Bundled Payment Models in Spine Surgery: Current Challenges and Opportunities, a Systematic Review

Reliable Prediction of Discharge Disposition Following Cervical Spine Surgery With Ensemble Machine Learning and Validation on a National Cohort

STUDY DESIGN

A retrospective cohort study.

OBJECTIVE

The purpose of this study is to develop a machine learning algorithm to predict nonhome discharge after cervical spine surgery that is validated and usable on a national scale to ensure generalizability and elucidate candidate drivers for prediction.

SUMMARY OF BACKGROUND DATA

Excessive length of hospital stay can be attributed to delays in postoperative referrals to intermediate care rehabilitation centers or skilled nursing facilities. Accurate preoperative prediction of patients who may require access to these resources can facilitate a more efficient referral and discharge process, thereby reducing hospital and patient costs in addition to minimizing the risk of hospital-acquired complications.

METHODS

Electronic medical records were retrospectively reviewed from a single-center data warehouse (SCDW) to identify patients undergoing cervical spine surgeries between 2008 and 2019 for machine learning algorithm development and internal validation. The National Inpatient Sample (NIS) database was queried to identify cervical spine fusion surgeries between 2009 and 2017 for external validation of algorithm performance. Gradient-boosted trees were constructed to predict nonhome discharge across patient cohorts. The area under the receiver operating characteristic curve (AUROC) was used to measure model performance. SHAP values were used to identify nonlinear risk factors for nonhome discharge and to interpret algorithm predictions.

RESULTS

A total of 3523 cases of cervical spine fusion surgeries were included from the SCDW data set, and 311,582 cases were isolated from NIS. The model demonstrated robust prediction of nonhome discharge across all cohorts, achieving an area under the receiver operating characteristic curve of 0.87 (SD=0.01) on both the SCDW and nationwide NIS test sets. Anterior approach only, age, elective admission status, Medicare insurance status, and total Elixhauser Comorbidity Index score were the most important predictors of discharge destination.

CONCLUSIONS

Machine learning algorithms reliably predict nonhome discharge across single-center and national cohorts and identify preoperative features of importance following cervical spine fusion surgery.

Baseline Frailty and Discharge to Advanced Care Facilities in Patients Undergoing Lumbar Interbody Fusion for Lumbar Degenerative Disease: A Multicenter Registry Analysis of 7153 Patient Cases Comparing the Risk Analysis Index to the 5-Factor Modified Frailty Index

OBJECTIVE

To evaluate the impact of frailty, as measured by the 5-factor modified Frailty Index (mFI-5) and the Risk Analysis Index (RAI), on advanced care facility discharge (FD) in patients who underwent lumbar fusion for lumbar degenerative spine disease.

METHODS

The American College of Surgeons National Surgical Quality Improvement Program (2012-2020) was queried for adults (≥18 years) undergoing lumbar fusion for lumbar degenerative disease. Descriptive statistics and univariate crosstabulation were used to assess baseline demographics, preoperative comorbidities, and postoperative outcomes. Receiver operating characteristic curve analysis was used to assess the discriminative threshold of the mFI-5 and RAI on FD within this population.

RESULTS

The median patient age in this study cohort (N = 7153) was 56 years and FD occurred in 7.3% of cases. Receiver operating characteristic curve analysis demonstrated that both the mFI-5 and the RAI accurately predicted FD (C-statistics: mFI-5: 0.627; RAI: 0.746). DeLong's test found that the RAI had superior discrimination when compared to the mFI-5 (P < 0.0001).

CONCLUSIONS

RAI is a reliable predictor of FD in lumbar degenerative disease patients who underwent lumbar interbody fusion and demonstrated superior discrimination compared to the mFI-5. Identification of patients at risk for FD may facilitate more precise risk stratification to enable better preoperative decision-making and help set more realistic expectations of care.

Implant Selection Strategies for Total Joint Arthroplasty: The Effects on Cost Containment and Physician Autonomy

BACKGROUND

With continued declines in reimbursement for total joint arthroplasty, health systems have explored implant cost containment measures to generate sustainable margins. This review evaluated how implementation of (1) implant price control programs, (2) vendor purchasing agreements, and (3) bundled payment models affected implant costs and physician autonomy in implant selection.

METHODS

PubMed, EBSCOhost, and Google Scholar were searched to identify studies that evaluated the efficacy of total hip or total knee arthroplasty implant selection strategies. The review included publications between January 1, 2002, and October 17, 2022. The mean Methodological Index for Nonrandomized Studies score was 18.3 ± 1.8.

RESULTS

A total of 13 studies (32,197 patients) were included. All studies implementing implant price capitation programs found decreased implant costs, ranging 2.2 to 26.1% and increased utilization of premium implants. Most studies found bundled payments models reduced total joint arthroplasty implant costs with greatest reduction being 28.9%. Additionally, while absolute single vendor agreements had higher implant costs, preferred single vendor agreements had reduced implant costs. When given price constraints, surgeons tended to select more premium implants.

CONCLUSION

Alternative payment models that incorporated implant selection strategies saw reduced costs and surgeon utilization of premium implants. The study findings encourage further research on implant selection strategies, which must balance the goals of cost containment with physician autonomy and optimized patient care.

LEVEL OF EVIDENCE

Level III.

Neurosurgeons Deliver Similar Quality Care Regardless of First Assistant Type: Resident Physician versus Nonphysician Surgical Assistant

OBJECTIVE

There are limited data evaluating the outcomes of attending neurosurgeons with different types of first assistants. This study considers a common neurosurgical procedure (single-level, posterior-only lumbar fusion surgery) and examines whether attending surgeons deliver equal patient outcomes, regardless of the type of first assistant (resident physician vs. nonphysician surgical assistant [NPSA]), among otherwise exact-matched patients.

METHODS

The authors retrospectively analyzed 3395 adult patients undergoing single-level, posterior-only lumbar fusion at a single academic medical center. Primary outcomes included readmissions, emergency department visits, reoperation, and mortality within 30 and 90 days after surgery. Secondary outcome measures included discharge disposition, length of stay, and length of surgery. Coarsened exact matching was used to match patients on key demographics and baseline characteristics known to independently affect neurosurgical outcomes.

RESULTS

Among exact-matched patients (n = 1402), there was no significant difference in adverse postsurgical events (readmission, emergency department visits, reoperation, or mortality) within 30 days or 90 days of the index operation between patients who had resident physicians and those who had NPSAs as first assistants. Patients who had resident physicians as first assistants demonstrated a longer length of stay (mean: 100.0 vs. 87.4 hours, P < 0.001) and a shorter duration of surgery (mean: 187.4 vs. 213.8 minutes, P < 0.001). There was no significant difference between the two groups in the percentage of patients discharged home.

CONCLUSIONS

For single-level posterior spinal fusion, in the setting described, there are no differences in short-term patient outcomes delivered by attending surgeons assisted by resident physicians versus NPSAs.

Drivers of Cost in Primary Single-Level Lumbar Fusion Surgery

STUDY DESIGN

Retrospective cohort.

OBJECTIVES

Allocating cost is challenging with traditional hospital accounting. Time-driven activity-based costing (TDABC) is an efficient method to accurately assign cost. We sought to characterize the variation in direct total hospital cost (THC) among both lumbar fusion approaches and surgeons.

METHODS

Patients were treated with single-level anterior interbody (ALIF), lateral interbody (LLIF), transforaminal interbody (TLIF), instrumented posterolateral (PLF) or in-situ fusion (ISF) for degenerative disease. Process maps were developed for preoperative, intraoperative and postoperative care. THC was composed of implant, medication, other supply, and personnel costs. Linear regression and descriptive statistics were used to analyze THC variation.

RESULTS

A total of 696 patients underwent surgery by 8 surgeons. Approximately 50% of THC variation was associated with procedure choice while patient characteristics explained 10%. Implants (including biologics) accounted for 45% of cost. With reference to PLF, THC ranged from 0.6x (ISF) to 1.7x (LLIF). Implant cost ranged from 2.5x reference (LLIF) to 0.1x (ISF). There was a 1.7x difference between the highest THC surgeon and the lowest. The fusion type with the highest THC variation was TLIF. The surgeon with the highest TLIF THC was 1.5x more expensive than the surgeon with the lowest.

CONCLUSIONS

Surgeon-based choices have the greatest effect on THC variation and represent the largest opportunities for cost savings. Primary single-level lumbar fusion THC is driven primarily by fusion type. Implants, including biologics, account for nearly half this cost. Future work should incorporate outcomes data to characterize the differential value conferred by higher THC fusions.

A Multi-Disciplinary Review of Time-Driven Activity-Based Costing: Practical Considerations for Spine Surgery

STUDY DESIGN

A multi-disciplinary review.

OBJECTIVES

To provide a roadmap for implementing time-driven activity-based costing (TDABC) for spine surgery. This is achieved by organizing and scrutinizing publications in the spine, neurosurgical, and orthopedic literature which utilize TDABC and related methodologies.

METHODS

PubMed and Google Scholar were searched for relevant articles. The articles were selected by two independent researchers. After article selection, data was extracted and summarized into research domains. Preferred reporting items for systematic reviews and meta-analyses (PRISMA) systematic review process was followed.

RESULTS

Of the 524 articles screened, thirty-five articles met the inclusion criteria. Each included article was examined and reviewed to define the primary research question and objective. Comparing different procedures was the most common primary objective. Direct observation along with one other strategy (surveys, interviews, surgical database, or EMR) was most commonly employed during process map development. Across all surgical subspecialties (spine, neurologic, and orthopedic surgery), costs were divided into direct cost, indirect cost, cost to patient, and total costs. The most commonly calculated direct costs included personnel and supply costs. Facility costs, hospital overhead costs, and utilities were the most commonly calculated indirect costs. Transportation costs and parental lost wages were considered when calculating cost to patient. The total cost was a sum of direct costs, indirect costs, and costs to the patient.

CONCLUSION

TDABC provides a common platform to accurately estimate costs of care delivery. Institutions embarking on TDABC for spine surgery should consider the breadth of methodologies highlighted in this review to determine which type of calculations are appropriate for their practice.

Predictive models to assess risk of extended length of stay in adults with spinal deformity and lumbar degenerative pathology: development and internal validation

BACKGROUND CONTEXT

Postoperative recovery after adult spinal deformity (ASD) operations is arduous, fraught with complications, and often requires extended hospital stays. A need exists for a method to rapidly predict patients at risk for extended length of stay (eLOS) in the preoperative setting.

PURPOSE

To develop a machine learning model to preoperatively estimate the likelihood of eLOS following elective multi-level lumbar/thoracolumbar spinal instrumented fusions (≥3 segments) for ASD.

STUDY DESIGN/SETTING

Retrospectively from a state-level inpatient database hosted by the Health care cost and Utilization Project.

PATIENT SAMPLE

Of 8,866 patients of age ≥50 with ASD undergoing elective lumbar or thoracolumbar multilevel instrumented fusions.

OUTCOME MEASURES

The primary outcome was eLOS (>7 days).

METHODS

Predictive variables consisted of demographics, comorbidities, and operative information. Significant variables from univariate and multivariate analyses were used to develop a logistic regression-based predictive model that use six predictors. Model accuracy was assessed through area under the curve (AUC), sensitivity, and specificity.

