The Burden of Chronic Diseases and Their Complications

By Tony Holbert
December 14, 2018

About six in ten Americans are living with one or more chronic diseases including hypertension, diabetes, asthma, chronic obstructive pulmonary disease (COPD), and congestive heart failure. Chronic diseases and their complications are the leading cause of direct medical spending and a major cause of indirect costs such as employee absences and decreased productivity.1 Take diabetes, for example, where the costs in the U.S. were $327 billion in 2017, with $237 billion in direct medical costs and $90 billion in reduced productivity, an amount that has increased 26% over the preceding five years.2

Chronic diseases

Chronic diseases

The burden of chronic disease in the U.S. and worldwide has sharply increased over the last two to three decades and is now so significant that both the Centers for Disease Control and Prevention3 and the World Health Organization4 consider both prevention of chronic disease and prevention of chronic disease complications to be critical goals in improving national and global health. While prevention of chronic disease is a multi-factorial problem that touches on changes that are occurring in demographics, lifestyles, and diet, the prevention of chronic disease complications is clearly within the domain of medical care. Our medical system, however, was formed in response to treating acute and episodic illnesses. Dr. Rifat Atun, director of the global health systems cluster at the Harvard T. H. Chan School of Public Health, commented to The New York Times that “The transition in terms of illness patterns has happened very quickly, but the health system transition has not.”5

The complications are important—they can lead to more morbidity and more cost than treatment of the underlying chronic disorder. Going back to our diabetes example, of the $237 billion in direct medical spend in the U.S., 30% was for inpatient care and 30% was for medications to treat diabetes complications, while only 15% was for drugs to treat diabetes and supplies like glucometers and glucose test strips, and 13% was for physician office visits.2 So a full 60% of the cost of diabetes care is due to potentially avoidable complications and hospitalizations. And these amounts don’t even factor in the personal impact of living with complications such as kidney failure, amputations, blindness, painful neuropathy, or lost time due to avoidable hospitalizations.

avoidable complications chronic disease

Tackling the problem

Why aren’t we doing better at preventing complications of chronic diseases? The sad irony is that we know of many interventions that work to avoid complications, we just aren’t implementing them. Some examples:

  • In 2006-2007, only 45.7% of patients with coronary artery disease who saw a doctor left with documented advice, instructions, or a prescription to take aspirin or another antiplatelet drug, a proven intervention that can prevent heart attacks.6
  • In 2005-2008, only 63.2% of patients who were prescribed medications to treat hypertension were taking them, and only 43.7% had their hypertension under optimal control to prevent complications.7
  • In 2008, only 33.4% of patients with asthma were given a written management plan with instructions on how to monitor and treat asthma exacerbations themselves to avoid preventable emergency room visits and hospitalizations.8
  • In 2009, only 77.1% of patients were referred to structured cardiac rehabilitation after having a heart attack.9
  • In 2005-2008, only 52.3% of patients with coronary artery disease and 33.4% of patients who had a previous stroke had LDL cholesterol levels treated to the optimal level for minimizing complications.10
  • In 2007, hospitalizations for heart failure exacerbations were alarmingly frequent—from 10 per 1000 for ages 65-74 to 38.7 per 1000 for over age 85.11 Despite often yielding significant reductions in heart failure admission rates,12 heart failure case management programs are not yet widespread.

These shortcomings have not been lost on health systems or insurers, who have been creating quality rating and incentive systems that try to move the needle on chronic disease care. Unfortunately, these legacy approaches to measuring care quality have run into concerning obstacles. Some measures ended up unfairly penalizing doctors who care for higher-risk patients13 (a failure of “risk adjustment,” which we’ll say more about later). Others have led to improvements in the things specifically measured, but with the side effect of reduced quality elsewhere as time and resources were redirected to meeting the measured items.14

In fact, measuring specific individual interventions turns out to be a “surrogate marker” for what we really care about, which is reduced complications. We would anticipate that a physician following all of the interventions we look at should have patients with fewer than expected complications, but this is not guaranteed. Minimal or half-hearted implementation of interventions may be less effective, or the measured interventions may crowd out unmeasured interventions that are equally or even more effective. When thinking about physician quality, it would be much better to measure the rate of complications, which is the outcome we really care about. This, however, is plagued by the risk-adjustment issue—we have to somehow account for the varying populations seen by each physician.

Grand Rounds’ approach

How does Grand Rounds find those doctors who guide their patients towards care that reduces chronic disease complications? We needed a new approach. Tracking raw metrics such as readmission rates suffers from the problem of risk adjustment—we don’t want to unfairly penalize doctors for caring for patients who are sicker or have barriers to care as can happen in that scenario. We don’t think that combing through documentation looking for specific interventions is the best approach either—the metric then focuses on what care was documented rather than what care was delivered. This leads to more administrative headaches and physician burnout as doctors strive to microscopically document everything that might be measured (and not much is worse for care quality than a burned out physician).

