Incorporating Health IT Into Primary Care Transformation

ABSTRACT

Objectives: To understand the role of health information technology (IT) vendors and health IT functionality in supporting advanced primary care.

Study Design: We synthesized multiple rounds of surveys and interviews (2017-2022) from a mixed-methods evaluation of Comprehensive Primary Care Plus (CPC+), a multipayer model developed by CMS. CPC+ was the first federal advanced primary care reform effort that formalized health IT vendors’ roles in supporting health IT implementation and specified detailed health IT requirements for practices.

Methods: We conducted content analysis to identify cross-cutting themes related to health IT for both practices and vendors, comparing similarities and differences across participants and (when possible) over time.

Results: Vendors and practices reported advances in registries and dashboards for improved information management within the practice as well as strengthened relationships between vendors and practices that supported health IT implementation. However, CPC+ practices noted several gaps or challenges using existing functionalities, and both vendors and practices reported broader challenges for more transformative health IT change, particularly the lack of interoperable health information exchange needed to support care management and care coordination. Key factors constraining vendors’ investment in further advances included long product development schedules, making it difficult to respond to rapidly evolving model requirements. Vendors also shared that CPC+ practices represented a small fraction of their client base, so investing in developing new functionality was not strategic unless it was more broadly relevant outside CPC+.

Conclusions: Continued collaboration among health IT vendors, practices, policy makers, and payers could support continued technological improvements, particularly related to information exchange and communication. Aligning requirements more closely with other federal and private models could also help mitigate the risk for vendors.

Am J Manag Care. 2024;30(1):e26-e31. https://doi.org/10.37765/ajmc.2024.89491

Takeaway Points

• Although health information technology (IT) requirements in Comprehensive Primary Care Plus fostered technological improvements and strengthened relationships between practices and vendors, there was less transformative health IT change than expected, particularly for interoperable health information exchange.
• CMS, payers, and policy makers may be able to encourage needed investment to improve interoperable health information exchange through future model requirements, although competing demands may constrain vendors’ and practices’ capacity for change.
• To minimize this constraint, practices, vendors, policy makers, and payers could build consensus about the highest-priority areas for improvement in functionalities and align future requirements across all applicable HHS funding programs, contracts, and policies to maximize their impact.

Electronic health records (EHRs) and related health information technology (IT) such as electronic patient registries can provide core information management functionality that allows practices to access patient data at the point of care and track patients’ progress and needs over time. EHRs and other health IT can also provide communication functionality that supports information sharing with patients and coordination with team members inside and outside the practice.^1,2 These activities are particularly important for primary care practices that have comprehensive, long-term responsibility for their patients and must manage the complex needs of their sickest patients as well as routine care for their overall panel. For example, physicians who reported using an EHR and participating in a patient-centered medical home were more likely to routinely engage in population management, quality measurement, and care coordination.³ However, many practices’ experiences with health IT do not live up fully to this potential.⁴ For example, population health functionality that can help practices identify all patients with shared care needs (such as those with a particular illness or outstanding preventive care visit) may still require manual effort to define and generate reports.⁴ Further, practices participating in the Comprehensive Primary Care (CPC) initiative, a multipayer advanced primary care model developed by CMS, identified difficulties recording data in and extracting data from EHRs and gaps in health information exchange as barriers to primary care transformation.⁵ CPC sought to leverage the EHR as a central support for care delivery changes but found key functionalities missing,⁵ partly because of the complex demands that value-based care placed on the tools designed for billing in a fee-for-service environment.⁶

To address these identified health IT barriers, CMS sought to strengthen relationships with health IT vendors and encourage development of new functionalities in a subsequent model, CPC Plus (CPC+). In addition to being the largest and most ambitious primary care payment and delivery reform effort ever tested in the United States, CPC+ was the first federal advanced primary care payment and delivery reform effort in which CMS formalized health IT vendors’ roles in supporting health IT implementation and specified detailed health IT requirements for practices. To formalize vendors’ roles, CMS encouraged vendors to provide participating practices with a letter of support and sign a memorandum of understanding with CMS describing their commitment to develop new or optimize existing advanced health IT capabilities.⁷ To support optimal development and use of health IT, CMS included requirements for health IT vendors and practices related to (1) information management within practices and (2) interoperability across practices.⁸

Our independent evaluation of CPC+ provided a unique opportunity to learn more about the roles of health IT vendors and health IT functionality in supporting advanced primary care. Examining lessons from CMS’ approach to health IT improvement may help clarify the roles policy makers and payers can play in fostering technical improvements to support care delivery changes. Using insights from practices and vendors involved in CPC+, we address the following research questions:

• What were health IT vendors’ perspectives on and experiences with providing functionalities to CPC+ practices?
• In the context of a formal partnership between vendors and practices, what were health IT vendors’ perspectives on and experiences with providing functionalities to CPC+ practices?

