Abstract
Background
Statin use is proven to be effective in lowering low-density lipoprotein cholesterol (LDL-C) levels and reducing risk of recurrent myocardial infarction, stroke, and mortality in individuals with established cardiovascular disease (CVD). We used medication dispensed data (e.g., SureScripts), which has been integrated with the electronic health record (EHR) to examine the factors and outcomes associated with adherence to statins.
Methods
This study is a secondary data analysis using longitudinal data between 1/1/2010–10/31/2021 (n = 1486,286 over nearly 12 years) from a large community-based health system on all primary care patients aged 35 years or older when diagnosed with the first CVD two years after their first primary care visit and had new statin prescriptions on or after CVD diagnosis. Multivariable logistic regression models were used to identify the factors associated with filling the prescription and statin adherence, respectively. Survival analysis was used to assess the association between statin adherence and LDL-C control.
Results
Of the 5155 patients with newly prescribed statins, a total of 3553 (68.9 %) were adherent, with insurance type, online patient portal use, race, age, statin intensity, and cardiologist visits emerging as significant predictors. Specifically, patients with PPO/FFS were less likely to fill statin prescriptions compared to those with HMO. Infrequent online patient portal use is associated with lower adherence. There is a disparity between patients race categories (Non-Hispanic Black (NHB) vs. Non-Hispanic White (NHW)) in filling the prescription and adhering to the filled prescription. Medication adherence is defined as proportion of days covered (PDC) of 80 % or greater. Adherence was positively associated with older age, high-intensity statins, and cardiologist visits. Having a visit with a cardiologist showed better adherence to the prescription and lowering of LDL values. Additionally, adhering to statins has shown a better outcome of lowering patients LDL values.
Conclusions
The findings emphasize demographic and healthcare factors in medication adherence and LDL control, suggesting tailored interventions for diverse populations, addressing disparities in insurance type, race, and online portal use, and involving cardiologists in medication management for improved medication adherence and clinical outcomes.
Keywords: Cardiovascular disease, Medication adherence, Statins
Central Illustration.
1. Introduction
Cardiovascular disease (CVD) continues to be one of the leading causes of deaths worldwide [1]. Statin use is proven to be effective in lowering low-density lipoprotein cholesterol (LDL-C) levels, reducing risk of recurrent myocardial infarction, stroke, and mortality in individuals with established CVD, and is recommended as part of a comprehensive secondary prevention strategy [[2], [3], [4], [5], [6]]. However, these benefits rely on a high level of medication adherence by patients [2]. The discrepancy in the clinical outcomes between care delivered and expected (or guideline-based) is partially explained by poor adherence to statins prescribed [7]. Poor adherence to statins has been associated with poor LDL-C control for individuals with CVD, leading to a higher risk of micro- or macro-cardiovascular events, hospitalization, and mortality [[7], [8], [9]]. Therefore, there has been a growing need for healthcare providers to incorporate adherence to statins in managing CVD treatment goals and outcomes. Healthcare providers, however, are usually not able to access concrete data to assess medication adherence in the electronic health record (EHR) during the office encounter [10,11]. In order to close this gap, we used medication dispensed data (e.g., SureScripts), which has been integrated with the EHR and readily available at the time of the office encounter, to examine the factors and outcomes associated with adherence to statins.
Although previous studies have investigated non-adherence to statins and links to patient social, economic, and psychological factors, as well as drug effects and adverse event profiles that limit its use, [[12], [13], [14]] they have largely focused on demographic and clinical factors, such as age, gender, and comorbidities, as predictors of non-adherence. There is a need for more research on modifiable factors that may influence adherence to statins, such as access to health care services, insurance, and characteristics of statins (e.g., intensity level). Moreover, many studies have been conducted in specific populations, such as veterans or Medicare beneficiaries, which may limit the generalizability of the findings. Research in diverse populations is essential to determine if the factors that influence adherence to statins are consistent across different patient groups and provide findings that are more generalizable to the entire population. Therefore, to fill these important gaps, the purpose of the study is two-fold: (1) to examine the factors associated with statin adherence in individuals with CVD at a large, community-based health system; and (2) to examine the impact of adherence on LDL-C control.
2. Methods
2.1. Study setting
The study was conducted at Sutter Health, a large, community-based health system serving 22 California counties (both urban and rural), including 10 counties in the San Francisco Bay area. Sutter Health provides comprehensive medical services across >100 ambulatory clinics and 24 acute care hospitals, serving to approximately 3.5 million individuals annually. In 2020, 45.6 % of Sutter patients were White, 15.6 % were Hispanic, 16.5 % were Asian, 4.7 % were Black/African American, and 17.4 % were other.
2.2. Study population
This study includes all primary care patients aged 35 years or older when diagnosed with CVD (see Appendix A for a list of ICD-9/ICD-10 codes for CVD) and had new statin prescriptions on or after CVD diagnosis between 1/1/2010–10/31/2021 (n = 1486,286 over nearly 12 years). A new prescription was defined as the patient having no prior fills for any statin before the date the prescription was ordered (index date/ INDEX_STATIN). Renewed prescriptions, transfers from outside pharmacies, verbal orders, prescriptions printed out as a hard-copy prescription in the doctor’s office, prescriptions from outside providers, and prescriptions cancelled by the prescriber before it was picked up by the patient were excluded. Statin prescriptions for female patients who are pregnant (based on gestation date) during 9 months before the index date and 12months after the index date were excluded.
