6. Research Gaps and Prioritization of Federally Funded HSR and PCR
One of the study’s key research questions focuses on identifying research gaps—understudied or underfunded areas in HSR and PCR that, if addressed, would move the fields forward and enhance the various research impacts as described in the previous chapter. While gaps indicate needs for research that HSR and PCR has yet to fully address, they also represent opportunities to find new ways to use research to solve problems and improve the health care system, and ultimately the nation’s collective health. The wide range of research gaps described in this chapter were elicited from both TEP and interview participants.
In this chapter, we first present key gaps in HSR and PCR as defined below, including crosscutting gaps in research approaches, as well as those specific to HSR and PCR, respectively. A comprehensive ranking of HSR and PCR gaps was beyond the scope of this study, especially given the broad range of gaps identified by participants. At the same time, study participants and Federal Advisory Group members emphasized the need to prioritize research gaps to more effectively and efficiently allocate limited research funding. Toward this end, the last section of the chapter offers potential criteria and mechanisms for prioritizing research gaps.
Overview of Gaps Identified by Technical Expert Panel and Interview Participants
Many gaps are related to specific “inputs” and “outputs” of health care services in the study’s research domain framework (Figure 2.1). For example, gaps related to health care inputs include gaps in understanding the Organization of Care, such as the needs to identify barriers to implementing new models of care or to understand the workforce needed for these models. Gaps related to health care outputs include lack of adequate quality-of-care measures for certain services or outcomes.
However, many gaps in HSR and PCR reflect the complexity of the health care system and the difficulty of understanding the linkages between health care inputs and outputs to produce important outcomes, for example, the need to better understand the effects of new financing and delivery models on costs and quality of care, or of interventions to address the social determinants of health on health outcomes, costs, and equity. Moreover, other critical gaps identified by study participants relate to research approaches and to the ability to generate and disseminate evidence in ways that positively impact real-world health care systems and practice.
Below, we highlight key gaps in these areas, that is, research gaps raised by multiple study participants within a stakeholder perspective or across stakeholder perspectives (Chapter 2). Study participants noted that some of these gaps have been the subject of research studies but that further research is needed, or different research approaches are needed to make a positive impact on health care delivery and health outcomes. Other gaps mentioned by individual study participants are shown in Appendix D.
Table 6.1 lists the key cross-cutting gaps in research approaches that impede the impact of HSR and PCR for improving real-world health care systems and practice.
Table 6.1. Key Cross-Cutting Gaps in Research Approaches to HSR and PCR Identified by Study Participants
| Research Domain | Gaps in HSR and PCR |
|---|---|
| Cross-Cutting |
|
Table 6.2 summarizes key research domain gaps raised by TEP and interview participants for HSR.
Table 6.2. Key HSR Gaps Identified by Study Participants
| Research Domain | Gaps in HSR on... |
|---|---|
| Organization of Care | |
| Financing of Care |
|
| Social Factors | |
| Personal Preferences and Behaviors | |
| Quality of Care |
|
| Access to Care |
|
| Cost and Utilization | |
| Equity |
|
a Similar gap included in key PCR gaps (Table 6.3).
Table 6.3 lists key research domain and cross-cutting gaps for PCR.
Table 6.3. Key PCR Gaps Identified by Study Participants
| Research Domain | Gaps in PCR on... |
|---|---|
| Organization of Care | |
| Financing of Care | |
| Social Factors | |
| Personal Preferences and Behaviors | |
| Quality of Care |
|
| Access to Care |
|
| Cost and Utilization | |
| Equity |
|
| Cross-Cutting PCR Gaps |
|
Similar gap included in key HSR gaps (Table 6.2).
The letter “a” in Tables 6.2 and 6.3 denotes gaps that were similar for both HSR and for PCR, though often exhibiting distinct aspects. These gaps are described together in the sections below. All entries in both tables should be read as areas that need more or better research (e.g., the first entry in table 6.2 refers to gaps in HSR on health care workforce needs, composition, and roles in new delivery models).
Cross-Cutting Gaps in Research Approaches to HSR and PCR
TEP and interview participants identified six key cross-cutting gaps in research approaches and dissemination of findings that limit the extent to which HSR and PCR can improve real-world health care systems and practice. The first two relate to the range of health care populations and settings covered by HSR and PCR. These are followed by gaps in the development and communication of results in ways that are actionable, and in the relevance and timeliness of results for improving health care delivery. Another relates to gaps in using theory to connect findings and advance knowledge on health care changes, and the last addresses gaps in leveraging and linking new sources of health care-related data afforded by advances in digital health and other technologies.
Examining Health Care Outcomes for Fuller Range of Populations and Settings
Effects of Health Care Interventions on Fuller Range of Patient Populations
TEP and interview participants noted that many health care interventions are tested on specific populations such as patients without major health conditions other than the one of interest, or elderly beneficiaries in Medicare. For an intervention to be broadly applicable to clinical practice, health care delivery systems and providers need evidence as to whether and how the intervention can be adapted so that it is effective for wider populations of patients. One study participant observed “so much research that’s conducted has so many exclusions that it doesn’t represent the real world.”
Change and Improvement in More Generalizable Health Care Settings
Similarly, multiple study participants commented that much evidence on change and improvement is generated from research in highly integrated health care systems, such as the VHA or Kaiser-Permanente, which are not typical of the settings in which most health care in the United States is delivered. In addition to research on and generalizable to the more fragmented private and commercial health care systems, other health care settings that participants identified as constituting large or important sources of care include safety net providers such as FQHCs and other community health centers, as well as hospitals serving low-income and vulnerable populations, in both rural and urban areas.
One participant called for more reporting of contextual factors irrespective of the setting in which research is being conducted. This participant argued that “most of things we’re studying . . . are context-dependent; we actually don’t expect them to work the same in one situation [versus] the other.”
