Developing implementation strategies to improve uptake of guideline-recommended treatments for individuals with familial hypercholesterolemia: a protocol

Summary

Findings from this study will help us understand how to improve the adoption of guideline-recommended treatment for individuals with FH. This should lead to implementation approaches with the potential to prevent early onset CV events and death. This study will fill a significant gap in the care of patients with FH because little is known about specific reasons why individuals with FH do not receive guideline-recommended care. Use of an implementation science framework enhances the rigor of the study and increases the likelihood that the resulting strategies can be translatable and scalable by other healthcare systems.

BACKGROUND

A significant, but vastly underdiagnosed, cause of ischemic heart disease is familial hypercholesterolemia (FH), a lipid disorder usually inherited in an autosomal dominant fashion.¹ Two methods of diagnosing FH are: 1) clinical diagnosis according to specific clinical diagnostic criteria; or 2) genetic diagnosis via the identification of a pathogenic or likely pathogenic variant in one or more of the FH-associated genes (LDLR, APOB, and PCSK9).² These diagnostic approaches may be used alone or in combination. A recent meta-analysis estimates the prevalence of heterozygous FH is about 1 in 250 individuals.³ Heterozygous FH is defined as having a causative variant in one of the FH-related genes and homozygous FH as having two variants in one of the genes. However, as many as 90% of individuals with FH are undiagnosed.⁴

Early identification of individuals with FH is key to reducing cardiovascular (CV) events. Individuals with FH have lifelong elevation of serum low-density lipoprotein cholesterol (LDL-C) concentrations that leads to an increased risk for early-onset CV events.^5–8 The presence of an FH gene mutation is an independent and additive risk factor to increased LDL-C levels and CV events.⁹ In addition, untreated individuals with FH and elevated cholesterol are up to 20 times more likely to have a premature CV event than individuals without FH and normal cholesterol.¹⁰ At Geisinger, individuals have been identified with pathogenic and likely pathogenic variants in one of the genes associated with FH via a population genomic screening initiative. This study has also found similar rates of inadequate treatment when compared with clinically identified FH populations.^1,11,12 The consequences of inadequate treatment are more profound for the genetically identified population due to the increased risk conferred by the variant(s).^1,11,12 Current FH guidelines recommend strategies for the treatment of individuals with FH for primary prevention that include at least a 50% reduction in LDL-C and ideally to reach an LDL-C goal of <100 mg/dL.^13,14 The FH-specific recommendations are consistent with the 2018 Guidelines on the Management of Blood Cholesterol by the American College and Cardiology and the American Heart Association.¹⁵ These guidelines recommend statins as the first-line therapy for FH because statins confer mortality and morbidity benefit in individuals with hypercholesterolemia.^13,14,16 Treatment should begin as early as 8 to 10 years of age in heterozygotes and at diagnosis in homozygotes.^14,17 Often, combination therapy with other lipid lowering therapies is necessary for individuals with FH to achieve their LDL-C target goals.^14,18,19 In addition, the guidelines recommend reduction in other atherosclerotic CV disease risk factors, promotion of nutrition and physical activity interventions, and cascade testing of at-risk first-degree relatives using either cholesterol levels or for the specific mutation if it has been identified in the index case.^13,20

Individuals with FH often require more aggressive treatment compared to those with multifactorial hypercholesterolemia due to the significantly higher risk of CV events. However, the adoption of guideline recommendations as routine clinical practice is slow for many conditions, including hypercholesterolemia.^21–25 As a result, despite lipid lowering therapies being widely available, only about half of individuals with FH receive treatment.^1,11,12 Inadequate treatment of FH represents a significant opportunity to improve the translation of evidence into practice. To date, strategies targeting patients and clinicians have focused on the translation of hypercholesterolemia treatment into clinical practice, but no strategy has specifically targeted individuals with FH.

Such general hypercholesterolemia patient-level interventions have included individualized risk communication and mailed outreach for primary CV disease prevention.^26,27 These approaches to increase disease prevention have yielded inconsistent results. While one study of direct-mailed outreach individualized to patients regarding the benefits of statin use resulted in increased statin prescriptions,²⁸ another study using direct-mailed outreach to patients to facilitate cholesterol testing had no effect on the uptake of cholesterol testing.²³ In another intervention, direct outreach by a community health worker to discuss individual CV risk led to increased discussion with primary care providers, but did not increase statin prescriptions for those eligible for primary CV disease prevention.²⁷ A different approach that used computerized decision aids to engage patients in conversations about CV disease risk and to support adherence to appropriate prevention methods found improved medication adherence and reduction in predicted CV risk.²⁹

Common barriers to general hypercholesterolemia care include suboptimal communication between the clinician and patient and misunderstandings about hypercholesterolemia and statin therapy.³⁰ One clinician-level intervention was to discuss CV risk with patients.^31,32 For example, when clinicians discuss individualized CV risk with their patients, there was an improvement in reaching target cholesterol levels.^31,32

It is unknown if these interventions will also help to improve uptake of treatment for individuals with FH. Preliminary experience at Geisinger suggests that the use of multidisciplinary care teams might lead to increase uptake of treatment,³³ but more evidence is needed to determine if this strategy effectively addresses barriers to care. This preliminary work, however, provides a basis for using the techniques and approaches, developed in the field of implementation science, to evaluate how to move guideline-recommended treatment into clinical practice.

