Feasibility of Lipid Screening During First Trimester of Pregnancy to Identify Women at Risk of Severe Dyslipidemia

Abstract

Background

Dyslipidemia is an important risk factor for atherosclerotic cardiovascular disease, especially when disease presents at a young age. Despite national screening guidelines to perform a lipid profile test in children and young adults, many reproductive‐age women have not undergone lipid screening. Our objective was to assess the feasibility of lipid screening during the first trimester of pregnancy as a strategy to increase lipid screening rates among women receiving prenatal care.

Methods and Results

A nonfasting lipid panel was incorporated into routine prenatal care among obstetricians at a single academic clinic. Educational materials and a clinical referral pathway were developed for patients with abnormal results. Over 6 months, 445 patients had a first prenatal care visit. Of the 358 patients who completed laboratory testing, 236 (66%) patients completed lipid testing. Overall, 59 (25%) patients had abnormal results. One patient with previously undiagnosed suspected familial hypercholesterolemia was identified. Barriers to ordering lipid tests included the burden of reviewing additional laboratory results and uncertainty about patient counseling.

Conclusions

Implementation of nonfasting lipid screening as part of routine prenatal care during the first trimester is feasible and may play a crucial role in timely diagnosis and management of lipid disorders in women of reproductive age. Future work should focus on optimizing health system workflow to minimize burden on clinical staff and facilitate follow‐up with appropriate specialists.

Nonstandard Abbreviations and Acronyms

EMR

electronic medical record

OB

obstetrician/gynecologist

Clinical Perspective.

What Is New?

• Lipid screening in the first trimester of pregnancy is a feasible approach to improve lipid screening rates of reproductive‐age women receiving prenatal care and to diagnose lipid disorders.

What Are the Clinical Implications?

• To improve uptake of lipid screening as part of routine first trimester prenatal laboratory exams, workflow patterns must be optimized within the health system to improve result management and timely patient referrals.

Atherosclerotic cardiovascular disease is the leading cause of morbidity and mortality among women in the United States. ¹ Dyslipidemia is a leading population‐adjusted risk factor among women, which begins early in life and is accentuated with aging. ² Early diagnosis of dyslipidemia can facilitate prompt treatment with lifestyle changes and, in some cases, medication to mitigate atherosclerotic cardiovascular disease events later in life. Despite national guidelines recommending screening of lipid disorders in childhood and early adulthood, less than half of reproductive‐age women have had lipid screening. ³ National guidelines recommend early initiation of pharmacotherapy for individuals with familial hypercholesterolemia, which is relatively common (1:250), as well as other less common lipid disorders, since these disorders may be identified by lipid screening before they cause individual harm to the patient.

Routine health screening for conditions that impact maternal and fetal health, such as testing for human immunodeficiency virus, viral hepatitis, and syphilis, is a standard component of prenatal care. ⁴ As many as 86% of women will have experienced at least 1 pregnancy by age 45 years and 44% of women receive their primary care from their obstetrician/gynecologist (OB). ⁵ As a result, pregnancy may provide a window of opportunity to improve screening for atherosclerotic cardiovascular disease risk factors that will impact long‐term maternal health.

Physiological changes during pregnancy impact lipid metabolism to aid in normal fetal growth and development. During pregnancy, plasma low‐density lipoprotein cholesterol (LDL‐C) levels increase by 25% to 50% and triglyceride levels can increase by 150% to 300%, with the greatest increases seen in the third trimester. ⁶ , ⁷ , ⁸ LDL‐C levels increase by <5% in the first trimester. ⁶ Since the majority of women initiate prenatal care beginning in the first trimester of pregnancy before significant changes in lipid metabolism have occurred, first trimester lipid screening approximates prepregnancy values.

We have previously demonstrated that the acceptability of lipid screening as part of routine prenatal care was high among pregnant patients. ⁹ Incorporating lipid screening into the routine assessment of first trimester prenatal laboratory exams would require changes to clinician workflow during pregnancy and potentially after delivery for those individuals with newly diagnosed dyslipidemia. Therefore, the objective of this study is to determine the feasibility and acceptability of lipid screening during the first trimester of pregnancy among OB clinicians providing prenatal care. Our hypothesis is that acceptability and feasibility of lipid screening will be high, however, we will identify barriers to implementation at the clinician and health system levels.