Following Change in Implementation and Outcomes of Health Care Change over Time
Several study participants commented that HSR and PCR studies are often neither designed nor funded for long enough periods to examine adaptations, sustainability, or outcomes of health care changes or interventions over time. As one participant noted, “Interventions are adapted and evolve over time as they’re implemented,” sometimes with the intention to better tailor the new intervention or practice to the local context but other times due to other operational, cultural, or policy pressures typically experienced in health care delivery organizations. Examining the nature of these pressures and natural adaptions that occur is important to understanding variation in fidelity to the initial intervention and its consequences for sustainability and effectiveness of the intervention over time. Another study participant gave the example of a project in which “we have no data about the fidelity of the intervention that the various grantees implemented. . . . So we are hard pressed to try to come up with an explanation for the rather minimal changes that we observed.” Other participants emphasized that doing so requires being attentive to the context in which a new health care intervention or practice is implemented. In the words of one study participant, “There is always a very local component to implementation.”
Communicating Results to Be Actionable by Diverse Health Care Stakeholders
Actionable results provide understanding of how to implement as well as what to implement. As described by a delivery system leader: “We know some [results], but it trickles out. Meanwhile, they keep inventing new stuff rather than helping us sort out the stuff we already have.” Delivery system leaders and other study participants expressed the need to articulate key steps in a change process, using terms like “pathway,” “playbook,” and “construct[ing] the workflow,” while “empowering practices to [each] figure out what it needs to do and how to do it better.” Research needs to offer actionable answers to hard questions like, as one participant put it, “So what do I do differently tomorrow? . . . I know that that’s true [the new intervention works], and now how do you structure it to be easy to do the new thing, or to override what I thought was the right way to go?”
Study participants also called for research studies to communicate findings so that they are actionable and can be implemented relatively quickly by diverse health care stakeholders, including not only health care delivery leaders and providers, but also payers, patients, and community leaders. Broader systems perspectives on research questions often lead to inclusion of wider sets of stakeholders.
Producing Relevant and Timely Results for Health Care Delivery Improvement
Multiple study participants pointed to gaps in the relevance and timeliness of HSR and PCR findings for improving health care delivery. They specifically called for the use and development of methods that better address these needs.
Several participants discussed the use of the randomized controlled trial (RCT), one of the hallmark study designs in health research in which subjects are randomly assigned to one or more intervention conditions and a control condition. The strength of RCT designs is the assurance they provide that differences in outcomes are the result of the interventions and not other causes. However, participants noted that, to maintain control over the study conditions, conventional RCTs are conducted in highly restricted environments compared to typical health care settings and exclude or ignore the complex dynamics of health care delivery. These aspects limit both generalizability to typical health care settings and understanding that stakeholders often need in order to guide implementation. As one researcher and health care delivery leaders described:
Many agencies are interested in funding randomized clinical trials that are very controlled, and there’s really very few that will fund research that’s in the midst of a complex, adaptive, health system that is messy. In terms of federal agencies, probably AHRQ has the best track record there, but also some of the private funders . . . because [they] want to see meaningful change. In really controlled settings, they’re often so controlled that they never translate into real meaningful change.
Pragmatic clinical trials. Study participants noted that, for these reasons, NIH and other federal agencies have increasingly funded pragmatic clinical trials, which compare health care interventions in more typical health care settings where patients receive their usual care, enroll more representative populations, and are designed to inform decisionmakers on the real-world implications of the intervention (Weinfurt, 2017).
Qualitative and mixed methods. Participants indicated the need for better integration of methods from both qualitative and quantitative approaches in order to better link health care context and process with outcomes and produce relevant evidence for stakeholders on effective health care interventions and change. Study participants also pointed to the need to expand the use of qualitative methods, such as interviews, focus groups, and observational methods, that more directly examine the context of health care delivery and process of change in health care systems.
Engaged research models. When focusing on the need of HSR and PCR to “actually improve what [patients] get,” one interview participant concluded that this “will be done much more quickly [and] with better results with people working very closely in an iterative fashion in the sort of messy, real world of care delivery.” Participants discussed three approaches to engaged research. The first two approaches in particular blur the lines between traditional health care quality improvement activities and HSR.
Embedded research models place health services researchers within service units to work collaboratively with health care delivery leaders and staff to develop research projects on issues relevant to the unit (Vindrola-Padros et al., 2017). By coproducing knowledge, researchers enhance the capacity of the health care delivery organization. One health service policy stakeholder noted the VHA as one agency that has invested in this model by “really pulling to get their research centers to partner with implementation . . . [and] working . . . on best practices in embedded research.” Outside the VHA, participants provided other examples of this model, such as developing long-standing partnerships between academia and delivery systems or employing academic researchers in delivery systems for short-term fellowships or sabbaticals.
Learning health systems likewise were described by participants as an approach to increasing the relevance and timeliness of research for health care delivery stakeholders. Learning health systems are an emerging set of research models in which health care delivery systems seek to develop improvements in care based on rigorous analysis of their own storehouses of EHR and other service data (McLachlan et al., 2018; Etheredge, 2014; Greene, Reid, and Larson, 2012). Others have noted the importance that research findings from local learning health systems are shared to improve care more broadly (Friedman, Wong and Blumenthal, 2010).
Study participants also mentioned participatory approaches, such as community-based participatory research. Although not necessarily more rapid than conventional research methods, community-based participatory research emphasizes relevance by engaging broader patient and community stakeholders in the research. As one research interview participant explained:
I do believe the whole learning health system research approach . . . and practice-based research and participatory models are really anchored in the reality of practice and health care organization operations, so that the research is really growing out of the questions that arise at the actual care delivery side of things and therefore are addressing questions that are going to be relevant and that can be directly applied.
Using Theory to Connect Findings and Advance Knowledge on Health Care Functioning and Change
Several study participants commented on the need to apply, develop, and refine theories of health care functioning and change in HSR and PCR, not for theory’s sake, but to help connect disparate findings and advance understanding of how to improve health care delivery. As one researcher and delivery leader summarized:
We need [to] figure out how to bring theory back so that it results in better questions and I think we might even have more impact because then some of these very diffused findings, they just hang out there because they’re not attached to a larger spread of understanding. And if we bring theory in then . . . [they] could [also] potentially bring new methods to us.
