Abstract
Introduction
Good medication adherence is associated with decreased healthcare expenditure; however, adherence is usually assessed for single medication. We aim to explore the associations of adherence levels to 23 chronic medications with emergency room (ER) visits and hospitalizations. The primary endpoints are ER visits and hospitalizations in internal medicine and surgical wards.
Methods
Individuals aged 50–74 years, with a diagnosis of diabetes mellitus or hypertension, treated with at least one antihypertensive or antidiabetic medication during 2017 were included. We determined personal adherence rates by calculating the mean adherence rates of the medications prescribed to each individual. Adherence rates were stratified into categories. We retrieved information about all the ER visits, and hospitalizations in internal medicine and surgical wards during 2016–2018.
Results
Of 268,792 persons included in the study, 50.6% were men. The mean age was 63.7 years. Hypertension was recorded for 217,953 (81.1%), diabetes for 160,082 (59.5%), and both diabetes and hypertension for 109,225 (40.6%). The mean number of antihypertensive and antidiabetic medications used was 2.2 ± 1.1. In total, 51,301 (19.1%) of the cohort visited the ER at least once during 2017, 21,740 (8.1%) were hospitalized in internal medicine wards, and 10,167 (3.8%) in surgical wards during 2017. Comparing the highest adherence category to the lowest, adjusted odds ratios were 0.64 (0.61, 0.67) for ER visits, 0.56 (0.52, 0.60) for hospitalization in internal wards, and 0.63 (0.57, 0.70) for hospitalization in surgical wards. Odds ratios were similar for the three consecutive years 2016–2018.
Conclusion
Better medication adherence was associated with fewer ER visits and hospitalizations among persons with diabetes and hypertension. Investing in improving medication adherence may reduce health costs and improve patients’ health.
KEY WORDS: medication adherence, hospitalizations, chronic care, diabetes mellitus, hypertension
INTRODUCTION
Adherence to medications has an important role in chronic disease management. Good adherence is associated with a long-term decrease in healthcare expenditure ,1 and medication non-adherence places a significant cost burden on healthcare systems.2 Adherence to over 80% of prescribed medications has been associated with fewer hospitalizations, emergency room (ER) visits, and outpatient visits among persons with various chronic diseases.3 In a retrospective study conducted in the USA, persons who did not adhere to oral antidiabetic medications were twice as likely to be hospitalized (all-cause and diabetes-related) as were those who adhered to their treatment.4 Lower adherence to oral antidiabetic medications was associated with a greater risk of diabetes-related hospitalizations/ER visits compared to perfect adherence.5 Among persons with hypertension, cardiovascular and all-cause mortality were inversely associated with antihypertensive medication adherence.6 Furthermore, poor antihypertensive medication adherence was associated with more preventable hospitalizations due to hypertension, and four-fold increase in costs, combated to high adherence.7
Adherence is commonly measured for single medication or medication group. However, among persons with diabetes, adherence has been shown to vary considerably across medication classes.8 Moreover, most persons with diabetes or hypertension use more than one medication, and adherence to different medications might differ.
The relationship between overall medication adherence to long-term treatment in patients with diabetes and hypertension, and ER visits and hospitalizations in a real-life setting has not been investigated.
We aimed to explore associations of adherence to antidiabetic and antihypertensive medications, with ER visits and hospitalizations.
METHODS
Setting
The local ethics committee of Meir Health Center approved the study. This is a cohort study conducted in Clalit Health Services (CHS), the largest of four health maintenance organizations (HMOs) in Israel. CHS serves over 52% of the population nationwide. Electronic medical records have been used for over 20 years to create an extensive healthcare database. Data includes demographic information that is updated directly from the population registry of the Israeli Ministry of Interior. CHS operates a chronic disease registry, which integrates information from patient files, hospital discharges, medication claims, laboratory results, prescriptions and medication claims, hospitalizations, and ER visits.9 Laboratory tests in CHS do not incur copayment; the results are recorded automatically in patients’ electronic medical records. Pharmacies, operated by or affiliated with CHS, record all medication claims electronically, and are linked to one central repository. CHS issues medications and requires nominal copayments of US $5–15 per medication; copayments are capped at US $75 per month. Members can buy chronic medications for up to 3 months at a time (documented as three separate purchases on the same date).
