Center variation in episode-of-care costs for adult spinal deformity surgery: results from a prospective, multicenter database

Abstract

Background Context

Adult spinal deformity (ASD) surgery is associated with significant resource utilization, costing more than $958 million in charges for Medicare patients and more than $1.7 billion in charges for managed care population in the last decade. Given the recent move toward bundled payment models, it is important to understand the various care components a patient receives over the course of a defined clinical episode, its associated cost, and the proportion of cost for each component toward the bundled payment.

Purpose

To examine the degree and determinants of variation in inpatient episode-of-care (EOC) cost, resource utilization, and patient-reported outcomes for patients undergoing ASD surgery across four spine deformity centers in the United States.

Study Design/Setting

Retrospective analysis of prospective, multicenter database.

Patient Sample

Consecutive patients enrolled in an ASD database from four spinal deformity centers.

Outcome Measures

Total in-patient EOC costs and Short Form (SF)-6D.

Methods

The study used a multicenter database of 210 consecutively enrolled operative patients from 2008 to 2013 at four participating centers in the United States. Demographic, surgical, and direct cost data, expressed in 2013 dollars, for the entire inpatient EOC were obtained from administrative databases from the respective hospitals. Mixed models and multivariable linear regression were used to evaluate the impact of center on total costs adjusting for patient characteristics, length of stay (LOS), and surgical factors.

Results

A total of 126 patients with complete baseline and 2-year follow-up data were included. The percentages of patients from each center were 36.5%, 7.1%, 24.6%, and 31.7%. Overall, the mean patient age was 58.4±12.6 years, 86% were women, and 94% were Caucasian. The proportion of total cost variation explained by the center at which the patient was treated was 17%. After adjusting for patient, LOS, and surgical factors the cost variation reduced to 4%. In multivariable analysis, each additional level fused increased total cost variation by $2,500, whereas recombinant human bone morphogenetic protein-2 (BMP) use and posterior-only surgical approach lowered total EOC costs by $10,500 and $9,400, respectively. No significant difference was observed in 2-year quality-adjusted life year across centers.

Conclusions

Total EOC costs for ASD surgery varied significantly by center. Levels fused, BMP use, and surgical approach were the primary drivers of cost variation across centers. Differences in resource utilization had no impact on 2-year quality-adjusted life year improvement across centers.

Introduction

Before implementing bundled payment models, it is important to identify the various care components and the associated cost of each component a patient receives over the course of a defined clinical episode, often called as “episode-of-care” (EOC). An in-depth understanding of the contribution of different care components during an EOC will not only help set a reasonable reimbursement rate, but also help to improve care coordination and reduce the use of duplicative and redundant services [1], [2]. A significant concern with regard to bundled payments is that the impact of variations in patient characteristics and surgical factors are often underestimated when setting payment structures. In general, patient characteristics such as age, body mass index (BMI), and comorbidities typically have significant influence on surgical outcomes [3], [4], [5], [6], whereas surgical factors such as type of intervention, need for revision or reoperation, use of implants or blood products, and length of stay influence resource utilization [7], [8], [9]. Collectively, these differences can introduce substantial variation in EOC costs within and across hospitals and placing more expensive hospitals at serious financial risk under bundled payment plans. Payers and providers need to identify and consider these sources of variation before agreeing to adopt bundled payment models.

In this context, we investigated the extent of variation in in-patient EOC cost, resource utilization, and health outcomes for patients undergoing adult spinal deformity (ASD) surgery across four spine deformity centers in the United States. Adult spinal deformity surgery is one of the most expensive procedures among the spine-related procedures, responsible for more than $958 million in charges for Medicare patients and more than $1.7 billion in charges for managed care population between the years 2000 and 2010 [10]. Because of the complex nature of spine procedures, a general lack of consensus exists among surgeons on surgical procedures for similar conditions, and disagreement on use and duration of postoperative acute care [7], [11], [12]. Despite the increasing number of cost-effectiveness studies in the ASD literature, few studies have examined the degree and sources of variation in EOC cost using actual direct costs instead of charges or generalized estimates such as Medicare Allowable rates. Hospital charges are often inaccurate proxies for cost, as they represent the amount that the hospital bills the insurance company, and do not necessarily represent the actual cost for the hospital or how much the hospital is paid. Direct costs reflect the actual costs from the hospital's administrative database of surgical supplies or implants, operating room time, hospital room stays, medications, laboratories, and imaging for a hospital and patient; these exclude indirect costs such as administration and facilities overhead.

In the present study, our objectives were the following: (1) to measure the degree and determinants of variation in total hospital direct costs for surgical treatment of ASD within and between four spine deformity centers, (2) to assess the degree of variation in total hospital costs that is explained by patient characteristics or surgical factors, and (3) to identify factors that are determinants of increased variation in total, 2-year EOC costs.