Optimizing low-density lipoprotein cholesterol (LDL-C) management – a US physician survey of barriers and burdens

Abstract

Background and Aims

Improving care of patients with hyperlipidemia requires an understanding of the barriers physicians perceive in prescribing low-density lipoprotein cholesterol (LDL-C)-lowering therapies. This study explores physicians’ perceptions of time and resource burdens, identify perceived patient adherence barriers, and examine factors influencing physicians’ decision-making in LDL-C management.

Methods

This is a non-interventional, cross-sectional, online survey of US-based primary care practitioners (PCP) and cardiologists who recommended or provided lipid-lowering therapy (LLT) to ≥50 adults per month, practiced for ≥2 years, and completed the survey in English. The survey comprised multiple-choice, constant sum, and numerical questions about physician decision-making, patient management, and perceptions of patient attitudes/behaviors regarding LDL-C management. Descriptive univariate analyses were conducted.

Results

200 PCPs and 200 cardiologists completed the survey. Most physicians reported prescribing lipid-lowering therapy (LLT) and that patients declined injectable proprotein convertase subtilisin/kexin type 9 inhibitors (PCSK9i). They attributed this refusal to cost/insurance, fear/discomfort taking injections, and a preference for oral therapies. Physicians viewed patients with a history of ASCVD, with LLT experience, and those with greater understanding of ASCVD risk to have higher LLT adherence compared to those without. Most physicians spent a median of 10 min in shared decision-making conversations, regardless of therapies they prescribed. They reported needing longer to instruct patients during adherence counseling for PCSK9is than for oral therapies.

Conclusions

Our findings suggest patient, clinician, and system barriers may all hinder LDL-C management and adherence. A greater understanding of the association between perceived barriers and real-world behaviors will help optimize lipid management.

Background

Cardiovascular disease (CVD) remains the leading cause of morbidity and mortality worldwide [[1], [2], [3], [4]]. Atherosclerotic cardiovascular disease (ASCVD), characterized by plaque build-up in artery walls, is considered a major contributor to the overall CVD burden [5].

Lowering low-density lipoprotein cholesterol (LDL-C) is central to reducing ASCVD risk [6,7], and statin therapy has long been the cornerstone of lipid-lowering therapy (LLT) [8]. Non-statin oral LLTs include ezetimibe and bempedoic acid, and injectable treatments include inclisiran and proprotein convertase subtilisin/kexin type 9 (PCSK9) monoclonal antibodies [[8], [9], [10]].

Guideline-concordant care includes various LDL-C-lowering treatment options [[10], [11], [12]] known to improve patient outcomes [13,14]. However, pharmacoepidemiologic data indicate that LDL-C management remains suboptimal in real-world clinical practice [6,[15], [16], [17], [18], [19]]. While use of combination therapy has increased in recent years, statin monotherapy remains the most widely prescribed LLT, and non-statin LLT underuse persists [16,17,[20], [21], [22]].

Complex guidelines, clinical inertia, physician time constraints, patient concerns about side effects, and the role of social and media influences are among the barriers to effective LDL-C management [[23], [24], [25], [26]].

Understanding the barriers and burdens physicians perceive in prescribing different LDL-C-lowering therapies plays an important role in identifying opportunities to improve care. In this study, we surveyed practicing primary care physicians (PCP) and cardiologists about their perspectives on the barriers to and burdens of prescribing and managing LDL-C-lowering therapies. We aimed to explore physicians’ perceptions of the time and resource burdens associated with optimized LDL-C management, identify their perceived patient adherence barriers, and examine factors influencing physicians’ decision-making in LDL-C management, including shared decision-making and the impact of patient characteristics, regimen complexity, and costs.

Methods

Study design

Our design was a non-interventional, cross-sectional, online survey. The study was conducted between April and June 2024. Potentially eligible participants completed online screening, received a downloadable document containing study information and a consent form, provided consent to participate, and completed the survey.

Ethics

The protocol, screening and consent forms, and survey were reviewed by an external institutional review board (Advarra) and the study was determined to be exempt from institutional review board oversight in February 2024.