METHODS

Study Design and Participants

This observational, prospective pilot study sought to assess the feasibility and acceptability of lipid screening among pregnant women receiving prenatal care from an urban academic obstetrics clinic within the University of Pennsylvania Health System between April 5, 2021 and October 5, 2021. This study was approved by the Institutional Review Board of the University of Pennsylvania as a quality improvement study and patient informed consent was not required. The data that support the findings of this study are available from the corresponding author upon reasonable request.

All OB physicians within the clinic (n=7) were asked to order a nonfasting lipid panel for patients seen for an initial prenatal visit in their first trimester (<14 weeks). Study team members met with OB physicians to explain the rationale for ordering the lipid panel and discuss ways to optimize workflow and return of results to patients. In response to these discussions, we operationalized several recommendations. First, the lipid panel order was incorporated into the first trimester laboratory order set that the provider panel uses in the electronic medical record (EMR). Second, we recommended ordering a lipid panel on all patients, regardless of prior lipid screening, to minimize the additional work of searching for prior laboratory results from outside the health system. Third, we created a clinical referral pathway for patients with abnormal results (Figure 1). We created EMR “smart phrases” that could be used to notify patients of abnormal results and referrals as appropriate. All physicians and clinical support staff were encouraged to contact study staff with questions about abnormal test results or difficulty referring patients according to the algorithm.

Figure 1. Algorithm for clinical referral for abnormal results.

LDL‐C indicates low‐density lipoprotein cholesterol.

After the pilot testing period was complete, the EMR was queried to identify all new prenatal care patients seen in the practice in the 6‐month window. Laboratory results and patient characteristics were abstracted and included: age, self‐reported race, self‐reported sex, height, weight at first prenatal visit, and medical comorbidities including chronic hypertension and diabetes. Prepregnancy body mass index was calculated using recorded height and weight at the initial prenatal visit.

After data collection was complete, we performed interviews with the participating obstetricians by phone or in‐person to elicit feedback about the acceptability of lipid screening in early pregnancy and to identify facilitators and barriers to incorporating lipid screening into routine prenatal care.

Outcomes

The primary implementation outcomes were feasibility and acceptability. ¹⁰ Feasibility, or the extent to which lipid screening can be successfully performed in the context of prenatal care, was defined as the proportion of prenatal patients who completed lipid testing among all patients who completed first trimester laboratory exams. Acceptability, or the perception among OB clinicians that lipid screening as part of prenatal care is agreeable or palatable, was assessed using interviews with participating clinicians. The secondary outcomes were the prevalence of dyslipidemia, as defined in our algorithm, and based on the 2018 American Heart Association/American College of Cardiology multisociety cholesterol guidelines, which defines dyslipidemia in nonpregnant adults as LDL‐C >130 mg/dL and/or triglycerides >150 mg/dL. ¹ There is no distinct definition of pregnancy‐associated dyslipidemia different from the nonpregnancy definition.

Statistical Analysis

We present descriptive characteristics of patients who did and did not complete lipid testing and summary statistics of lipid values. We analyzed continuous and categorical descriptive data using 2‐tailed t‐tests and Fisher exact tests, respectively. A P value <0.05 was considered statistically significant. Analyses were conducted with SAS, version 9.4 (SAS Inc., Cary, NC).

RESULTS

A total of 445 pregnant women were seen for a first prenatal care visit in the first trimester during the study period and 236 (53%) completed a lipid test. Among the 445 patients, 358 (80.4%) had any laboratory results documented in the EMR after the initial prenatal visit. Differences in baseline characteristics between women who did and did not have first trimester laboratory exams are presented in Table S1. Baseline characteristics of patients with first trimester prenatal laboratory exams are presented in Table 1. The mean age was 34.5 years and all patients had private health insurance. The majority of patients self‐identified as White (73.4%). The mean body mass index was 26.0 kg/m² and 17.3% of patients had a body mass index ≥30 kg/m². Few patients had chronic hypertension (4.1%) and pregestational diabetes (0.8%). Of these patients, 236 (65.9%) patients completed lipid testing. Patient characteristics were not significantly different between those who did (n=122) and did not (n=87) complete lipid testing, as demonstrated in Table 1.

