Patient Preferences Matter in Value-Based Care…But They are Missing from Measurement
“Value” may be the most-used buzzword in healthcare today. The US spends nearly three times as much per person for healthcare as any other Organization for Economic Cooperation and Development (OECD) country.¹ But the US has only mediocre outcomes in overall mortality, infant mortality, maternal mortality, deaths from cancer, and suicide.² Public and commercial payers are working hard to move from a siloed, fragmented fee-for-service system to value-based care.³ The desire to extract more value out of our $3.3 trillion healthcare system drives these efforts. But it is not always clear what “value” means in “value-based care,” “value-based payment,” or “value-based purchasing.”
To evaluate how well the system is accounting for non-cost components of value, I will consider definitions of value and examine how effectively non-cost components of value are currently measured.
One way to define value is “patient health outcomes achieved per dollar spent.”4 Other definitions are characterized as an equation: Value = Quality/Cost, or Value = (Quality + Service)/Cost.5 This looks like a straightforward equation, but a closer look suggests that it is not so simple. Players in the healthcare system define value differently, in part because patients, employers, government and private insurers, healthcare organizations, and healthcare workers receive different rewards, depending on the definition. Although each player contributes to the system, I endorse the assertion that “the creation of value for patients should determine the rewards for all other actors in the system.”4
Patients’ Healthcare Values
A 2018 University of Utah report5 demonstrated how perspectives on value differ. Patients’ views are multi-dimensional and complex. When asked to select five of 19 value statements, no single value statement was endorsed by more than half of the patient participants. Patients prioritized care access (affordability, timely appointments, convenient location, and reasonable wait time). They also prioritized provider expertise, relationship with the provider and staff, understanding test results, and improved health. In contrast, physicians’ values were more homogenous: 5 of the 19 value statements were endorsed by 49% of physician participants. There was some overlap between patient and physician values, but physicians did not prioritize any values associated with access. Only after a patient is in the office do patient and physician values begin to align. Both emphasized the physician/patient relationship and improved health.
Through a different lens, a recently published article I co-authored explored patient values using a qualitative study.6 The study’s purpose was to develop a taxonomy of patients’ healthcare goals when they are living with long-term services and support needs (e.g., assistance with feeding, bathing, getting dressed, and walking or getting from bed to a wheelchair). These patients are among the highest healthcare users because of their complex medical and social needs. More than half of the patients’ priorities were functional, such as being able to walk. Participants focused on wanting to live life, stay in their homes, spend time with family and friends, travel, and go to church. Some patients prioritized receiving the right medical care – neither too much nor too little. Managing pain and symptoms was a priority for many participants. Finally, many patients emphasized the need to be listened to and heard by their physician or other providers.
Together these two studies emphasize patients’ concerns for access, improved or stable health and function, and effective communications and relationships with clinicians and office staff.
If patients’ values are to be captured in value-based care, there needs to be an emphasis on what the value equation calls quality and service. Policies and law governing value-based care models within Medicare and Medicaid emphasize the cost side of the equation. Under the Patient Protection and Affordable Care Act (ACA),7 which created a range of value-based care programs, a program either had to maintain spending levels and improve care quality or decrease spending while maintaining quality. Patients’ “out-of-pocket costs,” is the only cost value patients endorsed, while all other values fall on the quality side (within the Centers for Medicare and Medicaid Services (CMS), quality includes service). For a value-based care system to reflect patient values in a manner consistent with the value equation above, the quality system would have to address and measure patients’ out-of-pocket cost burden and non-cost values.
I reviewed measures in the CMS Measures Inventory Tool8 to evaluate whether measures being used reflect patients’ values. Most performance measures in CMS programs are process measures. They measure how often a specific process is used under specific circumstances or within specific populations. Process measures are justified on the theory that processes linked to desired outcomes in research studies, if performed widely, will improve patient outcomes. For example, the Transitions of Care HEDIS® measure evaluates four processes associated with hospital discharge, including “Evidence of patient engagement (e.g., office visits, visits to the home, telehealth) provided within 30 days after discharge.”9 Certain transition processes are associated with better patient outcomes, but measure stewards are not required to provide evidence that implementing a process measure improves patient-prioritized outcomes, such as improving or stabilizing health and function. The Executive Director of the Medicare Payment Advisory Commission testified to Congress in 2014, “Current quality measures are overly process oriented and too numerous, they may not track well to health outcomes….”10 Six years later, process measures still dominate. Among 716 implemented measures found in the CMS Measures Inventory Tool, 53% are process measures, compared to 5% that are patient-reported outcome measures.
