson KK, McMahon JK, Connelly LM, Norvell DC. Factors affecting hospital length of stay following anterior cervical discectomy and fusion. Evid-Based Spine Care J . 2011; 2( 03): 11- 18. Google Scholar CrossRef Search ADS PubMed  23. Basques BA, Bohl DD, Golinvaux NS, Gruskay JA, Grauer JN. Preoperative factors affecting length of stay after elective anterior cervical discectomy and fusion with and without corpectomy. Spine . 2014; 39( 12): 939- 946. Google Scholar CrossRef Search ADS PubMed  24. Gruskay JA, Fu M, Basques B et al.   Factors affecting length of stay and complications following elective anterior cervical discectomy and fusion. Clin Spine Surg . 2017; 29( 1): E34- 42. Google Scholar CrossRef Search ADS   25. McGirt MJ, Godil SS, Asher AL, Parker SL, Devin CJ. Quality analysis of anterior cervical discectomy and fusion in the outpatient versus inpatient setting: analysis of 7288 patients from the NSQIP database. Neurosurg Focus . 2015; 39( 6): E9. Google Scholar CrossRef Search ADS PubMed  26. Kearon C, Hirsh J. Management of anticoagulation before and after elective surgery. N Engl J Med . 1997; 336( 21): 1506- 1511. Google Scholar CrossRef Search ADS PubMed  27. Schoenfeld AJ, Harris MB, Liu H, Birkmeyer JD. Variations in medicare payments for episodes of spine surgery. Spine J . 2014; 14( 12): 2793- 2798. Google Scholar CrossRef Search ADS PubMed  28. Akamnonu C, Cheriyan T, Goldstein JA, Lafage V, Errico TJ, Bendo JA. Unplanned hospital readmission after surgical treatment of common lumbar pathologies. Spine . 2015; 40( 6): 423- 428. Google Scholar CrossRef Search ADS PubMed  29. Bernatz JT, Anderson PA. Thirty-day readmission rates in spine surgery: systematic review and meta-analysis. Neurosurg Focus . 2015; 39( 4): E7. Google Scholar CrossRef Search ADS PubMed  30. Bosco JA 3rd, Karkenny AJ, Hutzler LH, Slover JD, Iorio R. Cost burden of 30-day readmissions following medicare total hip and knee arthroplasty. J Arthroplasty . 2014; 29( 5): 903- 905. Google Scholar CrossRef Search ADS PubMed  31. Taylor BE, Youngerman BE, Goldstein H et al.   Causes and timing of unplanned early readmission after neurosurgery. Neurosurgery . 2016; 79( 3): 356- 369. Google Scholar CrossRef Search ADS PubMed  32. Cutler DM, Ghosh K. The potential for cost savings through bundled episode payments. N Engl J Med . 2012; 366( 12): 1075- 1077. Google Scholar CrossRef Search ADS PubMed  33. Delisle DR. Big things come in bundled packages. Am J Med Qual . 2013; 28( 4): 339- 344. Google Scholar CrossRef Search ADS PubMed  34. Iorio R. Strategies and tactics for successful implementation of bundled payments: bundled payment for care improvement at a large, urban, academic medical center. J Arthroplasty . 2015; 30( 3): 349- 350. Google Scholar CrossRef Search ADS PubMed  35. Kazberouk A, McGuire K, Landon BE. A survey of innovative reimbursement models in spine care. Spine (Phila Pa 1976) . 2016; 41( 4): 344- 352. Google Scholar CrossRef Search ADS PubMed  36. Mechanic RE. Opportunities and challenges for episode-based payment. N Engl J Med . 2011; 365( 9): 777- 779. Google Scholar CrossRef Search ADS PubMed  37. Rossi VJ, Ahn J, Bohl DD, Tabaraee E, Singh K. Economic factors in the future delivery of spinal healthcare. World J Orthop . 2015; 6( 5): 409- 412. Google Scholar CrossRef Search ADS PubMed  38. Sood N, Huckfeldt PJ, Escarce JJ, Grabowski DC, Newhouse JP. Medicare's bundled payment pilot for acute and postacute care: analysis and recommendations on where to begin. Health Aff . 2011; 30( 9): 1708- 1717. Google Scholar CrossRef Search ADS   39. Bhashyam N, De la Garza Ramos R, Nakhla J et al.   Thirty-day readmission and reoperation rates after single-level anterior cervical discectomy and fusion versus those after cervical disc replacement. Neurosurg Focus . 2017; 42( 2): E6. Google Scholar CrossRef Search ADS PubMed  40. Veeravagu A, Cole T, Jiang B, Ratliff JK. Revision rates and complication incidence in single- and multilevel anterior cervical discectomy and fusion procedures: an administrative database study. Spine J . 2014; 14( 7): 1125- 1131. Google Scholar CrossRef Search ADS PubMed  41. Adamson T, Godil SS, Mehrlich M, Mendenhall S, Asher AL, McGirt MJ. Anterior cervical discectomy and fusion in the outpatient ambulatory surgery setting compared with the inpatient hospital setting: analysis of 1000 consecutive cases. J Neurosurg Spine  2016; 24( 6): 878- 884. Google Scholar CrossRef Search ADS PubMed  42. Asher AL, Speroff T, Dittus RS et al.   