Recruitment and eligibility

Physicians were identified and recruited via a validated national third-party medical panel, which includes verified, actively practicing US physicians. Screening questions confirmed current outpatient clinical activity and prescribing responsibilities for lipid-lowering therapy. An external vendor recruited study participants using the online panel and fielded the survey. Quota sampling was used to ensure a diverse target sample of 200 PCPs and 200 cardiologists. Based on published physician demographics [[27], [28], [29]], recruitment targets included a minimum of 10 % of PCPs practicing in rural or remote areas, and no >65 % and 75 % of PCPs and cardiologists, respectively, being male. Although recruitment quotas were not set by US region, geographic distribution was monitored throughout data collection. Eligible physicians were PCPs and cardiologists who self-reported to be recommending or providing LLT to at least 50 adults per month, had been practicing for at least 2 years, practiced within the US (50 states and District of Columbia), and were able to complete the survey in English. Physicians unwilling to consent to participate and who did not have experience in outpatient settings (e.g., solely in inpatient, urgent care, or emergency settings, or in a health maintenance organization-owned facility) were excluded.

Survey

The survey was based on expert input and a targeted literature review (Supplementary Table S1). A survey pilot targeting ∼10 % of the sample size was conducted to identify unexpected response distributions or write-in responses for inclusion in the final survey.

The questionnaire comprised multiple-choice, constant sum, and numerical questions about physician decision-making, current patient management, and perceptions of patient attitudes/behaviors (see Supplementary Materials for survey questionnaire). Participants were asked about their sociodemographic characteristics and prescribing practices, and about their patients’ characteristics. They were asked about their experience with recommending LLT to patients, their perceptions of patient attitudes and behaviors, and their ongoing patient management, including adherence counseling (specifically with regard to LLT). Before participants were asked how often they engaged in shared decision-making, they were provided with examples (such as educating patients, offering brochures, comparing options, addressing concerns, and allowing questions) of what the process entails.

Data analysis and statistics

All data were analyzed using SAS version 9.4. Descriptive univariate analyses were conducted on the survey data. For categorical variables (e.g., multiple choice questions and questions using Likert scales), frequencies and percentages (n, [ %]) were calculated. For continuous variables, distributional statistics were calculated. Bivariate comparisons were conducted to compare the results between PCPs and cardiologists using Pearson chi-square tests or Fisher’s exact tests.

Cross-tabulations were conducted between reasons for declining therapy, frequency of declining therapies, and time spent on different types of physician-patient counseling. For comparisons or cross-comparisons, Chi-square (categorical variables), Fisher’s exact (binary variables), and Wilcoxon rank-sum (continuous variables) statistical comparison tests were performed based on the variable type and distribution of the group.

Results

Physician sociodemographic characteristics

The pilot survey was conducted with 13 % of all physicians (n = 52). The survey was subsequently revised to consolidate two similar response options to one question and add an option to a multiple-choice question. Since changes to the survey were minimal, the pilot sample was included in the final analysis of 400 physicians.

Of the 400 physicians who completed the survey (see Supplementary Figure S1 for physician disposition), 200 were PCPs and 200 were cardiologists (Table 1). The mean (SD) age of the total sample was 49.3 (11.4) years. Most cardiologists (88.5 %) and PCPs (64.5 %) were male, more than half of all physicians were White (52.8 %), and just under one-third-were Asian (31.3 %), with similar racial distributions among PCPs and cardiologists. Few physicians (6.0 % PCPs; 6.5 % cardiologists) were of Hispanic, Latino, or Spanish ethnicity.

Table 1.

Physician-reported sociodemographic characteristics and patient caseload.

