Evaluating severe hypercholesterolemia in an uninsured, vulnerable population receiving care at a community clinic


 BackgroundIndividuals with LDL-C ≥190 mg/dL (severe hypercholesterolemia) require treatment with high-intensity statins and should be evaluated for underlying causes including Familial Hypercholesterolemia (FH). The characteristics of patients with LDL-C ≥ 190 mg/dL has been reported in routine healthcare settings, but limited data exist about uninsured populations.ObjectivesTo examine the scope of LDL-C ≥ 190 mg/dL in a clinic serving uninsured populations.MethodsPatients with LDL-C ≥190 mg/dL at North Dallas Shared Ministries, a charity, community clinic in Dallas, TX, were identified via query of the electronic health record, and data was extracted from medical charts. Patients were interviewed to obtain family history and assess knowledge of cholesterol levels.ResultsAmong 662 patients with lipid measurements, 27 had LDL-C ≥ 190 mg/dL (67% female, 78% Hispanic, mean age 52 years). Median pretreatment LDL-C was 210.5 mg/dL. Almost all (93%) were prescribed a statin, but only 33% were prescribed high-intensity statin. Treated LDL-C was 141 mg/dL. No patients had an ICD-10 diagnosis of FH, and only 31% had laboratory tests for secondary causes of hypercholesterolemia. Only 15% had any documentation of family history of hypercholesterolemia and/or ASCVD. After interviewing, we discovered previously undocumented family history of hypercholesterolemia in 54% and ASCVD in 31%.ConclusionsAmong the uninsured with LDL-C ≥ 190 mg/dL in a community clinic, the use of high-intensity statins was suboptimal, and most were not evaluated for causes of severe hypercholesterolemia. Future efforts to address gaps and education are warranted to improve care for severe hypercholesterolemia patients in this vulnerable population.Condensed AbstractFew studies evaluate severe hypercholesterolemia in uninsured populations. We queried the electronic health record at a charity community clinic to characterize individuals with severe hypercholesterolemia. Our results suggest more education is needed for patients and providers regarding severe hypercholesterolemia and FH to better care for this vulnerable population.


Introduction
Patients with severe hypercholesterolemia (de ned as LDL-C levels ≥ 190 mg/dL) have a high risk of developing premature atherosclerotic cardiovascular disease (ASCVD) and recurrent coronary events [1][2][3][4] . Clinical trial data have demonstrated that high intensity statins provide the greatest ASCVD risk reduction when compared to moderate-intensity statins or placebo for individuals with severe hypercholesterolemia and/or ASCVD 1,2,5,6 . For this reason, guidelines recommend maximally tolerated statin therapy for patients with severe hypercholesterolemia with a goal of reducing low-density lipoprotein cholesterol (LDL-C) to below 100 mg/dL, and such individuals should be evaluated for underlying causes of severe hypercholesterolemia including hypothyroidism, nephrotic syndrome, and cholestasis 1,5 .
Additionally, a signi cant portion of patients with LDL-C ≥ 190 mg/dl will have Familial Hypercholesterolemia (FH) 7 . FH is an autosomal dominant condition characterized by a substantial elevation of LDL-C 8 . Identifying and treating FH patients remains a top public health priority for effective ASCVD prevention 8, 9 . In fact, the United States Centers for Disease Control and Prevention (CDC) designated FH as one of three Tier 1 genomic conditions requiring early detection and early intervention 9,10 .
While the characteristics of patients with LDL-C ≥ 190 mg/dl has been reported in routine healthcare systems, limited data exist about such individuals in vulnerable populations such as the medically underserved at charity, community clinics 8, 11 . These clinics serve uninsured individuals who are unable to access care even from safety-net healthcare systems 12 . Furthermore, no data exists regarding the burden of FH and associated treatment gaps for this population. Here, we sought to begin addressing these gaps by describing the burden of severe hypercholesterolemia at a primary care, charity, community clinic.

Setting and Participants
The study was conducted at North Dallas Shared Ministries (NDSM) and was approved by the Institutional Review Board at UT Southwestern Medical Center. NDSM is a cooperative effort of 47 congregations that combine resources to deliver assistance to the poorest families and individuals in Dallas, TX. Among its many programs, NDSM includes a charity, community clinic that provides care to uninsured populations. The clinic rst opened in 2000 and grew rapidly to become a full-time primary care clinic, now open 7 AM -7 PM weekdays. As of 2019, over 7000 medically underserved individuals receive routine health care at NDSM annually 13 . These patients are uninsured and are unable to receive routine healthcare from safety-net healthcare systems. The electronic health records (EHR) used at NDSM is Athena Electronic Health Records (Watertown, MA) since 2016 14 .

