Current quality measures usually assess performance against a standard of care for a typical patient rather than the specific patient receiving care, much like value frameworks assess value for a typical patient. Changes to quality measures are needed to encourage consideration of patients’ goals for care and wider use of shared decision-making at the point of care. As patient preferences are captured through these processes, opportunities exist to aggregate this information, stratify patients by preference, and establish a more patient-centered approach to value assessment.
A defining aspect of personalized medicine is that patient preferences should be incorporated into care decisions.1 Care decisions that have typically been made by clinicians should, in a world of personalized medicine, also include patient assessments of value according to their preferences. But the integration of patient preference into health care decision-making should not end at the point of care. We agree with Dr Dubois that, in the new era of personalized medicine, value frameworks for health care services, therapies, and technologies should better capture value from the patient perspective.
Dr Dubois highlights the fundamental tension between patient-level personalization and population-level standardization of care that has emerged in the movement to value-based health care. Dr Dubois’ assertion that value framework methods should consider differences in value across groups of patients by their individual characteristics and preferences for care is well reasoned. Better capturing patient assessment of value in framework methods is particularly important as value assessment is increasingly driving value-based coverage and payment decisions.
This tension also impacts quality measurement of personalized medicine. Existing quality measures usually assess performance against a standard of care for a typical patient rather than the specific patient receiving care, much like value frameworks assess value for a typical patient. We assert that, as for value frameworks, modifications to the current approach to quality measurement are needed to address the tension between standardized quality measures and patient preferences.
In this commentary, we propose certain changes to quality measurement that would increase the extent to which patient preferences are elicited and factored into goal-setting and clinical decision-making. With increased capture of patient preferences, it would be possible to aggregate preferences and stratify patients into preference categories for value frameworks, in line with Dr Dubois’ vision.
How Quality Measures Need to Evolve
The current approach to quality measurement largely focuses on assessing whether a clinician met a standard of care, regardless of patient preference. Quality measurement has traditionally relied heavily on clinical guideline-driven process measures. Even outcome measures typically focus on clinician goals, such as decreasing mortality, often exclusive of patient interests in maintaining quality of life.
However, certain types of measures encourage elicitation of patient preferences in conversations between clinicians and patients. These include measures of patient goal setting and shared decision-making. In addition, measures of whether a clinician met a certain standard of care or achieved an outcome can be modified to incorporate patient preferences.
Measuring Patient Goal Setting, Care Planning, and Concordance
Measurement can be used to assess whether a clinician and patient discussed care goals and partnered to establish a care plan. Moreover, the care plan can be used over time to assess concordance of care with the jointly established goals.
Measures of whether a care plan was developed are becoming more common. The Centers for Medicare & Medicaid Services (CMS) Oncology Care Model includes this type of measure.2 However, measures rarely assess whether the care received is concordant with the plan. Measures of care plan concordance were recently proposed for quality measurement and value-based payment programs for the seriously ill.3 Use of these types of measures, attached to provider incentives, can help ensure that patients’ assessments of the value of care alternatives are influencing decision-making.
Measuring Shared Decision-Making
Measures can encourage wider use of shared decision-making. As Dr Dubois noted, shared decision-making is a way for clinicians to partner with patients to thoroughly consider the treatment options, the pros and cons of each option, and the potential outcomes so patients can make informed care decisions in concert with their clinicians. Ultimately, shared decision-making is a form of patient value assessment at the point of care for selecting the right care option for the patient.
A recent study showed that while shared decision-making has become more pervasive over time, it is still not standard practice.4 Measures of shared decision-making are included in some CMS payment programs, such as the Medicare Shared Savings Program (MSSP)5 and the Merit-Based Incentive Payment System (MIPS).6 Wider and more consistent use of shared decision-making measures with strong financial incentives attached can help establish it as the standard of care.
Allowance for Patient Preference in Other Measures
There are numerous quality measures that assess whether a certain recommended service or treatment was received by a patient. Some measures of this type, such as “Breast Cancer: Hormonal Therapy for Stage I (T1b)-IIIC Estrogen Receptor/Progesterone Receptor (ER/PR) Positive Breast Cancer,” make allowances for a patient refusing a certain treatment. In this breast cancer treatment measure, patients are excluded from the measure denominator if they refuse the treatment.7 However, these types of exclusions are not consistently used across measures, setting up a scenario where a clinician may be penalized in a value-based payment program for following a patient’s wishes.
Rather than simply assessing whether a service was received, additional measures can be modified to assess whether appropriate care was ordered by the clinician and exclude patients who refused the treatment. This would remove patients from the measure denominator based on their preference.
Toward a Patient-Centered Approach to Assessing Value
Provider quality measurement presents an opportunity to improve elicitation of patient goals and preferences across multiple conditions. Increasing the frequency of patient value assessments and aggregating the information would enable better understanding of population-level preferences for certain conditions. Based on that understanding, it would be possible to stratify patients into preference categories and establish a patient-centered, or “bottom-up,” approach to assessing value, in line with Dr Dubois’ vision for patient-driven value assessments.
This approach would require investment in new data capabilities. Currently, the outputs of patient goal-setting and shared decision-making processes are not deidentified and aggregated for other purposes. Additional validated tools for assessing goals and preferences within specific conditions are needed. The capture of this type of information in electronic health records also needs to be standardized, and application program interfaces (APIs) are needed so that the data can be translated and transferred to databases.
While investment would be required, this bottom-up approach to assessing value is essential for a patient-centered health care system. In the alternative scenario, patients who may see more benefit from a certain service or treatment than the typical patient may be denied access through one-size-fits-all standardized quality measurement, coverage, and value-based payment. Meeting the promise of personalized medicine will require this type of patient-driven approach to assessing value of care.
To read the initial viewpoint this article references, click here.
References
1. Prainsack. B. Personalized Medicine: Empowered Patients in the 21st Century? New York, NY: NYU Press; 2017.
2. Center for Medicare and Medicaid Innovation. Oncology Care Model Key Drivers & Change Package. https://innovation.cms.gov/Files/x/ocm-keydrivers-changepkg.pdf. Published August 4, 2016. Revised June 1, 2018. Accessed November 27, 2018.
3. Gordon and Betty Moore Foundation. Quality Measurement and Accountability for Community-Based Serious Illness Care. https://www.moore.org/docs/default-source/default-document-library/quality-measurement-and-accountability-for-community-based-serious-illness-care-final.pdf. Published November 2017. Accessed November 27, 2018.
4. Levine DM, Landon BE, Linder JA. Trends in patient-perceived shared decision making among adults in the United States, 2002-2014. Ann Fam Med. 2017;15(6):552-556.
5. Centers for Medicare & Medicaid Services (CMS). Medicare Shared Savings Program Accountable Care Organization (ACO) 2018 Quality Measures Narrative Specifications Document. https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/sharedsavingsprogram/Downloads/2018-reporting-year-narrative-specifications.pdf. Published January 20, 2018. Accessed November 27, 2018.
6. Centers for Medicare & Medicaid Services (CMS). Quality Payment Program, 2018 Quality Measures. qpp.cms.gov website. https://qpp.cms.gov/mips/explore-measures/quality-measures?py=2018. Accessed November 27, 2018.
7. Centers for Medicare & Medicaid Services (CMS) and Office of the National Coordinator for Health IT (ONC). Breast Cancer: Hormonal Therapy for Stage I (T1b)-IIIC Estrogen Receptor/Progesterone Receptor (ER/PR) Positive Breast Cancer. ecqi.healthit.gov website. https://ecqi.healthit.gov/ecqm/measures/cms140v5. Accessed November 27, 2018.
