A Scoping Review of US Insurers’ Use of Patient-Reported Outcomes

ABSTRACT

Objectives: To complete a scoping review of US health insurers’ use of health-related quality of life (HRQOL) patient-reported outcome measures (PROMs).

Study Design: Literature review.

Methods: A literature search was constructed for articles that contained an insurer-related term and an HRQOL-related term between 1999 and 2019 using the MEDLINE, Embase, Web of Science, Cochrane Database of Systematic Reviews, EconLit, and Business Source Complete databases.

Results: The search identified 14,253 unduplicated records, of which 2340 passed abstract screening and 350 were included in the review. The populations addressed in these studies included both populations with specific health conditions (eg, diabetes) and an entire member population. The most common purpose of the article was to evaluate a policy or program (n = 255; 72.9%); the range of interventions evaluated included federal policy, subgroup management strategies, and identification of individual patients. The most common insurance mechanism was Medicare (n = 205; 58.6%). The most common source of data was collected specifically for a research project (n = 172; 49.1%), and the least common source of data was collected by providers at the point of care (n = 34; 9.7%). The most commonly addressed age group was 65 years and older (n = 262; 74.9%), and the least commonly addressed was younger than 18 years (n = 36; 10.3%). The most commonly used PROMs were single-item self-rated health (n = 138; 47.1%) and activities of daily living (n = 88; 30.0%), with validated depression questionnaires (n = 56; 19.1%) being the most common disease-focused questionnaire.

Conclusions: This review found a wide variety of articles across insurance providers, health conditions, and uses of PROMs. There is a noted paucity of data in pediatric populations and little information about the use of data collected within health care settings that is transmitted to health insurers.

Am J Manag Care. 2022;28(6):e232-e238. https://doi.org/10.37765/ajmc.2022.89162

Takeaway Points

The 350 articles identified in this review covered a broad set of health conditions and purposes for which US health insurers use patient-reported outcomes (PROs) about health-related quality of life. More than half the articles addressed Medicare populations. The most common PRO was a single question about self-rated health across all studies and within all insurance subgroups except the Department of Veterans Affairs. Sixty articles used nonvalidated PROs. PROs were used as an analytic outcome 41% of the time. There were few articles about PRO use in pediatric populations or about the use of PRO data collected at the point of care. This review:

• Provides a catalog of documented uses of PROs about health-related quality of life by US health insurers over the past 20 years.
• Provides context for future initiatives, such as value-based purchasing with PROs.
• Identifies gaps in the current literature that should be addressed in future research.

Demonstrating the value of health care is a central goal of the US health care system.^1,2 As financiers of health care, health insurance providers have a vested interest in value, as market competition demands that they show value to current and potential purchasers of their products. There are many definitions of value within a health care context, including, but not limited to, member satisfaction, quality, member outcomes, access to services, total cost, and out-of-pocket member costs.³

One important definition of value in health care is health-related quality of life (HRQOL)—that is, an individual’s subjective assessment of their emotional, physical, and social functioning.⁴ Standardized measures of HRQOL have a long history in research, with more recent use in clinical practice, and are an important component of achieving patient-centered care.^5,6 These measures are conceptually part of a broader set of patient-reported outcomes (PROs) that also encompass measures of satisfaction, engagement with care, and health care experiences, also known as patient-reported experience measures (PREMs).⁷ Here we focus on standardized PRO measures (PROMs) of HRQOL that could support health insurance payers in many aspects of their missions, such as monitoring population health, evaluating internal and external programs, monitoring the quality of different insurance products, identifying high-risk individuals, and determining benefit coverage.

Multiple-item questionnaires can be categorized by their breadth; disease-specific questionnaires focus on the symptoms, sequelae, and treatment effects of a disease such as migraine, whereas generic questionnaires give an overall holistic view of HRQOL.⁸ Multiple-item questionnaires can also be categorized by their construction; psychometric measures, such as the Short Form-36 (SF-36), measure the amount of a construct, whereas econometric measures, such as the EuroQol 5-dimension instrument, measure the value of a construct.