Variable and statistic or category	PCPs (N = 200)	Cardiologists (N = 200)	P-value
Age+
N	200	200	0.128W
Mean (SD)	48.39 (11.42)	50.24 (11.42)
Median	48	50
Range	30.0 - 73.0	32.0 - 85.0
Gender, N ( %)
Female	68 (34.0)	21 (10.5)	<0.001E
Male	129 (64.5)	177 (88.5)
I prefer not to answer	3 (1.5)	2 (1.0)
Race, N ( %) *
White	107 (53.5)	104 (52.0)	0.524C
White and Black/AA	0 (0.0)	1 (0.5)
White and Other	1 (0.5)	0 (0.0)
Black/AA	9 (4.5)	5 (2.5)
Black/AA and Asian	1 (0.5)	0 (0.0)
American Indian/Alaska Native	1 (0.5)	1 (0.5)
Asian	64 (32.0)	61 (30.5)
Asian and Native Hawaiian or other Pacific Islander	1 (0.5)	0 (0.0)
Asian and Other	0 (0.0)	1 (0.5)
Other	5 (2.5)	8 (4.0)
Unknown	11 (5.5)	19 (9.5)
Ethnicity, N ( %)
Hispanic/Latino/Spanish	12 (6.0)	13 (6.5)	0.240E
Non-Hispanic/Latino/Spanish	176 (88.0)	182 (91.0)
I prefer not to answer	12 (6.0)	5 (2.5)
Years in practice post-residency/fellowship+
N	200	200	0.714W
Mean (SD)	16.53 (9.99)	17.01 (9.99)
Median	17	15.5
Range	2.0 - 41.0	2.0 - 43.0
Practice geography, N ( %)
Urban	72 (36.0)	100 (50.0)	<0.001C
Suburban	96 (48.0)	88 (44.0)
Rural	32 (16.0)	12 (6.0)
Practice setting, N ( %) *+
Outpatient solo practice	28 (14.0)	12 (6.0)	0.008C
Outpatient group practice – single specialty	64 (32.0)	79 (39.5)	0.118C
Outpatient group practice – multi-specialty	97 (48.5)	55 (27.5)	<0.001C
Hospital inpatient setting	25 (12.5)	74 (37.0)	<0.001C
Urgent care setting	8 (4.0)	3 (1.5)	0.126C
Outpatient VA facility	2 (1.0)	4 (2.0)	0.411C
Academic/university setting	25 (12.5)	76 (38.0)	<0.001C
HMO-owned facility	2 (1.0)	2 (1.0)	>0.999C
Integrated delivery network	9 (4.5)	11 (5.5)	0.646C
Hospital emergency setting	5 (2.5)	13 (6.5)	0.054C

Abbreviations: AA, African American; HMO, health maintenance organization; N, number of participants; LLT, lipid-lowering therapy; PCP, primary care practitioner; Q, quartile; SD, standard deviation, VA, Veterans Affairs. ^C Chi-square; ^E Exact Fisher; ^W Wilcoxon rank sum. * Respondents could select multiple responses;⁺Questions were asked as part of the eligibility screener.

Cardiologists were more likely to practice in an academic/university setting than PCPs (38.0 % vs 12.5 %). Cardiologists saw a significantly higher proportion of patients who were ≥65 years (57.9 % vs 42.6 %) and male (55.6 % vs 48.3 %) (Supplementary Table S2) than PCPs.

By US Bureau Census Bureau region, 34 % of participants were from the South, 27 % from the Northeast, 23 % from the West, and 16 % from the Midwest.

Physician prescribing behaviors

Most physicians (75.9 % of PCPs and 85.1 % of cardiologists) reported prescribing oral LLT to most of their patients with elevated LDL-C (Supplementary Figure S2). Approximately two-thirds (69.1 % of PCPs and 62.2 % of cardiologists) reported that most of their patients were not receiving any type of injectable therapy, and few (9.0 % of PCPs and 18.9 % of cardiologists) reported prescribing PCSK9 inhibitor (PCSK9i) injectables to their patients. About half (50.5 % of PCPs and 47.0 % of cardiologists) indicated that whether or not a patient was already prescribed an injectable therapy would not affect their decision to prescribe a PCSK9i injectable, while 42.5 % of PCPs and 46.5 % of cardiologists reported they would be more likely or much more likely to prescribe a PCSK9i to a patient already taking another injectable therapy.

Most physicians correctly – per the knowledge check – identified statins (98.5 % PCP; 96.5 % cardiologists) and ezetimibe (94.5 % PCPs; 96.0 % cardiologists) as LLTs (Supplementary Table S3), but fewer identified PCSK9i (84.0 % PCPs; 95 % cardiologists) and bempedoic acid (47.5 % PCPs; 85.5 % cardiologists) as LLTs.

Patient refusals of LDL-C-lowering therapies

Physicians reported that patients often refused LLTs. In general, PCSK9i injectables were more often refused than any LLT in general (Fig. 1A). Both LLT and PCSK9i refusals were more often reported by PCPs than cardiologists; 8.0 % of PCPs and 7.5 % of cardiologists reported their patients refused LLT more than half of the time, and 31.5 % of PCPs and 13.5 % of cardiologists reported their patients refused PCSK9i injectables more than half of the time.

Fig. 1.