The National Neurosurgery Quality and Outcomes Database (N2QOD). Spine . 2014; 39( 22 Suppl 1): S106- S116. Google Scholar CrossRef Search ADS PubMed  COMMENTS Based on a desperate need for national healthcare cost containment and the fact that by 2050 healthcare cost will consume over 20% of the GDP, the investigators of this submission have undertaken a prospective study on 445 patients with ACDF in order to answer 1 question: What drives the cost of ACDF over a 90-day period? From among over 30 appropriate variables, regression analysis indicated that the major drivers of cost in 1 episode of care were hospital length of stay and readmission (mainly due to complications). Surgery related factors were length and levels of surgery and follow-up drivers of cost included imaging studies. The findings are simple to understand but how to make them achievable seems tough. With the baby boomers getting older, cervical spondylotic radiculomyelopathies due to degenerative disc-osteophyte complex has become abundant. Such patients have comorbidities and are usually on anticoagulants, making their hospital course rocky and problematic. It is clear from this submission and many similar publications that attempts are needed to shorten hospital length of stay, length of surgery, and factors predisposing to complications such as venous thromboembolism, infections, and unnecessary postoperative imaging studies to name a few. Further studies should shed light on additional drivers of cost and better means to contain them. Bizhan Aarabi Baltimore, Maryland The authors argue that a progressive movement toward bundled payments for given procedures is well underway, making understanding of the particular drivers of variability of costs important. They investigated the factors that resulted in variation of costs for patients undergoing elective ACDF surgery using a prospective registry. As would be expected, factors that increased total costs included length of stay, treatment of additional spinal levels, and the need for hospital readmission. Surgeon contribution to cost was modest, as the major factors in cost variation were hospital based. An interesting finding was the magnitude of impact on overall costs for patients on preoperative anticoagulation therapy in whom there is a higher incidence of postoperative deep vein thrombosis and pulmonary embolus. Overall, the cost drivers of the model that emerged explained 62% of costs. The authors acknowledge that the impact of the study is diminished by the lack of granular data on the specific expenses such as type of biologics or instrumentation employed for a given patient. Nevertheless, the investigation is a meaningful step toward understanding of costs in spinal surgery. John Kenneth Houten Bronx, New York The authors have offered a retrospective analysis of the major factors driving cost in elective (nontrauma, nontumor) 1–3 level ACDFs from a single center. Four hundred forty-five patients were analyzed and a multivariate analysis was provided which explained 60% of the cost variability. Not surprisingly, factors such as surgical time, days in hospital, levels operated, complications, and readmissions were the primary contributors to cost. It was also discovered that preoperative anticoagulation conveyed a significant increase to the total cost. While the manuscript is limited by being a single institutional study and by 40% of costs not ascertained, it nevertheless sets the foundation for understanding procedural costs as the healthcare system moves toward more bundled payments. These analyses are vital for setting fair reimbursements. There will undoubtedly need to be more granularity within future studies (types of complications and costs, biologics/instrumentation costs), but this manuscript sets the stage for more in-depth investigations. Nathan E. Simmons Lebanon, New Hampshire Copyright © 2018 by the Congress of Neurological Surgeons This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

NeurosurgeryOxford University Press

Published: May 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off