We assessed the adherence to 11 antihypertensive medications, 8 antidiabetic medications, and 4 HMG-CoA reductase inhibitors. We examined the most commonly prescribed medications for hypertension and diabetes in Israel. Antihypertensive medications included hydrochlorothiazide, calcium channel blockers (nifedipine, amlodipine, and lercanidipine hydrochloride), beta blockers (atenolol metoprolol and bisoprolol), ARBs (valsartan and losartan), and ACEI (ramipril, captopril, and cilazapril), and HMG-CoA reductase inhibitors included rosuvastatin, atorvastatin, simvastatin, and pravastatin. The antidiabetic medications were metformin, glimepiride, repaglinide, empagliflozin, dapagliflozin, sitagliptin, vildagliptin, and linagliptin.
Study Population
The study population comprised CHS members aged 50–75 years with a documented diagnosis of diabetes mellitus or hypertension in the CHS chronic disease registry before January 1, 2016. To assure medication use throughout 2017, filling at least one prescription per year in the three consecutive years 2016–2018 for one or more of the study medications was an inclusion criterion. This approach enabled reliable calculation of medication adherence, and the exclusion of individuals whose treatment changed during the study period.
We analyzed all filled monthly prescriptions between January 1, 2017, and December 31, 2017. A filled prescription is an accurate and inexpensive proxy for actual medication use.10 We calculated the individual mean adherence rate for all the medications purchased, from the 23 medications included in the study.
Demographic information included age, gender, and socioeconomic status (SES). Patients with low SES were defined as those exempt from healthcare payments by the National Insurance Institute of Israel, based on their income. These individuals pay reduced copayments for chronic medications. We also retrieved the last available body mass index measurement of the participants, before January 1, 2017. We noted other relevant cardiovascular chronic diagnoses as recorded in the CHS chronic diseases register on January 1, 2017, Charlson comorbidity score (for adjustment of comorbidity among patients),11 and mean levels of LDL-cholesterol, HbA1c, creatinine, and hemoglobin recorded in 2017. We retrieved information about all the primary care visits, ER visits, and hospitalizations in internal medicine and surgical wards for the three consecutive years 2016–2018. Israel has universal health coverage. CHS has primary care clinics throughout the country and the majority of patient care is carried out within the community. Chronic care is provided by the family physician with the primary care clinic team and most specialist consultations are performed within the framework of community care. The cost of hospitalizations is fully covered by the HMO; ER visits have a copayment when there is no referral (depending on diagnosis). All hospitalizations and ER visits are documented within the Clalit database.
Statistical Analysis
We calculated overall adherence levels for each medication separately by dividing the actual fulfilled monthly prescriptions during 2017 by 12. This ratio is similar to proportion of days covered since all patients have started their medications before January 1 and continue taking them after December 31, 2017. We determined personal adherence rates by calculating the mean adherence rates of all the medications prescribed to each individual, since there appears to be no standardized method to measure multiple medication adherence,12 and from previous studies, we have learnt that adherence varies with different medications.13,14 We chose this method in order to assess the actual purchasing of the prescribed medication in chronic care where long-term purchasing is important.
Primary endpoints were ER visits and hospitalizations in internal medicine and surgical wards. We stratified adherence rates into 5 categories (1–20%, 21–40%, 41–60%, 61–80%, 81–100%); the lowest group was considered the reference group. We used logistic regression models to calculate odds ratios for ER visits and hospitalizations in internal and surgical wards, and adjusted the model for age, gender, SES, Charlson score, and family physician visits. We further analyzed the ER visits and hospitalizations in internal and surgical wards during 2016 and 2018 according to adherence categories during 2017. STATA 8.0 statistical software (Stata Corp. College Station, TX, USA) was used for statistical analysis.
RESULTS
Of 268,792 persons included in the study, 50.6% were men. The mean age was 63.7 years. Hypertension was recorded for 217,953 (81.1%) individuals and diabetes for 160,082 (59.5%). Diagnoses of both diabetes and hypertension were recorded for 109,225 (40.6%). The mean number of antihypertensive and antidiabetic medications used was 2.2 ± 1.1. Characteristics of the cohort are described in Table 1. The adherence rate was ≤ 20% for 11,331 (4.2%) persons, 21–40% for 20,911 (7.8%), 41–60% for 47,525 (17.7%), 61–80% for 74,648 (27.8%), and over 80% for 114,377 (42.5%).
Table 1.