Instead, Grand Rounds focuses on patient-oriented outcomes—the things that actually matter to patients and their loved ones. We ask: “Did this patient suffer a complication attributable to his or her chronic disease?” But to solve the risk adjustment problem, we also leverage the power of machine learning algorithms to look across a very large dataset for other patients with similar risk factors and backgrounds. Machine learning lets you consider an exhaustive set of risk factors and use the data to pick out the factors that actually do affect the risk, rather than the imprecise legacy approach of trying to assign risk adjustments manually, based on clinical intuition. Once you know what the expected complication rate is for a particular patient, you can see whether a doctor’s results for similar patients are above or below that expected complication rate15 and infer how effective their care systems are at preventing these complications.

This strategy allows for other advantages as well. Looking at patient-level outcomes lets us see into the entire care system directed by a physician, not just what they say or do in documented billing encounters. The generalizable data strategy that Grand Rounds uses (stay tuned for part two of this blog series where data scientist Peyton Rose explains further) also allows us to attribute a patient’s care among primary care doctors and appropriate specialists so we can use similar metrics to understand both primary care and specialist care quality.

And finally, the fact that these chronic disease complication models are separate allows us to adjust their importance based on what we know about a specific member. If a patient is coming to us looking for a cardiologist to treat their congestive heart failure, we can give added preference to those cardiologists who we know have lower heart failure complication rates. In contrast, if a diabetic patient needs help finding a primary care provider, we can give preferential weight to our models that predict diabetes-specific complications — the end result: a patient seeing a doctor who delivers exceptional care in the precise area they need, and a doctor seeing a patient like those they are already great at caring for.

Read more Clinical Insights blog posts here.

Source

    1. “Multiple Chronic Conditions in the United States,” Rand Corporation, https://www.rand.org/content/dam/rand/pubs/tools/TL200/TL221/RAND_
      TL221.pdf, May 2017
    2. “Economic Costs of Diabetes in the U.S. in 2017,” American Diabetes Association
      Diabetes Care, http://care.diabetesjournals.org/content/early/2018/03/20/dci18-0007, March 2018
    3.  “Healthy People 2020,” healthypeople.gov, https://www.healthypeople.gov/2020/topics-objectives
    4. “Noncommunicable diseases and their risk factors,” World Health Organization, http://www.who.int/ncds/en/
    5. “Lives Grow Longer, and Health Care’s Challenges Change,” The New York Times, https://www.nytimes.com/2015/07/17/upshot/lives-grow-longer-and-health-cares-challenges-change.html, July 2015
    6. National Ambulatory Medical Care Survey, as cited in Healthy People 2020
    7. National Health and Nutrition Examination Survey, as cited in Healthy People 2020
    8. National Health Interview Survey, as cited in Healthy People 2020
    9. ACTION registry of the ACC (now the Chest Pain MI Registry), as cited in Healthy People 2020
    10. National Health and Nutrition Examination Survey, as cited in Healthy People 2020
    11. National Hospital Discharge Survey, as cited in Healthy People 2020
    12. “A Literature Review of Heart Failure Transitional Care Interventions,” American Journal of Managed Care, https://www.ajmc.com/journals/ajac/2017/2017-vol5-n3/a-literature-review-of-heart-failure-transitional-care-interventions, September 2017
    13.  “The Value-Based Payment Modifier: Program Outcomes and Implications for Disparities,” Annals of Internal Medicine, http://annals.org/aim/article-abstract/2664654/value-based-payment-modifier-program-outcomes-implications-disparities, February 2018
    14. “Effects of Pay for Performance on the Quality of Primary Care in England,” New England Journal of Medicine, https://www.nejm.org/doi/full/10.1056/NEJMsa0807651, July 2009
    15. Sports stats nerds may recognize this as analogous to the “wins above replacement (WAR)” statistic, though the computations involved are decidedly much more complex due to the need for risk adjustment.
Share:

➞ Back to Blog

Share

Tony Holbert

Tony Holbert

Dr. Holbert, MD, MBA, is a board-certified Family Medicine physician and Clinical Quality Program Manager at Grand Rounds. He joined Grand Rounds in 2015 from private practice at an innovative NCQA-accredited accountable care organization (ACO). Throughout his training, Dr. Holbert sought out experiences working in diverse healthcare systems, including private urban and rural settings, a major county hospital district, the VA, and an integrated ACO. He developed an interest in understanding the factors which influence how effective (or ineffective) our healthcare system is at addressing the changing burden of disease. He completed his residency at Baylor College of Medicine, where he served as chief resident and concurrently completed a Leadership Executive MBA at the University of Houston. He earned a Doctor of Medicine with High Honors at University of Texas Medical Branch at Galveston. Prior to medical school, Dr. Holbert received a degree in chemistry from the University of Texas at Austin and an advanced degree in physical chemistry from Harvard University. He went on to spend several years at National Instruments in Austin leading a team designing and programming measurement, control and automation software for scientific and engineering applications—and even making his own small contribution to the software running the Large Hadron Collider at CERN.