Fig. 1 shows a flowchart of cohort selection. Of those with at least one internal medicine or family medicine office visit (n = 1486,286 over nearly 12 years), 153,034 patients were diagnosed with CVD and at least 35 years old at the time of the first CVD diagnosis. 69,576 patients had the first CVD (FIRST_CVD) diagnosis at least two years after their first PCP visit date. In this study, we focus on the patients with a new statin prescription after the CVD diagnosis in or after 2015 (N = 7572). We defined the INDEX_STATIN as the time of the first statin prescription and excluded those whose INDEX_STATIN was earlier than 2015 or before their CVD diagnosis (N = 62,004). Of these patients whose post-CVD new statins prescriptions were in or after 2015, 5720 patients had at least one-year of healthcare utilization (office or video visits) on or after the INDEX_STATIN. A total of 5460 patients had statin medication dispense data on or after the index date from dispense sources including Surescripts, and medication dispense claims data. Patients who switched in statin strength within 6 months (n = 269) of index date or were pregnant during 9 months prior to the index date and 12 months after (n = 36) were excluded from the study. As a result, a total of 5155 eligible patients were included in the analysis.
Fig. 1.
Flowchart diagram of eligible study patients.
2.3. Measurement
Surescripts, the source of medication dispense data in this study, is an electronic prescribing network that connects healthcare providers, pharmacies, and pharmacy benefit managers (PBMs) in the United States [11,15,16]. Surescripts data are available (within 24 h before a scheduled encounter) in the EHR.
Adherence to statins (primary outcome) was measured using the Proportion of Days Covered (PDC). PDC was calculated as the number of days with statin medication supply from SureScripts data divided by the total number of days in the observation period. The observation period starts from the index date (i.e., the date of the first statin prescription fill) and ends with 12 months after the index date. Patients were classified into three mutually exclusive groups based on their PDC: 1) primary non-adherent group includes patients who fail to fill their new statin prescription within 180 days of the index date; 2) secondary non-adherent group includes patients who filled their new statin prescription within 180 days of the index date, but the PDC was <80 %; 3) adherent group includes patients with PDC of 80 % or greater.
LDL-C control (secondary outcome) was defined as the post-index LDL-C below 70mg/dL. Baseline LDL-C was defined as the most recent LDL-C laboratory value during the 6-month period prior to the index date. Patients who had no LDL-C measured during the follow-up period were censored at 12-months.
Patient characteristics include age, gender, race/ethnicity, insurance, smoking status, body mass index (BMI), comorbidities, baseline statin intensity, non-statin lipid lowering therapy prescriptions, and access to health care. Non-statin lipid lowering therapy prescriptions include Proprotein Convertase Subtilisin/Kexin Type 9 Inhibitors, Bempedoic acid, Zetia, and Intestinal Cholesterol Absorption Inhibitors on or after CVD diagnosis. Access to health care was operationalized as having various types of visits (cardiology visit, emergency department (ED) visit, inpatient visit, patient portal use, primary care physician (PCP) visit, and telephone) during 12 months before index date (Yes/No for participant characteristics or median number of visits for data modeling). These characteristics were analyzed to explain the difference in primary and secondary outcomes.
2.4. Statistical analysis
Variables were presented as frequencies and percentages and were compared among the three statin intensity groups (low, moderate, high) as well as the three patient adherence groups (primary non-adherent, secondary non-adherent, adherent) using a Chi-square test.
Multivariable logistic regression models were used to identify the factors associated with 1) filling the prescription (i.e., primary non-adherent group vs. others); and 2) statin adherence (i.e., secondary non-adherent group vs. adherent group), respectively.
Survival analysis was used to assess the association between statin adherence and LDL-C control. We first plotted Kaplan-Meier curves to compare LDL-C control among different levels of statin adherence. We then utilized a Cox proportional hazard model to evaluate the relationship between statin adherence and LDL-C control., controlling for other baseline characteristics (e.g., age, gender, race/ethnicity, BMI, smoking status, insurance, comorbidities, baseline LDL, baseline statin intensity). Patients who had a baseline LDL-C but no subsequent LDL-C measurements or who did not achieve LDL-C control within 12 months were censored at 12 months. All statistical analyses were performed using SAS 9.4 at a significance level of 0.05, and all statistical tests were two-sided. This work was reviewed and approved by the Sutter Health Institutional Review Board and granted a waiver of Health Insurance Portability and Accountability Act Authorization and a waiver of informed consent as a data-only study.
3. Results
A total of 5155 patients were identified as eligible and included in the study (Fig. 1 and Table 1). The majority were male (53.2 %) and 65 years or older (61.1 %). Non-Hispanic Whites made up 70 % (3607) of the cohort; Hispanics, 9.9 % (511); non-Hispanic Asian, 8.9 % (461); non-Hispanic Blacks, 5.2 % (266) and other, 6 % (310). Over two-thirds (69.2 %) of these patients were overweight or obese. Nearly half of patients (46.1 %) had Medicare Fee for Service (FFS) insurance, and 25.7 % and 16.7 % had preferred provider organization (PPO)/FFS and Medicare health maintenance organization (HMO), respectively, and only 2.2 % were covered by Medicaid/Medi-Cal insurance. Within our sample, 9.4 % and 38.3 % were current and former smokers, respectively. Smoking status data were missing in 4.1 % of the sample (n = 209).
Table 1.