While a range of social and behavioral science theories can and have been applied in HSR and PCR, three sets of theories mentioned by study participants were systems science (a general field of related theories on the properties and interactions among components of social and/or technical systems), complexity science (a branch of systems science focused on understanding dynamics and change in complex systems), and implementation science (a field of theories and research developed in HSR and PCR focused on understanding the adoption, use, and spread of practices, interventions, and innovations in health care).
Leveraging Digital Health and Linking of Varied New Sources of Health Care–Related Data
TEP participants discussed an emerging gap in research on leveraging new sources of health care–related data afforded by advances in digital health and other technologies. Digital health includes a range of technologies, including social media, mobile apps, other platforms for patient-generated health information, “big data” analytics (including artificial intelligence, data mining), as well as EHRs maintained by health care providers. Participants noted that procedures for handling patient privacy and consent and infrastructure for linking these various sources of digital health data for analysis have yet to be developed.
Gaps Related to Specific Research Domains
In addition to the cross-cutting gaps in research approaches above, interview and TEP participants identified gaps related to the topical domains in the study’s research framework (Figure 2.1). Below we describe key gaps related to each of the health care input domains of Organization of Care, Financing of Care, Social Factors, and Personal Preferences and Behaviors; as well as to the health care output domains of Quality of Care, Access to Care, Cost and Utilization, and Equity.
These domains are discussed in the order shown in the framework and presented in Tables 6.2 and 6.3 for HSR and PCR gaps, respectively. Gaps that were similar for HSR and PCR (denoted with the letter “a” in the tables) are discussed under the same respective domain below, with PCR-specific aspects distinguished with italicized text. As mentioned previously, key gaps in each domain are those that were raised by multiple stakeholders within or across stakeholder perspectives. Additional research gaps identified by study participants can be found in Appendix D.
Organization of Care
Health care workforce needs, composition, and roles in new delivery models in primary care and other settings. In both interviews and TEP meetings, the importance of studying workforce-related issues through research was highlighted by participants from diverse perspectives, including delivery system leaders, state-level representatives, researchers, and others. One broad research gap related to workforce was the lack of knowledge around various configurations of the workforce (i.e., team-based structures and processes of care) and the effects such configurations could have on the health care system, including shortages. One state-level representative said: “Workforce is just a constant struggle for us, understanding that the health care system is transforming, and part of that transformation, undoubtedly, leads to the demand for health care providers with different skill sets along the way. And how do we equip those providers to adjust their practices in order to fit the demands of a transforming health care delivery system?” Other gaps related to workforce composition and roles included research to address the lack of specialists in the pediatrics workforce and optimal staffing for emergency mental health care. Particularly in primary care, a number of participants noted needed research on the emerging roles of nurse-practitioners, physician assistants, and clinical pharmacists in interprofessional team-based care. In addition, participants discussed a critical need for additional research in primary care to better understand and address provider stress and burnout.
Reducing burdens of HIT on health care providers, especially in primary care. Despite the expectation that HIT, in particular, EHR systems, will increase the efficiency and safety of health care delivery, study participants noted the widespread unintended consequences of these technologies, and the additional reporting tasks that they enable, in substantially increasing administrative and documentation burden on health care providers. One federal stakeholder asked, “How do they think about health IT as a facilitating tool, rather than an end-all and a be-all? In other words, how do they construct the workflow, . . . [which] clinical team member does what component of the work, how do they evaluate that, how does it come together?” Another federal stakeholder asked how HIT could be used in primary care “to improve the effectiveness and efficiency, quality, safety of primary care and not create more burdens and harm.” This participant felt that increasing HIT burden without improving its usability could “sink a ship that is barely staying afloat at the moment.”
Financing of Care
Effects of evolving models of financing on the range of health care outcomes, including in primary care. The need to better understand the effects of financing and payment mechanisms was highlighted by study participants for a range of health care outcomes. For example, payment reform came up as a means of supporting health care systems in addressing challenges with social factors and social determinants of health faced by their patients. Financing likewise featured extensively in the discussion of quality measurement and of the relationship between quality and payment to promote value. Participants also highlighted the need for risk adjustment methodologies that are more immune to health care delivery systems and providers “gaming the system,” for example, when they focus more on raising the reported risk profile of their patients rather than improving care. A health economics researcher noted that research on these gaps in financing mechanisms could be improved by taking into account the continuous, evolving nature of payment reform as it occurs in much of the health care system: “The question isn’t what’s going to be the best payment mechanism for all time . . . , but much more opportunity for continuous learning and improvement because let’s face it, payment models are impossible to get right the first time.”
One gap particular to PCR was the need to understand the role of comprehensive primary care within payment models such as ACOs. One interview participant noted that accurately attributing the value of primary care within an ACO is complex and challenging. The importance of this type of work was echoed by another interview participant, who noted that “there’s a need to get out of pure fee-for-service for primary care and to really understand the effect of alternative payment models and how those work.”
Effects of health care payment models on different patient populations across health care settings. Participants noted that more research is needed concerning the ways in which new payment models affect health care and health outcomes not only for the general population or elderly adults on which most studies have tended to focus, but also on vulnerable and special-needs populations. Examples given included payment reform for specialized care (e.g., end-stage renal disease) and for patients who require additional caregiver support (e.g., children and patients with dementia).
Social Factors
Role of health care systems in addressing social determinants of health in primary care and other settings. Study participants emphasized the importance of social and community factors in affecting the health outcomes especially for high-need and at-risk individuals. However, less clear for participants were the feasible and effective roles that health care delivery systems could play in addressing social determinants of health. Open questions raised by participants included the appropriate role of health care delivery systems, if any, in directly assisting patients with needs in housing and transportation, and effective models for health care systems to collaborate with other community resources (e.g., social service agencies, community coalitions, school-based health initiatives) to address these needs. One PCR interview participant described these issues in the context of specific types of patients requiring these supports (emphasis added):
[Patients] with the triad of diabetes, hypertension, and respiratory disease . . . [those who] smoke and they drink. These are the people who represent their own special universe of people who have all the chronic illnesses and all of the adverse social habits and an underlying, almost always, depression. What is the intensive intervention with those people that would make a difference? Is it financial support? Is it social support? When we look at their med list, they already have 20 medicines on their list. It’s not medication. You could call it the high-risk-score people who score high on chronic illness, habits, and social determinants of health. What is the effective response . . . to those people?