The purpose of this study is to promote the adoption of guideline-recommended treatment for FH. This will be accomplished by assessing barriers and facilitators to FH care, matching the identified barriers and facilitators to implementation strategies across three disparate healthcare systems, and piloting one implementation strategy in one healthcare system. To achieve this goal, this study has three specific aims:

Aim 1: Determine the barriers to and facilitators of treatment of FH.

Aim 2: Develop a list of potential implementation strategies to promote the adoption of guideline-recommended treatment of individuals with FH.

Aim 3: Pilot one implementation strategy from Aim 2 in one health care system to evaluate the implementation outcomes of the strategy.

Guiding framework

The Practical, Robust, Implementation, and Sustainability Model (PRISM) will guide this project, and it has been adapted to fit project needs (Table 1). PRISM was specifically chosen because it builds upon the Reach, Effectiveness, Adoption, Implementation, and Maintenance (RE-AIM)³⁴ framework by incorporating elements of quality improvement in healthcare systems. Multiple models and one framework influenced the development of PRISM including: Diffusion of Innovations,³⁵ the Chronic Care Model,³⁶ the Model for Improvement, and the RE-AIM framework. The combination of these models and the framework enable the use of PRISM to translate research into practice by capturing and assessing intervention design, external environment, organizational characteristics, and population level impact of the intervention.³⁷ This model is best suited for this project because it is able to incorporate perspectives of organizational stakeholders within a healthcare system as well as patient-identified barriers and facilitators to care. In addition, we can assess the external factors that are imposed on those systems. PRISM incorporates RE-AIM³⁴, a framework that we will use to assess the pilot implementation of our strategy and its impact on improving uptake of guidelines-recommended practice.

Table 1.

Study design and methods

Aim	Purpose	Method
Aim 1: Determine the barriers to and facilitators of treatment of FH	Understand barriers and facilitators to care of recipients	Semi-structured interviews with patients and organizational stakeholders
Aim 2: Develop a list of potential implementation strategies to promote the adoption of guideline-recommended treatment of individuals with FH	Develop a list of potential implementation strategies for the uptake of guideline-recommendation treatment of FH	Focus groups with patients and organizational stakeholders
Aim 3: Pilot one implementation strategy from Aim 2 in one health care system to evaluate the implementation outcomes of the strategy	Evaluate the implementation strategy	Qualitative/quantitative metrics relevant to the implementation strategy

^*Maintenance will not be measured in this study

METHODS

Overview

This is a developmental study with three aims employing two main study designs. The first aim will use qualitative methods to determine barriers and facilitators to care and conceptualize strategies to improve care. Guided by PRISM, we will conduct semi-structured interviews with individuals with a diagnosis of FH and organizational stakeholders from three healthcare systems. Previously published literature will be used to help develop the interview guides, which are also designed to capture information relevant to implementation that has not been previously identified. In Aim 2, we will again use qualitative methods and convene stakeholders, a subset of those who participated in Aim 1 interviews, to pair barriers and facilitators with implementation strategies using intervention mapping, which allows for a systematic approach to intervention planning, implementation, and evaluation using a six-step process.^38,39 These strategies will be presented to patients and organizational stakeholders to ensure that the strategies are consistent with participants’ needs. In the third aim, we will pilot one of the implementation strategies developed in Aim 2 in the Geisinger system. The pilot will be a single-site study to evaluate the impact of implementing the strategy on uptake of guideline-recommended care for FH. We will use the RE-AIM framework that is embedded in PRISM to measure specific outcomes related to the success of the pilot. This pilot study will utilize data available in the electronic health record (EHR), administrative data, survey data, and semi-structured interviews.

Setting

Three separate study sites will be involved in this research project: an integrated delivery system located in Pennsylvania, a community-based health system, and an academic medical center (the latter two are located in Saint Louis, Missouri). The integrated delivery system serves its patients with hypercholesterolemia through primary care clinics, a multidisciplinary lipid clinic, and a population-based genomic screening program. In 2019, it opened a multidisciplinary clinic (led by a clinician board-certified in lipidology) with the goal of improving lipid management for high-risk patients. The lipid clinic involves team-based care provided by nutritionists, pharmacists, and genetic counselors. The MyCode® Community Health Initiative⁴⁰ is a population-based screening initiative that is returning actionable genetic results to consented individuals and serves as the population for the genetically identified population. At the community-based health center, many individuals are treated within primary care clinics and cardiology specialty clinics. The academic medical center has a dedicated endocrinologist who has specialized in treating individuals with lipid disorders for the past 35 years.

