Journal Watch: “Toward value-based care using cost mining: cost aggregation and visualization across the entire colorectal cancer patient pathway.”
Cost-mining Parallels Activity Based Costing at the Patient Journey Level
This month’s journal watch comes from Australia regarding methods in value-based care in oncology, an increasingly costly area of healthcare. The article examines the case example of colorectal cancer (CRC), evaluating the entire journey (and their subsequent costs) of CRC patients from primary care to survivorship. A total of 4,246 unique cases of colon cancer and 4 million activities were analyzed to understand where costs arise in the cancer journey for these patients. The paper addresses a topic called “cost mining…as an approach to uncover high-cost pathways and specific cost drivers using real-world patient-level data.” They attempt to take process mining to the next level to determine the costliest events in the care journeys of colorectal cancer patients.
Their findings were that, not surprisingly, later stage cancer journeys cost more. Inpatient episodes were 99% of the cost burden while general practitioner visits, testing and prescriptions (excluding chemotherapy) were only 0.01% of the cost. Chemotherapy was about 6% of the costs, and they found significant variations in costs for the same chemotherapy regimens when patients were at different cancer stages. They note that “future research should utilize cost mining to investigate whether preventative interventions or earlier screening and diagnosis lead to quicker patient pathways or comparatively lower cost inpatient and chemotherapy episodes, given the significant correlation between cancer stage at the time of treatment and costs.” They note the advantage of cost mining is the ability to track cost of care at various activity and event levels. They describe their limitations in that the data was retrospective, descriptive and cannot be used for predictive value. However, the information could be used to help focus improvement efforts on events that have the highest cost, not just in preventing later stage cancers but in hospitalizations as well.
One seemingly paradoxical principle in value-based care for patients is to reduce unnecessary variation in care while maintaining the tenets of personalized care. Unnecessary variation in care means unwarranted differences in a process which result in achieving disparate health outcomes. At the basic level, if two different people with the same problem receive different care, and person A achieves a lesser health outcome and person B a better outcome, even with similar costs, person A has received less “value.” Visually, it is narrowing the width of the health outcome curve.
What is the relevance of this study to value-based healthcare? Recall the value equation: Value = Δoutcome/cost to create the outcome. To create value for patients, one needs to know which steps create better outcomes and the cost of those steps in the patient journey. In my perspective, the next best use for this method of cost-mining is to combine it with health outcome data. In this case, when patients had cost variation, did they also have outcome variation? We often don’t know which individual steps to adjust until the differences in the care journeys and the differences in outcomes between patients are elicited. The worst-case scenario is to have costly steps in the journey which worsen or do not improve the patient’s health. From a health equity perspective, it is also undesirable to have patients with low costs but whom also have low improvement in health outcomes.
I saw significant parallels in the authors’ concept of ‘cost-mining’ to the principles of Activity Based Costing (ABC). ABC is a tool for businesses to more accurately understand the origins of their aggregate costs by outlining a dollar-per-unit cost for each resource involved in the production of an item, service, or outcome. The authors essentially performed Activity Based Costing (ABC) at the level of the patient-journey across settings and stages of disease, rather just than at the level of a particular health setting (where ABC is often undertaken). Particularly for payers who are responsible for costs across settings in the patient journey, the concept of cost-mining has direct application. However, to be actionable, the costs must be correlated with measuring outcomes. One needs to know what you are “getting” for that cost. If one responds to costs by restricting or rationing useful care, outcomes will fall, and health-inequities will rise. Like we learn in basic math, zero divided by any number is still zero. Without improvement in the health of the patient, at any cost, the value is nil. Instead, payers and health systems should look for opportunities to reduce or eliminate costs in areas that do not improve outcomes. This article demonstrated how data can be used to understand costs of activities in the entire patient journey for a disease. The next stage is correlate these activity costs with patient-oriented outcomes to best understand the value of each activity.
Leusder M, Relijveld S, Demirtas D, Emery J, Tew M, Gibbs P, Millar J, White V, Jefford M, Franchini F, IJzerman M. Toward value-based care using cost mining: cost aggregation and visualization across the entire colorectal cancer patient pathway. BMC Med Res Methodol. 2024 Dec 27;24(1):321. doi: 10.1186/s12874-024-02446-5. PMID: 39730997; PMCID: PMC11681630.