Other participants pointed to the potential for collaboration with community partners and other disciplines to address “root causes” of these social determinants, such as the example provided by a state-level hospital association representative of a hospital working with demographers, land-use researchers, and local policymakers to reroute truck traffic and plant trees in neighborhoods with high frequencies of asthma-related ER visits.
Effects of social factors on demand for health care services, especially in primary care. Study participants also pointed to gaps in understanding the effects of social factors on the demand for health care services and, specifically, on primary care. This evidence would be useful for primary care and other health delivery systems to more clearly understand the impact of social determinants on services, but also how demand for services might change if these social needs were met.
Personal Preferences and Behaviors
Integrating patient preferences into care in primary care and other settings. Participants indicated that more evidence is needed on how best to integrate and engage the patient, their family, and their caregivers’ preferences into their care. Methods for engagement included tools, approaches for shared decisionmaking, measurement of care on dimensions important to patients, and incentives for patients. Broadly, participants mentioned the need to better understand what patients value in their health care. This research need was mentioned in relation to various types of patients, including those with chronic conditions and those utilizing genomic test results to inform their care.
Two PCR interview participants elaborated further on this gap. One participant indicated that yet-to-be-developed patient-oriented primary care quality measures would facilitate more engaged patient care, for example, in “prioritization of different conditions, [and] deescalating medications or treatments to meet people’s values” (see quality-of-care discussion below). Another PCR participant highlighted the gap from patient engagement to patient action: “How do we effectively engage patients to become more empowered for their own self-care from chronic conditions and multiple chronic conditions.”
Addressing health and health care misinformation, especially in primary care. TEP and interview participants highlighted the research gap centering on misinformation, including the need for research to address both the effect of “misinformation” on personal preferences and behaviors related to health care—most prominently the antivaccination movement—and ways to combat misinformation from various stakeholder perspectives.
Quality of Care
Developing harmonized measures to meaningfully, accurately, and feasibly assess quality of care across all health care settings. Overall, participants in interviews and both TEPs called for improvements in quality measurement. Significant inconsistency was noted across the health care system in quality measures and the way they are implemented. Participants called for research around prioritization of measures and greater harmonization of measures to reduce burden. One interview participant emphasized that the knowledge of how to create measures exists but research funding to create measures is lacking: “We currently don’t have valid measures of quality that everybody can use. . . . We know how to create good measures. The funding isn’t there, the testing isn’t there and the measurement infrastructure to develop, test and then promulgate measures isn’t there.”
Study participants noted that some areas of health care were particularly lacking in meaningful quality measures. As one PCR researcher commented, better primary care measures are “desperately” needed as a barometer of the key functions of primary care, while also not interfering with patient care. Another interview participant reported that primary care lacks non-disease- and non-organ-specific measures that adequately capture the quality of primary care in ways meaningful to both providers and patients:
These specific outcomes, blood pressure control, hemoglobin A1c control, . . . they just pile up in primary care. But they’re not what patients and doctors really care about. So how do we evolve to the next series of [measures]: “patients feel heard,” “patients feel safe and cared for,” “patients are functioning the way they want,” “they’re able to reach their life goals,” whatever those might be? . . . [new measures] make sense and then leads to things like prioritization of different conditions, deescalating medications or treatments to meet people’s values, coordination issues, integration with community services.
Ongoing research gaps were also noted in specific subareas, including (1) measures for pediatric populations, which have lagged behind quality measures for adults; (2) patient safety measures that look at all aspects of patient safety, including overuse or misuse of medical care and diagnostic errors; and (3) outcome-based measures—particularly in behavioral health.
Finally, patient-centered measures that capture what the patient is experiencing and cares about were highlighted as an area in need of more research. Further, although patient-reported measures (patient-reported outcome measures and patient-reported experience measures) have been developed, more research is needed on the practicality and feasibility of such measures, and how they can be implemented. One delivery systems leader noted that, while there is general consensus that such measures are valuable, more research is needed to understand how such measures can be implemented without the documentation burden that detracts from provider time delivering patient care. Study participants also stated that, after more patient-reported outcome and experience measures are developed, additional research will be needed on how to use such information to understand the impact of low-quality care and harm on patients.
Access to Care
Identifying both root causes and evidence-based solutions for barriers to access. Delivery leaders, federal agency leaders, researchers and other interview participants, as well as TEP participants, acknowledged the substantial evidence that has been accumulated to date on existing disparities in access. For example, one interview participant commented that “you’ve got communities in every city that are basically medical deserts.” A PCR participant indicated that this gap is not solely geographic, pointing to the ways that other factors, such as health insurance, can create barriers to access:
[For example, insurance product X] . . . the people that are picking this up . . . [they are the] “working poor,” people who don’t quite qualify for the Medicaid insurance, . . . but can’t quite afford the more expensive insurance policies, either. . . . premiums are cheap, but . . . providers are not willing to accept the plans, what you end up with is you end up with decreased access to care, because you have a limited network from which to choose, if you want to try to use this insurance.
Thus, study participants called for more research attention on the root causes of, and strategies for, addressing barriers to access, especially regarding the health care system’s potential role in addressing these causes. Despite noting important federally funded projects that have been conducted, such as AHRQ’s 2016 evidence review on telehealth, participants highlighted the need for further evidence on effective interventions to address barriers to access, including telehealth and other strategies such as virtual visits and remote monitoring.
Cost and Utilization
Challenge of lowering cost while improving care, including in primary care. In general, the challenges of the rising costs of care and utilization as a driver of costs were repeated research gaps discussed throughout the TEPs and interviews. An important theme was understanding how to improve care while lowering costs. One delivery systems leader said: “Thirty percent or more of what we spend money on adds no value, so, in today’s world, where increasingly not-for-profit health systems are economically challenged, we need research on what works and what doesn’t, so that we can be putting evidence-based improvements in place.” In discussing PCR specifically, one interview participant noted that understanding what proportion of the spending “pie” is related to primary care in different settings and geographies is important.