METHODS

Setting

CMS launched CPC+ in January 2017 in 14 regions and added 4 more regions in January 2018. Across these 18 regions, 68 EHR, population health, and other health IT vendors at the start of CPC+ committed to support 3070 primary care practices’ efforts to improve the care they provided to more than 17 million patients. Along with payment reforms to support primary care through enhanced and alternative payments, CPC+ practices were required to meet care delivery requirements (CDRs) within 5 primary care functions hypothesized to improve patient health and reduce costs: (1) access and continuity, (2) care management, (3) comprehensiveness and coordination, (4) patient and caregiver engagement, and (5) planned care and population health. In our analysis for this article, we focus primarily on the first 3 functions because they were most closely tied to required health IT functionalities.

Practices in CPC+ joined 1 of 2 tracks, with approximately the same number of practices in Track 1 and Track 2. Track 2 practices were required to meet more enhanced CDRs and more advanced health IT requirements. For example, although practices in both tracks were required to risk stratify their patients, Track 2 practices were required to do so using an established health IT–enabled algorithm.^9,10 Practices and health IT vendors could choose how to design and use required health IT functionalities, although Track 2 practices were required to formally partner with at least 1 health IT vendor that supported these required functionalities.

CPC+ CDRs and health IT requirements were initially informed by gaps that CMS identified during CPC and then evolved throughout CPC+. For example, the care management requirement originally included a 2-step risk-stratification process that included updating algorithm-based scores with the care team’s perception but later only required practices to ensure all empaneled patients were risk stratified.10 Similarly, when the health IT requirements were introduced, some had a 6- to 12-month timeline and others had to be completed within 24 months of the January 2017 model kickoff.⁸ In September 2018, CMS introduced changes that reworked or removed several required functionalities to reduce practice burden and to focus on functionalities that were higher priority or more straightforward to develop.

Data Collection and Analysis

The CPC+ evaluation team collected data from partnering health IT vendors and participating practices between 2017 and 2022 to better understand participation, available supports, and changes in care delivery. We do not analyze or report on the practices that joined CPC+ in 2018, as these practices account for only 5% of the total practices participating in CPC+ and their first-year implementation experiences were very similar to those of practices that joined CPC+ in 2017.¹¹ Here we describe our data collection and analysis efforts, which are also summarized in annual reports to CMS.^11-15 We identified cross-cutting themes related to health IT for both practices and vendors, comparing similarities and differences across participants and (when possible) over time.

Practice interviews. We conducted in-depth, semistructured interviews with physicians and staff from 100 unique CPC+ practices (we interviewed 81 practices in person in 2018, 59 practices by phone in 2019, 40 practices by phone in the first quarter of 2021, and 23 practices by phone in the last quarter of 2021). We interviewed 3 to 4 respondents per practice, except for the first year, when we interviewed 4 to 8 per practice. Typical respondents included medical practitioners, CPC+ coordinators, care managers, practice managers, health IT staff, and (when relevant) system-level representatives such as a chief medical officer or population health lead. Approximately 30 team members conducted practice interviews over the course of the 5-year evaluation, including 2 authors (G.C. and A.S.O.). Interviewers participated in a multisession training on interviewing best practices, the overarching research questions, and interview protocol questions to minimize bias and foster consistent, high-quality data collection.