Characteristics of the Study Cohort
| With hypertension (217,953) | With diabetes mellitus (160,082) | With hypertension and diabetes mellitus (109,225) | |
|---|---|---|---|
| Age, years (50–74) (mean ± SD) | 64.1 ± 6.1 | 63.5 ± 6.3 | 64.3 ± 6.1 |
| Gender (% women) | 49.7% | 48.0% | 48.0% |
| Low socioeconomic status* (%) | 40.0% | 45.8% | 45.6% |
| BMI (mean ± SD) | 30.2 ± 5.6 | 30.4 ± 5.6 | 30.9 ± 5.7 |
| Charlson score (mean ± SD) | 1.9 ± 1.8 | 2.5 ± 1.8 | 2.7 ± 1.9 |
| s/p MI (%) | 13.3% | 13.6% | 16.3% |
| s/p CVA (%) | 10.0% | 9.7% | 12.4% |
| Hemoglobin (mean ± SD) | 13.5 ± 1.4 | 13.4 ± 1.4 | 13.3 ± 1.5 |
| Creatinine (mean ± SD) | 0.9 ± 0.6 | 0.9 ± 0.6 | 0.9 ± 0.6 |
| LDL-cholesterol (mean ± SD) | 93.0 ± 30.5 | 87.4 ± 29.5 | 85.8 ± 29.3 |
| HbA1c (mean ± SD) | 7.3 ± 1.4 | 7.3 ± 1.4 | |
| Number of medications** (mean ± SD) | 2.2 ± 1.1 | 2.3 ± 1.2 | 2.6 ± 1.2 |
| Number of family physician visits in 2017 (mean ± SD) | 13.3 ± 8.7 | 13.8 ± 9.1 | 14.6 ± 9.5 |
*Low socioeconomic status was defined as exemption by the National Insurance Institute of Israel, from healthcare payments, based on income
**Total medications of the 23 medications examined
BMI, body mass index; MI, myocardial infarction; CVA, cerebrovascular accident; LDL, low-density lipoprotein; SD, standard deviation
During 2017, 51,301 (19.1%) of the cohort visited the ER at least once, 21,740 (8.1%) were hospitalized in an internal medicine ward, and 10,167 (3.8%) were hospitalized in a surgical ward.
Table 2 shows unadjusted and adjusted odds ratios for the three outcomes examined according to categories of medication adherence. The adjusted odds ratios were significant from the second category (21–40% adherence) for emergency room visits, and from the third category (41–60% adherence) for both hospitalization in an internal medicine ward and hospitalization in a surgical ward.
Table 2.
Odds Ratios and Adjusted Odds Ratios*(95% CI) for Emergency Room Visits and Hospitalization for the Whole Cohort During 2017
| Unadjusted odds ratios | Adjusted* odds ratios | |
|---|---|---|
| Emergency room visits | ||
| Adherence rate 21–40% | 0.96 (0.91, 1.01) | 0.90 (0.85, 0.95) |
| Adherence rate 41–60% | 0.89 (0.84, 0.93) | 0.79 (0.75, 0.99) |
| Adherence rate 61–80% | 0.82 (0.78, 0.86) | 0.71 (0.67, 0.74) |
| Adherence rate > 80% | 0.74 (0.71, 0.77) | 0.64 (0.61, 0.67) |
| Hospitalization in an internal medicine ward | ||
| Adherence rate 21–40% | 1.05 (0.97, 1.13) | 0.96 (0.88, 1.03) |
| Adherence rate 41–60% | 0.94 (0.87, 1.00) | 0.79 (0.74, 0.85) |
| Adherence rate 61–80% | 0.87 (0.81, 0.93) | 0.67 (0.62, 0.72) |
| Adherence rate > 80% | 0.74 (0.74, 0.79) | 0.56 (0.52, 0.60) |
| Hospitalization in a surgical ward | ||
| Adherence rate 21–40% | 1.06 (0.95, 1.19) | 0.99 (0.88, 1.11) |
| Adherence rate 41–60% | 0.97 (0.87, 1.07) | 0.85 (0.77, 0.95) |
| Adherence rate 61–80% | 0.91 (0.82, 1.00) | 0.75 (0.68, 0.84) |
| Adherence rate > 80% | 0.77 (0.70, 0.85) | 0.63 (0.57, 0.70) |
*Adjusted for age, gender, low socioeconomic status, Charlson score, and family physician visits
Table 3 shows odds ratios of the outcomes examined for the four top categories of medication adherence compared to the lowest category, for the three non-distinct groups: individuals with hypertension, individuals with diabetes mellitus, and individuals with both these conditions. Both ER visits and hospitalization in internal medicine wards decreased incrementally as medication adherence increased, from the third category of medication adherence (41–60% adherence). Hospitalization in surgical wards decreased as medication adherence increased, from the fourth category of medication adherence (61–80% adherence). For all three outcomes, the trends were similar for the three sub-groups examined.