Reducing waste in health care generally and primary care as well. Although the high prevalence of waste in health care is not a new issue, study participants suggested that a research gap still exists in the use of billing and other data to identify where waste can be reduced. Administrative waste was highlighted as an especially important area that needs more examination. Several study participants also suggested that utilization could be improved by closing the gap in understanding the appropriateness of care (what is the right sort of care for different types of patients) and the factors associated with utilization of inappropriate care. In PCR this was specifically framed by a TEP member as an issue that crosses all age groups and multiple aspects of health care, in the context of the right care at the right time: “This [. . .] crosses pediatrics, adult, aging, patient safety, whether it’s diagnosis, overutilization of treatment, underutilization of treatment, trying to build a better match between appropriate care and timely care, so that you decrease underutilization, you decrease overutilization.”
Costs of new care therapies and delivery models, including medical home models such as PCMH. Costs associated with new therapies or novel delivery approaches that could improve care were cited as underresearched. Specific examples included more precise understanding of increases in cost from new drugs, such as new cancer therapies (raised by both researchers and federal stakeholders), as well as the potential for decreases in cost through delivery approaches such as telemedicine. Participants also highlighted the need to better understand how different types of incentives and accountability built into new payment models relate to cost. In the area of PCR, evaluations of new models of care—such as the PCMH—often report health care costs, but typically inconsistently, which limit generalized assessments of the cost effects of such models.
Across both HSR and PCR, some study participants believed that better evidence on the effects of payment models on costs would also encourage health systems to engage in additional experimentation and research of their own on payment models. One researcher stated: “I think if we really do get there in terms of payment models that have accountability for better meaningful outcomes and lower costs, or at least close enough to it, I think systems will have incentives to do studies themselves of what is it that we can do that leads to better outcomes and lower costs, both in their own systems and hopefully sharing across others.” Researchers also pointed out the importance of understanding who bears the costs versus cost savings in these arrangements (patient, provider, delivery or payment organization, etc.) and how this could potentially perpetuate inequity in the health care system.
One additional research gap noted specifically for PCR was in the area of payment and care reform focusing on coordination of specific conditions between primary and specialized care (i.e., medical homes for end-stage renal disease). One interview participant suggested that, rather than using bundled payments for acute admissions and procedures, the focus should shift instead to prevention. Such models could be extended to other conditions such as degenerative joint disease or heart failure.
Reducing costs across the health care system. Finally, multiple participants stated that it was important to focus on costs across the health care system as an outcome. One participant said, “For me, the studies that stand out are the ones that actually look at total spend, not many of them do. . . . It’s just from experience: we know it’s way easier to move the quality needle than it is to move the cost needle.”
Others echoed the importance of including cost as an outcome in research and noted that it is reported too infrequently. Another participant suggested that research is particularly needed on how to reduce health care costs in non-Medicare populations, given the amount of evaluation that CMMI has conducted on costs for the Medicare population.
In the realm of PCR, in contrast, one interview participant noted that the cost of health care was largely driven by specialty and inpatient care, and the cost of primary care was not a gap that needed to be studied: The “main issues of high prices in this country are not that we’re paying too much for primary care.”
Equity
Solutions to reduce disparities in health and health care. Understanding and ultimately decreasing the effect of disparities on health care was also discussed. One delivery systems leader highlighted the need for more and better data: “The data is not even there to provide us with information to do the analysis, to monitor for disparities. So those are huge gaps in my mind.” However, other participants noted that the research gap is not simply the need to document disparities, but to design, test, and implement strategies to change how care is delivered so that disparities are reduced. As one interview participant stressed, given the evidence already generated on the existence of health care disparities, the most important question is, “How do we intervene?”
The role of primary care in addressing equity. A number of study participants raised the role particularly of primary care in addressing social factors related to health inequities. Similar to addressing social determinants of health generally, an interview participant framed the issue as helping providers identify roles in reducing disparities that are appropriate and feasible for the primary care system:
[There is] a huge need to understand how, what are the things that really can be done effectively in primary care to address health equity and to make that much more of a priority for the practice improvement work. What is the appropriate role for primary care practices in moving up strengths and addressing underlying fundamental social determinants of health? There is . . . not enough [research] in really understanding are there effective and efficient ways to do this? Or is it really that it’s just going to overburden busy primary care clinicians and that [it] really belongs in the public health and civic space of addressing poverty and educational status and structural racism and things like that.
Cross-Cutting Research Gaps Specific to Primary Care
Next we present two research gaps specific to PCR that cut across the topical research domains. The first addresses needs for greater research attention on the core functions of primary care. The second concerns the transformation and role of primary care in the wider health care system.
The Core Functions of Primary Care
Interview and PCR TEP participants discussed the importance of filling the gap in what they called the “basic science” of primary care. Researching the “basic science” of primary care involves, as described below, documenting “what’s actually going on in primary care” today and understanding and developing models of the core functions of primary care in addressing the holistic needs of patients for usual care, maintaining health, and coordinating services for other health and health-impacting services.
This “basic science” can then inform how to improve primary care. One interview participant highlighted how such knowledge can inform changes to service delivery, noting that the health system and primary care in particular have changed substantially since the time that the seminal research on these core primary care functions was conducted:
The number of acute patients we see in primary care offices has dramatically reduced, the complexity and chronicity has dramatically increased. . . . We need to understand that better because that would immediately change how we think about scheduling, how we integrate the use of virtual care and some of the telemedicine, and all of these other modalities.
Reflecting further about primary care patients, another participant called for study into the core functioning of primary care to include gaining a better understanding of the “ecosystem of primary care” from the standpoint of patient decisionmaking and behavior.
We actually don’t know what people do when they feel ill now; who do they see, what do they do, for what and when, where do they go? We’ve created so many options out there in the ecosystem of primary care that we don’t really understand what’s going on. And we probably know that that’s very variable depending on the nature of the population.
Stronger basic evidence on the current “ecosystem” of primary care would help to develop and refine theories of drivers and models of effective primary care that can be used to stimulate further improvements to primary care service delivery. For example, the same participant above asked whether the “four Cs” of primary care identified by Barbara Starfield and colleagues in the 1980s and 1990s (comprehensiveness, coordination, continuity, and first contact) operate differently in the current context of primary care than initially conceived.