We asked practice respondents about their experiences trying to implement care changes, probing about health IT among potential facilitators and barriers to their work for CPC+.¹⁶ Approximately 20 team members, including the first author (G.C.), coded practice interviews using a template analysis approach in NVivo 12 (QSR International). Coders applied 2 codebooks, one focused on components of CPC+ and the other focused on constructs related to the Consolidated Framework for Implementation Research, which includes a code for health IT. Coders participated in several trainings and compared coding across a few initial transcripts to foster consistent analysis with fidelity to the codebooks.¹⁷ The evaluation team analyzed interview data by thematic codes related to CPC+ primary care functions and factors drawn from the Consolidated Framework for Implementation Research.¹⁸

Because health IT questions were open ended and not asked in a standardized way in each interview round, we could not consistently analyze trends in experiences over time. Thus, we primarily assessed practices’ common experiences throughout the 4 rounds of interviews—only noting clearly apparent trends as reported in CPC+ annual reports.^11-15

Health IT interviews. We conducted in-depth, semistructured interviews with product development and policy experts at 13 health IT vendors in fall 2017, followed up with 10 of these vendors and interviewed 2 new vendors in winter 2019, and followed up with 10 of these vendors in winter 2021. We used structured data tables to summarize and synthesize details from professionally transcribed interview notes, identifying themes overall and by type of vendor. We also conducted semistructured interviews with 11 respondents from Track 2 practices that changed their vendor partnerships midway through CPC+, exploring practices’ motivation for this change, how vendor changes affected their ability to meet CPC+ goals, and their overall perspective on how vendor partnerships support delivery of comprehensive primary care.

RESULTS

Practices Reported Some Advancements and Ongoing Challenges With Using Health IT to Support Primary Care Functions

Across 3 rounds of interviews in 2018, 2019, and 2021, practices reported benefits as well as enduring challenges in developing and using advanced health IT in CPC+ to support primary care functions, noting it was a “work in progress” (see Table for full quote).

Primary care access and continuity. CPC+ required practices to improve patients’ timely use of needed care (access to care) from a care team that is cooperatively involved in a continuous relationship with the patients over the course of their health care management (continuity of care). Virtually all practices in our interviews provided patients with 24-hour access to a care team member with access to the EHR, and most practitioners in these practices reported that EHRs facilitated this availability through access to patient information when outside the office. Offsite EHR access also allowed for more consistent and comprehensive documentation of information in the patient’s record. Most practices reported that EHRs facilitated care continuity by allowing clinicians to view patients assigned to them in their patient panel. Practices sometimes reported errors in these assignments, which may have been due to EHR functionality, data entry processes, or workflow limitations. Several practices said that EHRs also supported continuity by consolidating patient information in a single searchable location, allowing practitioners to know which clinician in the practice had treated the patient and communicate with other relevant clinicians, which a physician described as “moving us toward better longitudinal care” (Table).

Care management. CPC+ practices were required to provide both shorter-term, episodic care management focused on acute care events such as emergency department visits, hospitalizations, and new diagnoses and longitudinal approaches for higher-risk patients who would benefit from ongoing, proactive care management. Many practices reported that EHRs facilitated care management through registries and dashboards, allowing care team members to look up past care and health history and update information on patients. Many practices reported having established relationships with local hospitals, but several noted that they experienced difficulties sharing information with at least some of the hospitals their patients visited. For example, several system-owned practices reported having formal relationships through which they received automated alerts in their EHR when a patient visited a system-affiliated hospital or an emergency department, but several of these practices also reported challenges obtaining discharge information from nonaffiliated hospitals due to interoperability issues.

Most practices risk stratified their patients for care management, but several noted challenges using EHRs to do so. Early on, these challenges included identifying appropriate risk-stratification algorithms, incorporating necessary data sources, lacking adequate EHR functionality to automate risk stratification, and implementing workflows to support systematic risk stratification. Throughout CPC+, practitioners and staff were uncertain about how automated risk scores in the EHRs were calculated and perceived that their practice had insufficient EHR functionality to support the risk-stratification process or lacked a clear process for updating risk scores based on clinical intuition. These concerns affected practitioners’ perceptions of the accuracy of risk scores and thus their buy-in to the value of assigning risk scores and using risk scores to identify patients who need care management, with one practitioner, for example, noting the risk score is “not valuable when the [patient] is sitting directly in front of you” (Table).

