• Center on Health Equity and Access
  • Clinical
  • Health Care Cost
  • Health Care Delivery
  • Insurance
  • Policy
  • Technology
  • Value-Based Care

Gauging the Experiences of the Seriously Ill in California: Analyzing Serious Illness Beyond Medicare Fee-for-Service

The American Journal of Accountable Care®June 2020
Volume 8
Issue 2

Significant populations of seriously ill individuals are insured by all lines of business and have meaningfully different needs and medical histories in California.

The American Journal of Accountable Care. 2020;8(2):22-25. https://doi.org/10.37765/ajac.2020.88439

Given the growing recognition of the need to improve care for individuals with serious illness, who frequently receive fragmented care1 of varying quality2 that does not reflect their goals, multiple payers have launched various serious illness programs. CMS has launched a Serious Illness Population track in the Primary Care First3 model and a High Needs Population track for its Direct Contracting4 model (described in a recently released request for applications5). Concurrently, Medicare Advantage (MA) plans are partnering with third-party firms6 and offering new supplemental benefits7 for services focused on their seriously ill members, whereas commercial insurers are establishing palliative care programs8 for their members.

However, most of the evidence driving these policies comes solely from the Medicare fee-for-service (FFS) program, largely due to restricted access to MA and commercial insurance claims data. This gives policy makers, providers, and health systems limited ability to understand how new initiatives are affecting overall care quality and individuals’ experience and offers few opportunities to share best practices and expand the spread of successful models.

Limited evidence has been a key driver in why health systems and provider organizations are still early in their initiatives to improve serious illness care. In a recent survey9 of accountable care organizations, 94% reported that they were “partially” or “widely” identifying their seriously ill populations, but far fewer (only 8%-21%) were widely implementing serious illness programs (eg, home-based palliative care, 24/7 clinical access).

To help improve overall understanding of the characteristics of the seriously ill and build the evidence base, the Duke Margolis Center for Health Policy and the Integrated Healthcare Association (IHA) explored the distribution of the seriously ill in California with the support of the Gordon and Betty Moore Foundation (and with a team of advisors listed in the Acknowledgments). This type of state-level examination can help inform national discussion and future analytical work by showing key trends in serious illness, identifying where further work is needed, and clarifying where policy priorities should focus going forward.

How We Identified the Seriously Ill

It is difficult to identify individuals with serious illness in common data sources, as few sources commonly capture whether someone has a functional limitation. To overcome this, several experts have developed methods that use surrogates to identify the seriously ill using a multipart10,11 definition: (1) diagnosis of a significant condition (eg, 1 serious condition [eg, cancer] or 3 or more serious chronic conditions), (2) recent health care utilization (as shown through a recent hospitalization), and (3) surrogates for functional limitations (eg, usage of specific durable medical equipment [eg, home oxygen] or home health care, evidence of limitations in activities of daily living).

Researchers use these definitions flexibly depending on the sensitivity and specificity needed for a particular research question. We adapted this definition for the available claims data from IHA’s multipayer claims database managed by Onpoint Health Data, which included more than 15 million Californians with commercial insurance (both employer-sponsored and individual insurance), Medicare FFS, and MA. Given the data source, we slightly altered the list of diagnoses that were used in the first part of the 3-part definition (as noted in the eAppendix [available at ajmc.com]). The third part of the definition—which includes surrogates of functional limitation—was difficult to operationalize, partially because the durable medical equipment and home health surrogates were not easily available in the available commercial claims data and partially because the presentation of serious illness may look different in a younger population. Because of these issues, we found that including the third part of the definition (surrogates of functional limitations) made the seriously ill percentage extremely small, so we primarily present results in this article without the third part of the definition. We analyzed the state in 3 levels of detail: covered California rating regions (n = 19), core-based statistical areas (n = 26), and counties (n = 58), with our primary analysis at the county level.

Do the Seriously Ill Look Different Across Insurance Plans?

This analysis examined the seriously ill for multiple insurance types and for different age cohorts (18-64 vs ≥65 years), with the results illustrated in Figure 1. Access to this source of broader data allowed us to get a more comprehensive look at serious illness across the state of California, as many prior analyses of serious illness care have been restricted to traditional Medicare FFS data.

Not surprisingly, rates of serious illness tend to be higher in the population 65 years and older. However, the age differences varied by insurance type. For Medicare FFS and MA, serious illness rates were similar between the younger and older age groups, likely reflecting the fact that the group aged 18 to 64 years in those plans was primarily composed of individuals with permanent disabilities or end-stage renal disease. In contrast, serious illness rates were lower overall for individuals with commercial insurance, with large differences between younger and older cohorts for both types of commercial insurance (preferred provider organization [PPO] and health maintenance organization [HMO]).