Aim 1: Determine the barriers to and facilitators of treatment of FH

Participants and recruitment strategy

Stakeholder participants, comprised of patients and organizational stakeholders, will be recruited from the integrated delivery, the community-based health system, and the academic medical center.

Patient stakeholders must have a clinical or genetic diagnosis of FH. A clinical diagnosis is defined as having either ≥ 2 outpatient encounters within the past 24 months associated with FH, ≥ 1 diagnosis code associated with FH on outpatient encounters, or inclusion of FH (ICD 10: E78.01) on the problem list of the medical record. A genetic diagnosis is defined as a pathogenic or likely pathogenic mutation in LDLR, APOB, or PCSK9.

We plan to recruit 10–15 individuals with FH at each of the three healthcare systems for a total of 30–45 participants. At the integrated delivery system and the community-based health system, a list of patients that meet at least one of the diagnostic criteria will be obtained from the EHR. Due to the possibility of miscoding of FH at the community-based health system, study personnel (TLS) will confirm the eligibility of individuals prior to recruitment. The academic medical center will recruit individuals with FH that have been seen in the lipid specialty clinic. In addition, we will recruit 10–15 individuals with a FH genetic diagnosis from a population-based genomic screening program (MyCode® Community Health Initiative⁴⁰) at the integrated delivery system. A total of 45–60 participants will be recruited for Aim 1. Figure 1 describes the method of recruitment and ascertainment of patient participants at each healthcare system.

Figure 1.

Recruitment and ascertainment locations for patient stakeholders across three healthcare settings. Patients with FH were recruited using two methods: clinical diagnosis based on ICD 10 coding or EHR encounters or genetic diagnosis via the MyCode Community Health Initiative.

Organizational stakeholders will be selected from setting in a consistent manner. Providers who are directly involved in the care of FH patients (cardiologists, endocrinologists, primary care providers, mid-level providers, pharmacists, genetic counselors), health insurance plan representatives, and administrative leaders who have relevant programmatic responsibilities will be eligible to participate. We will recruit 10–15 organizational stakeholder participants at each of the three sites for a total of 30–45 participants.

Data collection

All participants in Aim 1 will take part in semi-structured interviews. All interviews will be conducted in-person or by phone by one investigator (LKJ) using standardized interview guides. Guided by the PRISM framework, individualized interview guides have been created for each stakeholder group to discover barriers and facilitators to treatment (Supplemental Material 1). The proposed number of interviews should be sufficient, but interviews will be conducted as needed until thematic saturation occurs in each stakeholder group.⁴¹ Demographic data will be collected by a questionnaire administered during the interview. Stakeholders will be compensated for their time, where allowed.

Data analysis

Interviews will be digitally recorded, transcribed verbatim, and uploaded into Atlas.ti (www.atlasti.com) for analysis. A summary of each interview will be created by the interviewer after each interview. Key concepts from the stakeholder summaries allow for iteration of interview guides (if necessary), development of the initial codebook, and creation of an analytic framework.^42,43 Interpretive phenomenological analysis will be used when coding and interpreting interviews.⁴⁴ This analytic technique was chosen because of its focus on exploring the lived experience of the participant and the role of the researcher in interpreting the experience.⁴⁴ The first round of coding will use the codebook developed from the interview guides, stakeholder summaries, and PRISM constructs. If the pre-defined codes do not fully encompass emergent themes, then additional codes will be created from the text. Study team members will independently code 2–3 transcripts and then discuss their coding, adjust the codebook, and create a working analytic framework by grouping codes into categories or themes. This iterative coding process will continue until the code list is static, all transcripts are coded, and the analytic framework is finalized.

Aim 2: Develop a list of potential implementation strategies to promote the adoption of guideline-recommended treatment of individuals with FH

Participants and recruitment strategy

All stakeholders who participated in semi-structured interviews as part of Aim 1 will be eligible and asked to participate in an in-person focus group for Aim 2.

Data collection

We will conduct one focus group with each stakeholder group (one patient and one organizational stakeholder group) at each site, for a total of six stakeholder sessions. We will elicit feedback from each group on identified barriers and facilitators from Aim 1 and then develop and assess potential strategies that take advantage of facilitating factors while overcoming barriers (focus group guide available in supplemental material). A systematic review of implementation strategies to improve the uptake of statin therapy is being conducted and the results will be used to facilitate discussion during the stakeholder groups.