RESULTS

Of 8,866 patients met inclusion criteria. A saturated logistic model with all significant variables from multivariate analysis was developed (AUC=0.77), followed by generation of a simplified logistic model through stepwise logistic regression (AUC=0.76). Peak AUC was reached with inclusion of six selected predictors (combined anterior and posterior approach, surgery to both lumbar and thoracic regions, ≥8 level fusion, malnutrition, congestive heart failure, and academic institution). A cutoff of 0.18 for eLOS yielded a sensitivity of 77% and specificity of 68%.

CONCLUSIONS

This predictive model can facilitate identification of adults at risk for eLOS following elective multilevel lumbar/thoracolumbar spinal instrumented fusions for ASD. With a fair diagnostic accuracy, the predictive calculator will ideally enable clinicians to improve preoperative planning, guide patient expectations, enable optimization of modifiable risk factors, facilitate appropriate discharge planning, stratify financial risk, and accurately identify patients who may represent high-cost outliers. Future prospective studies that validate this risk assessment tool on external datasets would be valuable.

Machine Learning Can Accurately Predict Overnight Stay, Readmission, and 30-Day Complications Following Anterior Cruciate Ligament Reconstruction

PURPOSE

This study aimed to develop machine learning (ML) models to predict hospital admission (overnight stay) as well as short-term complications and readmission rates following anterior cruciate ligament reconstruction (ACLR). Furthermore, we sought to compare the ML models with logistic regression models in predicting ACLR outcomes.

METHODS

The American College of Surgeons National Surgical Quality Improvement Program database was queried for patients who underwent elective ACLR from 2012 to 2018. Artificial neural network ML and logistic regression models were developed to predict overnight stay, 30-day postoperative complications, and ACL-related readmission, and model performance was compared using the area under the receiver operating characteristic curve. Regression analyses were used to identify variables that were significantly associated with the predicted outcomes.

RESULTS

A total of 21,636 elective ACLR cases met inclusion criteria. Variables associated with hospital admission included White race, obesity, hypertension, and American Society of Anesthesiologists classification 3 and greater, anesthesia other than general, prolonged operative time, and inpatient setting. The incidence of hospital admission (overnight stay) was 10.2%, 30-day complications was 1.3%, and 30-day readmission for ACLR-related causes was 0.9%. Compared with logistic regression models, artificial neural network models reported superior area under the receiver operating characteristic curve values in predicting overnight stay (0.835 vs 0.589), 30-day complications (0.742 vs 0.590), reoperation (0.842 vs 0.601), ACLR-related readmission (0.872 vs 0.606), deep-vein thrombosis (0.804 vs 0.608), and surgical-site infection (0.818 vs 0.596).

CONCLUSIONS

The ML models developed in this study demonstrate an application of ML in which data from a national surgical patient registry was used to predict hospital admission and 30-day postoperative complications after elective ACLR. ML models developed performed well, outperforming regression models in predicting hospital admission and short-term complications following elective ACLR. ML models performed best when predicting ACLR-related readmissions and reoperations, followed by overnight stay.

LEVEL OF EVIDENCE

IV, retrospective comparative prognostic trial.

Posterior Cervical Fusion for Fracture Is Not the Same as Fusion for Degenerative Cervical Spine Disease: Implications for a Bundled Payment Model

STUDY DESIGN/SETTING

Retrospective.

OBJECTIVE

To understand differences in 30-day outcomes between patients undergoing posterior cervical fusion (PCF) for fracture versus degenerative cervical spine disease.

SUMMARY OF BACKGROUND DATA

Current bundled payment models for cervical fusions, such as the Bundled Payments for Care Improvement revolve around the use of diagnosis-related groups to categorize patients for reimbursement purposes. Though a PCF performed for a fracture may have a different postoperative course of care as compared with a fusion being done for degenerative cervical spine pathology, the current DRG system does not differentiate payments based on the indication/cause of surgery.

METHODS

The 2012-2017 American College of Surgeons-National Surgical Quality Improvement Program was queried using Current Procedural Terminology code 22600 to identify patients receiving elective PCFs. Multivariate analyses were used to compare rates of 30-day severe adverse events, minor adverse events, readmissions, length of stay, and nonhome discharges between the 2 groups.

RESULTS

A total 2546 (91.4%) PCFs were performed for degenerative cervical spine pathology and 240 (8.6%) for fracture. After adjustment for differences in baseline clinical characteristics, patients undergoing a PCF for a fracture versus degenerative pathology had higher odds of severe adverse events [18.8% vs. 10.6%, odds ratio (OR): 1.65 (95% CI, 1.10-2.46); P =0.015], prolonged length of stay >3 days [54.2% vs. 40.5%, OR: 1.93 (95% CI, 1.44-2.59); P <0.001], and nonhome discharges [34.2% vs. 27.6%, OR: 1.54 (95% CI, 1.10-2.17); P =0.012].

CONCLUSIONS

Patients undergoing PCFs for fracture have significant higher rates of postoperative adverse events and greater resource utilization as compared with individuals undergoing elective PCF for degenerative spine pathology. The study calls into question the need of risk adjustment of bundled prices based on indication/cause of the surgery to prevent the creation of a financial disincentive when taking care/performing surgery in spinal trauma patients.

The Divergence Between Hospital Charges and Reimbursements For Primary 1-2-level Lumbar Fusion Has Increased Over Time: A Medicare Administrative Claims Analysis

STUDY DESIGN

A retrospective cohort study was performed for patients undergoing 1-2-level lumbar fusion (1-2LF) from 2005 to 2014 using an administrative claims database.

OBJECTIVE

The objective of this study was to determine changes in: (1) annual charges; (2) annual reimbursement rates; and (3) annual difference (charges minus reimbursements) in patients undergoing 1-2LF.

SUMMARY OF BACKGROUND DATA

With implementation of value-based care in orthopaedics, coupled with the rise in number of patients undergoing 1-2LF, understanding the discordance in hospital charges and reimbursements is needed. The difference in hospital charges to reimbursements specifically for 1-2LF for degenerative disc disease has not been studied.

MATERIALS AND METHODS

A Medicare administrative claims database was queried for patients undergoing primary lumbar fusion using ICD-9 procedural code 81.04-81.08. Patients specifically undergoing 1-2LF were filtered from this cohort using ICD-9 procedural code 81.62. The query yielded 547,067 patients who underwent primary 1-2LF. Primary outcomes analyzed included trends in charges, reimbursement rates, and net difference in cost over time and per annual basis. Linear regression evaluated the change in costs over time with a P -value less than 0.05 considered significant.

RESULTS

From 2005 to 2014, total charges increased from $6,085,838,407 to $19,621,979,956 and total reimbursements increased from $1,677,764,831 to $4,656,702,685 (all P <0.001). Per patient charges increased 92.10% from 2005 to 2014 for patients undergoing primary 1-2LF from $129,992 to $249,697 ( P <0.001). Similarly, an increase in reimbursement per patient of 65.35% from $35,836 to $59,258 ( P <0.001) was noted. The annual difference in charges to reimbursements increased 102.26% during the study interval from $94,155 to $190,439 ( P <0.001).

CONCLUSIONS

Per patient charges and reimbursements both increased over the study period; however, charges increased 30% more than reimbursements. Further breakdown of hospital, surgeon, and anesthesiologist reimbursements for 1-2LF is needed.

LEVEL OF EVIDENCE

Level III.

An Ensemble Learning Approach to Improving Prediction of Case Duration for Spine Surgery: Algorithm Development and Validation

BACKGROUND

Estimating surgical case duration accurately is an important operating room efficiency metric. Current predictive techniques in spine surgery include less sophisticated approaches such as classical multivariable statistical models. Machine learning approaches have been used to predict outcomes such as length of stay and time returning to normal work, but have not been focused on case duration.

OBJECTIVE

The primary objective of this 4-year, single-academic-center, retrospective study was to use an ensemble learning approach that may improve the accuracy of scheduled case duration for spine surgery. The primary outcome measure was case duration.

METHODS

We compared machine learning models using surgical and patient features to our institutional method, which used historic averages and surgeon adjustments as needed. We implemented multivariable linear regression, random forest, bagging, and XGBoost (Extreme Gradient Boosting) and calculated the average R², root-mean-square error (RMSE), explained variance, and mean absolute error (MAE) using k-fold cross-validation. We then used the SHAP (Shapley Additive Explanations) explainer model to determine feature importance.

RESULTS

A total of 3189 patients who underwent spine surgery were included. The institution's current method of predicting case times has a very poor coefficient of determination with actual times (R²=0.213). On k-fold cross-validation, the linear regression model had an explained variance score of 0.345, an R² of 0.34, an RMSE of 162.84 minutes, and an MAE of 127.22 minutes. Among all models, the XGBoost regressor performed the best with an explained variance score of 0.778, an R² of 0.770, an RMSE of 92.95 minutes, and an MAE of 44.31 minutes. Based on SHAP analysis of the XGBoost regression, body mass index, spinal fusions, surgical procedure, and number of spine levels involved were the features with the most impact on the model.

CONCLUSIONS

Using ensemble learning-based predictive models, specifically XGBoost regression, can improve the accuracy of the estimation of spine surgery times.

A Practical Guide to Lower Extremity Nerve Blocks for the Sports Medicine Surgeon

BACKGROUND

Peripheral nerve blocks (PNBs) are vital in the administration of surgical analgesia and have grown in popularity for use in lower extremity arthroscopic procedures because of their capacity to safely and effectively control pain. The number and specificity of PNBs, however, have made choosing the best option for a procedure nebulous for orthopaedic surgeons.

PURPOSE

To present a narrative literature review of the PNBs available for arthroscopic hip and knee procedures that is adapted to an audience of orthopaedic surgeons.

STUDY DESIGN

Narrative literature review.

METHODS

A combination of the names of various lower extremity PNBs AND "hip arthroscopy" OR "knee arthroscopy" was used to search the English medical literature including PubMed, Cochrane Library, ScienceDirect, Embase, and Scopus. Placement technique, specificity of blockade, efficacy, and complications were assessed. Searches were performed through May 2, 2021.

RESULTS

A total of 157 studies were included in this review of lower extremity PNBs. Femoral nerve, lumbar plexus, sciatic nerve, and fascia iliaca compartment blocks were most commonly used in arthroscopic hip surgery, while femoral nerve, 3-in-1, and adductor canal blocks were preferred for arthroscopic knee surgery. Each block demonstrated a significant benefit (P > .05) in ≥1 of the following outcomes: intraoperative morphine, pain scores, nausea, and/or opioid consumption. Combination blocks including the lateral femoral cutaneous nerve block, obturator nerve block, quadratus lumborum block, and L1 and L2 paravertebral block have also been described. Complication rates ranged from 0% to 4.8% in those administered with ultrasound guidance. The most commonly reported complications included muscular weakness, postoperative falls, neuropathy, intravascular and intraneural injections, and hematomas.

CONCLUSION

When administered properly, PNBs were a safe and effective adjuvant method of pain control with a significant potential to limit postoperative narcotic use. While blockade choice varies by surgeon preference and procedure, all PNBs should be administered with ultrasound guidance, and vigilant protocols for the risk of postoperative falls should be exercised in patients who receive them.

Predictive Models for Length of Stay and Discharge Disposition in Elective Spine Surgery: Development, Validation, and Comparison to the ACS NSQIP Risk Calculator

STUDY DESIGN

A retrospective study at a single academic institution.