In addition to differences between age groups, serious illness prevalence also differed by insurance types (PPOs vs HMOs; Medicare FFS vs MA). First, seriously ill individuals are more likely to be insured by programs with less restrictive policies, networks, and protocols, and broader networks are more valued by older individuals12 and those with higher expected medical costs. The availability of the hospice benefit in Medicare FFS may also affect enrollment, causing the seriously ill to switch from MA to Medicare FFS to take advantage of that benefit.

Where Are the Seriously Ill Located?

Of individuals 65 years and older in California, 8.5% are individuals with serious illness (using our diagnosis and utilization definition). However, prevalence varies significantly across the state, ranging from 6% to 13%. As illustrated in Figure 2, serious illness rates vary by payer, with serious illness prevalence for individuals 65 years and older looking different for all payers (right) compared with Medicare FFS (left).

The geographic analysis helps identify where future serious illness interventions should be targeted, based on prevalence and the quality of care for those individuals. For example, among urban counties in California, Los Angeles County stands out for its high prevalence of serious illness among the population 65 years and older in Medicare FFS. This may be because Los Angeles County has a high prevalence of dual-eligible individuals enrolled in Medicare FFS, although there are likely other sociodemographic and care pattern factors that influence these rates. Further localized hotspotting could allow for additional clarity in how and where interventions should take place, especially for larger counties like Los Angeles County, which has a larger population than 41 states.

How Do the Seriously Ill Utilize Health Care?

In general, we found that individuals with serious illness overall do not use different health care services than those without serious illness, but rather use more of them. For example, the top 10 most frequently filled prescription drugs for both the seriously ill and the non—seriously ill are overwhelmingly low-cost chronic disease management drugs such as statins and blood sugar control medications. The seriously ill population differs in that it has more prescriptions, uses more higher-cost drugs, and uses a wider range of prescriptions than the non–seriously ill.

Similarly, the seriously ill are far more likely to use hospital emergency departments (EDs). Seriously ill individuals 65 years and older had 1167 ED visits per 1000 members in 2017, whereas the non—seriously ill elderly had 251 per 1000. The disparity for the population younger than 65 years was even greater; the seriously ill averaged 1844 visits per 1000 members, whereas the non–seriously ill averaged 159 per 1000. The top reasons for ED visits were similar except that the seriously ill were diagnosed with sepsis for 4% of visits. Although most of these visits were not avoidable, the seriously ill have a much higher percentage of ED visits that were avoidable and could have been managed by primary care or other outpatient clinic visits. From these data, interventions to treat urinary disorders may be an ideal place for health systems to start.

Readmissions could also be a critical area of focus. The seriously ill group 65 years and older had 2.5 times more readmissions than the non—seriously ill, whereas the seriously ill population aged 18 to 64 years had 5 times as many. Given the strong focus on readmissions by many health systems, improvements in serious illness care could be an important component of any broader readmission reduction program.

Policy Implications

Our analysis offers key implications for policy in 3 areas. First, our analysis provides a tool that CMS and other states could use to target, benchmark, and evaluate their serious illness initiatives. As noted earlier, the new Primary Care First—Serious Illness Population and Direct Contracting models offer new pathways for providers to improve serious illness care and could offer a bridge to more comprehensive approaches in the years to come. Future models could build on broader payment reform efforts13 to address long-term and expensive chronic care needs.

Secondly, this type of data could be useful for health systems and risk-bearing clinical entities as they seek to understand the seriously ill across all their patient populations (not just those with Medicare FFS). These could be cross-referenced with other data on population demographics, such as income levels, food insecurity, and environmental health. For example, a health system may identify that its seriously ill population consists mostly of individuals with serious cardiac conditions, such as stage D heart failure, and restructure its cardiology workflows and teams accordingly. It may focus on care management and reducing readmissions for this population and assess the impact on hospitalizations, ED visits, total cost of care, and readmissions. Smaller efforts would make it easier for organizations to make the business case, develop the necessary data infrastructure, and build an appropriate workforce necessary to deliver high-quality care.

Finally, studies like this one can use data to empower efforts to improve serious illness care in a particular county, which will require further study of what makes a county unique. Any number of reasons could contribute to challenges in serious illness care, including patient population, market dynamics, types of available providers, and geography. This will hopefully not just produce evidence but also be a first step in more extensive regional efforts to improve health care for the sickest populations.


This research was funded by the Gordon and Betty Moore Foundation. Karl Finison and Amy Kinner at Onpoint Health Data provided input on methodology and conducted the analytics to create a results file. The authors also thank collaborators who provided input and strategic guidance to the broader analysis, including Donald H. Taylor Jr, William Bleser, Jeffrey Clough, Arif Kamal, Gregory Daniel, Harriet Mather, Amy Kelley, David Muhlestein, Nathan Smith, Marissa Schlaifer, and Russ Montgomery.