Outside the United States, standardized HRQOL measures have been used in value-based purchasing, such as cost per quality-adjusted life-year thresholds used in coverage decision-making.⁹ Inside the United States, PROMs are increasing in prominence as a way to measure the quality and value of care, as evidenced by increased inclusion in the National Quality Forum metrics¹⁰ and as a requirement of bundled payment agreements with CMS.¹¹ Likewise, US value-based purchasing initiatives are emerging, such as the Institute for Clinical and Economic Review’s collaboration with CVS Health Corporation to determine pharmacy benefit plans and the American Society of Clinical Oncology’s value framework.^12-14

To the best of our knowledge, there is not a comprehensive review of the extent, range, or characteristics of HRQOL PROMs used by US health payers to put these newer initiatives in context. Here, we provide a scoping review of the literature to address this knowledge gap.

METHODS

We followed the methodology outlined by Arksey and O’Malley,¹⁵ followed PRISMA-ScR (Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews) guidelines,¹⁶ and registered the protocol on Open Science Framework before beginning analyses.¹⁷ Authors held conference calls every 2 weeks and utilized ad hoc email communication.

Identification of Articles

A professional health science librarian (R.T.) prepared the search strategy, and the review was managed within the DistillerSR (Evidence Partners) online systematic review software. The search terms required an insurer-related term (eg, health insurance, health plan, payer) and an HRQOL-related term (eg, patient-reported outcome, self-rated health, Patient Health Questionnaire-9). We searched MEDLINE (via PubMed), Embase (Embase.com), Web of Science (Clarivate), the Cochrane Database of Systematic Reviews (Wiley), EconLit, and Business Source Complete (EBSCO). The search was run on September 24, 2019, and was limited to studies published after 1999. The full search strategy is in the eAppendix Table (available at ajmc.com).

Level 1: Abstract Screening

All abstracts were screened for scope. Excluded abstracts were those that were exclusively outside the United States, did not study a health insurance member population, or used only PROs related to PREMs or wearables. If these characteristics could not be assessed from the abstract, it was included in level 2. Each abstract was screened by 2 of the authors who reviewed articles (J.H., A.M.C., B.G.G., P.M., E.C.S.S., and S.M.K.), and any discrepancies were discussed until consensus was reached.

Level 2: Full-Article Screening

All articles were screened to verify that the article focused on a health insurance member population within the United States. Articles were included if (1) the study participants were sampled in part or in whole because they were a member of an insurance product (eg, Medicare members) or members of more than 1 insurance product (eg, patients who were either Medicare or Department of Veterans Affairs [VA] members); (2) the unit of analysis was health care providers who provide care within a managed care or health maintenance organization entity (ie, being a provider for a managed care entity was part of the inclusion criteria); or (3) the source was a commentary, white paper, or opinion that specifically discusses health insurers’ use of HRQOL PROs.

We then screened in studies about an insurer’s management of its population while screening out studies in which the health insurance member population provided an accessible frame for descriptive or epidemiologic studies (eg, studies that used Surveillance, Epidemiology, and End Results–Medicare to evaluate associations in populations with cancer).¹⁸ Included articles focused on population health management, such as the application of data, care models, and financing strategies to improve the health and utilization of an entire group of patients/members. Studies were included if there was clear indication that insurers could use results to inform quality of care, preventive care, or care/disease management or coordination activities; to reduce or control health care utilization and/or costs; to evaluate the effect of an insurance activity such as providing a specific intervention or a financing model or trying to influence when or how members get care; or to measure/monitor member health status. We included only articles for management of physical and behavioral health; we excluded those articles that solely focused on other aspects of insurance, such as dental, supplemental or “gap,” or long-term supports and services insurance.

Private health insurer data are typically proprietary; therefore, any study that is done using a single private insurer’s data is almost certainly for the use of the insurer, in some way. Determining whether Medicare or Medicaid “used” PROs for population health management was more challenging than for private health insurers. Public CMS data set access is widespread and use in research is common, so a study that focused on Medicare participants could not automatically be assumed to meet our inclusion criteria unless language was very explicit (eg, “researchers from CMS conducted this study…”). For government insurance products, we included a publication if it either (1) evaluated an existing or future CMS- or government-directed population strategy, initiative, or intervention (eg, bundled payments, benefit expansion, modifications of quality measures, demonstration projects, value-based insurance design) or (2) would directly inform identification and targeting of strategies, initiatives, or interventions to specific populations.

An HRQOL PRO had to be in the article to be included.