It may be that primary care does something very different that may have to do more with prioritizing and integrating and personalizing—hmm, where [are those things] going on out there? Is primary care doing [those things]? Maybe [those things are] more what really drives the success of primary care and the four Cs simply facilitated them?
Multiple participants attributed the lack of research on the core functions of primary care to the lack of dedicated funding for core research on this central component of the health care system and the fragmentation of much federal PCR funding across agencies focused on other missions. PCR interview and TEP participants described having to “disguise” core research on the functioning and practice of primary care as organ- or disease-specific studies (e.g., screening for a particular disease in primary care) for NIH grant programs, or as patient safety or other health service topics for agencies like AHRQ. While appreciative for the resources afforded by these funding sources, participants noted that the aims of these programs can limit the ability to fully address core issues in primary care. As one researcher explained, agencies are not usually interested in funding research on such broad questions in primary care:
I think there’s been almost no funding for the things that I’ve brought up. . . . They come out of a deep understanding of what primary care is and needs to address. They’re not the questions that NIH institutes are going to ask. They’re not the questions the CDC is going to ask. They’re not what SAMHSA’s going to ask. And PCORI and AHRQ have very limited pieces. They ask some of them, but certainly not enough money to go around and they have to focus on some very specific assignments they were given with those limited resources.
Another interview participant particularly pointed to the lack of a central hub of funding for PCR:
So [PCR] is funded through a variety of agencies. The big challenge is none of them have a dedicated primary care funding mechanism or a specific component in their agency that’s dedicated to funding primary care research. It’s therefore harder to get research funded on things that are not disease specific or condition specific or fit a particular niche [but] that are more about the care of whole people and all their multi-morbidity complexity or about practice transformation issues, primary care community health interface, things like that.
Primary Care Transformation and Role in the Wider Health Care System
While research on the effects of new models of the primary care were discussed under the Organization of Care domain, study participants also identified broader research gaps on the development of new models and their role in the wider changing landscape of the health care system. As one participant described:
[We need] much more rigorous research on how do you actually re-engineer the primary care practice model to meet the needs of 21st century primary care? A lot of that is around team-based models. What are the different roles and tasks and how do you successfully implement team models? . . . How do patients respond to more team-based care models? Are they sustainable?
This same participant added that, to support primary care transformation, research would also be helpful on how the patient and community voices could be better integrated “as partner[s] in practice improvement . . . co-creating the new practice models . . . empower[ing] patients and communities [in] partnering with primary care clinicians and practices [on] how primary care should work and be more patient and community centered.”
Other participants highlighted the need for research on primary care as part of larger systems of care; for example, primary care’s role in successful ACOs and in the delivery of ever more complex care. On the latter point, a research interview participant provides the example of caring for a cancer survivor in primary care.
We start rapidly accelerating new innovations in the treatment of diseases . . . and we have more and more rapidly accelerating, highly-specialized knowledges—How are those being translated to the primary care clinicians so that they can then do the integrative work with a patient as a whole when the patients come back to them? [There’s] . . . almost no research on that and no active work going on in how to develop that.
Although research on these issues have been conducted, study participants indicated that “there is a lack of good and consistent studies on these topics.” For example, studies on the effects of various new models of primary care (e.g., PCMH) and value-based care programs (e.g., CPC+) were considered to report contradictory evidence on health outcomes and cost effects of these models. Participants noted that sorting out this evidence is needed to provide useful guidance on primary care transformation for health care systems.
Gaps Summary
This section first presented key gaps identified by study participants in research approaches and dissemination that impede the impacts of HSR and PCR for improving real-world health care systems and practices. The first two gaps related to the scope of HSR and PCR is the need to broaden in terms of health care populations and settings on a large scale, as well as assessing health care change over time. These were followed by gaps in the communication of results in ways that are actionable by health care delivery stakeholders, as well as the relevance and timeliness of those results for improving health care delivery. Another gap focused on using theory to connect findings and advanced knowledge on health care changes. The last gaps in that section addressed leveraging and linking new sources of health care–related data afforded by advances in digital health and other technologies.
HSR Gaps
The next sections of this chapter presented key gaps related to each of the topical domains in the study’s research framework introduced in Chapter 2. These gaps included needed research on specific topics within health care “input” domains (e.g., Organization of Care, Financing of Care, Social Factors, and Personal Preferences and Behaviors) and health care “output” domains (e.g., Quality of Care, Cost and Utilization of Health Care, Access to Care, and Equity). For example, gaps in HSR related to health care input domains included the need to identify barriers to implementing new models of care and to understand the workforce needed for these models. Gaps related to health care output domains included lack of adequate quality of care measures for certain services or outcomes.
PCR Gaps
A number of research gaps for PCR were similar to those for HSR (denoted by the letter “a” in Tables 6.2 and 6.3). The descriptions of these gaps highlighted their distinct PCR-specific aspects. For example, PCR participants noted the gap regarding the lack of adequate quality of care measures included the creation of metrics that adequately capture the quality of comprehensive and holistic primary care in ways meaningful to providers and patients. Several research gaps pertained only to PCR, including the role of primary care in addressing equity and two gaps that crossed research domains. The first PCR-specific, cross-cutting gap emphasized the need for research on the core functions of primary care, and the second concerned research on the transformation of primary care practice and its role as part of the wider health care system.
Framework for Prioritizing Gaps
As described to this point in the chapter, study participants identified a wide range and large number of research gaps in HSR and PCR. However there was also consensus among TEP and interview participants that identifying priorities, and the process of prioritization itself, is critically important. This view was echoed in meetings with the Federal Advisory Working Group. Study participants representing both HSR and PCR suggested a similar general approach to the prioritization process.
Prioritization Criteria
Overall, there was agreement among both TEP and interview participants that transparent and explicit criteria are important to the prioritization process. Possible criteria mentioned included the potential impact of the research, the potential to address a gap in a poorly funded research area, the potential to address foundational areas of research, and the timeliness of the research (Box 6.1). The interrelation of these criteria was often noted, as described in more detail below.