OBJECTIVE

The purpose of this study is to utilize machine learning to predict hospital length of stay (LOS) and discharge disposition following adult elective spine surgery, and to compare performance metrics of machine learning models to the American College of Surgeon's National Surgical Quality Improvement Program's (ACS NSQIP) prediction calculator.

SUMMARY OF BACKGROUND DATA

A total of 3678 adult patients undergoing elective spine surgery between 2014 and 2019, acquired from the electronic health record.

METHODS

Patients were divided into three stratified cohorts: cervical degenerative, lumbar degenerative, and adult spinal deformity groups. Predictive variables included demographics, body mass index, surgical region, surgical invasiveness, surgical approach, and comorbidities. Regression, classification trees, and least absolute shrinkage and selection operator (LASSO) were used to build predictive models. Validation of the models was conducted on 16% of patients (N=587), using area under the receiver operator curve (AUROC), sensitivity, specificity, and correlation. Patient data were manually entered into the ACS NSQIP online risk calculator to compare performance. Outcome variables were discharge disposition (home vs. rehabilitation) and LOS (days).

RESULTS

Of 3678 patients analyzed, 51.4% were male (n=1890) and 48.6% were female (n=1788). The average LOS was 3.66 days. In all, 78% were discharged home and 22% discharged to rehabilitation. Compared with NSQIP (Pearson R2 =0.16), the predictions of poisson regression ( R2 =0.29) and LASSO ( R2 =0.29) models were significantly more correlated with observed LOS ( P =0.025 and 0.004, respectively). Of the models generated to predict discharge location, logistic regression yielded an AUROC of 0.79, which was statistically equivalent to the AUROC of 0.75 for NSQIP ( P =0.135).

CONCLUSION

The predictive models developed in this study can enable accurate preoperative estimation of LOS and risk of rehabilitation discharge for adult patients undergoing elective spine surgery. The demonstrated models exhibited better performance than NSQIP for prediction of LOS and equivalent performance to NSQIP for prediction of discharge location.

Machine learning in clinical diagnosis, prognostication, and management of acute traumatic spinal cord injury (SCI): A systematic review

Background

Machine learning has been applied to improve diagnosis and prognostication of acute traumatic spinal cord injury. We investigate potential for clinical integration of machine learning in this patient population to navigate variability in injury and recovery.

Materials and methods

We performed a systematic review using PRISMA guidelines through PubMed database to identify studies that use machine learning algorithms for clinical application toward improvements in diagnosis, management, and predictive modeling.

Results

Of the 132 records identified, a total of 13 articles met inclusion criteria and were included in final analysis. Of the 13 articles, 5 focused on diagnostic accuracy and 8 were related to prognostication or management of traumatic spinal cord injury. Across studies, 1983 patients with spinal cord injury were evaluated with most classifying as ASIA C or D. Retrospective designs were used in 10 of 13 studies and 3 were prospective. Studies focused on MRI evaluation and segmentation for diagnostic accuracy and prognostication, investigation of mean arterial pressure in acute care and intraoperative settings, prediction of ambulatory and functional ability, chronic complication prevention, and psychological quality of life assessments. Decision tree, random forests (RF), support vector machines (SVM), hierarchical cluster tree analysis (HCTA), artificial neural networks (ANN), convolutional neural networks (CNN) machine learning subtypes were used.

Conclusions

Machine learning represents a platform technology with clinical application in traumatic spinal cord injury diagnosis, prognostication, management, rehabilitation, and risk prevention of chronic complications and mental illness. SVM models showed improved accuracy when compared to other ML subtypes surveyed. Inherent variability across patients with SCI offers unique opportunity for ML and personalized medicine to drive desired outcomes and assess risks in this patient population.

Robust Prediction of Non-home Discharge After Thoracolumbar Spine Surgery With Ensemble Machine Learning and Validation on a Nationwide Cohort

BACKGROUND

Delays in postoperative referrals to rehabilitation or skilled nursing facilities contribute toward extended hospital stays. Facilitating more efficient referrals through accurate preoperative prediction algorithms has the potential to reduce unnecessary economic burden and minimize risk of hospital-acquired complications. We develop a robust machine learning algorithm to predict non-home discharge after thoracolumbar spine surgery that generalizes to unseen populations and identifies markers for prediction.

METHODS

Retrospective electronic health records were obtained from our single-center data warehouse (SCDW) to identify patients undergoing thoracolumbar spine surgeries between 2008 and 2019 for algorithm development and internal validation. The National Inpatient Sample (NIS) database was queried to identify thoracolumbar surgeries between 2009 and 2017 for out-of-sample validation. Ensemble decision trees were constructed for prediction and area under the receiver operating characteristic curve (AUROC) was used to assess performance. Shapley additive explanations values were derived to identify drivers of non-home discharge for interpretation of algorithm predictions.

RESULTS

A total of 5224 cases of thoracolumbar spine surgeries were isolated from the SCDW and 492,312 cases were identified from NIS. The model achieved an AUROC of 0.81 (standard deviation [SD] = 0.01) on the SCDW test set and 0.77 (SD = 0.01) on the nationwide NIS data set, thereby demonstrating robust prediction of non-home discharge across all diverse patient cohorts. Age, total Elixhauser comorbidities, Medicare insurance, weighted Elixhauser score, and female sex were among the most important predictors of non-home discharge.

CONCLUSIONS

Machine learning algorithms reliably predict non-home discharge after thoracolumbar spine surgery across single-center and national cohorts and identify preoperative features of importance that elucidate algorithm decision-making.

Spine Surgery and COVID-19: The Influence of Practice Type on Preparedness, Response, and Economic Impact

STUDY DESIGN

Cross-sectional observational cohort study.

OBJECTIVE

To investigate preparation, response, and economic impact of COVID-19 on private, public, academic, and privademic spine surgeons.

METHODS

AO Spine COVID-19 and Spine Surgeon Global Impact Survey includes domains on surgeon demographics, location of practice, type of practice, COVID-19 perceptions, institutional preparedness and response, personal and practice impact, and future perceptions. The survey was distributed by AO Spine via email to members (n = 3805). Univariate and multivariate analyses were performed to identify differences between practice settings.

RESULTS

A total of 902 surgeons completed the survey. In all, 45.4% of respondents worked in an academic setting, 22.9% in privademics, 16.1% in private practice, and 15.6% in public hospitals. Academic practice setting was independently associated with performing elective and emergent spine surgeries at the time of survey distribution. A majority of surgeons reported a >75% decrease in case volume. Private practice and privademic surgeons reported losing income at a higher rate compared with academic or public surgeons. Practice setting was associated with personal protective equipment availability and economic issues as a source of stress.

CONCLUSIONS

The current study indicates that practice setting affected both preparedness and response to COVID-19. Surgeons in private and privademic practices reported increased worry about the economic implications of the current crisis compared with surgeons in academic and public hospitals. COVID-19 decreased overall clinical productivity, revenue, and income. Government response to the current pandemic and preparation for future pandemics needs to be adaptable to surgeons in all practice settings.

Reaching the medicare allowable threshold in adult spinal deformity surgery: multicenter cost analysis comparing actual direct hospital costs versus what the government will pay

STUDY DESIGN

Retrospective multicenter cost analysis.

OBJECTIVE

To (1) determine if index episode of care (iEOC) costs of Adult Spinal Deformity (ASD) surgeries are below the Medicare Allowable (MA) threshold, and (2) identify variables that can predict iEOC cases that are below MA. Previous studies have suggested that actual direct hospital cost of Adult Spinal Deformity (ASD) surgery is higher than Medicare Allowable (MA) rates, which has become the benchmark reimbursement target for hospital accounting systems.

METHODS

From a prospective, multicenter ASD surgical database, patients undergoing long instrumented fusions (> 5 level) with cost data were identified. iEOC cost was calculated utilizing actual direct hospital cost. MA rates were calculated using hospital specific, year-appropriate CMS Inpatient Pricer Payment System. Recursive partitioning identified potentially modifiable variables that can predict iEOC cost < MA.

RESULTS

Administrative direct cost data from 210 patients were obtained from 4 of 11 centers. Ninety-five (45%) patients had iEOC cost < MA. There was significant variation across the four centers in both iEOC cost ($56,788-$78,878, p < 0.0001) and reimbursement ($40,623-$91,351, p < 0.0001) across deformity-specific DRGs (453,454,456,457). Academic centers were more likely to have iEOC costs < MA (67.2% vs 8.9%, p < 0.0001). Recursive partitioning (r² = 0.309) identified rhBMP-2 use of < 24 mg, sagittal plane deformity, a combined anterior/posterior approach, and an SF36-MCS < 39 as predictive for iEOC cost < MA. Performing an anterior/posterior approach reimburses between 14.7% and 121.1% more (2.2-fold) than posterior-only approach. This change in DRG allows iEOC cost to be more likely below the MA threshold.

CONCLUSION

There is significant institutional (private vs academic) variation in ASD reimbursement. BMP use, deformity type, approach, and baseline mental health impact ASD surgery cost being below Medicare reimbursement. ASD surgeries with anterior/posterior approaches are in DRGs that can potentially reimburse 2.2-fold the posterior-only surgery, making it more likely to fall below the MA threshold.

LEVEL OF EVIDENCE

III.

Supervised Machine Learning for Predicting Length of Stay After Lumbar Arthrodesis: A Comprehensive Artificial Intelligence Approach

INTRODUCTION

Few studies have evaluated the utility of machine learning techniques to predict and classify outcomes, such as length of stay (LOS), for lumbar fusion patients. Six supervised machine learning algorithms may be able to predict and classify whether a patient will experience a short or long hospital LOS after lumbar fusion surgery with a high degree of accuracy.

METHODS

Data were obtained from the National Surgical Quality Improvement Program between 2009 and 2018. Demographic and comorbidity information was collected for patients who underwent anterior, anterolateral, or lateral transverse process technique arthrodesis procedure; anterior lumbar interbody fusion (ALIF); posterior, posterolateral, or lateral transverse process technique arthrodesis procedure; posterior lumbar interbody fusion/transforaminal lumbar interbody fusion (PLIF/TLIF); and posterior fusion procedure posterior spine fusion (PSF). Machine learning algorithmic analyses were done with the scikit-learn package in Python on a high-performance computing cluster. In the total sample, 85% of patients were used for training the models, whereas the remaining patients were used for testing the models. C-statistic area under the curve and prediction accuracy (PA) were calculated for each of the models to determine their accuracy in correctly classifying the test cases.

RESULTS

In total, 12,915 ALIF patients, 27,212 PLIF/TLIF patients, and 23,406 PSF patients were included in the algorithmic analyses. The patient factors most strongly associated with LOS were sex, ethnicity, dialysis, and disseminated cancer. The machine learning algorithms yielded area under the curve values of between 0.673 and 0.752 (PA: 69.6% to 80.1%) for ALIF, 0.673 and 0.729 (PA: 66.0% to 81.3%) for PLIF/TLIF, and 0.698 and 0.749 (PA: 69.9% to 80.4%) for PSF.

CONCLUSION

Machine learning classification algorithms were able to accurately predict long LOS for ALIF, PLIF/TLIF, and PSF patients. Supervised machine learning algorithms may be useful in clinical and administrative settings. These data may additionally help inform predictive analytic models and assist in setting patient expectations.