For study protocol articles or ClinicalTrials.gov sources, we performed a secondary search to determine whether there were published results. If so, we excluded the protocol sources. We also performed secondary searches when an article was excluded but clearly referenced a parent study or paper that met inclusion criteria. All articles were screened by 2 of the 3 team members (J.H., E.C.S.S., and S.M.K.) most experienced with insurance products, and any discrepancies were discussed until consensus was reached.

Level 3: Article Abstraction

We abstracted (1) type of research article (quantitative, qualitative, mixed-methods, protocol, white paper, commentary or editorial, review article or meta-analysis, or other); (2) years of data; (3) use of a comparison group; (4) sample size; (5) whether the insurer was part of an integrated delivery and financial system; (6) type of insurance plan (employer-sponsored, commercial, Medicare fee-for-service, Medicare Advantage, Medicaid, Children’s Health Insurance Plan [CHIP], Medicare Special Needs Plan, VA, or other; (7) a free-text description of the study population; (8) ages in the study population (0-17, 18-64, and/or ≥ 65 years); (9) study outcomes (HRQOL, physiologic outcomes [eg, health conditions, mortality], health care utilization, health care costs, quality of care, or other); (10) the HRQOL PROs used in the study; (11) what context the PRO was collected in (by a provider at the point of care, by an insurer in a routine health risk assessment, by an employer, for a separate research project, and/or other); (12) whether the PROs were an dependent variable, independent variable, both, or other; and (13) the purpose of the article (to evaluate an insurer policy, intervention, or strategy; to identify or stratify individual patients/members; to evaluate quality of care not related to a specific initiative; and/or other).

All articles were abstracted by a team member and 1 of the 2 most experienced team members (J.H. and S.M.K.) performed a quality assurance check. Any discrepancies were discussed until consensus was reached.

Analyses

Abstracted data were exported to Excel (Microsoft) and SAS 9.4 (SAS Institute) to complete data manipulation and analyses. Frequencies of PRO uses were counted for the quantitative and mixed-methods articles. For the nonquantitative articles, 2 reviewers (B.G.G. and P.M.) used Excel and Word (Microsoft) to create a summary of each study; then, major findings were synthesized to condense their themes.

RESULTS

We identified 14,253 unduplicated records, of which 2340 passed abstract screening and 350 passed full-article screening. Seven records were identified through secondary searching. Full details of the screening process are in the Figure.

The eAppendix Table includes the article title, journal, publication year, HRQOL PROs, and purpose(s) of the study. Table 1 presents summary statistics of the data abstracted from the articles. The majority of the articles (n = 299; 85.4%) were quantitative studies. Within these 299, 124 (41.4%) were retrospective analyses, 70 (23.4%) were prospective, 60 (20.1%) were randomized controlled trials, 40 (13.4%) were cross-sectional studies, and 4 (1.3%) were other.

Data used in these studies are most concentrated in the late 1990s and early 2000s, and publication dates are evenly distributed over this time period (data not shown). The most common data source was data collected specifically for a research project. A total of 103 articles (29.4%) discussed or used data routinely collected by insurance companies: 11 (10.6%) of these were Consumer Assessment of Healthcare Providers and Systems surveys,¹⁹ 9 (8.7%) were the Medicare Health Outcomes Survey,²⁰ and 37 (35.9%) were the Medicare Current Beneficiary Survey (MCBS).²¹ Thirty-eight studies (10.8%) used data from other sources, primarily data sets collected by state or federal agencies for multiple purposes, such as the Medical Expenditures Panel Survey and the Behavioral Risk Factor Surveillance Survey.^22,23 Just 34 (9.7%) articles discussed PROs or PROMs data collected by providers, and 10 of these were commentaries.

Seventy-five percent of the articles included individuals 65 years and older, but only 10% of the articles included children younger than 18 years. Of the 36 articles that included children, 22 were articles that addressed all ages, such as commentaries (n = 7) or reviews (n = 2).

Table 2 displays the validated PROMs that were used in at least 5 quantitative or mixed-method articles and how the PROM was used. Each article could include multiple PROMs, so the total observations are larger than the number of articles. In 11%, the PROM was only used to describe the study subjects. The most common use of self-rated health (SRH), activities of daily living (ADL), and instrumental ADL (IADL) was as an adjustor in other analyses, often using Medicare data such as the MCBS to evaluate a policy (eg, Medicare Part D). Other PROMs were more likely to be used as outcomes. Of the 307 quantitative and mixed-methods articles, 74 used validated instruments other than SRH, ADL, or IADL (see eAppendix Table); 56 used nonvalidated measures; and 15 exclusively used nonvalidated measures.