Patient characteristics by statin intensity.
| Statin Intensity |
All (N = 5155) |
P-value | ||||
|---|---|---|---|---|---|---|
| High (N = 2404) | Moderate (N = 2446) | Low (N = 305) | ||||
| N ( %) | N ( %) | N ( %) | N ( %) | |||
| Age group | 0.0001 | |||||
| 35–44 | 103 (4.3) | 68 (2.8) | 10 (3.3) | 181 (3.5) | ||
| 45–54 | 310 (12.9) | 258 (10.5) | 28 (9.2) | 596 (11.6) | ||
| 55–64 | 608 (25.3) | 556 (22.7) | 63 (20.7) | 1227 (23.8) | ||
| 65–74 | 728 (30.3) | 785 (32.1) | 108 (35.4) | 1621 (31.4) | ||
| ≥75 | 655 (27.2) | 779 (31.8) | 96 (31.5) | 1530 (29.7) | ||
| Female | 942 (39.2) | 1286 (52.6) | 182 (59.7) | 2410 (46.8) | <0.0001 | |
| Race/Ethnicity | 0.763 | |||||
| Hispanic | 237 (9.9) | 237 (9.7) | 37 (12.1) | 511 (9.9) | ||
| NH Asian | 221 (9.2) | 217 (8.9) | 23 (7.5) | 461 (8.9) | ||
| NH Black | 129 (5.4) | 119 (4.9) | 18 (5.9) | 266 (5.2) | ||
| NH Other | 144 (6.0) | 148 (6.1) | 18 (5.9) | 310 (6.0) | ||
| NH White | 1673 (69.6) | 1725 (70.5) | 209 (68.5) | 3607 (70.0) | ||
| BMI | <0.0001 | |||||
| Unknown | 164 (6.8) | 50 (2.0) | 5 (1.6) | 219 (4.2) | ||
| normal | 571 (23.8) | 707 (28.9) | 90 (29.5) | 1368 (26.5) | ||
| overweight | 879 (36.6) | 876 (35.8) | 110 (36.1) | 1865 (36.2) | ||
| obese | 790 (32.9) | 813 (33.2) | 100 (32.8) | 1703 (33.0) | ||
| Smoking status | <0.0001 | |||||
| Unknown | 156 (6.5) | 46 (1.9) | 7 (2.3) | 209 (4.1) | ||
| Non-smoker | 1114 (46.3) | 1229 (50.2) | 144 (47.2) | 2487 (48.2) | ||
| Former smoker | 881 (36.6) | 961 (39.3) | 130 (42.6) | 1972 (38.3) | ||
| Current smoker | 253 (10.5) | 210 (8.6) | 24 (7.9) | 487 (9.4) | ||
| Primary Insurance | <0.0001 | |||||
| Medicare HMO | 389 (16.2) | 401 (16.4) | 71 (23.3) | 861 (16.7) | ||
| Medicare FFS | 997 (41.5) | 1223 (50.0) | 155 (50.8) | 2375 (46.1) | ||
| Medicaid/Medi-Cal | 63 (2.6) | * | * | 111 (2.2) | ||
| HMO | 261 (10.9) | 186 (7.6) | 20 (6.6) | 467 (9.1) | ||
| PPO/FFS | 687 (28.6) | 583 (23.8) | 56 (18.4) | 1326 (25.7) | ||
| Self or unknown | 7 (0.3) | * | * | 15 (0.3) | ||
| Cardiology visit (Yes) | 1093 (45.5) | 1049 (42.9) | 121 (39.7) | 2263 (43.9) | 0.06 | |
| ED Visit (Yes) | 1151 (47.9) | 753 (30.8) | 82 (26.9) | 1986 (38.5) | <0.0001 | |
| Inpatient visit (Yes) | 1226 (51.0) | 635 (26.0) | 67 (22.0) | 1928 (37.4) | <0.0001 | |
| Patient portal use visit (Yes) | 1379 (57.4) | 1549 (63.3) | 183 (60.0) | 3111 (60.3) | 0.0001 | |
| PCP visit (Yes) | 2175 (90.5) | 2356 (96.3) | 293 (96.1) | 4824 (93.6) | <0.0001 | |
| Telephone visit (Yes) | 2239 (93.1) | 2342 (95.7) | 292 (95.7) | 4873 (94.5) | 0.0002 | |
| CHD (Yes) | 1272 (52.9) | 928 (37.9) | 107 (35.1) | 2307 (44.8) | <0.0001 | |
| PVD (Yes) | 281 (11.7) | 666 (27.2) | 82 (26.9) | 1029 (20.0) | <0.0001 | |
| HTN (Yes) | 1423 (59.2) | 1536 (62.8) | 197 (64.6) | 3156 (61.2) | 0.0167 | |
| Stroke (Yes) | 851 (35.4) | 852 (34.8) | 116 (38.0) | 1819 (35.3) | 0.5375 | |
| T2D (Yes) | 411 (17.1) | 487 (19.9) | 59 (19.3) | 957 (18.6) | 0.0392 | |
| Medication adherence | <0.0001 | |||||
| Adherence | 1764 (73.4 %) | 188 (61.6 %) | 1601 (65.5 %) | 3553 (68.9 %) | ||
| Secondary non-adherence | 502 (20.9 %) | 92 (30.2 %) | 671 (27.4 %) | 1265 (24.5 %) | ||
| Primary non-adherence | 138 (5.7 %) | 25 (8.2 %) | 174 (7.1 %) | 337 (6.5 %) | ||
| LDL level 6months pre index_statin | <0.0001 | |||||
| Unknown | 1503 (62.5 %) | 1172 (47.9 %) | 131 (43.0 %) | 2806 (54.4 %) | ||
| Uncontrolled | 713 (29.7 %) | 1090 (44.6 %) | 147 (48.2 %) | 1950 (37.8 %) | ||
| Controlled | 188 (7.8 %) | 184 (7.5 %) | 27 (8.9 %) | 399 (7.7 %) | ||
| Non-statin lipid lowering therapy prescriptions | 0.0001 | |||||
| No | 2279 (94.8 %) | 2377 (97.2 %) | 292 (95.7 %) | 4948 (96.0 %) | ||
| Yes | 125 (5.2 %) | 69 (2.8 %) | 13 (4.3 %) | 207 (4.0 %) | ||
Masked for privacy/security
Variables were compared among the three statin intensity groups (low, moderate, high) using a Chi-square test.