A. Physician-reported patient refusal rates for lipid-lowering therapies. Participants (n = 200 PCPs; n = 200 cardiologists) replied to the questions “Do some patients decline lipid-lowering therapy (any type) that you prescribe? If so, how often do they decline what you prescribe?” and “Do some patients decline the PCSK9i injectables that you prescribe? If so, how often do they decline what you prescribe?” Abbreviations: LLT, lipid-lowering therapy; PCP, primary care practitioner; PCSK9i, proprotein convertase subtilisin/kexin type 9 inhibitor. 1B Physician-reported reasons for patients refusing any lipid-lowering therapy (top) or refusing PCSK9i injectables (bottom). Physicians who reported that their patients declined therapy >10 % of the time (n = 138 PCPs; n = 110 cardiologists) replied to the questions “What are some of the reasons that patients provide when they refuse to initiate LLT that you prescribe? Please select all that apply.” (top panel) and “What are some of the reasons that patients provide when they refuse to initiate the PCSK9i injectables that you recommend? Please select all that apply.” (bottom panel). Abbreviations: LLT, lipid-lowering therapy; PCP, primary care practitioner; PCSK9i, proprotein convertase subtilisin/kexin type 9 inhibitor.

Among physicians who reported their patients declining any type of LLT >10 % of the time, the most common reasons for declining were preferences for lifestyle changes or natural remedies, concerns about potential side effects, and a general dislike of taking medications (Fig. 1B). The three most common reasons physicians reported for patients refusing PCSK9i therapy were cost/insurance, fear/discomfort taking injections, and a preference for oral therapies over injections (Fig. 1B).

Physicians who reported their patients refused any LLT <25 % of the time spent less time on first prescription counseling than those reporting patient refusal rates of ≥25 % (Supplementary Table S4).

Physicians who reported their patients refused PCSK9i injectables <25 % of the time and those reporting patient refusal rates of ≥25 % spent similar amounts of time on initial shared decision-making conversations and counseling sessions (Supplementary Table S5).

Shared physician-patient decision-making practices

Most physicians (PCPs: 93.0 %; cardiologists: 87.0 %) reported implementing shared decision-making practices “always or almost always” or “most of the time” when they first recommended LLT (Supplementary Figure S3). Most (PCPs: 95.0 %; cardiologists: 87.4 %) reported conducting these conversations themselves; the rest delegated conversations to another physician, physician assistant, nurse, nurse practitioner, or office staff.

Both PCPs and cardiologists spent a median [interquartile range; IQR] of 10 [5.0–15.0] minutes on initial shared decision-making conversations with patients. 95.5 % of PCPs and 91.5 % of cardiologists reported dedicating shared decision-making sessions to educating patients on treatment benefits, 68.0 % of PCPs and 68.5 % of cardiologists reported allowing time for questions, and 55.5 % of PCPs and 47.0 % of cardiologists reported asking about the patient’s fears or worries. Fewer than half of the physicians reported using these sessions to compare treatment options (reported by 46.0 % of PCPs and 47.5 % of cardiologists) or provide pamphlets or brochures (reported by 13.0 % of PCPs and 17.5 % of cardiologists).

Physicians spent similar amounts of time (median [IQR]) on initial shared decision-making conversations regardless of the frequency with which they conducted these conversations (10.0 [5.0 to 15.0] minutes) (Supplementary Figure S4; Supplementary Table S6). Physicians also spent a similar amount of time on initial shared decision-making conversations irrespective of whether they prescribed injectable therapies or not (10.0 [5.0 to 15.0] minutes) (Supplementary Table S7).

Physicians’ perceptions of patient adherence to LDL-C-lowering therapies and adherence counseling

PCPs and cardiologists had similar perceptions of patient adherence to LLT based on sociodemographic, clinical, and treatment characteristics (Fig. 2). They reported expecting experienced LLT users, patients with a history of ASCVD, those aware of their LDL-C goals, and those with a high baseline LDL-C to be more adherent than inexperienced LLT users, adults at risk for ASCVD, patients uninformed about their LDL-C goals, or those with near-target baseline LDL-C. Both PCPs and cardiologists believed patients who understood their ASCVD risk to be more adherent to LLT than those who did not. Perceptions of patient adherence to LLT did not significantly differ by physician age, sex, or race (Supplementary Table S8).

Fig. 2.

Physician perceptions of adherence. Participants (n = 200 PCPs; n = 200 cardiologists) "reported duration of adherence/compliance discussion at treatment initiation and follow up visits and" replied to the question “In your experience, who is more adherent to lipid lowering therapy?” Participants were provided with a set of patient sociodemographic and clinical characteristics and were asked to rank patient adherence matching each characteristic. Each characteristic had two opposites as anchors (e.g., LLT naïve vs LLT experienced), and each set of characteristics was ranked on a five-point scale (“much more adherent” or “slightly more adherent” for one anchor, to “no difference,” to “slightly more adherent” or “much more adherent” for the other anchor). Abbreviations: ASCVD, atherosclerotic cardiovascular disease; LDL-C, low-density lipoprotein cholesterol; LLT, lipid-lowering therapy; PCP, primary care practitioner.