Data analysis

Intervention mapping will be used to ensure that the strategies developed are grounded in a theoretical framework. The 6 steps of intervention mapping are to establish 1) assess the barriers and facilitators for FH care; 2) develop objectives and outcomes; 3) select evidence-based methods and implementation strategies that align with each objective; 4) design the implementation strategy; 5) develop a plan to implement the implementation strategy; and 6) develop an evaluation plan.^38,39 The study team will iteratively develop a list of potential strategies to implement and evaluate using the information obtained from Aim 1 and the information gathered from the focus groups in aim 2. We will clearly outline our implementation strategies based on domain (e.g., actor(s), action(s), target(s) of the action, temporality, dose, implementation outcome(s) affected, justification) set forth by Proctor and colleagues.⁴⁵ These strategies will be presented in as a list of potential strategies.

Aim 3: Pilot one implementation strategy from Aim 2 in one health care system to evaluate the implementation outcomes of the strategy.

Participants and recruitment strategy

Depending on strategy selected from Aim 2, eligible subjects will be obtained at the patient and stakeholder (cardiology, primary care, pharmacy, genetics) level.

Data collection

We will use the RE-AIM framework to measure implementation outcomes related to the success of the pilot of the implementation strategy (Table 2). To determine Reach, defined as individuals impacted by the implementation strategy, we will measure of the individuals targeted by the implementation strategy, how many are impacted by it. The Effectiveness, defined as the impact of the implementation strategy on health behavior or outcome, we will measure change in knowledge, attitude, and behavior to be impacted by the implementation strategy. The Adoption, defined as reach at the setting level, will be the measure of how many adopt it, of the setting targeted. Finally, Implementation, defined as fidelity to and cost of the implementation strategy, will be gauged by how consistently the implementation strategy is delivered over time or setting and administrative data to assess the cost (e.g., personnel (e.g., number of FTEs) and implementation costs (e.g., cost of additional equipment)) of the components needed for the intervention to be successful. Due to the short project timeline, maintenance will not be measured as part of this study.

Table 2.

RE-AIM framework, outcome measures, and data sources for Aim 3

RE-AIM Framework Dimension	Definition	Example measure	Example data source
Reach	Impacted by the implementation strategy	Of individuals targeted by the implementation strategy, how many are impacted by it?	EHR
Effectiveness	Impact of the implementation strategy on health behavior or outcome	Measure change in knowledge, attitude, and behavior to be impacted by the implementation strategy	Survey, EHR
Adoption	Reach at the setting level	Of setting targeted by the implementation strategy, how many adopt it?	TBD
Implementation	Fidelity to the implementation strategy	How consistently is the implementation strategy delivered over time/setting?	TBD
	Cost of implementation strategy	Personnel and implementation cost for the implementation strategy	Administrative data (e.g., number of FTEs, cost of additional equipment)

TBD – to be determined

Data analysis

This is a single-site study to collect discrete implementation outcomes. Descriptive summary statistics, including means and standard deviations, will be presented for continuous variables. Categorical variables will be characterized by frequency and proportions.

DISCUSSION

The management of FH is critically important because of the high risk of serious CV events in untreated or undertreated individuals. A variety of implementation strategies exist to improve the uptake of guideline-recommended treatment for individuals diagnosed with hypercholesterolemia, but few have been specifically tested in individuals with FH and none from an implementation perspective. These strategies provide important information to guide the development of new implementation strategies, but the lack of a comprehensive approach, coupled with incomplete engagement with relevant stakeholders presents barriers to their current application.

This research will promote understanding of the various barriers and facilitators that patients and organizational stakeholders experience while caring for themselves or individuals with FH across a variety of patient and health system settings. Previous research has focused on either patient or organizational stakeholder perspectives, but this approach incorporates both to develop a more robust strategy that will fulfill the needs of all relevant stakeholders.

Strengths of this study include a framework-based approach and inclusion of diverse patient and organizational stakeholder perspectives from three healthcare systems. The diversity of stakeholder groups that will participate in interviews and focus groups will increase the generalizability of the findings from this study.

This study has several potential limitations. We are only implementing one strategy at one healthcare system, but multiple strategies will likely be needed to improve the uptake of guideline-recommended treatment. In addition, we might design some strategies that require adaptation in order to be implemented into other healthcare systems. Another limitation is that medication adherence will be measured using a self-reported tool. The Shea scale has been shown to be accurate in measuring adherence to cholesterol-lowering regimens with a 73.7% sensitivity to detect poor adherers.⁴⁶ Systematic reviews of medication adherence scales have shown that this is an acceptable method but not necessarily the optimal method to use.⁴⁷

LIST OF ABBREVIATIONS

CV

cardiovascular

EHR

electronic health record

FH

familial hypercholesterolemia

LDL-C

low density lipoprotein cholesterol

LLT

lipid lowering therapy

RE-AIM

reach, effectiveness, adoption, implementation, maintenance

PRISM

practical, robust implementation and sustainability model