3.1. Adherence to statins
Out of 5155 patients, 3553 (68.9 %) patients were adherent (PDC≥80 %), while 337 (6.5 %) did not fill the prescription and 1265 (24.5 %) were not adherent to statins (PDC<80 %) (Table 2). Nearly three-quarters of patients aged 75 years or older (72.5 %) were adherent, while only 56.4 % and 63.1 % of patients aged 35–44 years old and 45–54 years old were adherent, respectively (P < 0.0001). Over 70 % of male vs. 67.1 % of female were adherent (P < 0.05). Adherence varied greatly among racial ethnic groups with the highest among NH Asian (71.8 %) and lowest among Hispanic (60.2 %); we also found significant variance by insurance, with the highest percentage of adherence patients for Medicare HMO (76.3 %) followed by Medicare FFS (69.6 %). Patients with access to health care were more likely to be adherent to statins (all P < 0.05 except for PCP visit). Seventy-three percent of patients with high intensity level of index statins were adherent, while 65.5 % and 61.6 % of patients with moderate and low intensity level of index statins were adherent, respectively(P < 0.0001).
Table 2.
Patient characteristics by medication adherence.
| Adherence to Statins |
All (N = 5155) |
||||
|---|---|---|---|---|---|
| Secondary non-adherence (N = 1265) | Primary non-adherence (N = 337) | Adherence (N = 3553) | P-value | ||
| N ( %) | N ( %) | N ( %) | N ( %) | ||
| Age group | <0.0001 | ||||
| 35–44 | 70 (38.7) | 9 (5.0) | 102 (56.4) | 181 (100.0) | |
| 45–54 | 166 (27.9) | 54 (9.1) | 376 (63.1) | 596 (100.0) | |
| 55–64 | 344 (28.0) | 76 (6.2) | 807 (65.8) | 1227 (100.0) | |
| 65–74 | 355 (21.9) | 107 (6.6) | 1159 (71.5) | 1621 (100.0) | |
| ≥75 | 330 (21.6) | 91 (5.9) | 1109 (72.5) | 1530 (100.0) | |
| Female | 637 (26.4) | 157 (6.5) | 1616 (67.1) | 2410 (100.0) | 0.0117 |
| Race/Ethnicity | <0.0001 | ||||
| Hispanic | 167 (33.1) | 34 (6.7) | 304 (60.2) | 505 (100.0) | |
| Multi-race | 19 (25.0) | 6 (7.9) | 51 (67.1) | 76 (100.0) | |
| NH Asian | 99 (22.0) | 28 (6.2) | 324 (71.8) | 451 (100.0) | |
| NH Black | 83 (32.5) | 6 (2.4) | 166 (65.1) | 255 (100.0) | |
| NH Other | 60 (20.3) | 29 (9.8) | 207 (69.9) | 296 (100.0) | |
| NH White | 837 (23.4) | 234 (6.6) | 2501 (70.0) | 3572 (100.0) | |
| BMI | 0.5509 | ||||
| Unknown | 56 (25.6) | 16 (7.3) | 147 (67.1) | 219 (100.0) | |
| Normal | 330 (24.1) | 100 (7.3) | 938 (68.6) | 1368 (100.0) | |
| Overweight | 440 (23.6) | 114 (6.1) | 1311 (70.3) | 1865 (100.0) | |
| Obese | 439 (25.8) | 107 (6.3) | 1157 (67.9) | 1703 (100.0) | |
| Smoking status | 0.2652 | ||||
| Unknown | 53 (25.4) | 16 (7.7) | 140 (67.0) | 209 (100.0) | |
| Non-smoker | 636 (25.6) | 166 (6.7) | 1685 (67.8) | 2487 (100.0) | |
| Former smoker | 451 (22.9) | 119 (6.0) | 1402 (71.1) | 1972 (100.0) | |
| Current smoker | 125 (25.7) | 36 (7.4) | 326 (66.9) | 487 (100.0) | |
| Primary Insurance | <0.0001 | ||||
| Medicaid/Medi-Cal | 36 (32.4) | 9 (8.1) | 66 (59.5) | 111 (100.0) | |
| Medicare FFS | 543 (22.9) | 179 (7.5) | 1653 (69.6) | 2375 (100.0) | |
| Medicare HMO | 187 (21.7) | 17 (2.0) | 657 (76.3) | 861 (100.0) | |
| HMO | 128 (27.4) | * | * | 467 (100.0) | |
| PPO/FFS | 364 (27.5) | 107 (8.1) | 855 (64.5) | 1326 (100.0) | |
| Other | 7 (46.7) | * | * | 15 (100.0) | |
| Cardiology visit (Yes) | 505 (22.3) | 126 (5.6) | 1632 (72.1) | 2263 (100.0) | <0.0001 |
| ED Visit (Yes) | 486 (24.5) | 93 (4.7) | 1407 (70.8) | 1986 (100.0) | <0.0001 |
| Inpatient visit (Yes) | 432 (22.4) | 104 (5.4) | 1392 (72.2) | 1928 (100.0) | 0.0002 |
| Patient portal use (Yes) | 736 (23.7) | 184 (5.9) | 2191 (70.4) | 3111 (100.0) | 0.0078 |
| PCP visit (Yes) | 1181 (24.5) | 313 (6.5) | 3330 (69.0) | 4824 (100.0) | 0.7787 |