LEVEL III

Diagnostic study, retrospective cohort study.

Postoperative Outcomes Following Posterior Lumbar Fusion in Patients With Multiple Sclerosis

SUMMARY OF BACKGROUND

Multiple sclerosis (MS) is a chronic inflammatory disease that can cause physical and neurological dysfunction. Patients with MS are living longer and undergoing more orthopedic procedures, but the risk of patients with MS undergoing posterior lumbar fusion (PLF) has not been studied in literature before.

OBJECTIVE

This study aims to (1) analyze the rates of postoperative complications of MS patients undergoing primary PLF and (2) analyze the economic burden associated with these surgeries in the MS population compared with patients without MS.

METHODS

A retrospective review of the Medicare database was conducted on all patients who underwent PLF and/or posterior lumbar interbody fusion between the years of 2006 and 2013. Cases involving same-day anterior or revision procedures and patients with a history of spine, infection, and trauma, or neoplasm were excluded from the study. Demographics, comorbidities, 90-day postoperative complications, cost, and length of stay were calculated. All outcomes of interest were analyzed using multivariate logistic regression, adjusting for age, sex, and comorbidity burden. Significance was defined as P-value <0.05.

RESULTS

There were 2363 patients with MS and 23,569 matched controlled patients. We found a significant increase in the risk of sepsis [odds ratio (OR)=1.85, P=0.034], urinary tract infection (OR=1.89, P<0.001), deep vein thrombosis (OR=1.4, P=0.044), 90-day emergency room visit (OR=1.14, P=0.027), and 90-day readmissions (OR=1.20, P=0.011) compared with patients without a history of MS. Patients with MS also incurred $4379 extra in total hospital charge, a $1679 increase in the cost of hospitalization, and an increase in length of stay (4.05 vs. 3.61, P<0.001).

CONCLUSIONS

A diagnosis of MS is associated with a significant increase in postoperative complications and higher costs after hospitalization. It is imperative for physicians to understand the risk factors of patients undergoing PLF and/or posterior lumbar interbody fusion with MS to better counsel them about postoperative complications before surgery.

LEVEL OF EVIDENCE

Level III.

90-Day Bundled Payment Simulation, Health Care Utilization, and Complications following Craniopharyngioma Resection in Adult Patients

Context Bundled payment and health care utilization models inform cost optimization and surgical outcomes. Economic analysis of payment plans for craniopharyngioma resection is unknown.

Objective This study aimed to identify impact of endocrine and nonendocrine complications (EC and NEC, respectively) on health care utilization and bundled payments following craniopharyngioma resection.

Design This study is presented as a retrospective cohort analysis (2000–2016) with 2 years of follow-up.

Setting The study included national inpatient hospitalization and outpatient visits.

Patients Patients undergoing craniopharyngioma resection were divided into the following four groups: group 1, no complications (NC); group 2, only EC; group 3, NEC; and group 4, both endocrine and nonendocrine complications (ENEC).

Interventions This study investigated transphenoidal or subfrontal approach for tumor resection.

Main Outcome Hospital readmission, health care utilization up to 24 months following discharge, and 90-day bundled payment performances are primary outcomes of this study.

Results Median index hospitalization payments were significantly lower for patients in NC cohort ($28,672) compared with those in EC ($32,847), NEC ($36,259), and ENEC ($32,596; p < 0.0001). Patients in ENEC incurred higher outpatient services and overall median payments at 6 months (NC: 38,268; EC: 49,844; NEC: 68,237; and ENEC: 81,053), 1 year (NC: 46,878; EC: 58,210; NEC: 81,043; and ENEC: 94,768), and 2 years (NC: 58,391; EC: 70,418; NEC: 98,838; and ENEC: 1,11,841; p < 0.0001). The 90-day median bundled payment was significantly different among the cohorts with the highest in ENEC ($60,728) and lowest in the NC ($33,089; p < 0.0001).

Conclusion ENEC following surgery incurred almost two times the overall median payments at 90 days, 6 months, 1 year. and 2 years compared with those without complications. Bundled payment model may not be a feasible option in this patient population. Type of complications and readmission rates should be considered to optimize payment model prediction following craniopharyngioma resection.

Is Uncontrolled Diabetes Mellitus Associated with Incidence of Complications After Posterior Instrumented Lumbar Fusion? A National Claims Database Analysis

BACKGROUND

Previous research has shown that diabetes mellitus (DM) is associated with postoperative complications, including surgical site infections (SSIs). However, evidence for the association between diabetes control and postoperative complications in patients with DM is mixed. Prior studies relied on a single metric for defining uncontrolled DM, which does not account for glycemic variability, and it is unknown whether a more comprehensive assessment of diabetes control is associated with postoperative complications.

QUESTIONS/PURPOSES

(1) Is there a difference in the incidence of SSI after lumbar spine fusion in patients with uncontrolled DM, defined with a comprehensive assessment of glycemic control, compared with patients with controlled DM? (2) Is there a difference in the incidence of other select postoperative complications after lumbar spine fusion in patients with uncontrolled DM compared with patients with controlled DM? (3) Is there a difference in total reimbursements between these groups?

METHODS

We used the PearlDiver Patient Records Database, a national administrative claims database that provides access to the full continuum of perioperative care. We included 46,490 patients with DM undergoing posterior lumbar fusion with instrumentation. Patients were required to be continuously enrolled in the database for at least 1 year before and 90 days after the index procedure. Patients were divided into uncontrolled and controlled DM cohorts, as defined by ICD-9 diagnostic codes. These are based on a comprehensive assessment of glycemic control, including consideration of patient self-monitoring of blood glucose levels, hemoglobin A1c, and the presence/severity of diabetes-related comorbidities. The cohorts differed only by age, insurance type, and Elixhauser comorbidity score. The primary outcome was the incidence of SSI, divided into superficial and deep, within 90 days postoperatively. Secondary complications included the incidence of cerebrovascular events, acute kidney injury, pulmonary embolism, pneumonia, urinary tract infection, blood transfusion, and total reimbursements. These are the sum of reimbursements occurring within 90 days of surgery, which capture the total professional and facility cost burden to the health payer (such as the insurer). We constructed multivariable logistic regression models to adjust for the effects of age, insurance type, and comorbidities.

RESULTS

After adjusting for potentially confounding variables including age, insurance type, and comorbidities, we found that patients with uncontrolled DM had an odds ratio for deep SSI of 1.52 (95% confidence interval 1.16 to 1.95; p = 0.002). Similarly, patients with uncontrolled DM had adjusted odds ratios of 1.25 (95% CI 1.01 to 1.53; p = 0.03) for cerebrovascular events, 1.36 (95% CI 1.18 to 1.57; p < 0.001) for acute kidney injury, 1.55 (95% CI 1.16 to 2.04; p = 0.002) for pulmonary embolism, 1.30 (95% CI 1.08 to 1.54; p = 0.004) for pneumonia, 1.33 (95% CI 1.19 to 1.49; p < 0.001) for urinary tract infection, and 1.27 (95% CI 1.04 to 1.53; p = 0.02) for perioperative transfusion. Patients with uncontrolled DM had higher median 90-day total reimbursements than patients with controlled DM: USD 27,915 (interquartile range 5472 to 63,400) versus USD 10,263 (IQR 4101 to 49,748; p < 0.001).

CONCLUSION

Our findings encourage surgeons to take a full diabetic history beyond the HbA1c value, including any self-monitoring of glucose measurements, time in acceptable range for continuous glucose monitors, and/or consideration of the presence/severity of diabetes-related complications before lumbar spine fusion, as HbA1c does not fully capture glycemic control or variability. We emphasize that uncontrolled DM is a clinical, rather than laboratory, diagnosis. Comprehensive diabetes histories should be incorporated into existing preoperative diabetes care pathways and elective surgery could be deferred to improve glycemic control. Future development of an index measure incorporating multidimensional measures of diabetes control (such as continuous or self-glucose monitoring, diabetes-related comorbidities) is warranted.

LEVEL OF EVIDENCE

Level III, therapeutic study.

The Impact of Non-Elective Admission on Cost of Care and Length of Stay in Anterior Cervical Discectomy and Fusion: A Propensity-Matched Analysis

STUDY DESIGN

Retrospective cohort study.

OBJECTIVE

To evaluate the impact of admission status on patient outcomes and healthcare costs in anterior cervical discectomy and fusion (ACDF).

SUMMARY OF BACKGROUND DATA

Undergoing ACDF non-electively has been associated with higher patient comorbidity burdens. However, the impact of non-elective status on the total cost of hospital stay has yet to be quantified.

METHODS

Patients undergoing ACDF at a single institution were placed into elective or non-elective cohorts. Propensity score-matching analysis in a 5:1 ratio controlling for insurance type and comorbidities was used to minimize selection bias. Demographics were compared by univariate analysis. Cost of care, length of stay (LOS), and clinical outcomes were compared between groups using multivariable linear and logistic regression with elective patients as reference cohort. All analyses controlled for sex, preoperative diagnosis, elixhauser comorbidity index (ECI), age, length of surgery, number of segments fused, and insurance type.

RESULTS

Of 708 patients in the final ACDF cohort, 590 underwent an elective procedure and 118 underwent a non-elective procedure. The non-elective group was significantly younger (53.7 vs. 49.5 yr; P = 0.0007). Cohorts had similar proportions of private versus public health insurance, although elective had higher rates of commercial insurance (39.22% vs. 15.25%; P < 0.0001) and non-elective had higher rates of managed care (32.77% vs. 56.78%; P < 0.0001). Operation duration was significantly longer in non-elective patients (158 vs. 177 minutes; P = 0.01). Adjusted analysis also demonstrated that admission status independently affected cost (+$6877, 95% confidence interval [CI]: $4906-$8848; P < 0.0001) and LOS (+4.9 days, 95% CI: 3.9-6.0; P < 0.0001) for the non-elective cohort. The non-elective cohort was significantly more likely to return to the operating room (OR: 3.39; 95% CI: 1.37-8.36, P = 0.0008) and experience non-home discharge (OR: 10.95; 95% CI: 5.00-24.02, P < 0.0001).

CONCLUSION

Patients undergoing ACDF non-electively had higher cost of care and longer LOS, as well as higher rates of postoperative adverse outcomes.Level of Evidence: 3.

Immediate Postpartum Long-Acting Reversible Contraception: Review of Insertion and Device Reimbursement Policies

BACKGROUND

Previous assessment of statewide policies on long-acting reversible contraception (LARC) indicate that an increasing number of states are implementing policies specifically for provision immediately postpartum, supported by current clinical guidelines. Less is known about how state policies describe payment methodologies for the insertion procedure and device costs.

METHODS

We conducted a systematic, web-based review of publicly available statewide policy language on immediate postpartum LARC among all 50 states. We examined the payor/s identified in the policy and policy type, if the policy included language on the global obstetric fee, whether providers and/or facilities were authorized to bill for procedure or device costs, and if the billing mechanism was identified as inpatient and/or outpatient services.