Similarly, Table 3 displays the validated measures in Medicare, Medicaid/CHIP, employer-sponsored/commercial, or VA insurance member populations. SRH (eg, “In general, my health is excellent, very good, good, fair, or poor”) was used in more than half of the Medicare and Medicaid articles; in Medicare articles, SRH, ADL, and IADL were often coadministered in the same surveys. In contrast, the dominant PROMs in studies of VA membership were from the RAND/SF/Veterans RAND (VR) family of measures (n = 16; 72.7%) (eg, the SF-36). There was not a PROM within the majority of employer-sponsored or commercial member studies, although the most common were SRH (n = 21; 33.3%) and the RAND/SF/VR family (n = 16; 25.4%). Studies of the VA membership were most likely to use a PROM about depression (n = 9; 40.9%), followed by studies of Medicare (n = 29; 20.1%), employer/commercial (n = 9; 14.3%), and Medicaid (n = 9; 14.1%) populations. The most common study purposes also varied to some degree by the insurer. For all insurers, the most common study purpose was to evaluate an insurer’s policy (70.8% of Medicare articles, 87.5% of Medicaid articles, 74.1% of VA articles, and 71.9% of employer/commercial articles). In the remaining articles, the VA was more focused on identifying individuals (18.5%) than evaluating quality (7.4%), whereas Medicare had a slight focus on identifying individuals (16.0%) over evaluating quality (13.2%). Medicaid articles had a slightly greater focus on evaluating quality (7.8%) than on identifying individuals (4.7%). Articles from employer/commercial plans were evenly split between evaluating quality (14.1%) and identifying individuals (14.1%).

There were 34 nonquantitative articles, of which 50% were commentary or editorial papers, 34% were mixed-methods design, and 12% were review articles; there was a single white paper (2.9%) and a single qualitative-only design (2.9%). Common themes across these nonquantitative articles included (1) case management for stress management, chronic disease management (eg, hypertension, diabetes), or mental health management (eg, depression, schizophrenia); (2) reducing cost by predicting health care utilization and cost management of health care without compromising quality of care; (3) commentaries and editorials advising health care payers and policy makers about the best way to use PROs and the need for policy changes in the context of value-based care principles; and (4) general guidelines for implementation of PROs in specific disease management.

DISCUSSION

This scoping review covered 20 years of literature about the use of HRQOL PROs in US health-insured populations and found a wide array of articles. These articles were published in journals with a focus on clinical practice, health services research, health economics, health policy, and managed care. The populations addressed in this work were also broad; they included such disparate groups as patients with cancer, osteoarthritis, migraine, depression, end-stage renal disease, schizophrenia, and asthma,^24-31 as well as general insurance members. The articles evaluated broad policy such as the impact of Medicare Part D; subgroup management strategies, such as telemonitoring; and the identification and management of individual patients, such as those with depression.^32-34

More than half (n = 205; 59%) of the articles addressed Medicare (83 [24%] mentioned Medicare in the title) and often used large, publicly available data sets (eg, the MCBS) to evaluate policy changes, such as the addition of Part D.^35,36 This scoping review highlights the common use cases of PROs by US health insurers as follows. First, private insurer data are much less available than government-sponsored data; although MCBS and many other government data sets are available to the public, data from other insured populations are largely proprietary business data or available via analytic data sets sold by companies such as IBM Watson Health.³⁷ Second, insurers do not have the same publish-or-perish mandates as academic researchers due to different professional incentives. That, coupled with the cost to obtain nongovernmental health insurer data, explains the relatively smaller number of studies in our search for these types of insured populations. Third, as noted in the Methods section, it was a challenge to determine the motivation of academic researchers’ use of government data sources and whether their results were used to inform insurers’ decision-making. For example, although CMS chose to collect and report PROs, it is difficult to infer how PROs are driving CMS’ programs.