Comprehensiveness and coordination. CPC+ practices were also encouraged to provide comprehensive and coordinated care meeting most of their patient population’s medical and behavioral health needs while playing a central role in helping patients and caregivers navigate a complex health care system. Many practices reported using their EHR to document and track their patients’ social needs—a key element in providing more comprehensive care, which one population health and CPC+ project manager described as “improving efficiencies and workflows” (Table)—although several others said their EHR lacked the functionality to support such tracking. Nearly all practices reported having access to inventories of social services resources, but most did not have this embedded in their EHR and kept these inventories on separate electronic or printed lists. Several practices also noted that their EHR supported behavioral health integration. These practices’ embedded behavioral health specialists had access to their EHR and were able to document patient information, which facilitated communication with primary care practitioners.

Practices varied in the extent to which EHRs supported their referral management activities. Some reported that EHRs supported the ability to track and coordinate referrals, especially when there was robust interoperability across provider settings, but others noted this EHR functionality was limited. Many practices reported improvements in communication with hospitals and specialists in the first 2 years of CPC+ due to improved relationships with these providers and enhancements to their EHRs. Practices affiliated with systems or those with local hospitals that had the same EHR tended to have better information sharing about care provided.

Despite Some Benefits From Formalizing CPC+ Partnerships, Vendors Did Not Substantially Increase Health IT Support for Practices During CPC+

Compared with their historically limited role providing informal support in other CMS advanced primary care models, vendors being formal partners with CMS and Track 2 practices in CPC+ introduced opportunities for collaboration as well as challenges managing expectations. In interviews, several vendors noted that formalized partnerships helped strengthen their relationships with practices by more clearly identifying which practices were participating in CPC+, enabling them to provide more targeted outreach around new functionalities. In contrast, a few vendors were uncomfortable that formal partnerships committed them to supporting health IT requirements that they were not involved in designing and that CMS reserved the right to change throughout the model. For example, one vendor noted it had invested resources to meet the 2015 edition certified EHR technology “electronic clinical quality measure (eCQM) filter” criterion rather than investing in other product improvements because it was originally a CPC+ requirement. When CMS removed the requirement, the vendor felt that investment had been wasted.

Health IT vendors reported primarily enhancing functionalities that were in place before CPC+ rather than creating new products specific to CPC+. For example, vendors reported adding practice site–level eCQM reporting to their standard reporting templates and improving the usability of displays such as care manager dashboards and health-related social needs assessments to better support CPC+ practices. Enhancing existing functionalities rather than creating new ones reflected several vendor considerations. First, vendors reported believing they had the functionalities necessary to support CPC+ practices and did not need to build new products. Second, vendors noted that their development schedules were set more than a year in advance and the evolving nature of requirements was not conducive to broader investments, noting such changes could “shift their entire road map” (Table). Third, CPC+ practices represented a small fraction of vendors’ client base, so it was not strategic to invest in developing new functionality unless it was relevant to practices outside CPC+. Vendors particularly identified this challenge in the context of more novel CPC+ requirements, such as care plans, that lacked a corresponding clinical or industry standard at the outset of CPC+ and for which CMS did not provide examples that would meet requirements. Several vendors suggested it would have been beneficial for CMS to align the CPC+ health IT requirements more closely with other federal and private models to mitigate the trade-offs they were asking vendors to make by partnering closely with CPC+ practices.

DISCUSSION

This article highlights the first time CMS (via the Center for Medicare and Medicaid Innovation) formally engaged health IT vendors in primary care transformation, including detailed requirements for their partnerships with practices. This is vendors’ only formal engagement with federal primary care reform to date because the successor to CPC+, Primary Care First, does not include vendor requirements. Our study findings identified several key benefits of this approach, with vendors and practices reporting advances in registries and dashboards for improved information management within the practice. Practices also reported increased support for health IT implementation through these partnerships. However, several challenges arose for developing and implementing more transformative health IT change, particularly for interoperable health information exchange needed to support care management and care coordination.

CMS, payers, and policy makers may be able to encourage needed investment to overcome these challenges through future model requirements, although competing demands from other requirements may constrain vendors’ and practices’ capacity for change. Greater consultation with vendors and practices when defining program requirements may make it more likely that they will undertake needed investment. For example, vendors partnering in CPC+ may have realized more progress with information management than information exchange functionalities because information management does not require as much buy-in from exchange partners at other health care delivery organizations and vendors. Moving forward, it is important for practices, vendors, policy makers, and payers to build consensus about the highest-priority areas for improvement in health IT functionalities. To the extent that future primary care models can help practices prioritize EHRs that better support clinical and population-based care, they will also help health IT vendors prioritize investments in these areas.