Potential Research Impact
Many of the interview and TEP participants highlighted the importance of prioritizing research by its potential for impact. Impact in this context was closely linked to the likelihood of actionable results. Multiple HSR TEP members agreed that impact could be described as research that might lead to opportunities to change practices or policies. One TEP member also related impact to timing, in the sense that research has to be done at the right time to inform legislation and must ultimately prove to be policy-relevant.
TEP members also conceptualized impact as having potential for “high gain,” with the idea that high-risk, high-gain projects should be considered for prioritization. One TEP member mentioned the NIH Pioneer Awards, noting that it is important to have a portfolio that specifically funds projects with higher risk of poor or null results but also greater potential for gain in knowledge and practice. A research interview participant noted that the potential for innovation often depends on a willingness to accept risk and expressed the wish that CMMI would be more “eager to do high-risk, high rewards kinds of projects and not fear failure . . . [which] is in direct competition with true innovation.”
In a later discussion, a TEP member suggested that a practical means to identify research of interest for a more high-risk portfolio would be to conceptualize the major changes or shifts that need to occur for the research to influence health outcomes and then to work backward to understand the first steps needed to make these changes.
Finally, in follow-up to one of the meetings, another TEP member described prioritization based on answers to a set of three questions about potential impact:
- Is the issue being addressed important?
- Is data needed to answer the question?
- If we had the data to answer the question, would action follow in a reasonable time frame?
This three-step process is appealing in its simplicity, though we note that defining “important” is in itself quite complex and requires explicit definition, as discussed in Chapter 5.
Box 6.1. Prioritization Criteria Suggested by Study Participants
Participants identified four prioritization criteria for consideration:
|
Potential to Address a Gap in an Underfunded Research Area
Several participants suggested prioritizing research areas that address a knowledge gap that is poorly funded or not receiving adequate attention. A member of the HSR TEP suggested an exercise to compare research gaps against both research that is proposed and research that is already funded in order to disentangle whether “people are proposing things in these gap areas that aren’t getting funded or are they not proposing them at all.”
Potential to Address Foundational Areas of Research
Some interview and TEP participants raised the idea of prioritizing funding for research that is foundational or that serves as a building block to solve core health care system issues and make progress in other research areas. For example, several participants considered research on “new workforce configurations” to be foundational, noting that it can help ensure that primary care remains a sustainable profession in the future since, in the words of one TEP member, “without a workforce, we’d go out of business.”
There was also recognition that, within the field of HSR, foundational research that moves the field forward may not always be as well understood as in the natural or physical sciences:
Now I’ve always felt in health services research, we don’t have that same sense of what is basic or fundamental. Because health services research itself is largely an applied field. It’s sort of part of the definition. But if we could look back and say, “Well, what are some of the fundamental building blocks that if we had good research, that would provide answers to questions in those fundamental areas, it may then serve as a springboard to solution across a range of these gap areas.”
Timeliness of the Proposed Research
Both interviews and TEP discussions related prioritization to the timing of addressing research gaps, in particular the ability to align with policy timelines whether in the near or longer term. Participants highlighted the difference between short-, medium-, and long-term needs for research, and how these horizons affect prioritization.
Some participants noted the difficulty of prioritizing research gaps that need to be addressed quickly in order to influence policy. A specific example was study of the Affordable Care Act, with relevant research ideally becoming available in 6 to 12 months from implementation of a major policy change. Prioritizing funding—or even obtaining funding—to perform such work was noted to be a major challenge.
Other participants cautioned against deprioritizing gaps simply because the research might take a long time to perform. One interview participant noted that “you can look at that and say, well, that’s a longer-term horizon enough you don’t have to pay attention to it, but if we don’t pay attention to it now, then the horizon gets further away.” Conversely, another interview participant explicitly stated that chronic issues should be given lower priority than more pressing needs. Federal Advisory Group members also noted that, within federal agencies, an important criterion for prioritization is the alignment of an issue with the mission of an agency and its comparative advantage and expertise in funding research on the topic.
During the HSR TEP, there was a lengthy discussion of the tension between funding time-sensitive issues versus funding important, but more chronic, problems. One member proposed that funders of HSR may need to sort research funding into separate “buckets” or categories to ensure that research with varying time horizons is funded. One proposed research category was long-standing, chronic problems in health care. Another was emerging problems that need to be solved, such as the current opioid crisis, or HAIs. A third category focused on how to solve problems and who can identify the solutions, which relates to methods, modeling, technology, and other related fields.
Frameworks to Guide the Prioritization Process
During our discussions of prioritization, some participants identified frameworks or processes that could explicitly guide the prioritization process. Below we list some of these for consideration. Regardless of the framework or process chosen, there was general consensus that any such process should be transparent and explicit (Box 6.2).
Box 6.2. Principles for the Prioritization Process Suggested by Study Participants
|
In general, TEP participants were in favor of using known frameworks for possible prioritization, as outlined below:
- The most commonly mentioned framework was the Triple or Quadruple Aim. The Triple Aim was first developed by the Institute for Healthcare Improvement and defined as “applying integrated approaches to simultaneously improve care, improve population health, and reduce costs per capita” (Institute for Healthcare Improvement, undated). Since the development of the Triple Aim, some have suggested expanding the framework to the Quadruple Aim, which adds the goal of improving the work life of health care providers, including clinicians and staff (Bodenheimer and Sinsky, 2014). One advantage of using the Triple or Quadruple Aim, as noted by TEP members, is that it is widely recognized and thus may be easily understood.
- The IOM definition of quality—with its six dimensions of safety, effectiveness, patient-centeredness, timeliness, efficiency, and equitability (Institute of Medicine, 2001)—was mentioned by members in both TEPs. These dimensions—as well as others, such as access—have been incorporated into definitions of a high performing health system, such as that described by the Commonwealth Fund, which referred to a system that “. . . achieves better access, improved quality, and greater efficiency, particularly for society’s most vulnerable, including low-income people, the uninsured, minority Americans, young children, and elderly adults” (The Commonwealth Fund, 2019). Using these definitions as part of a prioritization framework would mean ascertaining whether a given study contributes to a better understanding of quality of care, affordability, and of the equitable allocation of health care resources and health in the population.