An area of concern, from the measurement community’s perspective, is that many insurance agencies lack the infrastructure and expertise to select validated measures and evaluate their internal programs with PROMs. Sixty articles used “homegrown” questionnaires for concepts that have existing validated questionnaires available (eg, pain, physical symptoms, self-efficacy, sleep). These studies include several studies of employer and insurer programs such as worksite health promotion programs, telephone and web interventions, and identification of high-risk members.^38-40 Many also evaluated the impact of state-level Medicaid expansions.^41,42 HRQOL measurement has a long history with questionnaires that are carefully constructed to ensure that PROMs actually capture the concept they purport to capture.⁴³ Also, new next-generation measures use modern measurement techniques, such as item response theory, which are highly flexible and adaptable to different needs, but they require some measurement background to be used to their full potential.⁴⁴ Thus, when insurers use PROs to evaluate programs and measure the health of their large member populations, validated PROMs are likely more accurate and reliable tools with which to perform meaningful and useful evaluation.

From this review, it appears that the insurance community has been a leader in using “generic” PROMs such as SRH, RAND/SF/VR measures, and ADLs. Generic measures are intended to provide an overall picture of someone’s HRQOL, agnostic to their health conditions, unlike “disease-specific” measures that intentionally focus on a single condition.⁴⁵ Generic measures are appropriate for population-level management, such as an insured population, because they apply to all survey respondents. In contrast, disease-specific measures have had higher uptake in research and clinical settings that focus on particular diseases, where it is easier to ensure that targeted questionnaires are assigned correctly. When relevant, payers should engage with providers to select measures that are clinically actionable. This can ensure that proven measures are used and reduce the overall burden of questionnaires for patients. In this review, we found a handful of insured population studies using disease-specific measures for conditions such as asthma, migraine, and joint replacement.^{23,25,26,28-30} The Patient Reported Outcomes Measurement Information System, initially supported by the National Institutes of Health, is among the efforts currently underway to develop generic measures for research and clinical use.⁴⁶ Health insurers may have important insights about the use of generic measures to share with their clinical partners.

This review found 2 important areas with a paucity of articles. The first is pediatric populations; only 36 articles considered pediatric populations (often implicitly in broad discussions of PROMs), and only 14 articles explicitly focused on pediatric populations alone. Although measurement of HRQOL in pediatric populations is more difficult than in adult populations because of proxy responders,⁴⁷ pediatric populations deserve the same attention to their QOL. The second is articles that use data collected by health care providers at the point of care (eg, collected routinely in clinical waiting rooms or within the electronic health record). Only 34 (10%) of articles discussed using data collected by providers, and 12 of these were commentaries, reviews, or protocols, leaving 22 (6%) using data from providers. The study of information collected at the point of care is particularly important because many alternative payment models, such as bundled payments with CMS,^48-50 require collection of PROMs by the health service provider and a substantial integration of information from providers and payers to support these alternative models.

Limitations

This study was limited by difficulties in finding and identifying this literature. Despite the large number of included search terms and databases, we found few “grey literature” sources such as industry newsletters, so this search may be incomplete. As outlined in the Methods section, determining whether an article was truly addressing population health management from an insurer’s perspective was complex and difficult to standardize across reviewers, so some articles may have been inappropriately excluded. Additionally, the intended use of academic articles often needed to be inferred by the reviewers.

CONCLUSIONS

The thoughtful use of PROMs can support health insurers’ goals of measuring and demonstrating value. Supporting the use of validated questionnaires would improve the accuracy of value assessments. Despite the interest in alternative payment models that include PROMs, there is substantial need for guidance about how to collect and share these data among providers and insurers using patient-centered and rigorous methods. This review demonstrates wide interest in and use of PROMs among US health insurers, which provides a foundation for these future advancements. This review also provides a foundation for future studies to quantify the costs and values of PROMs for patients, providers, and payers.

Acknowledgments

The authors would like to thank Eric Chan, PhD, and Emre Yücel, PhD, for their contributions, which provided insight and expertise that greatly assisted the research.

Author Affiliations: General Internal Medicine, University of Pittsburgh (JH), Pittsburgh, PA; University of Utah (AMC), Salt Lake City, UT; University of Washington (BGG), Seattle, WA; Washington State University (BGG), Spokane, WA; University of Pittsburgh (PM, RT), Pittsburgh, PA; UPMC Center for High-Value Health Care (ECSS, SMK), Pittsburgh, PA.