| Telephone visit (Yes) | 1202 (24.7) | 307 (6.3) | 3364 (69.0) | 4873 (100.0) | 0.0149 |
| CHD (Yes) | 530 (23.0) | 149 (6.5) | 1628 (70.6) | 2307 (100.0) | 0.0537 |
| HTN (Yes) | 727 (23.0) | 203 (6.4) | 2226 (70.5) | 3156 (100.0) | 0.0048 |
| PVD (Yes) | 267 (25.9) | 63 (6.1) | 699 (67.9) | 1029 (100.0) | 0.458 |
| Stroke (Yes) | 468 (25.7) | 125 (6.9) | 1226 (67.4) | 1819 (100.0) | 0.2179 |
| T2D (Yes) | 250 (26.1) | 74 (7.7) | 633 (66.1) | 957 (100.0) | 0.0786 |
| Index Statin intensity | <0.0001 | ||||
| high | 502 (20.9) | 138 (5.7) | 1764 (73.4) | 2404 (100.0) | |
| moderate | 671 (27.4) | 174 (7.1) | 1601 (65.5) | 2446 (100.0) | |
| low | 92 (30.2) | 25 (8.2) | 188 (61.6) | 305 (100.0) | |
| LDL level 6 months pre index_statin | 0.002 | ||||
| Unknown | 708 (25.2) | 173 (6.2) | 1925 (68.6) | 2806 (100.0) | |
| Uncontrolled | 491 (25.2) | 138 (7.1) | 1321 (67.7) | 1950 (100.0) | |
| Controlled | 66 (16.5) | 26 (6.5) | 307 (76.9) | 399 (100.0) | |
| Non-statin lipid lowering therapy prescriptions | 0.3484 | ||||
| No | 1206 (24.4 %) | 326 (6.6 %) | 3416 (69.0 %) | 4948 (100.0 %) | |
| Yes | 59 (28.5 %) | 11 (5.3 %) | 137 (66.2 %) | 207 (100.0 %) | |
Masked for privacy/security
Variables were compared among the three patient adherence groups (primary non-adherent, secondary non-adherent, adherent) using a Chi-square test.
3.2. Factors associated with filling the prescription
In the multivariable logistic regression model (Fig. 2), patients with PPO/FFS (OR= 0.60; 95 % CI, 0.37–0.96) or other insurance plan (OR= 0.20; 95 % CI, 0.05–0.79) were significantly less likely to fill their statin prescription compared to those with HMO. Patients with less frequent use of online patient portal (OR= 0.65, 95 % CI, 0.51–0.84) were also less likely to fill the statins prescription. Black patients were more likely to fill the prescription than non-Hispanic White patients (OR= 3.33; 95 % CI, 1.45–7.64), obese patients were more likely to fill the prescription (OR=1.37; 95 % CI, 1.01–1.88), patients with recent ED visits were more likely to fill the prescription (OR= 1.64; 95 % CI, 1.21–2.22), and patients seeing Cardiologist were also significantly more likely to fill the statins prescription (OR= 1.32; 95 % CI, 1.04–1.68).
Fig. 2.
Adjusted Odds Ratio (OR) for filling the statin prescription among patients with ASCVD.
Description: * 1 year on or before index statin order date; ∼ median income by patient's zip code, Q1=$55,443, Q3=$86,961; ^ LDL <70 is controlled, ≥70 is uncontrolled; PCP median visits=3; Online patient portal median visits=2
3.3. Factors associated with statin adherence
In the multivariable logistic regression model (Fig. 3), older patients were more likely to adhere (OR varies from 1.53 to 2.12; 95 % CI varies from 1.06–2.20 to 1.43–3.15). Patients receiving high intensity statins prescriptions (OR=1.46; 95 % CI, 1.26–1.69), being diagnosed with hypertension (OR, 1.17; 95 % CI, 1.01–1.35), or having visits to a cardiologist (OR, 1.19; 95 % CI, 1.04–1.37) were more likely to adhere. Female (vs. male: OR, 0.86; 95 % CI,0.75–0.99), Hispanic (vs. non-Hispanic white (NHW): OR= 0.71; 95 % CI, 0.57–0.87), patients with other insurance plan (OR= 0.29; 95 % CI, 0.09–0.95), and less frequent use of online patient portal (OR, 0.82, 95 % CI, 0.71–0.95) were significantly less likely to adhere. Patients with uncontrolled (OR, 0.66; 95 % CI, 0.49–0.89) or unknown baseline LDL-C (OR, 0.6; 95 % CI, 0.45–0.8) were less likely to adhere compared to those whose baseline LDL-C were under control at the time of the statin prescription.
Fig. 3.
Adjusted Odds Ratio (OR) for statin adherence among patients with ASCVD.