Table 3.
Odds Ratios (95% CI) for Emergency Room Visits and Hospitalizations Among Individuals with Diabetes Mellitus and Hypertension During 2017 (Crude)
| With hypertension (217,953) | With diabetes mellitus (160,082) | With hypertension and diabetes mellitus (109,225) | |
|---|---|---|---|
| Emergency room visits | |||
| Adherence rate 21–40% | 0.97 (0.91, 1.03) | 1.01 (0.94, 1.09) | 1.04 (0.95, 1.14) |
| Adherence rate 41–60% | 0.87 (0.82, 0.92) | 0.93 (0.87, 0.99) | 0.91 (0.83, 0.99) |
| Adherence rate 61–80% | 0.80 (0.76, 0.85) | 0.87 (0.82, 0.92) | 0.82 (0.76, 0.90) |
| Adherence rate > 80% | 0.72 (0.68, 0.76) | 0.77 (0.72, 0.82) | 0.72 (0.67, 0.78) |
| Hospitalization in an internal medicine ward | |||
| Adherence rate 21–40% | 1.06 (0.97, 1.15) | 1.07 (0.97, 1.18) | 1.05 (0.94, 1.18) |
| Adherence rate 41–60% | 0.92 (0.85, 0.99) | 1.00 (0.92, 1.09) | 0.94 (0.85, 1.04) |
| Adherence rate 61–80% | 0.82 (0.76, 0.89) | 0.94 (0.86, 1.02) | 0.81 (0.73, 0.89) |
| Adherence rate > 80% | 0.69 (0.64, 0.74) | 0.77 (0.70, 0.83) | 0.63 (0.57, 0.69) |
| Hospitalization in a surgical ward | |||
| Adherence rate 21–40% | 1.11 (0.98, 1.26) | 1.05 (0.92, 1.20) | 1.10 (0.93, 1.30) |
| Adherence rate 41–60% | 0.95 (0.85, 1.07) | 1.00 (0.88, 1.13) | 0.95 (0.82, 1.11) |
| Adherence rate 61–80% | 0.90 (0.80, 1.00) | 0.91 (0.81, 1.03) | 0.85 (0.73, 0.98) |
| Adherence rate > 80% | 0.75 (0.68, 0.84) | 0.74 (0.66, 0.84) | 0.68 (0.57, 0.78) |
The same trend was observed for ER visits and hospitalizations in 2016, 2017, and 2018 (Table 4). The risk for ER visits or hospitalization was higher among men, among persons with low SES, persons with higher Charlson score, and those who attended more visits in the primary care clinic during 2017.
Table 4.
Adjusted Odds Ratio* (95% CI) for Emergency Room Visits and Hospitalization for the Whole Cohort in 2016, 2017 (Index Year), and 2018
| 2016 | 2017 | 2018 | |
|---|---|---|---|
| Emergency room visits | |||
| Adherence rate 21–40% | 0.88 (0.83, 0.93) | 0.90 (0.85, 0.95) | 0.89 (0.84, 0.94) |
| Adherence rate 41–60% | 0.80 (0.76, 0.84) | 0.79 (0.75, 0.99) | 0.77 (0.73, 0.81) |
| Adherence rate 61–80% | 0.75 (0.71, 0.78) | 0.71 (0.67, 0.74) | 0.70 (0.66, 0.73) |
| Adherence rate > 80% | 0.67 (0.64, 0.70) | 0.64 (0.61, 0.67) | 0.63 (0.60, 0.66) |
| Hospitalization in internal wards | |||
| Adherence rate 21–40% | 0.84 (0.77, 0.91) | 0.96 (0.88, 1.03) | 0.82 (0.76, 0.88) |
| Adherence rate 41–60% | 0.78 (0.73, 0.84) | 0.79 (0.74, 0.85) | 0.65 (0.61, 0.70) |
| Adherence rate 61–80% | 0.74 (0.69, 0.80) | 0.67 (0.62, 0.72) | 0.55 (0.55, 0.63) |
| Adherence rate > 80% | 0.63 (0.60, 0.66) | 0.56 (0.52, 0.60) | 0.51 (0.48, 0.54) |
| Hospitalization in surgical wards | |||
| Adherence rate 21–40% | 0.90 (0.80, 1.01) | 0.99 (0.88, 1.11) | 0.83 (0.74, 0.92) |
| Adherence rate 41–60% | 0.86 (0.77, 0.95) | 0.85 (0.77, 0.95) | 0.76 (0.69, 0.84) |
| Adherence rate 61–80% | 0.82 (0.74, 0.91) | 0.75 (0.68, 0.84) | 0.70 (0.64, 0.77) |
| Adherence rate > 80% | 0.76 (0.69, 0.84) | 0.63 (0.57, 0.70) | 0.65 (0.59, 0.71) |
Adjusted for age, gender, low socioeconomic status, Charlson score, and family physician visits
DISCUSSION
This study of individuals with hypertension and diabetes demonstrated a negative association between medication adherence and hospitalizations, where increased adherence was associated with a decreased probability of ER visits and inpatient hospitalizations. Notably, this phenomenon was consistent over time.