- TEP participants also referenced the IOM definition of primary care: “The provision of integrated, accessible health care services by clinicians who are accountable for addressing the large majority of personal health care needs, developing a sustained partnership with patients, and practicing in the context of family and community” (Institute of Medicine, 1996). Priorities for PCR that could be extracted from this definition would include research on integrated care, access to care, management and coordination of care, and patient- and relationship-centered care.
Prioritization Processes
Two TEP members also suggested examples of existing methods and processes through which prioritization could occur after framework criteria and principles are established:
- In the HSR TEP, one member highlighted the Child Health and Nutrition Research Initiative, which is a systematic yet flexible method for setting research priorities for global child health. The method consists of 15 steps, with the goal of informing those who invest in research about the risks associated with their investments, with the target audience being international agencies, large research funding donors, national governments, and policymakers (Rudan et al., 2008).
- Another process mentioned was the RAND’s Delphi method (RAND Corporation, undated), which is a structured communication method that relies on a panel of experts to assign ratings, discuss ratings, and then reconsider their original ratings and come to a consensus on prioritization.
Regardless of which framework or process is chosen, one PCR TEP member emphasized the importance of reexamining and refining criteria periodically to ensure that changes in health care system priorities are incorporated as needs and the research evolve.
Stakeholder Engagement in Prioritization
Similarly, regardless of the framework or process used, there was consensus among both TEP and interview participants that it is critically important to engage stakeholders throughout the prioritization process (Box 6.3).
Box 6.3. Stakeholder Engagement in the Prioritization of Gaps Suggested by Study Participants
|
Relevant Stakeholders
Both interview and TEP participants identified key stakeholders for inclusion in the prioritization process. One TEP member noted that steering and other advisory committees—whose makeup might be similar to that of a prioritization group—tend to include mostly researchers and experts, and that broader representation should be considered.
Important groups included researchers, providers, payers, patients and consumers, state-level representatives, and employers. However, one interview participant noted that representatives from such groups should be chosen to include those who have demonstrated experience thinking through relevant health delivery issues and ways in which they could be improved. Elected officials were mentioned as key stakeholders, although the concern was cited that elected officials may have a separate agenda whose goals may not align with those of other stakeholders.
Another TEP participant suggested that inclusion of other stakeholders such as industry representatives (e.g., pharmaceutical or device manufacturers) may raise conflict-of-interest issues. This suggestion was offered with the caveat that, while such perspectives should be included, it would be important to ensure that they do not “hijack” the prioritization process. Ensuring a balance among perspectives was considered critical, along with strong disclosure requirements and mechanisms to mitigate conflicts-of-interest.
Patients and Families
One group that received particular attention in study participant discussions was patients, families, and other caregivers. Involvement of patients and families was seen as extremely important, as voiced by the patient/consumer representatives as well as other TEP and interview participants, who believed that obtaining such input is important to ensure that research funding fills gaps in a meaningful way for patients. TEP members also considered the manner in which such input is obtained was important. For example, one TEP member noted that panels whose advisory group included only a single patient stakeholder representative were typically ineffective since it can be difficult for patients to speak up when the representation and power differential may influence their comfort level.
Study participants also emphasized that selecting patient representatives for participation in prioritization must be a thoughtful process. Such patient representatives should have some experience and understanding of the health care system and ideally would have served in the role on similar groups. However, little preparation exists to prepare patients, family members, and caregivers for these roles, who often must learn through participation. Developing a pool of patient representatives who can serve in the prioritization process might be accomplished via the use of toolkits (e.g., PCORI’s Patient Engagement Toolkit) (Patient-Centered Primary Care Collaborative, 2019) and trainings. Engaging patients and families throughout the process is equally important, as one participant noted:
What we’ve seen in health care systems is that you can get farther if you have stakeholders together at the same time in the same group. . . . Patients and families can understand challenges and why things can’t be implemented or aren’t a good idea and help come up with creative solutions. But without that involvement, they don’t hear the pushback and can’t be truly part of the process.
Prioritization Summary
While a comprehensive ranking of HSR and PCR gaps was beyond the scope of this study, TEP and interview participants discussed potential criteria and mechanisms for prioritizing research gaps. Study participants emphasized that criteria for prioritization need to be transparent and explicit, and that the range of relevant stakeholders need to be engaged in the prioritization process—including patients and families. Possible criteria for prioritization include the potential impact of the research, the potential to address a gap in an underfunded research area, the potential to address foundational areas of research, and the timeliness of the research. Federal Advisory Group members also noted that, within federal agencies, an important criterion for prioritization is the alignment of an issue with the mission of an agency and its comparative advantage, and expertise in funding research on the topic.
Chapter Summary
TEP and interview participants identified a wide range of pressing research gaps in HSR and PCR. These gaps are driven by the complexity and rapidly changing landscape of the U.S. health care system. Many of these gaps are related to specific “inputs” and “outputs” of health care services in the study’s research domain framework. However, many of the gaps also reflect the difficulty of understanding the linkages between health care inputs and outputs to produce important outcomes, and the limitations of currently used research approaches to generate and disseminate evidence in ways that positively impact real-world health care systems and practices. At the same time, these gaps also represent opportunities to find new ways to use research to solve problems and improve the health care system, and ultimately the nation’s collective health.
In this chapter, we highlighted key gaps, that is, research gaps that were raised by multiple study participants within a stakeholder perspective or across stakeholder perspectives. Study participants noted that many of these gaps have been the subject of research studies sponsored by federal agencies and other funders, but that further research is needed, or different research approaches are required to make a positive impact on health care delivery and health outcomes.
Study participants and Federal Advisory Group members emphasized the need to prioritize research gaps to effectively and efficiently allocate limited research funding. Criteria for prioritization suggested by study participants included the potential impact of the research, the potential to address a gap in an underfunded research area, the potential to address foundational areas of research, and the timeliness of the research. Federal Advisory Group members also noted that, within federal agencies, an important criterion for prioritization is the alignment of an issue with the mission of an agency and its comparative advantage and expertise in funding research on the topic.