Just under half of the physicians (47.0 % of PCPs and 43.5 % of cardiologists) reported “always or almost always” discussing adherence when starting a new treatment (Table 2). Over half (63.5 % of PCPs and 51.0 % of cardiologists) reported conducting discussions on adherence themselves, and about a quarter delegated such discussions to a nurse or nurse practitioner. These initial adherence conversations reportedly took 6.0 (5.0 to 10.0) minutes when conducted by PCPs and 5.0 (5.0 to 10.0) minutes when conducted by cardiologists.

Table 2.

Adherence discussions.

Variable	Statistic or Category	PCP (N = 200)	Cardiologist (N = 200)	P-value
HCP doing adherence discussion, NA ( %)	Nurse	37 (18.5)	40 (20.0)	0.027C
	Nurse practitioner	8 (4.0)	22 (11.0)
	Office staff	21 (10.5)	18 (9.0)
	Pharmacist	2 (1.0)	3 (1.5)
	Physician - Self	127 (63.5)	102 (51.0)
	Physician assistant	4 (2.0)	8 (4.0)
	Other (Please specify:)	0 (0.0)	2 (1.0)
	Another physician	1 (0.5)	5 (2.5)
How often is the discussion about adherence as long as you think the patient really needs?, N ( %)	≥90 % of the time	39 (19.5)	56 (28.0)	0.304C
	75–89 % of the time	56 (28.0)	49 (24.5)
	50–74 % of the time	60 (30.0)	48 (24.0)
	25–49 % of the time	29 (14.5)	32 (16.0)
	10–24 % of the time	13 (6.5)	14 (7.0)
	1–9 % of the time	3 (1.5)	1 (0.5)
Time for comprehensive counseling about oral therapies when first prescribing	Mean (SD)	9.97 (7.07)	9.90 (9.46)	0.299W
	Median (Q1 to Q3)	9.5 (5.0 to 15.0)	7.0 (5.0 to 10.0)
	Range	1.0 to 40.0	1.0 to 80.0
Time for comprehensive counseling about oral therapies at follow-up visit	Mean (SD)	5.88 (5.20)	6.33 (8.17)	0.283W
	Median (Q1 to Q3)	5.0 (3.0 to 5.0)	5.0 (2.0 to 5.0)
	Range	0.0 to 30.0	1.0 to 70.0
Time for comprehensive counseling about PCSK9i injectables when first prescribing	Mean (SD)	13.44 (9.30)	12.89 (9.73)	0.372W
	Median (Q1 to Q3)	10.0 (7.0 to 15.0)	10.0 (6.0 to 15.0)
	Range	0.0 to 50.0	0.0 to 80.0
Time for comprehensive counseling about PCSK9i injectables at follow-up visit	Mean (SD)	7.20 (6.31)	7.53 (8.13)	0.715W
	Median (Q1 to Q3)	5.0 (4.0 to 10.0)	5.0 (3.0 to 10.0)
	Range	0.0 to 45.0	0.0 to 70.0
Top reason comprehensive counseling for PCSK9i injectables takes longer than oral therapy?, N ( %)a	N	141	138
	Educating on administration of therapy	81 (57.4)	76 (55.1)	0.022E
	Disease progression	3 (2.1)	16 (11.6)
	Risks and benefits of the therapy	36 (25.5)	32 (23.2)
	Adherence discussion	18 (12.8)	12 (8.7)
	Other (please specify)	3 (2.1)	2 (1.4)
Notifications: starts on a new statin prescription, N ( %)	Yes, usually	54 (27.0)	52 (26.0)	0.862C
	Sometimes	52 (26.0)	55 (27.5)
	No, usually not	87 (43.5)	83 (41.5)
	Don’t know	7 (3.5)	10 (5.0)
Notifications: fails to refill a statin prescription that you prescribed, N ( %)	Yes, usually	36 (18.0)	47 (23.5)	0.474C
	Sometimes	82 (41.0)	70 (35.0)
	No, usually not	75 (37.5)	77 (38.5)
	Don’t know	7 (3.5)	6 (3.0)
Notifications: starts on a self-injected LLT, and fails to fill the first prescription, N ( %)	Yes, usually	38 (19.0)	51 (25.5)	0.100C
	Sometimes	53 (26.5)	63 (31.5)
	No, usually not	90 (45.0)	75 (37.5)
	Don’t know	19 (9.5)	11 (5.5)
Notifications: has ongoing treatment with an PCSK9i injectable, and the patient fails to refill/no-shows, N ( %)	Yes, usually	35 (17.5)	58 (29.0)	0.004C
	Sometimes	64 (32.0)	70 (35.0)
	No, usually not	80 (40.0)	64 (32.0)
	Don’t know	21 (10.5)	8 (4.0)
Follow-up: starts on a new statin prescription, N ( %)	N	106	107
	Very likely	39 (36.8)	41 (38.3)	0.808E
	Likely	47 (44.3)	51 (47.7)
	Unlikely	17 (16.0)	13 (12.1)
	Very unlikely	3 (2.8)	2 (1.9)
Follow-up: fails to refill a statin prescription that you prescribed, N ( %)	N	117	118
	Very likely	56 (47.9)	38 (32.2)	0.084E
	Likely	45 (38.5)	62 (52.5)
	Unlikely	14 (12.0)	16 (13.6)
	Very unlikely	2 (1.7)	2 (1.7)
Follow-up: starts on a self-injected LLT, and fails to fill the first prescription, N ( %)	N	114	91
	Very likely	66 (57.9)	36 (39.6)	0.018C
	Likely	34 (29.8)	44 (48.4)
	Unlikely	14 (12.3)	11 (12.1)
Follow-up: has ongoing treatment with an PCSK9i injectable, and the patient fails to refill/no-shows, N ( %)	N	128	99
	Very likely	62 (48.4)	41 (41.4)	0.463E
	Likely	57 (44.5)	48 (48.5)
	Unlikely	8 (6.3)	10 (10.1)
	Very unlikely	1 (0.8)	0 (0.0)