Data Collection
We queried the EHR at NDSM for patients who met our eligibility criteria since 2016 (when HER was implemented at NDSM). Inclusion criteria included any recorded LDL-C ≥ 190 mg/dL. We did not exclude any individuals among those with LDL-C ≥ 190 mg/dL. The charts were manually reviewed to collect information on demographics, past medical and family medical history, untreated and treated LDL-C levels, comorbidities, and use of lipid lowering drugs. Physical exam ndings were reviewed for documentation of arcus senilis and xanthomas (for FH). We searched charts (both clinic notes and laboratory results) for evidence of secondary causes of severe hypercholesterolemia, including hypothyroidism (Thyroid Stimulating Hormone (TSH) level), cholestasis (bilirubin), and nephrotic syndrome (urine protein). For those without evidence of secondary causes, we sought to establish whether they ful ll FH diagnostic criteria for Simon Broome and/or Dutch Lipid Clinic Network (DLCN) 15,16 (Appendix A and B). Additionally, Make Early Diagnosis to Prevent Early Dealths (MEDPED) criteria was used to determine the FH status of patients using age and lipid levels 17,18 (Appendix C).
All charts that met inclusion criteria were reviewed by JP and BW, with supervision by a University of Texas Southwestern (UTSW) FH clinic physician, ZA. Study data was stored within the Research Electronic Data Capture (REDCap) database and the study was approved by the UTSW Institutional Review Board.

Patient Interviews
Patients who met quali cation criteria were called via telephone for two purposes: we gathered further past medical and family history (not previously recorded EHR) needed to assess whether they ful ll Simon Broome and/or DLCN FH diagnostic criteria; and patients were asked a series of questions about their knowledge of cholesterol and hypercholesterolemia (Appendix D). The interviews were conducted in the patient's native language by uent interviewers. All patients were free to answer or not answer any question that was posed to them.

Statistics
Continuous variables were summarized as means with standard deviation and/or medians with associated interquartile ranges (IQR's). Categorical variables were summarized with percentages. The interviews with patients were conversational and from the information gathered, summary statistics were created.

Study Population
Between 2016 and 2020, LDL-C values were measured for 662 unique patients at NDSM clinic. Twentyseven of these patients met inclusion criteria, having at least one measured LDL-C ≥ 190 mg/dL (Table   1). Population mean age was 52 years with a range of 38-79 years as well as a female predominance (67%). Most patients self-identi ed as Hispanic (74%) and white (63%), which re ects the proportion at the NDSM clinic overall. Sixty-three percent of patients also had hypertension and 30% had also been diagnosed with type 2 diabetes mellitus. Three patients (11%) had hypothyroidism, a possible secondary cause of hypercholesterolemia, but over half (59%) did not have all the laboratory tests needed to exclude secondary causes of severe hypercholesterolemia (TSH, urine protein, and/or bilirubin).

Lipid levels
Of the 27 patients who met inclusion criteria, 26 were untreated LDL-C levels, 21 patients also had treated LDL-C levels available (Figure 1), and six had multiple measurements of LDL-C ≥ 190 mg/dL. The mean time period between pre-and post-treatment lipid measurements was 0.98 years.

Statin prescriptions
Almost all patients (93%) were prescribed a statin medication ( Figure 2). After treatment, 14% of patients achieved LDL-C levels below 100 mg/dL and 10% had LDL-C levels that remained above 190 mg/dL.
Only 33% were prescribed a high intensity statin, de ned as atorvastatin 40-80 mg or rosuvastatin 20-40 mg daily. Most patients that were prescribed atorvastatin were only taking 10-20 mg daily. Other statins prescribed included simvastatin (15%) and lovastatin (26%). Two patients had not been prescribed a statin medication, relying solely on lifestyle modi cation per EHR provider notes.
Family history Documentation about family history was sparse, with 19% having any documentation of hypercholesterolemia in their family and 15% having any documentation of ASCVD in their family.

FH diagnosis
No patients had an ICD-10 diagnosis of FH, and no patients had any mention of FH or FH family screening anywhere within the medical record.
Because of missing family history documentation, determining FH status from the Simone Broome or DLCN criteria, which includes family history, was incomplete. No patients ful lled criteria for "De nite FH" for either Simon Broome or DLCN. One patient met criteria for "possible FH" for the Simone Broome Criteria. All 27 patients met DLCN criteria for "possible FH." Using MEDPED, no patients met criteria for FH diagnosis.