RESULTS

Three-fourths of states (76%; n = 38) had statewide policies on immediate postpartum LARC. All policies identified Medicaid as the payor, although two also included non-Medicaid plans. Language allowing for reimbursement separate from the global obstetric fee for insertion procedures was present in 76% of states; 23 states permit it and 6 do not. Device cost reimbursement separate from the fee was identified in more state policies (92%); 31 states allow it and 4 do not. More policies included inpatient or outpatient billing mechanisms for device costs (82%; n = 31) than insertion procedures (50%; n = 19).

CONCLUSIONS

Medicaid reimbursement policies for immediate postpartum LARC services vary by state reimbursement process, type, and mechanism. Observed differences indicate payment methodologies more often include the cost of the device than provider reimbursement (31 states vs. 23 states). Fewer than one-half of states offer reimbursement for provider insertion fees, a significant systems barrier to contraceptive access for women who choose LARC immediately postpartum.

The Impact of Physical Therapy Following Cervical Spine Surgery for Degenerative Spine Disorders: A Systematic Review

STUDY DESIGN

Systematic review.

OBJECTIVE

To characterize the effects of postoperative physical therapy (PT) after surgery for cervical spondylosis on patient-reported outcomes and impairments. Secondarily, to identify associated complications, adverse effects, and health care costs with postoperative PT, and to describe the content, timing, and duration of the PT.

SUMMARY OF BACKGROUND DATA

Cervical spine surgery is common; however, it is unclear if the addition of postoperative PT leads to improved patient outcomes and decreased health care costs.

MATERIALS AND METHODS

PubMed, Embase, Cochrane Central Register of Controlled Trials, Physiotherapy Evidence Database, and Web of Science were searched until July 2019. All peer-reviewed articles involving cervical spine surgery with postoperative PT for cervical spondylosis were considered for inclusion. Risk of bias was assessed using the Revised Cochrane risk-of-bias tool for randomized trials. Findings were described narratively, and GRADE approach was used to define the quality of evidence.

RESULTS

A total of 10,743 studies were screened. Six studies met inclusion criteria; 2 randomized controlled trials and 4 subsequent follow-up studies containing study arms that included postoperative PT after cervical spine surgery. Meta-analysis was not performed due to study heterogeneity and no study compared PT+surgery to surgery alone. PT treatment included exercise therapy, cognitive behavioral therapy, and optional vestibular rehabilitation. Included studies indicated PT appeared to have positive effects on patient outcomes, however, there were no treatment control groups and the quality of evidence was very low to low. Timing, duration, and content of PT programs varied. No studies reported complications, adverse effects, or cost-effectiveness relating to PT after surgery.

CONCLUSIONS

Current literature prevents a definitive conclusion regarding the impact of postoperative PT, given the lack of treatment control groups. PT treatment was limited to exercise therapy, cognitive behavioral therapy, and optional vestibular rehabilitation in the included studies. PT treatment varied, limiting consistent recommendations for content, timing, and treatment duration. Controlled trials are needed to determine the effectiveness of the addition of postoperative PT following cervical spine surgery for cervical spondylosis.

LEVEL OF EVIDENCE

Level II.

Recent Trends in Medicare Utilization and Reimbursement for Orthopaedic Procedures Performed at Ambulatory Surgery Centers

BACKGROUND

As part of a market-driven response to the increasing costs of hospital-based surgical care, an increasing volume of orthopaedic procedures are being performed in ambulatory surgery centers (ASCs). The purpose of the present study was to identify recent trends in orthopaedic ASC procedure volume, utilization, and reimbursements in the Medicare system between 2012 and 2017.

METHODS

This cross-sectional, national study tracked annual Medicare claims and payments and aggregated data at the county level. Descriptive statistics and multivariate regression models were used to evaluate trends in procedure volume, utilization rates, and reimbursement rates, and to identify demographic predictors of ASC utilization.

RESULTS

A total of 1,914,905 orthopaedic procedures were performed at ASCs in the Medicare population between 2012 and 2017, with an 8.8% increase in annual procedure volume and a 10.5% increase in average reimbursements per case. ASC orthopaedic procedure utilization, including utilization across all subspecialties, is strongly associated with metropolitan areas compared with rural areas. In addition, orthopaedic procedure utilization, including for sports and hand procedures, was found to be significantly higher in wealthier counties (measured by average household income) and in counties located in the South.

CONCLUSIONS

This study demonstrated increasing orthopaedic ASC procedure volume in recent years, driven by increases in hand procedure volume. Medicare reimbursements per case have steadily risen and outpaced the rate of inflation over the study period. However, as orthopaedic practice overhead continues to increase, other Medicare expenditures such as hospital payments and operational and implant costs also must be evaluated. These findings may provide a source of information that can be used by orthopaedic surgeons, policy makers, investors, and other stakeholders to make informed decisions regarding the costs and benefits of the use of ASCs for orthopaedic procedures.

Cost analysis of primary single-level lumbar discectomies using the Value Driven Outcomes database in a large academic center

BACKGROUND CONTEXT

Improving value is an established point of emphasis to reduce the rapidly rising health care costs in the United States. Back pain is a major driver of costs with a substantial fraction caused by lumbar radiculopathy. The most common surgical treatment for lumbar radiculopathy is microdiscectomy. Research is sparse regarding variables driving cost in microdiscectomies and often limited by cost data derived from payer-based Medicare data.

PURPOSE

To identify targets for cost reduction by determining variables associated with significant cost variation in microdiscectomies, using cost data derived from the Value Driven Outcomes tool and actual system costs.

STUDY DESIGN

Single-center, retrospective study of prospectively collected registry data.

PATIENT SAMPLE

Six hundred twenty-two patients identified by CPT code and manually screened for initial, unilateral, single-level lumbar discectomy performed between 2014 and 2018 at a single institution.

OUTCOME MEASURES

Primary outcome measures include total direct cost, clinical length of stay, and OR minutes. Total Direct Cost was further differentiated into facility and nonfacility costs.

METHODS

Univariate and multivariate generalized linear models (GLM) were used to identify variables associated with variation in primary outcome measures. Costs were normalized by mean cost for patients with normal body mass index (BMI) and a healthy American Society of Anesthesiologists (ASA) classification. Average marginal effects were reported as percentage of normalized costs.

RESULTS

Advanced age, male gender, Hispanic, black, unemployment, obesity, higher ASA class, insurance status, and being retired were positively associated with costs in univariate analysis. Asian, Native American, outpatient procedures, and being a student were associated with decreases in costs. In multivariate analysis, we found that obesity led to higher average marginal total direct (9%), total facility (15%), and facility OR costs (22%), as well as 24 more OR minutes per surgery. While being overweight was not associated with greater total direct costs, it was associated with higher total facility (8%), and facility OR costs (12%), with 11 more OR minutes per surgery. Age was associated with a longer LOS but not with OR costs. As expected, outpatient surgical costs, LOS, and OR time were significantly lower than inpatient procedures. Severe systematic disease was associated with greater total and nonfacility costs. In addition, Medicare patients had higher facility costs (14%) compared to privately insured patients.

CONCLUSIONS

Significant drivers of total direct cost in multivariate GLM analysis were obesity, severe systemic disease and inpatient surgery. Average LOS was increased due to age and inpatient status, conversely it was decreased by unemployment and retirement. Significant variables in OR time were male sex, Hispanic race and both obese and overweight BMIs.

The Influence of Body Mass Index in Obese and Morbidly Obese Patients on Complications and 30- and 90-day Readmissions Following Lumbar Spine Fusion

STUDY DESIGN

Retrospective cohort study using the 2013-2017 National Readmission Database.

OBJECTIVE

The aim of this study was to quantify the influence of body mass index (BMI) on complication and readmission rates following lumbar spine fusion.

SUMMARY OF BACKGROUND DATA

Compared to controls, patients with BMI ≥35 had greater odds of readmission, infection, and wound complications following lumbar spine fusion.

METHODS

Patients who underwent elective lumbar spine fusion within the population-based sample were considered for inclusion. Exclusion criteria included nonelective lumbar spine fusions, malnourished, anorexic, or underweight patients, and surgical indications of trauma or neoplasm. Patients were grouped by BMI: 18.5 to 29.9 (controls), 30 to 34.9 (obesity I), 35 to 39.9 (obesity II), and ≥40 (obesity III). Multivariate regression was performed to analyze differences in complications and readmissions between groups. Predictive modeling was conducted to estimate the impact of BMI on 30- and 90-day infection, wound complication, and readmissions rates.

RESULTS

A total of 86,697 patients were included for analysis, with an average age of 58.9 years and 58.9% being female. The obesity II group had significantly higher odds of infection (odds ratio [OR]: 1.82, 95% confidence interval [CI]: 1.28-2.62, P = 0.001), wound dehiscence (OR: 3.08, 95% CI: 1.70-6.18, P = 0.0006), and 30-day readmission (OR: 1.32, 95% CI: 1.11-1.58, P = 0.002), whereas the obesity III group had significantly higher odds of acute renal failure (OR: 2.14, 95% CI: 1.20-4.06, P = 0.014), infection (OR: 2.43, 95% CI: 1.72-3.48, P < 0.0001), wound dehiscence (OR: 3.76, 95% CI: 2.08-7.51, P < 0.0001), 30-day readmission (OR: 1.62, 95% CI: 1.36-1.93, P < 0.0001), and 90-day readmission (OR: 1.53, 95% CI: 1.31-1.79, P < 0.0001) compared with controls. Predictive modeling showed cumulative increases of 6.44% in infection, 3.69% in wound dehiscence, and 1.35% in readmission within 90-days for each successive BMI cohort.

CONCLUSION

Progressively higher risks for infection, wound complications, and hospital readmission were found with each progressive BMI level.Level of Evidence: 3.

Prevalence and Outcomes of Major Psychiatric Disorders Preceding Index Surgery for Degenerative Thoracic/Lumbar Spine Disease

The relationship between preexisting major psychiatric disorders and outcomes of spine surgery for degenerative thoracic/lumbar disease remains unclear. A 5% subset of inpatients was randomly selected from the Taiwan National Health Insurance Research Database. A total of 10,109 inpatients aged 18 years or over with degenerative thoracic/lumbar disease and underwent spine surgery met inclusion criteria. Major psychiatric disorders diagnosed by psychiatrists preceding index surgery, including anxiety disorder, depression disorder, bipolar disorder, schizophrenia and dementia, were identified. The prevalence of psychiatric disorders, and their differential risks on in-hospital and post-discharge outcomes were examined. 10.4% had major psychiatric disorders, of which depression (6.6%) and anxiety (4.9%) were most common. Logistic regression revealed increased risks of ventilator use in depression (OR = 1.62, 95% CI = 1.04-2.54, p < 0.05), extended hospitalization length in bipolar (OR = 1.77, 95% CI = 1.08-2.89, p < 0.05), and higher rehabilitation utilization in depression (OR = 1.25, 95% CI = 1.06-1.47, p < 0.01) and bipolar (OR = 1.69, 95% CI = 1.04-2.76, p < 0.05). Those patients with anxiety had a decreased risk of longer hospitalization duration (OR = 0.77, 95% CI = 0.60-0.98, p < 0.05), while those with dementia and schizophrenia had no change in risks. Preoperative recognition of major psychiatric disorders for risk and treatment assessment is suggested as people with preexisting depression or bipolar disorder have worse outcomes after spine surgery.

[Cost coverage for complex paediatric spinal deformities with preoperative halo traction-an economic problem?]