Description: * 1 yr on or before index statin order date; ∼ median income by patient's zip code, Q1=$55,443, Q3=$86,961; ^ LDL <70 is controlled, ≥70 is uncontrolled; PCP median visits=3; Online patient portal median visits=2
3.4. Impact of adherence on LDL-C control
Over a median follow-up of 1 year, there were 2035 (39.5 %) patients achieved who LDL-C control (control rate, 1.42/1000 person-years). Specifically, the time to achieve LDL-C control for the adherent group is shorter than that of primary non-adherence group, whose time to achieve LDL-C control is shorter than that of secondary non-adherence group (Fig. 4).
Fig. 4.
Kaplan–Meier Estimates of Time to LDL Control by Statin Adherence among Patients with Uncontrolled LDL at Baseline, 1/2015–10/2021.
The Cox models yield similar results (Fig. 5). The primary non-adherence group had poorer LDL-C control than adherence group (hazard ratio (HR)=0.65; 95 % CI, 0.53–0.79), while LDL-C control for the secondary non-adherence group was much lower (HR=0.48; 95 % CI, 0.43–0.55). Patients receiving an index statin order of high or moderate intensity level (high vs. low: HR=2.75; 95 % CI, 2.11–3.59; moderate vs. low: HR=1.94; 95 % CI, 1.49–2.53), seeing a Cardiologist (HR=1.17; 95 % CI, 1.07–1.28), and having mild comorbidities (vs. no comorbidity: HR=1.16; 95 % CI, 1.01–1.33) had a significantly higher LDL-C control rate. Female (HR=0.7; 95 % CI, 0.64–0.77), Black patients (vs. NHW: HR=0.71; 95 % CI, 0.56–0.9), patients with Medicaid/Medi-Cal or Medicare (Medicaid/Medi-Cal vs. HMO: HR=0.58; 95 % CI, 0.39–0.87; Medicare vs. HMO: HR=0.82; 95 % CI, 0.68–0.98), patients with recent ED visits (vs. no ED visit: HR=0.83; 95 % CI, 0.74–0.92), less frequent use of online patient portal (vs. >median: HR=0.89; 95 % CI, 0.81–0.98), and uncontrolled (vs. controlled: HR=0.67; 95 % CI, 0.58–0.77), or unknown (vs. controlled: HR=0.69; 95 % CI, 0.59–0.79) baseline LDL-C had a significantly lower LDL-C control rate.
Fig. 5.
Hazard ratios of patient LDL-C control, within 12 months after the index statin prescription, 1/2015 - 10/2021.
Description: * 1 yr on or before index statin order date; ∼ median income by patient's zip code, Q1=$55,443, Q3=$86,961; ^ LDL <70 is controlled, ≥70 is uncontrolled; Medication adherence 12 months after the index statin prescription ++
4. Discussion
The study aimed to explore factors influencing adherence of newly prescribed statins in individuals with cardiovascular disease (CVD) within a large community-based health system. Notably, 68.9 % of the 5155 patients were adherent, with insurance type, online patient portal use, race, age, statin intensity, and cardiologist visits emerging as significant predictors.
In contrast to existing literature [17], patients with PPO/FFS or other insurance were less likely to fill statin prescriptions compared to those with HMO. Black patients exhibited higher rates of filling a statin prescription but lower adherence rates than non-Hispanic White patients, further advancing some prior studies where Black Americans, low educational attainment, and high poverty levels are positively associated with nonadherence [[18], [19], [20]]. Additionally, the positive association between infrequent online patient portal use and lower adherence is consistent with the literature [21,22] emphasizing the positive impact of digital health tools. A potential explanation is that some functionalities of the portal including laboratory tests and imaging, medical notes, messaging with providers, medication refill, and current medication list may be helpful to improve medication adherence.
Older age [18], high-intensity statin prescriptions [23], and cardiologist visits [24,25] correlated positively with adherence, aligning with prior research . However, the study provides new insights, such as the impact of recent ED visits and the significant association of cardiologist involvement on adherence.
We should be cautious in interpreting the result that primary non-adherence to statins had better LDL-C control than secondary non-adherence to statins. Since 2018, non-statin lipid-lowering therapy (e.g., ezetimibe or proprotein convertase subtilisin/kexin type 9 (PCSK9)) has shown significantly lower LDL-C, when used alone or added to statins [26], and new non-statin medications might affect statin adherence since patients may switch to non-statin lipid-lowing medication. Our results highlight the importance of tracking statin usage and non-statin lipid-lowering medications in future work to fully understand treatment effects on patients with newly diagnosed CVD. In our analysis, having non-statin lipid-lowering therapy prescriptions was not statistically significantly associated with filling the statin prescription or statin adherence.
The findings underline the importance of considering demographic and healthcare-related factors when addressing medication adherence. To enhance adherence in patient management, strategies tailored to diverse populations should be implemented. For instance, improving access to healthcare resources for those with PPO/FFS or other insurance and promoting online patient portal engagement might enhance medication adherence.
Strengths of the study include its large sample size, community-based setting, and comprehensive examination of factors associated with adherence and LDL-C control. This study has several limitations that must be considered. First, it was conducted within a single healthcare system, which may introduce geographic or healthcare system-related biases that could affect the generalizability of the findings to other regions or systems. However, the uniqueness of Sutter Health is that Sutter serves a diverse population, in particular Asians and Hispanics are especially well-represented. Other limitations include the reliance on prescription fill data as a proxy for adherence and potential unmeasured confounders affecting the results. There are a variety of other factors, such as socioeconomic status, medication cost, patient education, and healthcare access, that can influence statin adherence but were not assessed in this study. A more comprehensive evaluation of these variables, along with a deeper exploration of the complex interplay between them, is needed to gain a more thorough understanding of the factors affecting statin adherence and its impact on patient outcomes.