Given these potential benefits of increased collaboration, health systems, practices, and their relevant specialty societies (eg, the American Academy of Family Physicians, the Society of General Internal Medicine, the American College of Physicians, the American Board of Internal Medicine) can reflect on recent experiences with CPC+ and other advanced primary care programs to identify the most important primary care health IT functionalities. A recent National Academies of Sciences, Engineering, and Medicine report outlining key primary care digital health functions may also be a helpful input for this assessment.⁴ As part of their work setting model requirements, payers and policy makers can consult with health IT vendors individually or through a trade association such as the Healthcare Information and Management Systems Society Electronic Health Record Association to identify technical challenges of developing new functionalities along with realistic timelines for working toward these goals. Finally, payers and policy makers can align these requirements “across all applicable HHS funding programs, contracts, and policies” to maximize their impact, as specified in the new HHS health IT alignment policy.¹⁹

Limitations

Findings from this study offer several novel insights about the role of health IT vendors and functionality in supporting advanced primary care, but it also has several limitations. First, we analyzed data that were originally collected to understand practices’ implementation experiences that included but did not primarily focus on health IT. Rather, practice interview questions primarily focused on CDRs, and health IT topics emerged out of responses to these questions rather than to questions about health IT. Second, by focusing on identified themes explicitly related to health IT rather than reanalyzing original interview data, we may not have captured all the ways in which health IT facilitated or impeded practices’ work on primary care transformation. Finally, our results are likely not generalizable to all primary care practices, as practices that applied to participate in CPC+ may have been uniquely motivated or positioned to make care delivery changes. However, these experiences could still be relevant to similarly motivated practices, including the 1360 practices that applied to but were not selected for CPC+. These limitations suggest that future research could identify additional knowledge around using health IT for advanced primary care, building on the experiences of practices and vendors working together in CPC+.

CONCLUSIONS

CPC+ reflects the next stage in the evolution of advanced primary care, building on findings from CPC that health IT limitations challenged practices’ success. CPC+ required more explicit partnership between practices and vendors, which facilitated vendors providing support that some practices found useful. Moreover, vendors and practices successfully developed and used some enhanced information management functionality. However, interoperable health information exchange proved more challenging; vendors reported few changes, and practices reported ongoing limitations related to care management and care coordination. These results indicate that continued collaboration among health IT vendors, practices, policy makers, and payers could support continued technological improvements, particularly related to information exchange and communication. 

Author Affiliations: Mathematica, Washington, DC (GC, TL, ASO), Ann Arbor, MI (MH), and Princeton, NJ (KG).

Source of Funding: HHS, CMS under contract HHSM-500-2014-00034I/HHSM-500-T0010. Payments under the contract were made by CMS to Mathematica. The contents of this paper are solely the responsibility of the authors and do not necessarily represent the official views of HHS or any of its agencies.

Author Disclosures: The 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 (GC, TL, MH, ASO, KG); acquisition of data (GC, MH); analysis and interpretation of data (GC, TL, MH, ASO, KG); drafting of the manuscript (GC, TL, KG); critical revision of the manuscript for important intellectual content (GC, TL, MH, ASO); administrative, technical, or logistic support (TL, MH); and supervision (KG).

Address Correspondence to: Genna Cohen, PhD, Mathematica, 1100 1st St NE, 12th Floor, Washington, DC 20002-4221. Email: GCohen@mathematica-mpr.com.

REFERENCES

1. Finkelstein J, Barr MS, Kothari PP, Nace DK, Quinn M. Patient-centered medical home cyberinfrastructure current and future landscape. Am J Prev Med. 2011;40(5 suppl 2):S225-S233. doi:10.1016/j.amepre.2011.01.003

2. Kellerman R, Kirk L. Principles of the patient-centered medical home. Am Fam Physician. 2007;76(6):774-775.

3. King J, Patel V, Jamoom E, DesRoches C. The role of health IT and delivery system reform in facilitating advanced care delivery. Am J Manag Care. 2016;22(4):258-265.