Table 1.

Demographics and Clinical Characteristics of Patients With Laboratory Results After New OB Visit According to Status of Lipid Panel Completed

Characteristics	Total cohort	Lipid panel not completed	Lipid panel completed	P value
	N=358	n=122	n=236
Age, y, mean (SD)	34.5 (4.4)	34.5 (4.4)	34.5 (4.3)	0.916
Age categories
18–34 y	177 (49.4%)	65 (53.2%)	112 (47.4%)	0.317
≥35 y	181 (50.5%)	57 (46.7%)	124 (52.5%)
Race or ethnicity
NH White	263 (73.4%)	87 (72.1%)	176 (75.8%)	0.529
NH Black	38 (10.6%)	14 (12.2%)	24 (10.1%)	0.719
Asian/East Indian	35 (9.7%)	15 (12.2%)	20 (8.4%)	0.264
Hispanic	9 (2.5%)	4 (3.2%)	5 (2.1%)	0.497
Other	13 (3.6%)	2 (1.6%)	11 (4.6%)	0.233
BMI, kg/m2, mean (SD)	26.0 (5.9)	25.7 (5.4)	26.3 (6.1)	0.269
BMI ≥30 kg/m2	62 (17.3%)	17 (13.9%)	45 (19.0%)	0.242
Hypertension	15 (4.1%)	2 (1.6%)	12 (5.0%)	0.152
Diabetes	3 (0.8%)	1 (0.8%)	2 (0.8%)	1.0

Data presented as n (%) unless otherwise indicated.

BMI indicates body mass index; NH, Non‐Hispanic; and OB, obstetrician/gynecologist.

Among the 236 patients who completed lipid testing, 59 (25%) had abnormal results, none of whom had severe abnormalities requiring urgent referral (Table 2 and Figure 2). Overall, 25 (10.6%) patients had triglyceride levels >150 mg/dL, 24 (10.2%) patients had LDL‐C >130 mg/dL, and 10 (4.2%) patients had both elevated triglyceride and LDL‐C levels. All patients with abnormal results were counseled to follow‐up with their primary care doctor postpartum. The mean value of lipid levels are presented in Table S2. One patient had an LDL‐C of 265 mg/dL with a family history of premature coronary artery disease. This patient was referred to a preventive cardiologist postpartum for close follow‐up and management.

Table 2.

Prevalence of Dyslipidemia Identified During First Trimester Screening

Dyslipidemia category	n (%)
Patients meeting any dyslipidemia criteria	59 (25.0%)
LDL‐C ≥130 mg/dL	24 (10.2%)
LDL‐C ≥160 mg/dL	6 (2.5%)
LDL‐C ≥190 mg/dL	1 (0.4%)
Elevated triglycerides
Triglycerides ≥150 mg/dL	25 (10.5%)
Mixed hyperlipidemia
LDL‐C >130 mg/dL and triglycerides ≥150 mg/dL	10 (4.2%)

Data presented as n (%). LDL‐C indicates low‐density lipoprotein cholesterol.

Figure 2. Feasibility and clinical findings of first trimester lipid screening among pregnant patients.

FH indicates familial hypercholesterolemia.

Of the 7 participating physicians, lipid testing was ordered 100% of the time for 2 physicians, most of the time (88%–95%) for 3 physicians, and majority of the time for 2 physicians (66%–69%). Physician interviews identified that the most common reasons for not ordering lipid screening included feeling overwhelmed with reviewing laboratory results, especially in the context of increased demands introduced by the COVID‐19 pandemic, uncertainty about how to counsel patients with abnormal results, and seeing that patients had prepregnancy lipid results. Overall, physicians reported that the workflow was acceptable but preferred to have improved support for handling abnormal results.

DISCUSSION

In our cohort of pregnant women receiving prenatal care as part of an urban academic health system, 65.9% of patients completed lipid screening in early pregnancy as part of our pilot intervention. Patient characteristics of those who did and did not complete lipid testing were similar.