Source of Funding: Funding for Dr Hanmer and Ms McCracken was provided by the Wolff Center at UPMC. Funding for Dr Kinsky and Ms Swart was provided by UPMC Insurance Services Division.

Author Disclosures: Dr Hanmer, Ms McCracken, Ms Swart, and Dr Kinsky are employed by UPMC. Dr Cizik is a grant consultant for the University of Washington on a National Institutes of Health grant studying the use of patient-reported outcomes to determine phenotypes in low back pain and is the principal investigator funded on an Orthopaedic Research and Education Foundation grant using patient-reported outcomes in trauma patients. The remaining authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (JH, AMC, BGG, RT, SMK); acquisition of data (AMC, BGG, PM, RT, SMK); analysis and interpretation of data (JH, AMC, BGG, PM, ECSS, SMK); drafting of the manuscript (JH, AMC, BGG, PM, ECSS, RT, SMK); critical revision of the manuscript for important intellectual content (JH, AMC, BGG, PM, SMK); statistical analysis (JH); provision of patients or study materials (AMC, BGG); obtaining funding (JH); administrative, technical, or logistic support (PM, ECSS, SMK); and supervision (JH, SMK).

Address Correspondence to: Janel Hanmer, MD, PhD, General Internal Medicine, University of Pittsburgh, 230 McKee Pl, Ste 600, Pittsburgh, PA 15213. Email: hanmerjz@upmc.edu.

REFERENCES

1. Porter ME. A strategy for health care reform—toward a value-based system. N Engl J Med. 2009;361(2):109-112. doi:10.1056/NEJMp0904131

2. Epstein RM, Street RL Jr. The values and value of patient-centered care. Ann Fam Med. 2011;9(2):100-103. doi:10.1370/afm.1239

3. Porter ME. What is value in health care? N Engl J Med. 2010;363(26):2477-2481. doi:10.1056/nejmp1011024

4. Wilson IB, Cleary PD. Linking clinical variables with health-related quality of life: a conceptual model of patient outcomes. JAMA. 1995;273(1):59-65. doi:10.1001/jama.1995.03520250075037

5. McHorney CA. Health status assessment methods for adults: past accomplishments and future challenges. Annu Rev Public Health. 1999;20:309-335. doi:10.1146/annurev.publhealth.20.1.309

6. Basch E. Patient-reported outcomes—harnessing patients’ voices to improve clinical care. N Engl J Med. 2017;376(2):105-108. doi:10.1056/NEJMp1611252

7. Coulter A. Measuring what matters to patients. BMJ. 2017;356:j816. doi:10.1136/bmj.j816

8. McDowell I. Measuring Health: A Guide to Rating Scales and Questionnaires. 3rd ed. Oxford University Press; 2006. Accessed August 18, 2021. https://oxford.universitypressscholarship.com/view/10.1093/acprof:oso/9780195165678.001.0001/acprof-9780195165678

9. Marsh K, van Til JA, Molsen-David E, et al. Health preference research in Europe: a review of its use in marketing authorization, reimbursement, and pricing decisions—report of the ISPOR Stated Preference Research Special Interest Group. Value Health. 2020;23(7):831-841. doi:10.1016/j.jval.2019.11.009

10. Patient reported outcomes (PROs) in performance measurement. National Quality Forum. January 10, 2013. Accessed February 2, 2021. https://www.qualityforum.org/WorkArea/linkit.aspx?LinkIdentifier=id&ItemID=72537

11. Bundled Payments for Care Improvement (BPCI) initiative: general information. Updated January 18, 2022. Accessed February 2, 2021. https://innovation.cms.gov/innovation-models/bundled-payments

12. Unützer J, Chan YF, Hafer E, et al. Quality improvement with pay-for-performance incentives in integrated behavioral health care. Am J Public Health. 2012;102(6):e41-e45. doi:10.2105/AJPH.2011.300555

13. Wong SE, Everest L, Jiang DM, Saluja R, Chan KKW, Sridhar SS. Application of the ASCO value framework and ESMO magnitude of clinical benefit scale to assess the value of abiraterone and enzalutamide in advanced prostate cancer. JCO Oncol Pract. 2020;16(2):e201-e210. doi:10.1200/JOP.19.00421