BACKGROUND

Spinal surgery is largely reimbursed in a differentiated manner via the DRG system. For treatments of complex paediatric deformities with increased pre and postoperative effort due to special treatment approaches, it seems that the costs for the treatment are not fully covered.

MATERIALS AND METHODS

All paediatric cases with surgical treatment of the spine that were treated in a single spine centre from 2018-2020 were considered. The subgroup of patients with inpatient halo-gravity traction (halo group) before surgery was compared with all other cases treated in terms of economic and demographic factors.

RESULTS

There were 86 cases that were treated surgically without halo traction and 6 cases with halo traction. The groups did not differ significantly in age (p = 0.41) or Patient Clinical Complexity Level (PCCL, p = 0.76). The average length of hospital stay in the halo group was significantly longer than in the other cases (84.2 ± 40.1 d vs. 11.0 ± 6.4 d; p = 0.001). Due to DRG grouping and long-stay surcharges, the mean revenue per case was significantly higher in the halo group than in the other cases (€ 63,615 ± 45,138 vs. € 16,836 ± 9356) (p = 0.003). The contribution margin for the period of the long-term surcharges varied between 11,394 and 9766 €. The high additional costs due to the necessary medical devices of halo traction were not sufficiently reflected in the reimbursement.

CONCLUSION

Paediatric spine surgery can be challenging in special cases. In particular, severe deformities of the spine may require additional procedures. The subgroup of patients requiring preoperative halo traction is not adequately compensated by the DRG system.

Using machine learning methods to predict nonhome discharge after elective total shoulder arthroplasty

Background

Machine learning has shown potential in accurately predicting outcomes after orthopedic surgery, thereby allowing for improved patient selection, risk stratification, and preoperative planning. This study sought to develop machine learning models to predict nonhome discharge after total shoulder arthroplasty (TSA).

Methods

The American College of Surgeons National Surgical Quality Improvement Program database was queried for patients who underwent elective TSA from 2012 to 2018. Boosted decision tree and artificial neural networks (ANN) machine learning models were developed to predict non-home discharge and 30-day postoperative complications. Model performance was measured using the area under the receiver operating characteristic curve (AUC) and overall accuracy (%). Multivariate binary logistic regression analyses were used to identify variables that were significantly associated with the predicted outcomes.

Results

There were 21,544 elective TSA cases identified in the National Surgical Quality Improvement Program registry from 2012 to 2018 that met inclusion criteria. Multivariate logistic regression identified several variables associated with increased risk of nonhome discharge including female sex (odds ratio [OR] = 2.83; 95% confidence interval [CI] = 2.53-3.17; P < .001), age older than 70 years (OR = 3.19; 95% CI = 2.86-3.57; P < .001), American Society of Anesthesiologists classification 3 or greater (OR = 2.70; 95% CI = 2.41-2.03; P < .001), prolonged operative time (OR = 1.38; 95% CI = 1.20-1.58; P < .001), as well as history of diabetes (OR = 1.56; 95% CI = 1.38-1.75; P < .001), chronic obstructive pulmonary disease (OR = 1.71; 95% CI = 1.46-2.01; P < .001), congestive heart failure (OR = 2.65; 95% CI = 1.72-4.01; P < .001), hypertension (OR = 1.35; 95% CI = 1.20-1.52; P = .004), dialysis (OR = 3.58; 95% CI = 2.01-6.39; P = .002), wound infection (OR = 5.67; 95% CI = 3.46-9.29; P < .001), steroid use (OR = 1.43; 95% CI = 1.18-1.74; P = .010), and bleeding disorder (OR = 1.84; 95% CI = 1.45-2.34; P < .001). The boosted decision tree model for predicting nonhome discharge had an AUC of 0.788 and an overall accuracy of 90.3%. The ANN model for predicting nonhome discharge had an AUC of 0.851 and an overall accuracy of 89.9%. For predicting the occurrence of 1 or more postoperative complications, the boosted decision tree model had an AUC of 0.795 and an overall accuracy of 95.5%. The ANN model yielded an AUC of 0.788 and an overall accuracy of 92.5%.

Conclusions

Both the boosted decision tree and ANN models performed well in predicting nonhome discharge with similar overall accuracy, but the ANN had higher discriminative ability. Based on the findings of this study, machine learning has the potential to accurately predict nonhome discharge after elective TSA. Surgeons can use such tools to guide patient expectations and to improve preoperative discharge planning, with the ultimate goal of decreasing hospital length of stay and improving cost-efficiency.

Predictors of Readmission and Prolonged Length of Stay After Cervical Disc Arthroplasty

STUDY DESIGN

Retrospective cohort study.

OBJECTIVE

The aim of the study was to assess which factors increase risk of readmission within 30 days of surgery or prolonged length of stay (LOS) (≥2 days) after cervical disc arthroplasty (CDA).

SUMMARY OF BACKGROUND DATA

Several studies have shown noninferiority at mid- and long-term outcomes after cervical disc arthroplasty (CDA) compared to anterior cervical discectomy and fusion ACDF, but few have evaluated short-term outcomes regarding risk of readmission or prolonged LOS after surgery.

METHODS

Demographics, comorbidities, operative details, postoperative complications, and perioperative outcomes were collected for patients undergoing single level CDA in the National Surgical Quality Improvement Program (NSQIP) database. Patients with prolonged LOS, defined as >2 days, and readmission within 30 days following CDA were identified. Univariable and multivariable logistic regression models were used to identify risk factors for prolonged LOS and readmission.

RESULTS

A total of 3221 patients underwent single level CDA. Average age was 45.6 years (range 19-82) and 53% of patients were male. A total of 472 (14.7%) experienced a prolonged LOS and 36 (1.1%) patients were readmitted within 30 days following surgery. Predictors of readmission were postoperative superficial wound infection (odds ratio [OR] = 73.83, P < 0.001), American Society of Anesthesiologists (ASA) classification (OR = 1.98, P = 0.048), and body mass index (BMI) (OR = 1.06, P = 0.02). Female sex (OR = 1.76, P < 0.001), diabetes (OR = 1.50, P = 0.024), postoperative wound dehiscence (OR = 13.11, P = 0.042), ASA class (OR = 1.43, P < 0.01), and operative time (OR = 1.01, P < 0.001) were significantly associated with prolonged LOS.

CONCLUSION

From a nationwide database analysis of 3221 patients, wound complications are predictors of both prolonged LOS and readmission. Patient comorbidities, including diabetes, higher ASA classification, female sex, and higher BMI also increased risk of prolonged LOS or readmission.Level of Evidence: 3.

Bundled Payment Models in Spine Surgery

STUDY DESIGN

The following is a narrative discussion of bundled payments in spine surgery.

OBJECTIVE

The cost of healthcare in the United States has continued to increase. To lower the cost of healthcare, reimbursement models are being investigated as potential cost saving interventions by driving incentives and quality improvement in fields such a spine surgery.

METHODS

Narrative overview of literature pertaining to bundled payments in spine surgery synthesizing findings from computerized databases and authoritative texts.

RESULTS

Spine surgery is challenging to define payment modes because of high cost variability and surgical decision-making nuances. While implementing bundled care payments in spine surgery, it is important to understand concepts such as value-based purchasing, episodes of care, prospective versus retrospective payment models, one versus two-sided risk, risk adjustment, and outlier protection. Strategies for implementation underscore the importance of risk stratification and modeling, adoption of evidence based clinical pathways, and data collection and dissemination. While bundled care models have been successfully implemented, challenges facing institutions adopting bundled care payment models include financial stressors during adoption of the model, distribution of risks, incentivization of treating only low risk patients, and nuanced variation in procedures leading to variation in costs.

CONCLUSION

An alternative for fee for service payments, bundled care payments may lead to higher cost savings and surgeon accountability in a patient's care.

Comparison of Surgical Outcomes of Microdiskectomy Procedures by Patient Admission Status

OBJECTIVE

We sought to compare the cost and in-hospital outcomes following lumbar microdiskectomy procedures by admission type.

METHODS

Patients undergoing lumbar microdiskectomy at a single institution from 2008 to 2016 following an elective admission (EL) were compared against those who were admitted from the emergency department (ED) or from elsewhere within or outside the hospital system (TR) for their perioperative outcomes and cost. Multivariable modeling controlled for age, sex, self-reported race, Elixhauser comorbidity score, payer type, number of segments, and procedure length.

RESULTS

Of the 1249 patients included in this study, 1116 (89.4%) were admitted electively while 123 (9.8%) were admitted from the ED and 10 (0.8%) were transferred from other hospitals. EL patients had significantly lower comorbidity burdens (P < 0.0001). Univariate and multivariable analyses revealed that transfer admission patients experienced significantly longer hospitalizations (ED: +1.7 days; P < 0.0001; TR: +5.3 days; P < 0.0001) and higher direct costs (ED: $1889; P < 0.0001; TR: $7001; P < 0.0001) compared with EL patients. Despite these risks, ED and TR patients only had increased odds of nonhome discharge compared with EL patients (ED: 3.4; P = 0.002; TR: 7.9; P = 0.02).

CONCLUSIONS

Patients admitted as transfers and from the ED had significantly increased hospitalization lengths of stay and direct costs compared with electively admitted patients.

Strategies for spinal surgery reimbursement: bundling in the working-age population

Introduction

Bundled payments for spine surgery, which is known for having high overall cost with wide variation, have been previously studied in older adults. However, there has been limited work examining bundled payments in working-age patients. We sought to identify the variation in the cost of spine surgery among working age adults in a large, national insurance claims database.

Methods

We queried the TRICARE claims database for all patients, aged 18–64, undergoing cervical and non-cervical spinal fusion surgery between 2012 and 2014. We calculated the case mix adjusted, price standardized payments for all aspects of care during the 60-, 90-, and 180-day periods post operation. Variation was assessed by stratifying Hospital Referral Regions into quintiles.

Results

After adjusting for case mix, there was significant variation in the cost of both cervical ($10,538.23, 60% of first quintile) and non-cervical ($20,155.59, 74%). Relative variation in total cost decreased from 60- to 180-days (63 to 55% and 76 to 69%). Index hospitalization was the primary driver of costs and variation for both cervical (1st-to-5th quintile range: $11,033–$19,960) and non-cervical ($18,565–$36,844) followed by readmissions for cervical ($0–$11,521) and non-cervical ($0–$13,932). Even at the highest quintile, post-acute care remained the lowest contribution to overall cost ($2070 & $2984).

Conclusions

There is wide variation in the cost of spine surgery across the United States for working age adults, driven largely by index procedure and readmissions costs. Our findings suggest that implementing episodes longer than the current 90-day standard would do little to better control cost variation.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12913-021-06112-0.

Recent trends in Medicare utilization and surgeon reimbursement for shoulder arthroplasty

BACKGROUND

Recent efforts to contain health care costs and move toward value-based health care have intensified, with a continued focus on Medicare expenditures, especially for high-volume procedures. As total shoulder arthroplasty (TSA) volume continues to increase, especially within the Medicare population, it is important for orthopedic surgeons to understand recent trends in the allocation of health care expenditures and potential effects on reimbursements. The purpose of this study was to evaluate trends in annual Medicare utilization and provider reimbursement rates for shoulder arthroplasty procedures between 2012 and 2017.