Table 2. Adherence discussions. Abbreviations: HCP, healthcare practitioner; LLT, lipid-lowering therapy; N, number; PCP, primary care practitioner; PCSK9i, proprotein convertase subtilisin/kexin type 9 inhibitor; Q, quartile; SD, standard deviation.; Q, quartile; SD, standard deviation. ^C Chi-square; ^E Exact Fisher; ^W Wilcoxon rank sum. ^a Only asked of those who said that discussions for PCSK9i injectables take longer than oral therapy; ^C Chi-square; ^E Exact Fisher; ^W Wilcoxon rank sum.

The physicians reported that comprehensive first prescription counseling took longer for PCSK9i injectables (PCPs: 10.0 [7.0 to 15.0] minutes and cardiologists: 10.0 [6.0 to 15.0] minutes) than for oral LLTs (PCPs: 9.5 [5.0 to 15.0] minutes and cardiologists: 7.0 [5.0 to 10.0] minutes). More than half of the physicians (PCPs: 57.4 %; cardiologists: 55.1 %) attributed the longer time needed for comprehensive counseling for PCSK9i injectables to time spent educating patients on therapy administration.

Few physicians reported “usually” receiving notifications when their patients failed to fill or refill LLT prescriptions. Around a fifth (18.0 % of PCPs and 23.5 % of cardiologists) reported “usually” receiving notifications when patients failed to refill statin prescriptions, 19.0 % of PCPs and 25.5 % of cardiologists reported “usually” receiving notification when patients due to start self-injected LLT failed to fill prescriptions, and 17.5 % and 29.0 % reported “usually” receiving notifications when patients with ongoing PCSK9i injectable treatment failed to refill prescription. When physicians did receive pertinent notifications, most were likely or very likely to follow up on that information (Table 2).

Discussion

ASCVD constitutes a major clinical and economic burden [[3], [4], [6], [5]]. Elevated LDL-C levels, which represent a leading ASCVD risk factor (5, 7), remain sub-optimally managed despite the availability of safe and effective LLTs.

Recent studies show that many patients eligible for LLT per major US and European treatment guidelines [[10], [11], [12],30] are not receiving treatment [6,[15], [16], [17], [18], [19], [20], [21],30,31]. Patient, practitioner, and health system barriers likely all contribute to LDL-C guideline-discordant care [26,31]. Although guidelines [10,11] stipulate risk-based LDL-C thresholds and make recommendations on how they may be achieved through non-statin therapies, current performance metrics from the National Committee for Quality Assurance Healthcare Effectiveness Data and Information Set and the Centers for Medicare & Medicaid Services are based solely on use of statin therapy [32]. The fact that these performance metrics are process measures rather than clinically-oriented outcomes may add to confusion and possibly clinical inertia and guideline-discordant management.