Implications for research and practice involve the need for targeted interventions addressing disparities in adherence based on insurance type, race, and online portal use. Further investigation into the impact of recent ED visits on adherence is warranted. Implementing strategies that involve cardiologists in medication management could prove beneficial. Overall, these findings contribute valuable insights to the ongoing efforts to enhance medication adherence in diverse patient populations. Understanding the factors that influence adherence to statins can help healthcare providers identify potential barriers, tailor interventions to meet the specific needs of patients, and improve overall health outcomes. Furthermore, LDL-C is a primary target for lipid-lowering therapy in individuals with CVD. Maintaining adequate control of LDL-C is essential for reducing the risk of cardiovascular events. By examining the relationship between statin adherence and LDL-C control, this study can help healthcare providers and patients better understand the importance of medication adherence in achieving optimal LDL-C control and improving clinical outcomes. Overall, the study has the potential to improve clinical practice and inform healthcare policies aimed at improving adherence to statins in individuals with CVD, ultimately leading to better health outcomes and reduced healthcare costs.
5. Conclusions
The study explored statin adherence in patients with cardiovascular disease, revealing that insurance type, online portal use, race, age, statin intensity, and cardiologist visits may be significant predictors. Unique findings include lower adherence in PPO/FFS patients, a discrepancy between patients race categories (Black vs. Non-Hispanic White patients) in filling the prescription and adhering to the filled prescription, and the positive association of healthcare utilization such as online portal use and cardiologist visits, on adherence. Positive correlations were found with older age, high-intensity statins, and cardiologist visits. The study emphasizes demographic and healthcare factors in medication adherence, suggesting the potential for tailored interventions for diverse populations, addressing disparities in insurance type, race, and online portal use, and involving cardiologists in medication management for improved adherence. Among CVD patients, having access to cardiologists may be an important factor for improving patient medication adherence and achieving better outcomes, and highlights the importance of care coordination between primary and specialty care.
Sources of financial support
This study was funded by Sutter Health.
Availability of data and materials
The datasets generated and/or analyzed during the current study are not available for replication because they contain patient health information that could compromise the privacy of research participants but are available from the corresponding author on reasonable request.
CRediT authorship contribution statement
Jiang Li: Writing – review & editing, Writing – original draft, Validation, Methodology, Formal analysis, Conceptualization. Satish Mudiganti: Writing – review & editing, Writing – original draft, Validation, Formal analysis, Data curation, Conceptualization. Hannah Husby: Writing – review & editing, Project administration. J B Jones: Writing – review & editing, Conceptualization. Xiaowei Yan: Writing – review & editing, Writing – original draft, Methodology, Formal analysis, Conceptualization.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Footnotes
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.ajpc.2025.101028.
Contributor Information
Jiang Li, Email: Jiang.Li2@sutterhealth.org.
Satish Mudiganti, Email: Satish.Mudiganti@sutterhealth.org.
Hannah Husby, Email: Hannah.HusbyAlvarado@sutterhealth.org.
J B Jones, Email: james.jones2@sutterhealth.org.
Xiaowei Yan, Email: Sherry.Yan@sutterhealth.org.
Appendix. Supplementary materials
References
- 1.World Health Organization; 2022. World health statistics 2022: monitoring health for the SDGs, sustainable development goals.https://www.who.int/data/gho/publications/world-health-statistics Accessed March 22, 2023. [Google Scholar]
- 2.Shalev V., Chodick G., Silber H., Kokia E., Jan J., Heymann A.D. Continuation of statin treatment and all-cause mortality: a population-based cohort study. Arch Intern Med. 2009;169(3):260–268. doi: 10.1001/archinternmed.2008.552. [DOI] [PubMed] [Google Scholar]
- 3.Mills E.J., Wu P., Chong G., et al. Efficacy and safety of statin treatment for cardiovascular disease: a network meta-analysis of 170,255 patients from 76 randomized trials. QJM Mon J Assoc Physicians. 2011;104(2):109–124. doi: 10.1093/qjmed/hcq165. [DOI] [PubMed] [Google Scholar]
- 4.Hansen M.R., Hróbjartsson A., Pottegård A., et al. Postponement of cardiovascular outcomes by statin use: a systematic review and meta-analysis of randomized clinical trials. Basic Clin Pharmacol Toxicol. 2021;128(2):286–296. doi: 10.1111/bcpt.13485. [DOI] [PubMed] [Google Scholar]
- 5.Zhong P., Wu D., Ye X., et al. Secondary prevention of major cerebrovascular events with seven different statins: a multi-treatment meta-analysis. Drug Des Devel Ther. 2017;11:2517–2526. doi: 10.2147/DDDT.S135785. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Yang X.H., Zhang B.L., Cheng Y., Fu S.K., Jin H.M. Statin use and the risk of CVD events, stroke, and all-cause mortality in patients with diabetes: a systematic review and meta-analysis. Nutr Metab Cardiovasc Dis NMCD. 2022;32(11):2470–2482. doi: 10.1016/j.numecd.2022.07.018. [DOI] [PubMed] [Google Scholar]