METHODS

This study tracked annual Medicare claims and payments to shoulder arthroplasty surgeons via publicly available databases and aggregated data at the county level. Descriptive statistics were used to evaluate trends in procedure volume, utilization rate (per 10,000 Medicare beneficiaries), and reimbursement rate. We used adjusted multiple linear regression models to examine associations between county-specific variables (ie, urban or rural, average household income, poverty rate, percentage Medicare population, and race and ethnicity demographics) and procedure volume, utilization rate, and reimbursement rate.

RESULTS

Between 2012 and 2017, there was an 81.3% increase in primary TSA volume and 55.5% increase in primary TSA utilization. The Midwest and South had higher utilization rates than the Northeast and West (P < .001). TSA utilization rates in metropolitan areas were significantly higher than in rural areas (P < .001). Utilization rates for primary TSA procedures also had a significant negative association with poverty rate (P < .001). Regarding reimbursements, the Medicare payment per TSA case decreased from 2012 to 2017, with overall inflation-adjusted decreases of 7.1% and 11.8% for primary and revision cases, respectively. TSAs performed in metropolitan areas received significantly higher reimbursements per case than TSAs performed in rural areas ($1108.05 and $1066.40, respectively; P = .002). Furthermore, reimbursements per case were on average higher in the Northeast and West than in the South and Midwest (P < .001).

CONCLUSIONS

Our study confirms that although TSA volume and per capita utilization have increased dramatically since 2012, Medicare Part B reimbursements to surgeons have continued to fall even after the adoption of bundled-payment models for orthopedic procedures. Cost-containment efforts continue to focus on Medicare reimbursements to surgeons, although other expenditures such as hospital payments and operational and implant costs must also be evaluated as part of an overall transition to value-based health care.

A Cost-Effectiveness Analysis of Smoking-Cessation Interventions Prior to Posterolateral Lumbar Fusion

BACKGROUND

Smoking cessation represents an opportunity to reduce both short and long-term effects of smoking on complications after lumbar fusion and smoking-related morbidity and mortality. However, the cost-effectiveness of smoking-cessation interventions prior to lumbar fusion is not fully known.

METHODS

We created a decision-analytic Markov model to evaluate the cost-effectiveness of 5 smoking-cessation strategies (behavioral counseling, nicotine replacement therapy [NRT], bupropion or varenicline monotherapy, and a combined intervention) prior to single-level, instrumented lumbar posterolateral fusion (PLF) from the health payer perspective. Probabilities, costs, and utilities were obtained from published sources. We calculated the costs and quality-adjusted life years (QALYs) associated with each strategy over multiple time horizons and accounted for uncertainty with probabilistic sensitivity analyses (PSAs) consisting of 10,000 second-order Monte Carlo simulations.

RESULTS

Every smoking-cessation intervention was more effective and less costly than usual care at the lifetime horizon. In the short term, behavioral counseling, NRT, varenicline monotherapy, and the combined intervention were also cost-saving, while bupropion monotherapy was more effective but more costly than usual care. The mean lifetime cost savings for behavioral counseling, NRT, bupropion monotherapy, varenicline monotherapy, and the combined intervention were $3,291 (standard deviation [SD], $868), $2,571 (SD, $479), $2,851 (SD, $830), $6,767 (SD, $1,604), and $34,923 (SD, $4,248), respectively. The minimum efficacy threshold (relative risk for smoking cessation) for lifetime cost savings varied from 1.01 (behavioral counseling) to 1.15 (varenicline monotherapy). A PSA revealed that the combined smoking-cessation intervention was always more effective and less costly than usual care.

CONCLUSIONS

Even brief smoking-cessation interventions yield large short-term and long-term cost savings. Smoking-cessation interventions prior to PLF can both reduce costs and improve patient outcomes as health payers/systems shift toward value-based reimbursement (e.g., bundled payments) or population health models.

LEVEL OF EVIDENCE

Economic Level II. See Instructions for Authors for a complete description of levels of evidence.

Impact of Surgical Approaches on Complications, Emergency Room Admissions, and Health Care Utilization in Patients Undergoing Lumbar Fusions for Degenerative Disc Diseases: A MarketScan Database Analysis

OBJECTIVE

To identify the impact of different surgical approaches for lumbar degenerative disc disease (DDD) on complications, reoperations/readmissions, and health care utilization.

METHODS

We used International Classification of Diseases, Ninth Revision and Tenth Revision and Current Procedural Terminology codes to extract data from MarketScan. Patients were divided into 6 groups: single-level anterior only (sA), single-level anterior + posterior (sAP), single-level posterior (sP), multilevel anterior (mA), multilevel anterior + posterior (mAP), and multilevel posterior only (mP). Outcomes of interest were cumulative complication rates, reoperation rates, readmission, and health care utilization at 6, 12, and 24 months.

RESULTS

Of 148,499 patients, 3% had sA fusion and 54% had mP procedures. Patients in the mAP cohort incurred higher cumulative complication rates (21%) compared with sA (13%), sAP (15%), sP (14%), mA (18%), and mP (18%). Emergency room admissions within 30 days were highest in the mA cohort (14%) followed by mAP (11%) and mP (8%). At 12 and 24 months, patients with mA procedures were most likely to have either new fusion or refusion (8% and 12%) followed by sA (7% and 10%), sAP (4% and 7%), mAP (4% and 8%) mP (4% and 7%), and sP (3% and 7%). Compared with the mP cohort, patients in the mA cohort incurred 1.2 times the overall median payments, whereas mAP and sA incurred 1.1 times the payments at 12 months. This difference was further reduced at 24 months.

CONCLUSIONS

mAP procedures are associated with higher cumulative complications and health care utilization compared with other procedures and the difference in health care utilization tends to decrease over 12 and 24 months.

Development and validation of a novel scoring tool for predicting facility discharge after elective posterior lumbar fusion

BACKGROUND CONTEXT

Discharge to acute/intermediate care facilities is a common occurrence after posterior lumbar fusion and can be associated with increased costs and complications after these procedures. This is particularly relevant with the growing popularity of bundled payment plans, creating a need to identify patients at greatest risk.

PURPOSE

To develop and validate a risk-stratification tool to identify patients at greatest risk for facility discharge after open posterior lumbar fusion.

STUDY DESIGN

Retrospective cohort study.

PATIENT SAMPLE

Patients were queried using separate databases from the institution of study and the American College of Surgeons National Surgical Quality Improvement Program (NSQIP) for all patients undergoing open lumbar fusion between 2011 and 2018.

OUTCOME MEASURES

Discharge to intermediate care and/or rehabilitation facilities.

METHODS

Using an 80:20 training and testing NSQIP data split, collected preoperative demographic and operative variables were used in a multivariate logistic regression to identify potential risk factors for postoperative facility discharge, retaining those with a p value <.05. A nomogram was generated to develop a scoring system from this model, with probability cutoffs determined for facility discharge. This model was subsequently validated within the NSQIP database, in addition to external validation at the institution of study. Overall model performance and calibration was assessed using the Brier score and calibration plots, respectively.

RESULTS

A total of 11,486 patients (10,453 NSQIP, 1,033 local cohort) were deemed eligible for study, of which 16.1% were discharged to facilities (16.7% NSQIP, 9.6% local cohort). Utilizing training data, age (p<.001), body mass index (p<.001), female sex (p<.001), diabetes (p=.043), peripheral vascular disease (p=.001), cancer (p=.010), revision surgery (p<.001), number of levels fused (p<.001), and spondylolisthesis (p=.049) were identified as significant risk factors for facility discharge. The area under the receiver operating characteristic curve (AUC) indicated a strong predictive model (AUC=0.750), with similar predictive ability in the testing (AUC=0.757) and local data sets (AUC=0.773). Using this tool, patients identified as low- and high-risk had a 7.94% and 33.28% incidence of facility discharge in the testing data set, while rates of 4.44% and 16.33% were observed at the institution of study.

CONCLUSIONS

Using preoperative variables as predictors, this scoring system demonstrated high efficiency in risk-stratifying patients with an approximate four to fivefold difference in rates of facility discharge after posterior lumbar fusion. This tool may help inform medical decision-making and guide reimbursement under bundled-care repayment plans.

Health Care Utilization in Patients Undergoing Repeat Stereotactic Radiosurgery for Vestibular Schwannoma with 5-Year Follow-up: A National Database Analysis

Background Stereotactic radiosurgery (SRS) has been used as an alternative to microsurgery in patients with small vestibular schwannoma (VS). We compare health care utilization metrics in patients undergoing repeat-SRS (re-SRS) and no repeat SRS (nr-SRS) at long-term follow-up.

Materials and Methods We queried the MarketScan database using International Classification of Diseases, Ninth Revision and Current Procedural Terminology 4, from 2000 to 2016. We included adult patients who had diagnosis of VS and treatment with SRS with at least 5 years of continuous enrollment after the procedure. Outcomes were hospital admissions, outpatient services, and medication refills.

Results Of 1,047 patients, 5.1% (n = 53) had repeat SRS. Majority of re-SRS (74%, n = 39) were done within 2 years of index procedure and 51% were within 1-year of initial procedure. Patients who required re-SRS incurred higher hospital readmission rate, outpatient services, and had higher payments compared with those who did not require re-SRS at 6 months, 1 year, and 2 years following the initial procedures. Re-SRS received 3.0- and 3.1-times higher payments at 1 and 2 years compared with nr-SRS cohort. At 5 years following SRS, median combined payments for re-SRS cohort was $105286 (interquartile range [IQR] $70999, $156569) compared with $44172 (IQR $22956, $84840) for nr-SRS cohort.

Conclusion More than half of the re-SRS procedures were noted within first year of initial SRS for VS. Overall payments at 5 years for repeat SRS was more than double that for nr-SRS. Repeat SRS was also associated with more re-admissions and outpatient services at annual follow-up time points.

Variability in Hospital Costs of Adult Spinal Deformity Care

STUDY DESIGN

Retrospective, single-center analysis.

OBJECTIVE

To calculate the total clinical hospital cost of the Adult Spinal Deformity (ASD) care trajectory, to explain cost variability by patient and surgery characteristics, and to identify areas of process improvement opportunities.

SUMMARY OF BACKGROUND DATA

ASD is associated with a high financial and clinical burden on society. ASD care thus requires improved insights in costs and its drivers as a critical step toward the improvement of value, i.e., the ratio between delivered health outcome and associated costs.

METHODS

Patient characteristics and surgical variables were collected following ethical approval in a cohort of 139 ASD patients, treated between December, 2014 and January, 2018. Clinical hospital costs were calculated, including all care activities, from initial consultation to 1 year after initial surgery (excl. overhead) in a university hospital setting. Multiple linear regression analysis was performed to analyze the impact of patient and surgical characteristics on clinical costs.

RESULTS

75.5% of the total clinical hospital cost (&OV0556;27,865) was incurred during initial surgery with costs related to the operating theatre (80.3%), nursing units (11.9%), and intensive care (2.9%) being the largest contributors. 57.5% of the variation in total cost could be explained in order of importance by surgical invasiveness, age, coronary disease, single or multiple-staged surgery, and mobility status. Revision surgery, unplanned surgery due to complications, was found to increase average costs by 87.6% compared with elective surgeries (&OV0556; 44,907 (± &OV0556; 23,429) vs. &OV0556; 23,944 (± &OV0556; 7302)).