14. Dubois RW. CVS to restrict patient access using cost-effectiveness: too much, too soon. Health Affairs. September 17, 2018. Accessed February 8, 2021. https://www.healthaffairs.org/do/10.1377/forefront.20180913.889578/full/

15. Arksey H, O’Malley L. Scoping studies: towards a methodological framework. Int J Soc Res Methodol. 2005;8(1):19-32. doi:10.1080/1364557032000119616

16. Tricco AC, Lillie E, Zarin W, et al. PRISMA extension for scoping reviews (PRISMA-ScR): checklist and explanation. Ann Intern Med. 2018;169(7):467-473. doi:10.7326/M18-0850

17. Hanmer J. Scoping review of PRO uses by healthcare payers. Open Science Framework. October 14, 2019. Accessed February 8, 2021. https://osf.io/bf3jz

18. SEER-Medicaid linked data resource. National Cancer Institute Healthcare Delivery Research Program. Accessed February 4, 2021. https://healthcaredelivery.cancer.gov/seermedicaid/

19. CAHPS Patient-Centered Medical Home (PCMH) item set.Agency for Healthcare Research and Quality. Updated November 2021. Accessed February 8, 2021. https://www.ahrq.gov/cahps/surveys-guidance/item-sets/PCMH/index.html

20. Jones N III, Jones SL, Miller NA. The Medicare Health Outcomes Survey program: overview, context, and near-term prospects. Health Qual Life Outcomes. 2004;2:33. doi:10.1186/1477-7525-2-33

21. Adler GS. A profile of the Medicare Current Beneficiary Survey. Health Care Financ Rev. 1994;15(4):153-163.

22. Cohen JW, Cohen SB, Banthin JS. The Medical Expenditure Panel Survey: a national information resource to support healthcare cost research and inform policy and practice. Med Care. 2009;47(7 suppl 1):S44-S50. doi:10.1097/MLR.0b013e3181a23e3a

23. Behavioral Risk Factor Surveillance System. CDC. 2019. Accessed February 4, 2021. https://www.cdc.gov/brfss/index.html

24. Rao AV, Hsieh F, Feussner JR, Cohen HJ. Geriatric evaluation and management units in the care of the frail elderly cancer patient. J Gerontol A Biol Sci Med Sci. 2005;60(6):798-803. doi:10.1093/gerona/60.6.798

25. MacLean C. Value-based purchasing for osteoarthritis and total knee arthroplasty: what role for patient-reported outcomes? J Am Acad Orthop Surg. 2017;25(suppl 1):S55-S59. doi:10.5435/JAAOS-D-16-00638

26. Solomon GD, Hu HX, Simmons R, Conboy K, Jeddeloh R, Eastburn M. Development of a migraine disease management initiative in a managed care plan. Dis Manage Health Outcomes. 2002;10(3):167-173. doi:10.2165/00115677-200210030-00004

27. Waxmonsky JA, Thomas M, Giese A, et al. Evaluating depression care management in a community setting: main outcomes for a Medicaid HMO population with multiple medical and psychiatric comorbidities. Depress Res Treat. 2012;2012:769298. doi:10.1155/2012/769298

28. Moss AH, Davison SN. How the ESRD quality incentive program could potentially improve quality of life for patients on dialysis. Clin J Am Soc Nephrol. 2015;10(5):888-893. doi:10.2215/CJN.07410714

29. Dickey B, Normand SLT, Hermann RC, et al. Guideline recommendations for treatment of schizophrenia: the impact of managed care. Arch Gen Psychiatry. 2003;60(4):340-348. doi:10.1001/archpsyc.60.4.340

30. Buchner DA, Butt LT, De Stefano A, Edgren B, Suarez A, Evans RM. Effects of an asthma management program on the asthmatic member: patient-centered results of a 2-year study in a managed care organization. Am J Manag Care. 1998;4(9):1288-1297.

31. Bierman AS, Bubolz TA, Fisher ES, Wasson JH. How well does a single question about health predict the financial health of Medicare managed care plans? Eff Clin Pract. 1999;2(2):56-62.

