Abstract
The most common reason for acute care utilization for individuals with sickle cell disease (SCD) is Vaso-Occlusive crisis (VOC). Patients typically seek out care through the emergency department (ED) At these locations patients often have long waits to get care and often receive sub optimal pain management, which Results in over 40% of patients requiring hospital admission. Johns Hopkins Hospital has implemented a new model of service for people with SCD; an outpatient Sickle Cell Infusion Clinic (SCIC) that was opened in 2008 as an alternative source of urgent care for patients having VOC. The purpose of this study is to determine the net financial benefit of implementing the sickle cell infusion clinic model.
A cost-benefit analysis is conducted from the payer’s perspective focusing on direct medical cost (procedures, drugs, tests, etc.) of SCD patients and excludes indirect medical costs (patients’ productive changes). Health care costs and utilization data was available for the last 3 out of 5 years that the SCIC was opened. A literature review was conducted to determine the costs of individual components of the total costs for SCD patients: inpatient hospitalization, ED visit, primary care and secondary care visits, and other healthcare costs for patients with SCD. The overall and average visit cost of the SCIC was determined from the 2012-2013 budget and visits. The billing data for a subset of patients seen in the SCIC that were insured by one of the Medicaid’s MCO was used to determine utilization and costs of healthcare services for 2010, 2011, and 2012. The baseline utilization of healthcare services before the SCIC was implemented was estimated from the literature. As we did not have baseline data we estimated that the 2010 utilization of healthcare services reflected a 20% decrease in hospitalizations and a 40% decrease in ED visits. The overall cost of care for patients with SCD was determined from the above-mentioned sample of billing claims data and utilization estimates. The costs of running the SCIC was added to the overall costs of care. Finally, net savings for the SCIC was determined by calculating the difference in overall cost and savings per beneficiary per month (PBPM). All values are reported in 2012 inflation-adjusted dollars.
For the subset of patients covered by the Medicaid MCO, the SCIC model resulted in a 7.6% ($676 PBPM) cost savings in the first year (estimated baseline compared to 2010) with savings of 29.2% ($2598 PBPM) when comparing estimated baseline to 2012. The total medical costs for the subset of patients using the SCIC in 2010 was $3,492,339 with an average cost of $94,388 per patient. The SCIC had 1,428 visits by 246 unique patients in FY2012 with an average cost per visit of $434 or $203 PBPM. Other costs in this patient population include: inpatient hospitalization ($3,985 PBPM), ED visits ($326 PBPM), primary and secondary care visits ($26 PBPM), and pharmacy ($493 PBPM). The total cost of care for the same number of sickle cell patients as in our sample who did not utilize the SCIC would have been $3,779,588, with an average cost of $102,151per patient. The SCIC model resulted in cost savings primarily due to a decrease in hospitalizations and ED visits. The number of hospitalization decreased 52.0% (2.88 HPY) and the number of ED visits decreased 48.4% (2.32 visits VPY) in the fifth year of operating the infusion clinic model (2012). The average cost of a hospitalization and an ED visit was $10,797 and $1,024 respectively. These values did not change with the implementation of the SCIC. If we extrapolate the cost savings seen in the subset of patients using the more conservative 7.6% cost savings, to the entire patient cohort this would result in a cost savings of $1.9 million.
Preliminary cost-benefit analysis shows that the SCIC model resulted in significant cost savings that increased significantly in successive years. Cost savings was driven by two major factors: 1) decrease in inpatient hospitalizations and 2) decrease in ED visits. Additional analysis to include actual baseline data is planned along with a sensitivity analysis to identify if there is a certain threshold population density for which this model would be most cost effective.
Lanzkron: GlycoMimetics, Inc.: Research Funding.
Author notes
Asterisk with author names denotes non-ASH members.
