By Anupam Goel, Chief Health Information Officer, UnitedHealth Group
In 2017, each American spent $10,739 on healthcare costs. All of us in healthcare look to identify opportunities to increase value before we become disrupted by someone outside the industry. Several medical professional organizations have supported initiatives to reduce low-value care (disclaimer: I’ve been a member of the American College of Physicians and certified by the American Board of Internal Medicine since completing my residency). Unfortunately, the evidence supporting the contention that these educational initiatives reduce low-value care is weak.
Health systems and payers may be more comfortable deploying these patient-centered tactics once a patient develops an acute or chronic condition.
Health insurance companies use prior authorization to ensure high-value healthcare delivery. In some cases, prior authorization reduces inappropriate medication use (e.g., opioid abuse and overdose among Medicaid beneficiaries, antimicrobial stewardship). In other cases, prior authorization increases costs without reducing inappropriate care (e.g., low back pain with increasing costs due to more spinal injections and hospital admissions,) or impairs healthy behavior change (e.g., reducing tobacco use). In January 2018, several organizations, including the American Medical Association, America’s Health Insurance Plans and the American Pharmacists Association, issued a consensus statement suggesting five methods to improve the prior authorization process.
Some groups have tried other approaches to increase high-value healthcare delivery without resorting to prior authorization. National guidelines delivered through real-time decision support can replace traditional prior authorization in ordering chemotherapy and high-cost imaging. One study found a high-deductible health plan linked to value-based pharmacy benefits with free chronic disease medications increased medication adherence rates among patients with initially low levels of adherence and higher socioeconomic status. But these interventions, like prior authorization, do not address the primary drivers of chronic disease (tobacco use, poor diet and physical activity linked to obesity, excessive alcohol consumption, uncontrolled high blood pressure and hyperlipidemia) in America.
If behavior change is the root cause of chronic disease, how might we support patients make behavioral changes to reduce their risk of chronic disease, the ultimate high-value intervention? In 2009, Polly Ryan outlined the Integrated Theory of Health Behavior Change. The behavioral model suggests a patient’s knowledge, beliefs and social facilitation impact self-regulation skill and ability. The self-regulation skill and ability affect engagement in self-management behavior which subsequently impacts health status. Patient engagement precedes changes in health status. Dixon-Fyle et al. outline a multi-level paradigm using technology to support patients change health behavior with peers, caregivers and clinicians. The group suggests cognitive biases, habits, and social norms with a focus on the patient rather than a specific disease.
Even under the most advanced value-based arrangements available today, behavior change seems only tangentially related to commercial healthcare insurance premiums or healthcare system payments. Payors receive higher payments for having more patients with well-controlled chronic diseases rather than preventing patients from developing chronic disease. Employees often switch health insurance plans to meet changing needs in their lives, reducing the incentive for any health insurance company to support behavior change with expected improvements five or 10 years in the future. Of all the different members of the healthcare landscape, Medicare and Medicaid may be the entities most aligned to help patients support behavior change to prevent chronic disease.
Considering Ryan’s theory of health behavior change and Dixon-Pyle’s model that acknowledges our irrational behaviors, what tools might be useful to help patients support behavior change to prevent chronic disease? Assessing patient preferences and current health behaviors would allow patients and other interested parties gauge each individual’s readiness to change and possible interventions to support healthy behavior change. Framing the risks and benefits of medical decisions around disease prevention and health promotion in patient-friendly terms (e.g., overall mortality, missed days of work, quality of life, out-of-pocket costs) instead of disease-centric ones (e.g., disease-specific mortality, blood pressure reduction) could help patients connect specific actions and their corresponding outcomes.
Health systems and payers may be more comfortable deploying these patient-centered tactics once a patient develops an acute or chronic condition. Now that I have a diagnosis of knee pain, eliciting my preferences and functional status can help providers and payers suggest specific interventions to consider (e.g., physical therapy, acupuncture, non-prescription medications, joint injections) rather than force a specific sequence of interventions as suggested by some prior authorization workflows. This paradigm supports the possibility of healthcare mass-customization consistent with what other industries have already done.
Regardless of the specific health behavior target, Ryan and Dixon-Pyle suggest peer support for new behaviors. With various public and private social media networks, patients could be connected with others who have similar challenges or health states to determine what normative behaviors exist in that network. Rather than relying on local friends, a patient could compare their behaviors and health state against others across the country. The next-level of engagement would be linking specific health behaviors with corresponding health states.
Doing more of what we’ve always done to provide higher healthcare value (e.g., education, prior authorization) seems unlikely to meaningfully bend the healthcare cost curve. Engaging patients around their own health states and preferences could support a generalizable model of health behavior change for primary prevention and chronic disease management with opportunities for healthcare to deliver personalized diagnostic or treatment choices. The patient-level information could then be aggregated across similar populations to leverage social networks to nudge different health behaviors. Although this work may require a fundamental redesign of how we interact with patients across health systems and payer entities, the pivot represents an opportunity to simultaneously engage patients and deliver higher-value healthcare as defined by them.