In this study, we surveyed PCPs and cardiologists about their perspectives on the barriers to and burdens of prescribing and managing LDL-C-lowering therapies. Most physicians reported prescribing LLTs, typically oral statins, to their patients. However, in line with previous findings highlighting key knowledge gaps in dyslipidemia management [19,33], we demonstrate persistent barriers to optimizing LDL-C management, especially with injectables.

Physicians often reported patients declining PCSK9i injectables. Physicians attributed PCSK9i injectable refusal by patients to cost/insurance, fear/discomfort taking injections, and a preference for oral therapies. This broad range of refusal reasons may suggest the need for physicians to spend more time engaging with patients through education to help them better understand their ASCVD risk, and the potential benefits and risks of therapy, thereby facilitating shared decision making and informed treatment choices [34].

Physicians perceived non-Black, non-Hispanic, White, or Asian patients to be slightly more adherent than Black or Hispanic patients, suggesting that racial/ethnic disparities in LLT utilization (previously described in [20]) may be linked to preconceived ideas around adherence. For example, Black and Hispanic patients are more likely to be eligible for LLT but less likely to receive recommended treatment compared to non-Hispanic, White, and Asian individuals [[22], [35]].

The lower refusal rates reported by cardiologists may in part be attributable to patient case mix (e.g., cardiologists seeing more higher-risk patients) and/or may reflect referral dynamics as prior research suggests that generalists may use patient referrals to further patient adherence and trust in treatment recommendations [36].

Shared decision-making is a strong determinant of confidence and trust between PCPs and patients [37]. Here, most physicians reported dedicating a median of 10 min, around half the standard US PCP consultation time [38], to shared decision-making conversations, regardless of whether they prescribed oral or injectable therapies. However, the physicians reported that they needed longer to instruct patients during adherence counseling when administering PCSK9i injectables than when using oral LLT. Almost half of the physicians also reported being more likely to prescribe a PCSK9i injectable if a patient had prior experience of other injectable therapies. To enhance accuracy, these self-reported time estimates could be checked against objective data gathered from time and motion studies.

Few physicians reported usually receiving notifications regarding failure to fill prescriptions for LLT. However, when physicians did report receiving pertinent notifications, they were likely to follow up on them. Efforts to enhance coordinated communication and information exchange between healthcare practitioners, such as clinicians and pharmacists, could improve continuity of care and lead to better outcomes for patients with elevated LDL-C [[39], [40], [41]]. Integration of other members of the healthcare team into adherence discussions, clear guidance on effective communication for ASCVD prevention, and patient-friendly lipid-management information may also help [19,42].

There are some limitations to our study. Participants recruited from online panels may not be representative of the overall US PCP or cardiologist population. We attempted to minimize this risk by establishing soft recruitment quotas. The 400 participants were representative of the overall US PCP and cardiologist populations in terms of key demographics. While the quota of no >75 % of cardiologists being male was exceeded, the high percentage of male participants in the study population reflects the high percentage of male practitioners in general and male cardiologists in particular; in 2016, only 13 % of all US cardiologists were female [29].

However, with a mean age of 49.3 (11.4) years, the sample may have been skewed toward younger and more digitally engaged physicians compared with the overall US physician population (average age: 54.4 years [43]), potentially limiting generalizability.

The study was conducted only with English-speaking participants based in the US; our findings may thus not necessarily represent the lived experience of non-English speaking US practitioners and their patients. Further investigation to better understand the patient case mix of these physicians would therefore add additional value. Furthermore, the study included only physicians and did not capture the perspectives of advanced practice providers, who may contribute substantially to outpatient cardiovascular care.

Other potential limitations of this study relate to the screening process (self-reported data) and to recall and social desirability biases, which constitute limitations of any primary data collection during survey studies. To minimize the potential impact of social desirability bias, however, the survey questionnaire was self-administered and anonymous, and the survey instructions emphasized the anonymous nature of the study.

