Lipid Testing Trends in the US Before and After the Release of the 2013 Cholesterol Treatment Guidelines

Figures & data

Figure 1 Study flow chart. Inclusion and exclusion criteria were applied to 15,607,733 patients in the MarketScan^® and Medicare Supplemental claims databases to select the four study cohorts.

Table 1 Characteristics of Study Cohort Patients Using All Available Data Leading Up to and Including the Date of the Index Event (Medication Initiation or CVD Hospitalization Event)

Baseline Characteristic (%)	Any Statin Initiation, n = 3,874,048	High-Intensity Statin Initiation, n = 1,259,819	Ezetimibe Initiation, n = 323,510	Very High CVD Risk, n = 41,179
Demographics
Age group
18–39	7.7	4.4	4.2	0.7
40–49	20.6	15.7	14.6	4.3
50–64	54.7	55.9	58.0	30.2
65+	17.1	23.9	23.2	64.8
Sex
Male	51.9	59.4	51.3	58.1
Female	48.1	40.6	48.7	41.9
Region*
Northeast	15.4	14.8	14.1	19.5
North Central	25.0	25.4	26.5	34.5
South	41.9	42.4	44.5	31.9
West	14.7	15.0	12.9	13.3
Type of insurance benefit plan†
Comprehensive	7.3	10.8	9.8	32.5
HMO	13.4	13.9	11.3	13.8
PPO	61.2	58.0	63.7	43.3
Other	14.6	14.6	11.7	8.3
History of ASCVD events
Any ASCVD event	15.3	32.9	25.4	100.0
Most recent time of ASCVD event‡
<1 month	9.2	19.7	10.8	99.1
1–6 months	3.4	7.7	8.8	0.9
7–12 months	1.3	2.7	3.1	0.0
>12 months	1.4	2.8	2.7	0.0
No ASCVD event	84.8	67.1	74.6	0.0
Aneurysm	0.8	1.5	1.1	8.0
Carotid/vertebral/basilar stenosis	2.0	4.4	3.4	38.4
Carotid endarterectomy	0.2	0.5	0.4	5.8
Cerebrovascular disease	3.9	7.4	5.3	66.0
Coronary atherosclerosis/angina/old MI	9.6	25.2	20.4	78.2
CABG/PCI	4.0	11.5	6.5	41.5
Carotid/vertebral/basilar stenting	0.0	0.1	0.1	1.8
Endovascular stent graft	0.0	0.1	0.1	0.6
Ischemic stroke	2.2	3.7	1.7	62.1
Myocardial infarction	3.8	10.0	4.9	71.0
Peripheral vascular disease	2.4	5.1	4.1	32.3
Peripheral artery disease	0.4	1.0	0.8	7.4
Transient ischemic attack	1.6	2.5	1.5	32.5
Unstable angina	2.1	5.5	3.5	30.7
Other comorbidities and medication use
End-stage renal disease or hemodialysis	0.3	0.5	0.3	6.4
Chronic kidney disease, all stages	4.4	8.0	6.1	46.9
Heart failure	2.2	4.4	3.4	38.7
Diabetes	20.4	30.6	27.0	52.9
Cancer§	5.0	6.3	5.7	17.6
Cognitive impairment	2.0	2.8	1.8	36.5
Hypertension	44.3	57.4	52.3	91.4
Dyslipidemia/hyperlipidemia	13.9	23.2	25.0	34.3
Muscle events (rhabdomyolysis/myositis)	2.9	3.3	4.6	7.0
Obesity	2.2	3.1	2.1	7.1
COPD or asthma	6.9	9.0	8.0	26.9
Hepatic disorders	0.8	1.0	1.1	4.5
Frailty	25.2	31.3	29.5	76.3
Concomitant lipid lowering medication	8.6	15.3	18.9	18.2
Prior LDL-C testing	78.9	79.6	80.0	63.8

Notes: *Percentages do not sum to 100 due to missing data (≤3% for all cohorts); ^†Percentages do not sum to 100 due to missing data (≤4% for all cohorts). Other insurance types were exclusive provider organization (EPO), point-of-service plan (POS), POS with capitation, consumer-driven health plan (CDHP), and high-deductible health plan (HDHP); ^‡The very high CVD risk group qualified for the study cohort at the time of discharge from the most recent MI or IS hospitalization. Therefore, by definition, the most recent ASCVD event for all patients in this cohort occurred a short period of time before their hospitalization discharge. In rare instances (<1%), the most recent MI or IS event occurred over one month prior to study entry at discharge, suggesting a long hospitalization; ^§The cancer variable excluded non-melanoma skin cancer.

Abbreviations: HMO, health maintenance organization; PPO, preferred provider organization; CABG/PCI, coronary artery bypass grafting or percutaneous coronary intervention; COPD, chronic obstructive pulmonary disease.

Figure 2 LDL-C testing rates (with 95% CI) per 1,000 persons per calendar year quarter from 2008 to 2016, by study cohort. The unadjusted observed rates are plotted in black; the time-series model fit to the pre-guideline data is plotted in blue; and the model predictions extrapolated to the post-guideline period are plotted in red. The dashed line corresponds to the date of the guideline change, and the quarter during which the guideline change occurred (10/01/2013–12/31/2013) was treated as the last pre-guideline quarter.

Figure 3 Differences between observed rates and model-predicted rates of LDL-C tests per 1,000 persons per calendar year quarter, by study cohort. Differences were taken by subtracting the model-predicted rate from the observed rate for each calendar year quarter following the guideline change.

Figure 4 Rate ratios (with 95% CI) for LDL-C testing associated with predictor variables. Predictors included demographics, history of ASCVD events, comorbidities and medication use. Their association with LDL-C testing (following index event) was estimated separately by study cohort and was not computed if prevalence of the predictor was <1%.