The Effect of Telehealth on Patterns of Health Care Utilization and Medication Prescription in Patients with Diabetes or Hypertension During COVID-19: A Nationwide Study

Abstract

Background:

In response to the coronavirus disease-19 pandemic, audio-based telehealth services for consultation and medication prescription were temporarily introduced in Korea. This study investigated the impact of telehealth services on patterns of health care utilization and medication prescription in patients with hypertension or diabetes.

Methods:

The 2019 to 2021 Health Insurance Review and Assessment Service claims data were used. The difference-in-difference approach was used to investigate the effect of telehealth services between the case and control group before and after the intervention period. The pre-intervention period was from February 24, 2019, to February 23, 2020, and the post-intervention period from February 24, 2020, to February 23, 2021. The control group included individuals who used in-person outpatient services and the case group those who utilized both telehealth and in-person services.

Results:

A total of 250,640 patients with hypertension and 154,212 patients with diabetes were included. The use of telehealth services was associated with an increase in outpatient visits in those with hypertension (0.07, p = 0.0027) and diabetes (0.32, p < 0.0001). A decrease in hospitalizations (−0.2%, p = 0.0007) and emergency department visits (−0.11%, p = 0.0016) was found in individuals with hypertension. Policy implementation also resulted in an increase in medication possession ratio (MPR) and the proportion of appropriate prescription in patients with hypertension (MPR: 3.0%, p < 0.0001, prescription: 3.1%, p < 0.0001) and diabetes (MPR: 3.4%, p < 0.0001, prescription: 1.7%, p < 0.0001).

Conclusions:

The findings confirm a relationship between implementing telehealth services and improved patterns of health care utilization and medication prescription, suggesting the potential benefit of telehealth in managing chronic diseases.

Introduction

Telehealth refers to the use of audio or video technology in the delivery of health care, including patient consultation, monitoring, and education.^1,2 Early applications of telehealth were introduced to patients facing difficulties in accessing basic medical care due to various reasons such as geographic rurality.^1,3 With the advancement of information technology and a greater emphasis on patient and quality outcomes, telehealth has become more widely adopted as an efficient means to deliver health care services.⁴ Telehealth has the potential to decrease unnecessary health care spending by preventing medication misuse and emergency department visits, while enhancing access to medical care at a comparable quality, thereby potentially increasing efficiency and patient satisfaction.^1,4 Hence, studies have recommended the use of telehealth as a supplement to in-person health care visits.^2,5

The emergence of the coronavirus disease-19 (COVID-19) has led to abrupt disturbances in the daily life of individuals as many governments adopted strong social distancing or isolation measures to contain its transmission.⁶ Specifically, such policies have limited interpersonal interactions by changing work patterns, suspending schools, and restricting the use of community centers.⁷ Patterns of health care utilization have been no exception as a decrease in the use of medical services has been found due to the concern of infection.⁸ Specifically, individuals residing in rural or underserved areas and with health-related problems, including patients with chronic diseases, may be particularly vulnerable as they face difficulties in receiving essential medical services. Unsurprisingly, the need to ensure safe and continuous medical use has been emphasized as studies have shown that patients with hypertension are more likely to suffer from severe symptoms when infected with COVID-19.⁹

In response to the increased need for remote health care services during the COVID-19 pandemic, the Korean government temporarily introduced telehealth services in 2020. Hospitals and clinics were allowed to provide medical consultation and prescribe medication through audio-only based methods and were reimbursed at levels similar to outpatient visits by the National Health Insurance (NHI) accordingly. The purpose of this study was to investigate the effect of the temporary introduction of telehealth services on patterns of health care utilization and medication prescription in individuals diagnosed with hypertension or diabetes mellitus. The hypothesis was that individuals who received in-person and telehealth services concurrently would show a more favorable pattern of health care utilization and medication prescription compared to those who received only in-person outpatient services.

Methods

DATA AND STUDY POPULATION

This study used the 2019 to 2021 claims data from the Health Insurance Review and Assessment Service (HIRA) of the Republic of Korea. All individuals are covered by the NHI or Medical Aid in Korea, in which the NHI comprises ∼97% of the entire population. Under the NHI, HIRA collects all medical claims data from hospitals as it assesses the appropriateness of claimed medical treatments for reimbursement.

The study population included individuals diagnosed with and with records of medication prescription for hypertension (International Classification of Diseases 10th Revision [ICD-10] code: I10) or diabetes mellitus (ICD-19 code: E10–14). Patients with hypertension or diabetes were selected as these individuals were analyzed to have used telehealth services most frequently after policy introduction. Individuals who were newly diagnosed within the study period were excluded as they may show inconsistent patterns of treatment and prescription. The control group included patients without records of telehealth visits, despite having records of filing medical claims for hypertension or diabetes. The case group included those with records of telehealth visits and medical prescription, in addition to ordinary outpatient visits.