As our study was based on physician perceptions, our findings should be interpreted as indicative of perceived practices and behaviors, rather than objectively measured physician or patient behaviors. Physician recall and interpretation of patient behavior are subject to known biases; Hines and Stone demonstrated that cardiologists’ recollections of which patients were nonadherent differed markedly from patients’ own reports, illustrating the potential for misalignment between provider perceptions and actual patient behaviors [44].

Conclusions

Our findings are based on physicians’ self-reported experiences and reveal consistent and actionable trends in physician-perceived barriers and burdens of optimizing LDL-C management. The multifactorial reported barriers and burdens suggest opportunities to strengthen LDL-C management through clearer physician-patient communication and education, especially about the potential benefits and risks of potential treatments, broader integration of non-statin therapies, and improved continuity of care. Addressing perceived challenges may help align practice with guideline-directed therapy, close the existing LLT gap, and improve patient adherence and outcomes. Further research is needed to develop targeted strategies to address these barriers, which may help physicians promote patient understanding of and adherence to LLT, facilitate shared decision-making, and ensure that patients receive appropriate LLT to meet their needs.

Contributors

Medical writing was provided by Louisa F. Ludwig-Begall, PhD, and Stephen Gilliver, PhD (PPD clinical research business of Thermo Fisher Scientific) in accordance with Good Publication Practice guidelines and was funded by Merck Sharp & Dohme LLC., a subsidiary of Merck & Co., Inc., Rahway, NJ, USA.

Funders

This study was supported by Merck Sharp & Dohme LLC., a subsidiary of Merck & Co., Inc., Rahway, NJ, USA.

Prior presentations

Results included in this paper were previously presented at the 2025 National Lipid Association conference.: Karmarkar T, Bash LD, Exter J, Schmier JK, Jayade SP, Roney KC, Simpson RJ, Baum S, Leiter LA. The Burdens And Barriers Of Optimizing LDL-C Management: A Survey Of Physicians In The United States. NLA, Miami, FL, USA: May 2025.

Ethics approval and consent to participate

The Advarra institutional review board determined the study to be exempt from institutional review board oversight. All study participants provided electronic informed consent.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

CRediT authorship contribution statement

Lawrence A. Leiter: Writing – review & editing, Methodology, Conceptualization. Taruja Karmarkar: Writing – review & editing, Project administration, Methodology, Conceptualization. Lori D. Bash: Writing – review & editing, Methodology, Conceptualization. Jason Exter: Writing – review & editing, Conceptualization. Jordana K. Schmier: Writing – review & editing, Project administration, Methodology, Investigation, Data curation. Sayeli P. Jayade: Writing – review & editing, Methodology, Investigation, Formal analysis, Data curation. Kyle C. Roney: Writing – review & editing, Visualization, Software, Formal analysis, Data curation. Ross J. Simpson: Writing – review & editing, Methodology, Conceptualization. Seth J. Baum: Writing – review & editing, Methodology, Conceptualization.

Declaration of competing interest

At the time this study was conducted, Taruja Karmarkar, Lori D. Bash, and Jason Exter are/were employees of Merck & Co., Inc. Rahway, NJ, USA, which funded this study. Jordana K. Schmier, Sayeli P. Jayade, and Kyle C. Roney are employees of OPEN Health, which received funding from Merck & Co., Inc. Rahway, NJ, USA in connection with the development of this manuscript, study design, management, and statistical analysis for the study. Jason Exter is an employee of Alnylam Pharmaceuticals. Seth J. Baum has received consulting fees from Altimmune, Amgen, Beren Therapeutics, Eli Lilly, Ionis, Madrigal, Merck, Novartis, and Regeneron and has received honoraria for serving on scientific advisory boards for Altimmune, Amgen, Boehringer Ingelheim, Eli Lilly, Esperion, Ionis, Novartis, and Regeneron. He has served as Chairman of Scientific Advisory Board, National Triglyceride Alliance (payments received) and Chairman of the Board of Directors, Family Heart Foundation (FHF) (no payments received). Lawrence A. Leiter has received payments from Amgen, Eli Lilly, HLS, Merck, Novartis, and Regeneron for providing CME. He has served on advisory boards (payments received) for Amgen, Eli Lilly, HLS, Merck, Novartis, and Regeneron. He has served as an Executive/Steering Committee member (payments received) for Amgen, Eli Lilly, and Novartis. Ross J. Simpson Jr has received consulting fees from Merck & Co., Inc. Taruja Karmarkar owns Merck & Co., Inc stock. Jason Exter owns Merck & Co., Inc and Alnylam Pharmaceuticals stock.