- 7.De Vera M.A., Bhole V., Burns L.C., Lacaille D. Impact of statin adherence on cardiovascular disease and mortality outcomes: a systematic review. Br J Clin Pharmacol. 2014;78(4):684–698. doi: 10.1111/bcp.12339. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Banach M., Stulc T., Dent R., Toth P.P. Statin non-adherence and residual cardiovascular risk: there is need for substantial improvement. Int J Cardiol. 2016;225:184–196. doi: 10.1016/j.ijcard.2016.09.075. [DOI] [PubMed] [Google Scholar]
- 9.Herttua K., Martikainen P., Batty G.D. Kivim äki M. Poor adherence to statin and antihypertensive therapies as risk factors for fatal stroke. J Am Coll Cardiol. 2016;67(13):1507–1515. doi: 10.1016/j.jacc.2016.01.044. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Bosl W., Mandel J., Jonikas M., Ramoni R.B., Kohane I.S., Mandl K.D. Scalable decision support at the point of care: a substitutable electronic health record app for monitoring medication adherence. Interact J Med Res. 2013;2(2):e13. doi: 10.2196/ijmr.2480. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Toscos T., Drouin M., Pater J.A., et al. Medication adherence for atrial fibrillation patients: triangulating measures from a smart pill bottle, e-prescribing software, and patient communication through the electronic health record. JAMIA Open. 2020;3(2):233–242. doi: 10.1093/jamiaopen/ooaa007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Ingersgaard M.V., Helms Andersen T., Norgaard O., Grabowski D., Olesen K. Reasons for nonadherence to statins - a systematic review of reviews. Patient Prefer Adherence. 2020;14:675–691. doi: 10.2147/PPA.S245365. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Fung V., Graetz I., Reed M., Jaffe M.G. Patient-reported adherence to statin therapy, barriers to adherence, and perceptions of cardiovascular risk. PLoS ONE. 2018;13(2) doi: 10.1371/journal.pone.0191817. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Erickson S.R., Bravo M., Tootoo J. Geosocial factors associated with adherence to statin medications. Ann Pharmacother. 2020;54(12):1194–1202. doi: 10.1177/1060028020934879. [DOI] [PubMed] [Google Scholar]
- 15.Sperl-Hillen J.M., Anderson J.P., Haapala J., et al. 1180-P: better care strategies are necessary to address poor medication adherence. Diabetes. 2020;69(Supplement_1):1180. doi: 10.2337/db20-1180-P. [DOI] [Google Scholar]
- 16.Blecker S., Adhikari S., Zhang H., et al. Validation of EHR medication fill data obtained through electronic linkage with pharmacies. J Manag Care Spec Pharm. 2021;27(10):1482–1487. doi: 10.18553/jmcp.2021.27.10.1482. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Tong S.T., Sabo R.T., Hochheimer C.J., et al. Uptake of statin guidelines to prevent and treat cardiovascular disease. J Am Board Fam Med JABFM. 2021;34(1):113–122. doi: 10.3122/jabfm.2021.01.200292. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Hope H.F., Binkley G.M., Fenton S., Kitas G.D., Verstappen S.M.M., Symmons D.P.M. Systematic review of the predictors of statin adherence for the primary prevention of cardiovascular disease. PLOS ONE. 2019;14(1) doi: 10.1371/journal.pone.0201196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Ofori-Asenso R., Jakhu A., Curtis A.J., et al. A systematic review and meta-analysis of the factors associated with nonadherence and discontinuation of statins among people aged ≥65 years. J Gerontol A Biol Sci Med Sci. 2018;73(6):798–805. doi: 10.1093/gerona/glx256. [DOI] [PubMed] [Google Scholar]
- 20.Abbass I., Revere L., Mitchell J., Appari A. Medication nonadherence: the role of cost, community, and individual factors. Health Serv Res. 2017;52(4):1511–1533. doi: 10.1111/1475-6773.12547. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Han H.R., Gleason K.T., Sun C.A., et al. Using patient portals to improve patient outcomes: systematic review. JMIR Hum Factors. 2019;6(4) doi: 10.2196/15038. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Carini E., Villani L., Pezzullo A.M., et al. The impact of digital patient portals on health outcomes, system efficiency, and patient attitudes: updated systematic literature review. J Med Internet Res. 2021;23(9) doi: 10.2196/26189. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Rezende Macedo do Nascimento R.C., Mueller T., Godman B., et al. Real-world evaluation of the impact of statin intensity on adherence and persistence to therapy: a Scottish population-based study. Br J Clin Pharmacol. 2020;86(12):2349–2361. doi: 10.1111/bcp.14333. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Colantonio L.D., Rosenson R.S., Deng L., et al. Adherence to statin therapy among US adults between 2007 and 2014. J Am Heart Assoc. 2019;8(1) doi: 10.1161/JAHA.118.010376. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Rehman H., Ahmed S., Akeroyd J., et al. Association between cardiology visit and evidence-based statin prescription and statin adherence patterns: findings from the veterans affairs national database. J Am Coll Cardiol. 2019;73:1709. doi: 10.1016/S0735-1097(19)32315-0. [DOI] [Google Scholar]
- 26.Grundy S.M., Stone N.J., Bailey A.L., et al. 2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA guideline on the management of blood cholesterol: executive summary: a report of the american college of cardiology/american heart association task force on clinical practice guidelines. J Am Coll Cardiol. 2019;73(24):3168–3209. doi: 10.1016/j.jacc.2018.11.002. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets generated and/or analyzed during the current study are not available for replication because they contain patient health information that could compromise the privacy of research participants but are available from the corresponding author on reasonable request.