Patient interviews
Of the 27 patients (8 English speakers, 19 Spanish speakers) that met inclusion criteria, we were able to interview 13 (4 English speakers, 9 Spanish speakers) ( Table 2). Over half (54%) of the patients had family history of high cholesterol and 31% had a family history of cardiovascular disease, neither of which were previously documented in the EHR. With this new information, nine (69%) now met criteria for "Possible FH" based on the Simone Broome criteria.
Approximately half (54%) of the patients claimed to have never heard of the terms LDL-C (low density lipoprotein-cholesterol) and HDL-C (high density lipoprotein-cholesterol). The patients that had heard of these terms commonly understood them as "one is bad, and one is good" however being unsure which was which. Almost all (92%) patients could recall being told they had high cholesterol by a health care provider either at NDSM or another clinic. Additionally, 15% of patients recalled hearing the term "Familial Hypercholesterolemia" but no patients could give a detailed description of the disease.
All patients reported knowing the importance of lowering cholesterol levels, citing reasons such as "eliminating heart attack and stroke" and "to live... [not wanting] to die." In addition, 92% cited ways that they knew to lower cholesterol including "exercise, eating properly, less stress" and "staying away from certain [fried] foods", but only one patient stated taking medication as an action that could be taken.
Of the 13 patients contacted, only 4 reported currently adhering to statin therapy. The reasons given by patients for not being on statin therapy ranged from intolerance (headaches, cramps), di culty in obtaining prescription re lls, and patients stopping medication after their cholesterol levels "normalized" (Table 3). Yes, family history of heart disease 4 31% Yes, current or remote tobacco use 5 38% Unfamiliar with the terms "LDL" and "HDL" 7 54% Remember being told that they have high cholesterol 12 92% Currently taking a statin drug 4 31% Familiar with the term "Familial Hypercholesterolemia" 2 15% Claim they know how to lower their cholesterol 12 92% Believe lowering cholesterol is important for their health 13 100% "I have heard of HDL and LDL, but they were not explained in the clinic."

Discussion
In an uninsured population receiving care at a charity community clinic, we identi ed several gaps: only 10% of all patients had had lipid levels measured; few patients with severe hypercholesterolemia were on high intensity statin drugs; and family history was missing in nearly all patients as was laboratory workup for secondary causes of severe hypercholesterolemia. FH was not mentioned in any of their medical records. Subsequently, after calling the severe hypercholesterolemia patients to gather more information, many had previously unknown family histories of premature ASCVD and hypercholesterolemia.
While many studies have looked at FH and severe hypercholesterolemia within large health systems, prior studies have not looked at the burden of disease at a charity, community clinic serving underserved, uninsured populations who are unable to access care even from safety-net healthcare systems 12,19,20 . An estimated 26 million uninsured patients rely on one of 1,400 charity, community clinics across the United States 21 . Such patients suffer from disparities in care, often getting fewer treatments or treatments not adherent to guidelines, and the uninsured have higher morbidity and mortality relative to the general population 21 .
Our ndings at NDSM are consistent -albeit more pronounced -with patient cohorts in various, large health care systems. Within our study, the population mean age of 52 and female predominance (67%) are similar to the demographics of previous studies of patients with severe hypercholesterolemia and possible FH 20,22,23 . At NDSM, only 33% were prescribed a high-intensity statin compared to 42% across patients in the CASCADE-FH registry which was conducted across 11 different US lipid clinics 20 . In addition, the treated LDL-C levels with statin therapy were higher in our study than comparative studies both in the US and abroad. 20,23−25 In our study population, 30% of patients were previously diagnosed with Type II Diabetes Mellitus (T2DM). This is a higher rate than the national age adjusted prevalence of 13.0% and much higher than other cohorts of patients with severe hypercholesterolemia and FH 20,26 29 . We believe this may be due to the higher number of uninsured patient's that are treated at NDSM relative to the 97% insured patient population that was interviewed by Jones et al. Although the reasons for provider under prescribing and patient nonadherence are complex, these anecdotes perhaps give insight into speci c opportunities at charity, community clinics and may be an area of improvement.
Our ndings suggest that there is not enough knowledge about FH and severe hypercholesterolemia in the patient population as well as from providers. Future efforts should address these issues, with patient education on hypercholesterolemia within charity, community clinics as well as provider education about lipid screening and assessing patients with severe hypercholesterolemia.

Limitations
There are several limitations to the study. First, we were only able to obtain LDL-C values from patients at NDSM who had been given a lipid screening. While the 2018 AHA/ACC Cholesterol Guidelines recommends that all patients between the ages of 40-75 should be screened for lipid abnormalities annually, it is likely there are patients at the clinic with elevated LDL-C who have simply not been screened 1,9,30 . Because all patient data was taken from the NDSM Athena EHR, our study is limited to patients who had an untreated lipid screening after NDSM switched to Athena's EHR in 06/2016. Therefore, patients with LDL-C ≥190 mg/dL prior to this time who had been given a statin would likely not be represented in this study. Further, our study has all the limitations that come with limited sample size. A greater understanding of the populations knowledge of cholesterol could be gleamed if we were able to have a larger sample size and speak to a greater proportion of patients.

Conclusions
Among a medically underserved community with LDL-C ≥ 190 mg/dL in a charity community clinic, we found use of high intensity statins was suboptimal and most patients were not evaluated for underlying causes of severe hypercholesterolemia. Future efforts to improve care and education are warranted for both providers and uninsured patients. Mean LDL-cholesterol levels before and after statin therapy.
The bars represent the mean values of LDL-cholesterol among the study population prior to and after treatment with a statin medication. The mean treatment period between untreated and treated values was approximately one year.

Figure 2
Statins prescribed to target patient population at NDSM clinic (LDL-C ≥190 mg/dL) The bars represent the number of patients who were prescribed each form of therapy. Black bars represent therapy that is considered high intensity and grey bars indicate low or moderate intensity therapy. Total patients n = 27.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. Supplementarymaterial.docx