4. McCauley L, Phillips RL Jr, Meisnere M, Robinson SK, eds. Implementing High-Quality Primary Care: Rebuilding the Foundation of Health Care. The National Academies Press; 2021. doi:10.17226/25983

5. Peikes D, Dale S, Ghosh A, et al. The Comprehensive Primary Care initiative: effects on spending, quality, patients, and physicians. Health Aff (Millwood). 2018;37(6):890-899. doi:10.1377/hlthaff.2017.1678

6. Cimino JJ. Improving the electronic health record—are clinicians getting what they wished for? JAMA. 2013;309(10):991-992. doi:10.1001/jama.2013.890

7. Comprehensive Primary Care Plus: information for health IT vendors. CMS. Accessed December 12, 2023. https://www.cms.gov/priorities/innovation/files/slides/cpcplus-healthvendoroverview-slides.pdf

8. CPC+ certified health IT requirements (revised). CMS. Accessed December 12, 2023. https://www.cms.gov/priorities/innovation/files/x/cpcplus-hit-track2reqs.pdf

9. Final 2019 CPC+ health IT requirements. CMS. November 21, 2018. Accessed December 12, 2023. https://www.cms.gov/priorities/innovation/files/x/cpcplus-hit-py2019.pdf

10. CPC+ care delivery requirements crosswalk. CMS. December 2019. Accessed December 12, 2023. https://www.cms.gov/priorities/innovation/files/x/cpcplus-practicecaredlvreqs.pdf

11. Anglin G, Peikes D, Petersen D, et al. Independent Evaluation of Comprehensive Primary Care Plus (CPC+): Second Annual Report. Mathematica; July 2020. Accessed December 12, 2023. https://www.cms.gov/priorities/innovation/data-and-reports/2020/cpc-evaluation-annual-report-2

12. Peikes D, Anglin G, Harrington M, et al. Independent Evaluation of Comprehensive Primary Care Plus (CPC+): First Annual Report. Mathematica Policy Research; April 2019. Accessed December 12, 2023. https://downloads.cms.gov/files/cmmi/cpcplus-first-ann-rpt.pdf

13. Peikes D, Swankoski K, O’Malley A, et al. Independent Evaluation of Comprehensive Primary Care Plus (CPC+): Third Annual Report. Mathematica; January 2021. Accessed December 12, 2023. https://www.cms.gov/priorities/innovation/data-and-reports/2021/cpc-plus-third-anual-eval-report

14. Swankoski K, O’Malley A, Tu H, et al. Independent Evaluation of Comprehensive Primary Care Plus (CPC+): Fourth Annual Report. Mathematica; May 2022. Accessed December 12, 2023. https://www.cms.gov/priorities/innovation/data-and-reports/2022/cpc-plus-fourth-annual-eval-report

15. O’Malley A, Singh P, Fu N, et al. Independent Evaluation of Comprehensive Primary Care Plus (CPC+): Final Report. Mathematica; December 2023. Accessed December 15, 2023. https://www.cms.gov/priorities/innovation/data-and-reports/2023/cpc-plus-fifth-annual-eval-report

16. Anglin G, Peikes D, Harrington M, et al. Independent Evaluation of Comprehensive Primary Care Plus (CPC+): First Annual Report: Supplemental Volume. Mathematica Policy Research; April 2019. Accessed
December 12, 2023. https://www.mathematica.org/download-media?MediaItemId={08B2EA53-6CEA-499F-A7C4-A53553C1E0A3}

17. Keith RE, Heitkamp S, Little J, et al. Challenges and lessons learned conducting a formative evaluation of a multicomponent care delivery intervention. Am J Eval. 2023;44(4):649-662. doi:10.1177/10982140221116096

18. Keith RE, Crosson JC, O’Malley AS, Cromp D, Taylor EF. Using the Consolidated Framework for Implementation Research (CFIR) to produce actionable findings: a rapid-cycle evaluation approach to improving implementation. Implement Sci. 2017;12(1):15. doi:10.1186/s13012-017-0550-7

19. Tripathy M, Posnack S. E pluribus unum. Health IT Buzz. August 5, 2022. Accessed December 12, 2023. https://www.healthit.gov/buzz-blog/interoperability/e-pluribus-unum