The importance of medication adherence among persons with hypertension or diabetes is well known. A meta-analysis reported a 10% lower rate of hospitalization events among persons with diabetes who had good medication adherence.15 In another meta-analysis, 31% of treatment resistance to hypertension was potentially explained by poor adherence, while some of the studies included suggested a considerably greater effect.16 In a study by Durand et al., persons with hypertension with poor medication adherence were 1.5 times more likely to experience all-cause hospitalization than were persons with good medication adherence.17 The reported adherence rates varied substantially among studies, and most of them investigated a single medication or a single medication group. Our study examined associations of adherence to comprehensive treatment rather than to a single medication or medication group. Moreover, the trend of lower risk for ER visits and hospitalization was consistent over adherence categories and persisted during the years before and after the index year. Since the medications examined were all for chronic treatment, and were purchased the year before and the year after the index year, the findings may indicate that medication use over time did not change significantly.
In our cohort, 42.5% of the individuals adhered to over 80% of their antidiabetic and antihypertensive medications. This proportion is lower than the 67% of individuals with diabetes reported as at least 80% adherent to medications, according to a meta-analysis.18 In studies in which we assessed adherence separately for commonly used antihypertensive and hypoglycemic medications, the values were in the range of 52–79%.13,14 Thus, the relatively low rate of good adherence (above 80%) in the current study is presumably due to the averaging of the adherence rate to all the antidiabetic and antihypertensive medications taken.
Strengths and Limitations
This is a data-based study. In the absence of data of actual medication use, we considered medication adherence according to medication purchasing. To mitigate this limitation, we assessed medication claims over a 3-year period, and included only individuals with consistent data. The high quality and completeness of the databases are strengths of the study.
We had no information regarding various factors that could affect medication adherence and deterioration in patient conditions, and consequently lead to the need for hospitalization. The healthy adherer effect may be a marker for overall healthier behavior. However, the consistent negative relationship between adherence category and ER visits and hospitalizations and the persistence of this phenomenon the year before and the year after the medication adherence was measured may indicate that adherence did not change over time and therefore cannot explain the specific change before hospitalization. Overall health behavior changes little over time and therefore should be regarded as an internal patient character rather than a bias.
Our study created an opportunity to evaluate overall adherence since each patient’s adherence was calculated based on adherence to a number of commonly used medications in a very large cohort of patients aged 50–75. Moreover, most studies defined good adherence as a dichotomous measure, generally as medication adherence greater than 80%. However, in everyday life, medication adherence is not a dichotomous but rather a continuous phenomenon influenced by various variables.
CONCLUSION
We found an association of better medication adherence with fewer ER visits and hospitalizations among persons with diabetes and hypertension aged 50–75. Investing in improving medication adherence may reduce the use and costs of health resources, and improve patients’ health.