Overall, we found that 26% of patients had evidence of dyslipidemia and only 1 patient out of 236 was identified to have severe lipid abnormalities (probable familial hypercholesterolemia), which is consistent with population‐based expectations. ¹¹ Since up to 90% of patients with familial hypercholesterolemia remain undiagnosed and early treatment can reduce the risk of premature atherosclerosis, screening patients during routine prenatal care has the potential to improve diagnosis rates for patients and their families. ¹² Earlier detection of nonsevere dyslipidemia may motivate early adoption of lifestyle changes to prevent premature subclinical and clinical atherosclerosis, especially among individuals with adverse pregnancy outcomes, as recommended by the American Heart Association and the American College of Obstetricians and Gynecologists. ¹³ Finally, increasing evidence suggests that earlier treatment may be warranted in adults with high lifetime risk of atherosclerotic cardiovascular disease, an area of active investigation. ¹⁴

Overall acceptance of lipid screening by obstetricians was high, though we identified some barriers to the feasibility of implementing this program more broadly, as demonstrated by only 65.9% of patients completing lipid testing. Specifically, the 2 physicians with the highest patient volume had the lowest rates of ordering lipid screening, highlighting the need for improved results management workflow. Optimizing and systematizing lipid testing and follow‐up of abnormal results will facilitate broader acceptability and adoption. In this pilot study, individual physicians and their nurses followed up results using current practices for other laboratory results. Automated EMR‐triggered workflows could be used to identify individuals at risk based upon abnormal lipid results. The responsibility for addressing these abnormalities can be transferred to staff dedicated to address the issue within health systems or individual care sites, such as a nurse or advanced level practitioner within a primary care or preventive cardiology clinic. Further education with obstetric providers about rationale for testing, approaches to treatment and cardiovascular risk reduction could also help improve comfort level with reviewing and discussing laboratory results with patients.

While hyperlipidemia is commonly seen in the second half of pregnancy as a physiologically acquired mechanism, it is not fully understood. Hormonal changes greatly contribute to lipid abnormalities. Since changes in lipid levels in the first trimester are small in comparison, first trimester laboratory exams are a reasonable estimate of prepregnancy lipid levels. A small risk of overestimation of hyperlipidemia may occur as a result of lipid testing during pregnancy. The potential harm from this approach (worry about an abnormal lipid profile that is not being actively managed during pregnancy) is likely outweighed by the potential benefit of detecting familial hypercholesterolemia, identifying individuals who may require triglyceride monitoring during pregnancy, or identifying those who are predisposed to long‐term metabolic conditions associated with hyperlipidemia, such as insulin resistance or diabetes.

Our findings should be interpreted in light of several limitations. First, our study is an observational pilot study to assess feasibility and may not broadly represent the challenges to implementing prenatal lipid screening in a larger and more diverse population of clinicians and patients. Second, we are unable to confirm how abnormal results were managed after delivery, since many patients in our cohort have primary care doctors outside of our health system and may not follow‐up with primary care doctors until later in the postpartum period. Third, 19.6% of patients did not have any first trimester laboratory values documented, regardless of lipid testing. This may have been a result of patients not completing recommended laboratory testing or completed laboratory testing not being scanned into the EMR. Furthermore, we did not evaluate acceptability of lipid testing among patients in this cohort, though we have previously demonstrated high acceptability of this approach in a similar population of prenatal women. ⁶ Finally, our patient population was privately insured which limits generalizability.

CONCLUSIONS

Incorporating nonfasting lipid screening into routine prenatal care as part of first trimester laboratory exams is feasible and may improve the diagnosis of lipid disorders in women receiving prenatal care. Future work should expand upon the best strategies to implement lipid screening into the prenatal care workflow of obstetric providers in diverse settings, improve the communication of results to patients, and aid in transitions of care in the postpartum period to promote cardiovascular health across the life course.

Sources of Funding

This work was supported by Tom and Debbie Whitesell.

Disclosures

Dr Soffer has received grants to the University of Pennsylvania from Akcea, Amgen, Amryt, AstraZeneca, Ionis, Novartis, Regeneron, RegenXBio, Verve and served as consultant for Ionis, Akcea, Novartis, Amgen. The remaining authors have no disclosures to report.

Supporting information

Tables S1–S2