32. Mahmoudi E, Jensen GA. Has Medicare Part D reduced racial/ethnic disparities in prescription drug use and spending? Health Serv Res. 2014;49(2):502-525. doi:10.1111/1475-6773.12099

33. Weinstock RS, Teresi JA, Goland R, et al; IDEATel Consortium. Glycemic control and health disparities in older ethnically diverse underserved adults with diabetes: five-year results from the Informatics for Diabetes Education and Telemedicine (IDEATel) study. Diabetes Care. 2011;34(2):274-279. doi:10.2337/dc10-1346

34. Taylor JK, Schoenbaum M, Katon WJ, Pincus HA, Hogan DM, Unutzer J. Strategies for identifying and channeling patients for depression care management. Am J Manag Care. 2008;14(8):497-504.

35. Pak TY, Kim G. The impact of Medicare Part D on cognitive functioning at older ages. Soc Sci Med. 2017;193:118-126. doi:10.1016/j.socscimed.2017.09.037

36. Liu FX, Alexander GC, Crawford SY, Pickard AS, Hedeker D, Walton SM. The impact of Medicare Part D on out-of-pocket costs for prescription drugs, medication utilization, health resource utilization, and preference-based health utility. Health Serv Res. 2011;46(4):1104-1123. doi:10.1111/j.1475-6773.2011.01273.x

37. Truven Health Analytics. IBM. Accessed February 5, 2021. https://www.ibm.com/watson-health/about/truven-health-analytics

38. McCarty CA, Scheuer D. Lessons learned from employee fitness programs at the Marshfield Clinic. WMJ. 2005;104(5):61-65.

39. Glasgow RE, Nelson CC, Kearney KA, et al. Reach, engagement, and retention in an internet-based weight loss program in a multi-site randomized controlled trial. J Med Internet Res. 2007;9(2):e11. doi:10.2196/jmir.9.2.e11

40. Wells AR, Guo X, Coberley CR, Pope JE. Integrating well-being information and the multidimensional adaptive prediction process to estimate individual-level future health care expenditure levels. Popul Health Manag. 2016;19(6):429-438. doi:10.1089/pop.2015.0184

41. Adams EK, Markowitz S, Dietz PM, Tong VT. Expansion of Medicaid covered smoking cessation services: maternal smoking and birth outcomes. Medicare Medicaid Res Rev. 2013;3(3):mmrr.003.03.a02. doi:10.5600/mmrr.003.03.a02

42. Lind A, Kaplan L, Berg GD. Evaluation of an asthma disease management program in a Medicaid population. Dis Manage Health Outcomes. 2006;14:151-161. doi:10.2165/00115677-200614030-00004

43. Streiner DL, Norman GR, Cairney J. Health Measurement Scales: A Practical Guide to Their Development and Use. Oxford University Press; 2015.

44. Hays RD, Morales LS, Reise SP. Item response theory and health outcomes measurement in the 21st century. Med Care. 2000;38(9 suppl):II28-II42. doi:10.1097/00005650-200009002-00007

45. Patrick DL, Deyo RA. Generic and disease-specific measures in assessing health status and quality of life. Med Care. 1989;27(3 suppl):S217-S232. doi:10.1097/00005650-198903001-00018

46. Cella D, Yount S, Rothrock N, et al; PROMIS Cooperative Group. The Patient-Reported Outcomes Measurement Information System (PROMIS): progress of an NIH Roadmap cooperative group during its first two years. Med Care. 2007;45(5 suppl 1):S3-S11. doi:10.1097/01.mlr.0000258615.42478.55

47. Pickard AS, Knight SJ. Proxy evaluation of health-related quality of life: a conceptual framework for understanding multiple proxy perspectives. Med Care. 2005;43(5):493-499. doi:10.1097/01.mlr.0000160419.27642.a8

48. Snyder CF, Wu AW, Miller RS, Jensen RE, Bantug ET, Wolff AC. The role of informatics in promoting patient-centered care. Cancer J. 2011;17(4):211-218. doi:10.1097/PPO.0b013e318225ff89

49. Gensheimer SG, Wu AW, Snyder CF; PRO-EHR Users’ Guide Steering Group; PRO-EHR Users’ Guide Working Group. Oh, the places we’ll go: patient-reported outcomes and electronic health records. Patient. 2018;11(6):591-598. doi:10.1007/s40271-018-0321-9

50. Trombley MJ, McClellan SR, Kahvecioglu DC, et al. Association of Medicare’s Bundled Payments for Care Improvement initiative with patient-reported outcomes. Health Serv Res. 2019;54(4):793-804. doi:10.1111/1475-6773.13159