CONCLUSION

This study identified opportunities for process improvement by calculating the total clinical hospital costs. In addition, it identified patient and treatment characteristics that predict 57.5% of cost variation, which could be taken into account when developing a payment system. Future research should include outcome data to assess variation in value.

LEVEL OF EVIDENCE

4.

A pain relieving reimbursement program? Effects of a value-based reimbursement program on patient reported outcome measures

Background

Value-based reimbursement programs have become increasingly common. However, little is known about the effect of such programs on patient reported outcomes. Thus, the aim of this study was to analyze the effect of introducing a value-based reimbursement program on patient reported outcome measures and to explore whether a selection bias towards less complicated patients occurred.

Methods

This is a retrospective observational study with a before and after design based on the introduction of a value-based reimbursement program in Region Stockholm, Sweden. We analyzed patient level data from inpatient and outpatient care of patients undergoing lumbar spine surgery during 2006–2015. Patient reported outcome measures used was Global Assessment, EQ-5D-3L and Oswestry Disability Index. The case-mix of surgically treated patients was analyzed using medical and socioeconomic factors.

Results

The value-based reimbursement program did not have any effect on targeted or non-targeted patient reported outcome measures. Moreover, the share of surgically treated patients with risk factors such as having comorbidities and being born outside of Europe increased after the introduction. Hence, the value-based reimbursement program did not encourage discrimination against sicker patients. However, the income was higher among patients surgically treated after the introduction of the value-based reimbursement. This indicates that a value-based reimbursement program may contribute to increased inequalities in access to healthcare.

Conclusions

The value-based reimbursement program did not have any effect on patient reported outcome measures. Our study contributes to the understanding of the effects of a value-based reimbursement program on patient reported outcome measures and to what extent cherry-picking arises.

A Comparison of the Elixhauser and Charlson Comorbidity Indices: Predicting In-Hospital Complications Following Anterior Lumbar Interbody Fusions

OBJECTIVE

The Elixhauser Comorbidity Index (ECI) and Charlson Comorbidity Index (CCI) are commonly used measures that use administrative data to characterize a patient's comorbidity burden. The purpose of this study was to compare the ability of these measures to predict outcomes following anterior lumbar interbody fusion.

METHODS

The National Inpatient Sample was queried for all ALIF procedures between 2013 and 2014. The area under the receiver operating curve (AUC) was used to compare the ECI and CCI in their ability to predict postoperative complications when incorporated into a base model containing age, sex, race, and primary payer. Percent superiority was computed using AUC values for ECI, CCI, and base models.

RESULTS

A total of 43,930 hospitalizations were included in this study. The ECI was superior to the CCI and baseline models in predicting minor (AUC 71 vs. 0.66, P < 0.0001) and major (AUC 0.74 vs. 0.67, P < 0.0001) complications. When evaluating individual complications, the ECI was superior to the CCI in predicting airway complications (65% superior, AUC 0.85 vs. 0.72, P = 0.0001); hemorrhagic anemia (83% superior, AUC 0.71 vs. 0.66, P < 0.0001); myocardial infarction (76% superior, AUC 0.86 vs. 0.67, P < 0.0001); cardiac arrest (75% superior, AUC 0.85 vs. 0.67, P < 0.0001); pulmonary embolism (105% superior, AUC 0.91 vs. 0.71, P < 0.0001); and urinary tract infection (43% superior, AUC 0.76 vs. 0.73, P = 0.046).

CONCLUSIONS

The ECI was superior to the CCI in predicting 6 of the 15 complications analyzed in this study. Combined with previous results, the ECI may be a better predictive model in spine surgery.

Refining Risk-Adjustment of 90-Day Costs Following Surgical Fixation of Ankle Fractures: An Analysis of Medicare Beneficiaries

As the current healthcare model transitions from fee-for-service to value-based payments, identifying cost-drivers of 90-day payments following surgical procedures will be a key factor in risk-adjusting prospective bundled payments and ensuring success of these alternative payment models. The 5% Medicare Standard Analytical Files data set for 2005-2014 was used to identify patients undergoing open reduction and internal fixation (ORIF) for isolated unimalleolar, bimalleolar, and trimalleolar ankle fractures. All acute care and post-acute care payments starting from day 0 of surgery to day 90 postoperatively were used to calculate 90-day costs. Patients with missing data were excluded. Multivariate linear regression modeling was used to derive marginal cost impact of patient-level (age, sex, and comorbidities), procedure-level (fracture type, morphology, location of surgery, concurrent ankle arthroscopy, and syndesmotic fixation), and state-level factors on 90-day costs after surgery. A total of 6499 patients were included in the study. The risk-adjusted 90-day cost for a female patient, aged 65 to 69 years, undergoing outpatient ORIF for a closed unimalleolar ankle fracture in Michigan was $6949 ± $1060. Individuals aged <65 or ≥70 years had significantly higher costs. Procedure-level factors associated with significant marginal cost increases were inpatient surgery (+$5577), trimalleolar fracture (+$1082), and syndesmotic fixation (+$2822). The top 5 comorbidities with the largest marginal cost increases were chronic kidney disease (+$8897), malnutrition (+$7908), obesity (+$5362), cerebrovascular disease/stroke (+$4159), and anemia (+$3087). Higher costs were seen in Nevada (+$6371), Massachusetts (+$4497), Oklahoma (+$4002), New Jersey (+$3802), and Maryland (+$3043) compared with Michigan. With the use of a national administrative claims database, the study identifies numerous patient-level, procedure-level, and state-level factors that significantly contribute to the cost variation seen in 90-day payments after ORIF for ankle fracture. Risk adjustment of 90-day costs will become a necessity as bundled-payment models begin to take over the current fee-for-service model in patients with fractures.

Primary single-level lumbar microdisectomy/decompression at a free-standing ambulatory surgical center vs a hospital-owned outpatient department-an analysis of 90-day outcomes and costs

BACKGROUND CONTEXT

While free-standing ambulatory surgical centers (ASCs) have been extolled as lower cost settings than hospital outpatient facilities/departments (HOPDs) for performing routine elective spine surgeries, differences in 90-day costs and complications have yet to be compared between the two types of treatment facilities.

PURPOSE

We carried a comprehensive analysis to report the differences on payments to providers and facilities as a reflection of true costs to patients, employers and health plans for patients undergoing primary, single-level lumbar microdiscectomy/decompression at ASC versus HOPD.

STUDY DESIGN

Retrospective review of Medicare advantage and commercially insured enrollees from the Humana dataset from 2007 to 2017Q1.

OUTCOME MEASURES

To understand the differences in 90-day complications, readmissions, emergency department visits and costs for patients undergoing primary, single-level lumbar microdiscectomy/decompressions at an ASC versus HOPD.

METHODS

The Humana 2007 to 2017Q1 was queried using Current Procedural Terminology codes to identify patients undergoing primary, single-level lumbar microdiscectomy/decompressions. Patients undergoing two-level surgery, open laminectomies, fusions, revision discectomies, and/or deformities were excluded. Service Location codes for HOPD (Location Code 22) and free-standing ASC (Location Code 24) were used to determine surgery treatment facilities. Using propensity scoring, we matched two groups who had surgery performed in ASCs or HOPDs based on age, gender, race, region and Elixhauser comorbidity index. Multivariable logistic regression analyses were performed on matched cohorts to assess for differences in 90-day outcomes between facilities, while controlling for age, gender, race, region, plan, and Elixhauser comorbidity index.

RESULTS

A total of 1,077 and 10,475 primary single-level decompressions were performed in ASCs and HOPDs, respectively. Following a matching algorithm with propensity scoring, the two cohorts were comprised of 990 patients each. Observed differences in 90-day complication rates were not statistically or clinically significant (ASC=9.1% vs. HOPD=10.3%; p=.362) nor were readmissions (ASC=4.5% vs. HOPD=5.3%; p=.466). On average, performing surgery in an ASC versus HOPD resulted in significant cost savings of over $2,000/case in Medicare Advantage ($5,814 vs. $7,829) and over $3,500/case ($10,116 vs. $13,623) in commercial beneficiaries.

CONCLUSION

Performing single-level decompression surgeries in an ASC compared with HOPDs was associated with approximately $2,000 to $3,500 cost-savings per case with no statistically significant impact on complication or readmission rates.

Epidemiology and Socioeconomic Trends in Adult Spinal Deformity Care

Adult spinal deformity (ASD) has gained significant attention over the past decade with improvements in diagnostic tools, classification schemes, and surgical technique. The demographics of the aging population in the United States are undergoing a fundamental shift as medical care advances and life expectancy increases. The “baby boomers” represent the fastest growing demographic in the United States and by 2050, the number of individuals 65 yr and older is projected to reach 89 million, more than double its current size. Based on current prevalence estimates there are approximately 27.5 million elderly individuals with some form of spinal deformity, which will place a significant burden on our health care systems. Rates of surgery for ASD and case complexity are both increasing, with concomitant increase in the cost of deformity care. At the same time, patients are more medically complex with increasing number of comorbidities that result in increased surgical risk and complication profiles. This review aims to highlight recent trends in the epidemiology and socioeconomic patterns in surgery for ASD.

Predictive modeling in spine surgery

As the cost of healthcare in the United States increases at an unsustainable rate, health-policy leaders are looking towards innovative ways to maximize value in delivery of care. Incorporating technology, such as artificial intelligence/machine-learning, to assist physicians in decision-making and predicting outcomes, on a real-time basis, is a major topic of discussion. While machine learning is gradually pulling traction in the medical community, it still remains a nascent field in the realm of spine surgery. The current review aims to gather current literature discussing the validity and applicability of machine-learning models in spine surgery.

Feasibility of Bundled Payments in Anterior, Middle, and Posterior Cranial Fossa Skull Base Meningioma Surgery: MarketScan Analysis of Health Care Utilization and Outcomes

BACKGROUND

The aim of our study was to compare the health care utilization and outcomes after surgery for anterior cranial fossa skull base meningioma (AFM), middle cranial fossa skull base meningioma (MFM), and posterior cranial fossa skull base meningioma (PFM) across the United States.

METHODS

We queried the MarketScan database using International Classification of Diseases, Ninth Revision and Current Procedural Terminology 4, from 2000 to 2016. We included adult patients who had at least 24 months of enrollment after the surgical procedure. The outcome of interest was length of hospital stay, disposition, complications, and reoperation after the procedure.

RESULTS

A cohort of 1191 patients was identified from the database. Less than half of patients (43.66%) were in the AFM cohort, 32.24% were in the MFM cohort, and only 24.1% were in the PFM cohort. Patients who underwent surgery for PFM had longer hospital stay (P = 0.0009), high complication rate (P = 0.0011), and less likely to be discharged home (P = 0.0013) during index hospitalization. There were no differences in overall payments at 12 months and 24 months among the cohorts. There was no significant difference in 90-day median payments among the groups ($66,212 [AFM] vs. $65,602 [MFM] and $71,837 [PFM]; P = 0.198). Male gender, commercial insurance (compared with Medicare), and higher comorbidity scores (score 3 compared with score 0) were associated with higher 90-day payments in the PFM cohort.

CONCLUSIONS

Overall payments (at 12 months and 24 months) and 90-day payments were not different among the cohorts. Patients with PFM had longer hospital stay and higher complication rate and were less likely to be discharged home with higher utilization of outpatient services at 12 months and 24 months.