Researchers at Weill Cornell’s Department of Healthcare Policy and Research and the University of Florida’s Department of Health Outcomes & Biomedical Informatics developed a new patient taxonomy with clinically meaningful categories for high-cost Medicare patients.
The full study, published in Healthcare presents a novel patient taxonomy that developed 10 clinically meaningful categories to describe and identify high-cost patients which could help health systems tailor interventions to individuals who account for a disproportionate share of healthcare spending and utilization.
There is a great deal of interest in understanding high-cost patients given the resources needed to care for them. Our current understanding of this group is that they are not homogenous; rather, they have diverse healthcare needs and current approaches to categorizing these patients have several limitations. Foremost is the fact that existing categories or taxonomies seek to map patients to one segment, which does not account for the complex interplay between varying health needs (e.g. someone presenting with opioid misuse may also have mental illness). Second, most studies focus exclusively on medical conditions and do not account for mental health conditions. Third, they primarily rely on administrative data such as claims and do not incorporate patients’ social needs.
With these limitations in mind, researchers from Weill Cornell Medical College and the University of Florida developed a new taxonomy with 10 overlapping patient categories to understand the medical, behavioral, and social complexity of high-cost Medicare patients. The categories were developed through a literature review, focus groups, and interviews with patients, physicians, health systems leaders, and health policy experts. Preliminary categories were analyzed against 428,024 of 2013 Medicare beneficiaries in the New York metropolitan area. The 10 categories were: (1) multiple chronic conditions; (2) seriously ill; (3) frail; (4) serious mental illness; (5) single condition with high pharmacy cost; (6) chronic pain; (7) end-stage renal disease; (8) single high-cost chronic condition; (9) opioid use disorder; and (10) socially vulnerable.
Initial findings confirmed current thinking around the demographics of high-cost patients. Compared to non-high-cost patients, high-cost patients tended to be older, male, African American, and have more chronic conditions. From a medical standpoint, 97.4% of all high-cost patients had multiple chronic conditions, 53.7% were seriously ill, and 48.9% were frail. In terms of quantifiable cost, high-cost patients accounted for Medicare spending per beneficiary 8 times higher than non-high-cost patients. The authors also found that 72.7% of high-cost patients could be mapped into multiple categories, with the highest concentration being patients who were both frail and seriously ill.
These findings identify areas of need that can be met with targeted interventions by health systems. Two apparent examples are those relating to beneficiaries suffering from behavioral conditions and opioid misuse disorder: more than 30% of high-cost patients were found to have serious mental illness, and those with opioid misuse disorder had a high likelihood of being a high-cost patient. These findings suggest that behavioral conditions are highly prevalent among high-cost Medicare beneficiaries and that interventions are needed to better manage them.
To identify which interventions may be of use to beneficiaries, the ten categories were further distilled into five domains: medical care, behavioral health services, social services, palliative care, and pharmaceutical pricing policies.
The categories, as presented, are highly actionable. For example, socially vulnerable patients likely require services from non-medical organizations such as transportation and housing. Thus, healthcare organizations may want to harness local social services organizations to address food and housing insecurity in conjunction with clinical care. For patients with end-stage renal disease who often have multiple comorbid chronic conditions, institutions may seek care managers to better coordinate their intensive medical care.
Some limitations of the study are that the patient population used to develop the taxonomy was located in the New York metropolitan area and may not be representative of other regions and municipalities. There may also be differences in drivers of high-cost and high-utilization patients between commercial payer and Medicaid populations. For example, commercial payer populations tend to have fewer chronic conditions and drivers of high utilization are more often due to acute conditions. In contrast, behavioral and social conditions may drive higher level of utilizations in Medicaid populations
Ultimately, the categories developed by the researchers can play a pivotal role in helping health systems develop a strategy to care for high-cost, high-utilization patients. The framework provided may help clinicians and health systems better understand their patient populations and improve care models targeted to the unique needs of these patients.
The full article “Developing an actionable patient taxonomy to understand and characterize high-cost Medicare patients” written by Yongkang Zhang, Zachary Grinspan, Dhruv Khullar, Mark Aaron Unruh, Elizabeth Shenkman, Andrea Cohen and Rainu Kaushal was published in Volume 8, Issue 1 of Healthcare: The Journal of Delivery Science and Innovation and can be found on ScienceDirect.
Through a partnership with Elsevier, MedCity News will present a news article each month on a notable topic and will link to the full study that will be available for the general public for a limited period of time. The goal is to highlight important research being undertaken as the industry is undergoing a fundamental shift. The article will be available to the public until Aug. 31.
Photo: JamesBrey, Getty Images