Declarations
Conflict of Interest
The authors declare that they do not have a conflict of interest.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Simon-Tuval T, Triki N, Chodick G, Greenberg D. The association between adherence to cardiovascular medications and healthcare utilization. Eur J Health Econ. 2016;17(5):603–10. doi: 10.1007/s10198-015-0703-z. [DOI] [PubMed] [Google Scholar]
- 2.Cutler RL, Fernandez-Llimos F, Frommer M, Benrimoj C, Garcia-Cardenas V. Economic impact of medication non-adherence by disease groups: a systematic review. BMJ Open. 2018;8(1):e016982. doi: 10.1136/bmjopen-2017-016982. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Roebuck MC, Kaestner RJ, Dougherty JS. Impact of Medication Adherence on Health Services Utilization in Medicaid. Med Care. 2018;56(3):266–273. doi: 10.1097/MLR.0000000000000870. [DOI] [PubMed] [Google Scholar]
- 4.Pednekar P, Heller DA, Peterson AM. Association of Medication Adherence with Hospital Utilization and Costs Among Elderly with Diabetes Enrolled in a State Pharmaceutical Assistance Program. J Manag Care Spec Pharm. 2020;26(9):1099–1108. doi: 10.18553/jmcp.2020.26.9.1099. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lo-Ciganic WH, Donohue JM, Jones BL, et al. Trajectories of Diabetes Medication Adherence and Hospitalization Risk: A Retrospective Cohort Study in a Large State Medicaid Program. J Gen Intern Med. 2016;31(9):1052–60. doi: 10.1007/s11606-016-3747-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kim S, Shin DW, Yun JM, et al. Medication Adherence and the Risk of Cardiovascular Mortality and Hospitalization Among Patients With Newly Prescribed Antihypertensive Medications. Hypertension. 2016;67(3):506–12. doi: 10.1161/HYPERTENSIONAHA.115.06731. [DOI] [PubMed] [Google Scholar]
- 7.Will JC, Zhang Z, Ritchey MD, Loustalot F. Medication Adherence and Incident Preventable Hospitalizations for Hypertension. Am J Prev Med. 2016;50(4):489–499. doi: 10.1016/j.amepre.2015.08.021. [DOI] [PubMed] [Google Scholar]
- 8.McGovern A, Tippu Z, Hinton W, Munro N, Whyte M, de Lusignan S. Comparison of medication adherence and persistence in type 2 diabetes: A systematic review and meta-analysis. Diabetes Obes Metab. 2018;20(4):1040–1043. doi: 10.1111/dom.13160. [DOI] [PubMed] [Google Scholar]
- 9.Rennert G, Peterburg Y. Prevalence of selected chronic diseases in Israel. Isr Med Assoc J. 2001;3:404–8. [PubMed] [Google Scholar]
- 10.Forbes CA, Deshpande S, Sorio-Vilela F, et al. A systematic literature review comparing methods for the measurement of patient persistence and adherence. Curr Med Res Opin 2018:1-13. doi: 10.1080/03007995.2018.1477747. [DOI] [PubMed]
- 11.Sharabiani MT, Aylin P, Bottle A. Systematic review of comorbidity indices for administrative data. Med Care. 2012;50(12):1109–18. doi: 10.1097/MLR.0b013e31825f64d0. [DOI] [PubMed] [Google Scholar]
- 12.Pednekar PP, Ágh T, et al. Methods for Measuring Multiple Medication Adherence: A Systematic Review-Report of the ISPOR Medication Adherence and Persistence Special Interest Group. Value Health. 2019;22(2):139–156. doi: 10.1016/j.jval.2018.08.006. [DOI] [PubMed] [Google Scholar]
- 13.Shani M, Lustman A, Vinker S. Adherence to oral antihypertensive medications, are all medications equal? J Clin Hypertens (Greenwich). 2019;21(2):243–248. doi: 10.1111/jch.13475. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Shani M, Lustman A, Vinker S. Diabetes medication persistence, different medications have different persistence rates. Prim Care Diabetes. 2017;11(4):360–364. doi: 10.1016/j.pcd.2017.03.006. [DOI] [PubMed] [Google Scholar]
- 15.Khunti K, Seidu S, Kunutsor S, Davies M. Association Between Adherence to Pharmacotherapy and Outcomes in Type 2 Diabetes: A Meta-analysis. Diabetes Care. 2017;40(11):1588–1596. doi: 10.2337/dc16-1925. [DOI] [PubMed] [Google Scholar]
- 16.Durand H, Hayes P, Morrissey EC, et al. Medication adherence among patients with apparent treatment-resistant hypertension: systematic review and meta-analysis. J Hypertens. 2017;35(12):2346–2357. doi: 10.1097/HJH.0000000000001502. [DOI] [PubMed] [Google Scholar]
- 17.Wu PH, Yang CY, Yao ZL, Lin WZ, Wu LW, Chang CC. Relationship of blood pressure control and hospitalization risk to medication adherence among patients with hypertension in Taiwan. Am J Hypertens. 2010;23(2):155–60. doi: 10.1038/ajh.2009.210. [DOI] [PubMed] [Google Scholar]
- 18.Iglay K, Cartier SE, Rosen VM, et al. Meta-analysis of studies examining medication adherence, persistence, and discontinuation of oral antihyperglycemic agents in type 2 diabetes. Curr Med Res Opin. 2015;31(7):1283–96. doi: 10.1185/03007995.2015.1053048. [DOI] [PubMed] [Google Scholar]