The Problem
Health Care costs in the United States continue to balloon absorbing larger portions of our Gross Domestic Product (GDP). In 2016, health care spending increased by 4.3% and reached $3.3 trillion, 17.9% of GDP.1 The federal government and the state of North Carolina have passed legislation in an attempt to reduce rising health care costs and improve outcomes. Since 2013, Duke Connected Care has evolved from a non-risk-bearing Accountable Care Organization (ACO) to a revenue-generating, risk-bearing ACO for primary care providers and their enrollees. Although Duke has participated in national efforts to improve quality, such as the National Surgical Quality Improvement Plan (NSQIP). There has not been a comprehensive effort to capture the true value provided to patients. Michael Porter defines ‘value’ in health care as “patient outcomes achieved per dollar expended”.2
Duke Health, the state of North Carolina, and our nation do not currently understand or harness existing technology to appropriately capture the cost of treating a patient for a given condition. As Michael Porter states in Harvard Business Review, “We need to abandon the idea that charges billed or reimbursements paid in any way reflect costs”.2 We currently have the technological capability to switch from a charge-based analysis to a true cost-based analysis, but we must carefully combine previously compartmented datasets to do so.In 2016, the University of Utah followed the principles laid out by Porter and piloted a project that reduced costs and optimized outcomes amongst patients with lower extremity joint replacements.3
Duke has large clinical and administrative datasets, but they are not currently coordinated and synced in a way that would allow for a robust analysis of cost and outcome. We must build the information technology infrastructure to facilitate these analyses so that we can move beyond billing and reimbursement and begin to understand our health system in terms of the value it generates.
Our Solution
We would like to understand the value of the care we provide for our patients within the Duke Health System, and ductal carcinoma of the breast is an ideal first use case. We will attempt to tabulate cost of management on a per-patient basis from diagnosis to completion of the multidisciplinary treatment course. This would typically include interfaces with Diagnostic Radiology, Pathology, Surgery, Medical Oncology, and Radiation Oncology. This calculation can be done through existing datasets that have not previously been combined within Duke Health. Rather than representing cost as an arbitrary percentage of charge as has been done historically, we propose employment the time-driven activity-based costing (TDABC) system described by Porter. This system requires only two parameters: “the cost of each of the resources used in the process and the quantity of time the patient spends with each resource.” Regardless of the nature of the resource—including service providers themselves–, capacity cost rate (CCR) can be calculated with the below equation.
???????????? = ???????????????????????????????? ???????????????????????????????????????????????? ???????? ???????????????????????????????? ÷ ???????????????????????????????????? ???????????????????????????????? ???????? ????????????????????????????????
This methodology represents a paradigm shift that will not only change the way we view the cost of treating breast cancer but will hopefully cause us to view our patients and the value of the care they receive in a more comprehensive manner. There are a few progressive systems that have applied these principles with great success. The University of Utah built infrastructure to reduce costs and improve quality in septic patients and patients undergoing total hip replacement.3,4 Other institutions such as the Head and Neck Center at MD Anderson, the Cleft Lip and Palate Program at Children’s Hospital in Boston, and the knee replacement unit at Brigham & Women’s Hospital have lowered costs and improved outcomes using these methodologies as well.