To focus on the adequacy of outcome measures in use, I searched the CMS Measures Inventory Tool for outcome measures currently in use in hospital quality programs. I focused on implemented and final measures classified as “outcome,” “patient-reported outcome,” or “patient engagement/experience” within hospital-based measure programs. The search netted 48 measures. This included all 13 outcome measures in the Hospital Value Based Purchasing program. I chose hospital measures because the subset of all measures captured was feasible to analyze in a brief survey for this blog, and, while not identical, the types of measures used in hospital measurement programs align with those used across other CMS programs. Where relevant, I also highlight similarities with other measurement programs, such as clinician and group measures.
Hospital outcome measures that measure the quality component of the value equation fall into broad categories: 1) Consumer Assessment of Healthcare Providers and Systems (CAHPS) patient experience measures, 2) preventable healthcare harm (safety), 3) admission, readmission, and length of stay, and 4) other. Twelve are safety measures that detect sub-par care that results in patient harm, such as hospital-acquired infection, operating on the wrong site, pressure ulcers, and burns. More than half of the 43 hospital “outcome” measures address admissions, readmissions, and length of stay. The remaining measures include factors such as time in the emergency department before in-patient admission, patient perception of discharge and understanding of discharge instructions, stroke education, and ventilator liberation rate.
CAHPS measures, which are based on a survey completed by the patient or a surrogate after receiving care, most directly address values associated with service and access across care delivery settings. The Hospital CAHPS captures being treated with courtesy and respect, providers who listen, and managing care after discharge.11 The Clinician and Group CAHPS survey instruments capture patient perceptions of respect and listening, along with access values such as getting appointments as quickly as needed and follow-up with test results.12
Safety is critically important. Most safety measures cover “never events”: those that are not acceptable under any circumstances. Examples include burns, falls, or operating on the wrong site. There is widespread support for these measures among experts. Patients tend not to prioritize avoiding harm because many people do not appreciate the risks associated with receiving healthcare but avoiding harm in hospital encounters is certainly aligned with improving or maintaining health and function.
The 23 utilization measures I reviewed, including admission, readmission, emergency department visits, and length of stay, are in place to avoid unnecessary risk, decrease cost, and deliver care in the most appropriate setting. These are laudable goals, but the measures are based on overall rates, not on quality at the individual level. So an “avoidable readmission” rate of 30% is higher than patient safety experts think is acceptable, but measures do not differentiate avoidable from appropriate readmissions. Take the case of a patient who is critically ill and dies or has functional decline because she was diverted from a needed readmission. The diverted readmission saves the system money and improves the hospital’s readmission rate, but the patient did not receive value-based care. If a hospital decreased its readmission rate from 30% to 15%, there is no way to know whether the avoided readmissions were unnecessary and therefore avoidable. Because current measures do not link readmissions or avoided admissions to patient health and function, there is no way to know whether the measure is improving or maintaining patient health and function.
Among the remaining seven “outcome” measures, three were process-focused, not outcome-focused: time from emergency department (ED) admission to decision to admit a patient to an in-patient unit, time from ED admission to in-patient admission, and whether stroke education was provided. The Care Transition Measure is a survey that captures patient discharge experience. Only three measures addressed patient health and function: improved visual function 90 days after cataract surgery, healthy term newborn, and ventilator liberation rate.
In other words, for most hospitalized patients, there are no hospital quality measures that evaluate improved health or function.
The state of outcome measurement in other settings aligns with hospital outcome measures, except that Skilled Nursing Facility and Home Health measure programs track some functional outcomes. While some health and functional status measures are included in the physician quality measurement program (MIPS) there are incentives to report claims-based process measures and avoid outcome measures.
Patients value access, relationships, and improving or maintaining health and function. CAHPS surveys capture some aspects of access and relationships. But only a small subset of hospital measures, and measures across other programs, capture health and function outcomes that are of central importance to patients.
To move toward quality measurement that captures patients’ values, I recommend system-wide changes to capture health and functional status outcomes, and a systematic linking of outcome measures to value-based payment models. This recommendation echoes the call to action for outcome measures in oncology care published as this blog was finalized.13 This is critical: when a value equation fails to account for patient values in quality measures, then value-based care is not value-based, it’s just cheaper. In some cases, that is better for patients, but in many cases, it is not.
CVP Health experts have researched and established healthcare quality policy and led measure development efforts. We come to measurement with a pragmatic understanding of measurement challenges, and, as this blog demonstrates, a patient-focused view of value. We use data to evaluate the efficacy of current measurement approaches, such as risk adjustment. With this approach and expertise, CVP contributes to creating a path to a patient-centered, value-based care system that benefits all Americans.