Author Affiliations: Duke University Margolis Center for Health Policy, Durham, NC (DA), and Washington, DC (MJ, RS); Integrated Healthcare Association (DY), Oakland, CA.

Source of Funding: Gordon and Betty Moore Foundation.

Author Disclosures: Mr Anderson presented at the September 2019 Rise West conference on risk adjustment. The remaining authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (DA, MJ, RS); acquisition of data (DY); analysis and interpretation of data (DA, DY, RS); drafting of the manuscript (DA, MJ, RS); critical revision of the manuscript for important intellectual content (DA, MJ, RS); provision of study materials or patients (DA); obtaining funding (DY); and administrative, technical, or logistic support (MJ).

Send Correspondence to: David Anderson, MSPPM, Duke University Margolis Center for Health Policy, 100 Fuqua Dr, Box 90120, Durham, NC 27708-0120. Email: david.m.anderson@duke.edu.


1. Meghani SH, Hinds PS. Policy brief: the Institute of Medicine report Dying in America: Improving quality and honoring individual preferences near the end of life. Nurs Outlook. 2015;63(1):51-59. doi:10.1016/j.outlook.2014.11.007

2. Aldridge MD, Bradley EH. Epidemiology and patterns of care at the end of life: rising complexity, shifts in care patterns and sites of death. Health Aff (Millwood). 2017;36(7):117-1183. doi:10.1377/hlthaff.2017.0182

3. Primary Care First Model Options. CMS. Updated April 8, 2020. Accessed December 20, 2019. https://innovation.cms.gov/innovation-models/primary-care-first-model-options

4. Direct Contracting Model Options. CMS. Updated April 20, 2020. Accessed December 20, 2019. https://innovation.cms.gov/innovation-models/direct-contracting-model-options

5. Direct Contracting Model: global and professional options: request for applications. CMS. November 25, 2019. Accessed December 20, 2019. https://innovation.cms.gov/files/x/dc-rfa.pdf

6. Olson A, Harker M, Saunders R, Taylor DH Jr. Innovations in Medicare Advantage to improve care for seriously ill patients. Duke Margolis Center for Health Policy. July 2018. Accessed December 20, 2019. https://healthpolicy.duke.edu/sites/default/files/atoms/files/innovations_in_medicare_advantage_to_improve_care_for_seriously_ill_patients.pdf

7. Improving serious illness care in Medicare Advantage: new regulatory flexibility for supplemental benefits. Duke Margolis Center for Health Policy. December 9, 2019. Accessed December 20, 2019. https://healthpolicy.duke.edu/publications/improving-serious-illness-care-medicare-advantage-new-regulatory-flexibility

8. Blue Shield of California expands palliative care program, offers home-based care statewide. News release. BlueCross BlueShield; March 15, 2018. Accessed December 20, 2019. https://www.bcbs.com/press-releases/blue-shield-of-california-expands-palliative-care-program-offers-home-based-care

9. Bleser WK, Saunders RS, Winfield L, et al. ACO serious illness care: survey and case studies depict current challenges and future opportunities. Health Aff (Millwood). 2019;38(6):1011-1020. doi:10.1377/hlthaff.2019.00013

10. Kelley AS, Bollens-Lund E. Identifying the population with serious illness: the “denominator” challenge. J Palliat Med. 2018;21(S2):S7-S16. doi:10.1089/jpm.2017.0548

11. Kelley AS, Covinsky KE, Gorges RJ, et al. Identifying older adults with serious illness: a critical step toward improving the value of health care. Health Serv Res. 2017;52(1):113-131. doi:10.1111/1475-6773.12479

12. Drake C. What are consumers willing to pay for a broad network health plan? evidence from Covered California. J Health Econ. 2019;65:63-77. doi:10.1016/j.jhealeco.2018.12.003

13. O’Donnell J, Saunders RS, Japinga M, et al. Expanding payment reforms to better incentivize chronic care for degenerative joint disease. Health Affairs. April 23, 2018. Accessed December 20, 2019. https://www.healthaffairs.org/do/10.1377/hblog20180416.346268/full/

Related Videos
Dr Jeffrey Sippel
Lalan Wilfongd, MD, US Oncology Network
Ben Urick, PharmD, PhD
Bincy Abraham, MD
Screenshot of Laura Wingate in a Zoom interview
Screenshot of Haley Naik, MD, smiling
Screenshot of Byron Lam, MD
Screenshot of Charles Wykoff, MD, PhD
Related Content
© 2023 MJH Life Sciences
All rights reserved.