The previous article in this series explained some of the ways payers, providers, and digital health companies demonstrate that they have made a positive difference in their patients’ lives.
Jim Wallace, CEO of DecisionRx, writes to me that, “Payer incentives for wellness are largely stuck at the water’s edge, so to speak: few payers have gotten a lot of people to go in for a swim.” Their incentives (health club memberships, subsidies for over-the-counter healthcare products) haven’t changed their members’ behavior much. Wallace thinks that the wellness programs will have to engage much more intimately with members through digital monitoring and nudges: “This is the holy grail of wellness.”
This article looks at the measurements and reporting strategies used by some specific companies. Their interventions, goals, and measures depend a lot on who’s paying for the service.
Wondr Health Gives Employers What They Ask For
Wondr Health, like so many companies in this article and in the value-based healthcare space generally, works with employers and payers. Ultimately, for many employers, the purpose of engaging Wondr Health’s behavioral interventions is to improve employee productivity.
According to chief medical officer Dr. Tim Church, some of Wondr Health’s clients focus on cutting down the cost of health care claims, which can require one to two years of measurement. Some are looking for clinical improvements, and some even for culture change. The employer’s goals determine the interventions that follow.
Wondr Health does ask for employee self-assessments. Many employees using the service report sleeping better, feeling more confident and energetic, and other quality-of-life changes.
Church also pointed out that interventions that might work in direct-to-consumer services or other contexts might not work in employer programs. He says that monetary rewards usually don’t work and are rarely offered by their clients. Lasting change usually comes from self-motivation and is ingrained within company culture.
DecisionRx Runs a Study With Specific Measures
The importance of carefully choosing what you measure can be illustrated by DecisionRx. Their platform combines pharmacogenomics and medical history analytics to determine the risks of medications, which can be raised by interactions with other medications (perhaps unnoticed by the doctor doing the prescribing), and patient vulnerabilities related to age and medication conditions. The company cites studies such as this one that followed patients over several years to assess the benefits of pharmacogenomics programs.
Two types of benefits were noted: medical cost savings and “healthcare resource utilization.” The latter measure, a process indicator, doesn’t tell us the patients’ health or feelings of wellness. But I can readily accept that patients are better off if they don’t go to the emergency room and spend fewer days as inpatients.
Wellframe Matches Control Subjects to its Members
Wellframe, a health management platform with many applications used by health plans, faces a measurement dilemma typical of many digital platforms, according to executive vice president and general manager Jake Sattelmair. The company can’t engage in the gold standard of randomized clinical trials for many reasons: the difficulty of finding subjects, the time required, and the ethical quandary of denying care to half of the subjects. Wellframe is strongly committed to validating its platform, so it uses a different path to causal inference.
The technique involves finding data from public sources such as Centers for Medicare & Medicaid Services (CMS). Patients in the Wellframe program are matched with control subjects from the outside source, based on typical demographic criteria such as clinical trials use: age, gender, clinical conditions at baseline, the subject’s historical use of health care, etc. Basically, patients are matched with others who have similar risk, after being assigned a rating though a process called propensity matching. The technique controls for unknown variables that affect the digital health intervention, such as social or environmental factors.
Wellframe can then compare outcomes, as a clinical trial does. One of their published papers illustrates their research. Some of the measures they use include care utilization and cost, process indicators (engagement, medication adherence, retention), and outcomes (changes in biometrics, changes in self-reported quality of health).
Sattelmair says that Wellframe uses claims data to determine patient outcomes, but that doing so has drawbacks. It might take twelve to eighteen months to get the claims data. Therefore, they base initial analyses of their interventions on process indicators, which they can obtain quickly, and then incorporate outcome data later.
Sattelmair also sees weaknesses in before-and-after assessments, because you can’t prove that your intervention caused a change. People undergoing treatment are changing in other ways too; they may choose to start an exercise program or undergo a financial change, for instance.
Resilient Healthcare Employs Extensive Data Analytics
Resilient Healthcare provides in-home care for conditions that have traditionally required a hospital stay. Doing so presents a challenge, not just to reproduce the clinical functions that a well-furnished hospital can offer—nursing, infusions, medications, even meals—but to monitor the patient for safety reasons and measure their progress. Payers are naturally interested in evidence that Resilient Healthcare can provide comparable care and reduce readmissions.
Jackleen Samuel, president and CEO, says that they analyze patient demographics, clinical information, and process indicators such as medication compliance and service utilization. For instance, how often do patients call in to their provider, demonstrating that they’re engaged in their health? The system tries to measure whether patients are learning more about their illness, and are becoming more comfortable with their physicians. She says there is “lots of data to slice and dice.”
The results? Data shows that in-home care leads to a shorter course of treatment, lower utilization of the health care system, and better medication adherence.
The pandemic has significantly boosted in-home hospital care, just as it did telemedicine. Samuel says that before the pandemic, there were just three hospitals in the U.S. offering in-home treatment. Medicare launched a payment system for in-home treatment during the pandemic, and now more than 200 hospitals have been approved by Medicare for the service.
DUOS Adopts Measures Wanted by Medicare Advantage Plans
DUOS serves Medicare Advantage plans, connecting their patients to health resources ranging from primary care visits to transportation. Karl Ulfers, co-founder and vice president of growth, says that the Medicare Advantage insurers focuses on three measures: activation (that is, getting patients to sign up for DUOS), benefit use, and Star ratings.
Activation is an obvious measure because the service can’t help members if they don’t sign up. By measuring activation, the programs give DUOS an incentive to make their service enticing and engaging. They have achieved 85% activation among plan members.
Benefit use is defined by the Medicare Advantage programs to include their own benefits and benefits offered by government programs. The reason the programs like to measure this is that the patient can exploit these benefits without paying extra, and therefore rate the programs more highly. DUOS members achieve 80% compliance on getting in-home visits or primary care.
Star compliance, described in the previous article in this series, covers measures such as whether the members have gotten health screenings recommended by their physicians. DUOS reaches 70% on these measures, less than benefit use because the measures are more complex and depend on a variety of activities, such as seeing a medical specialist.
TimeDoc Health Uses Measures for Both Treatment and Evaluation
TimeDoc Health also has a rigorous process for measuring outcomes. The company is a virtual platform for care coordination, including remote care.
Vice president of clinical strategy Sarah Cameron says that the platform first helps the provider and patient choose a treatment plan. Measurable goals, such as reducing the A1C of a diabetic patient or signing up for a screening, are set as part of the plan.
Measurements of progress toward these goals are uploaded to the EHR, as structured data when possible and otherwise as a PDF. By checking progress (or lack thereof) toward the patient’s goals, a care coordinator or clinician can reach out to find out how to help the patient.
For instance, did the patient have trouble getting medication? And was that due to cost, transportation problems, or time pressures? And why didn’t the patient schedule a colonoscopy? If they’re leery of such procedures (many people are!), can their provider approve a substitute?
The administrators can then exploit the measurements to earn incentive payments and better Star ratings.
One study conducted by TimeDoc Health demonstrated significant improvement in one outcome (remission of symptoms) and one process indicator (adherence to treatment).
The final article in this series will look at a psychology, a discipline where objective measurement is difficult, and at standards in wellness.