OUTCOME VARIABLES

The outcome variable was health care utilization and continuity of medication prescription. Health care utilization was measured based on the number of outpatient visits and the percentage of patients who underwent hospitalizations or visited the emergency department for hypertension, diabetes, and their related complications. The number of outpatient visits referred to visits made for the primary treatment of hypertension or diabetes mellitus, which were identified based on the ICD-10 codes of the main diagnosis. The percentage of patients who underwent hospitalizations or visited the emergency department was also examined because hypertension, diabetes, and their related complications are categorized as ambulatory care sensitive conditions (ACSC).¹⁰ Hospitalizations and emergency department visits for these conditions can be regarded as being potentially preventable through the provision of quality primary health care services.¹¹

Continuity of medication prescription was measured using the medication possession ratio (MPR) and appropriate medication prescription, in which results were expressed based on percentage increases in MPR and appropriate medication prescription. Medical prescription is an important indicator in chronic disease patients as adherence to the provided prescription guidelines is essential considering that maintaining treatment continuity is needed to prevent unnecessary hospitalizations and disease exacerbation.¹²

Hence, prescription continuity is an important indicator to determine the extent to which medications are taken steadily in the treatment of chronic diseases and has been measured using MPR and appropriate medication prescription.¹³ The MPR was calculated by dividing the sum of the days' supply for all drug fills by the number of days in the period and multiplying it by 100 to obtain a percentage. Adequate medication adherence is generally defined as an MPR of 0.8 or above.¹⁴ Appropriate medication prescription was defined as having an MPR of 80–110%. Previous studies have reported that an MPR of at least 80% is needed to achieve the targeted effect of treatment, whereas an upper limit of 110% can be applied to reduce unnecessary costs and hospitalizations resulting from the adverse effects of medication.^15,16

STUDY DESIGN AND STATISTICAL ANALYSIS

This study aimed to evaluate how the temporary introduction of audio-based telehealth service impacted patterns of health care utilization and medication prescription in patients with hypertension or diabetes mellitus using a difference-in-difference (DID) approach. The DID method can be utilized to assess the before and after effect of policy implementation through the modeling of an outcome before intervention.

The differences between groups can be calculated to obtain a counterfactual, beyond which any additional effect can be attributed to the introduced policy and interpreted as the average effect of the policy on the case group.^17
–19 As the audio-only telehealth service was temporarily implemented on February 24, 2020, the pre-intervention period was from February 24, 2019, to February 23, 2020, and the post-intervention period from February 24, 2020, to February 23, 2021. The control group included patients diagnosed with hypertension or diabetes, who did not receive telehealth services (only in-person outpatient visits), whereas the case group included those who received both telehealth and in-person outpatient services, in which all individuals in the case group received remote prescription. The DID specification of the analyses were the following: $O u t c o m e m e a s u r e_{i t} = b_{0} + b_{1} P o l i c y + b_{2} C a s e + b_{3} (P o l i c y * C a s e) + b_{4} X_{i t} + e .$

To assure the internal validity of the study population, the propensity score matching method was applied in the selection of controls. This method enables an observational study to be designed and analyzed in a way that imitates the characteristics of a randomized controlled trial.²⁰ The propensity score was generated based on five variables, namely sex (male or female), age group (groups of 5 years), type of health care insurance (NHI or Medical Aid), the Charlson Comorbidity Index (0, 1, 2, or 3+),²¹ and region (the 17 administrative districts of Korea). The greedy matching technique was utilized, which conducts propensity score matching within the caliper range of 0.25.^22,23 Results were expressed based on the adjusted difference in change and p-values. p-Values were considered significant at p < 0.05. All analyses were conducted using the SAS software, version 9.4 (Cary, NC, USA).

ETHICAL APPROVAL

This study was approved by the Institutional Review Board of the Health Insurance Review & Assessment Service (IRB No.: 2021072-001).

Results

The general characteristics of the study population are presented in Tables 1 and 2. After propensity score matching, the final study population included 250,640 patients with hypertension, in which 125,320 individuals belonged to the control and case groups. Similarly, a total of 154,212 patients with diabetes were included, in which 77,106 individuals were assigned to the control and case groups.

Table 1.

General Characteristics of the Study Participants with Hypertension

VARIABLES	BEFORE MATCHING ( N = 4,923,841)				AFTER MATCHING ( N = 250,640)
VARIABLES	CASE ( N = 126,323)		CONTROL ( N = 4,797,518)		CASE ( N = 125,320)		CONTROL ( N = 125,320)
	N	%	N	%	N	%	N	%
Sex
Male	60,336	47.8	2,345,501	48.9	59,824	47.7	59,821	47.7
Female	65,987	52.2	2,452,017	51.1	65,496	52.3	65,499	52.3
Age
<30	260	0.2	10,650	0.3	254	0.1	251	0.1
31–35	571	0.5	20,294	0.4	564	0.5	564	0.5
36–40	1,921	1.5	68,308	1.4	1,890	1.5	1,891	1.5
40–45	4,648	3.7	152,209	3.2	4,594	3.7	4,596	3.7
46–50	10,089	8	321,775	6.7	10,014	8	10,014	8
51–55	15,934	12.6	503,556	10.5	15,782	12.6	15,782	12.6
56–60	20,911	16.6	703,269	14.7	20,756	16.6	20,756	16.6
61–65	20,671	16.4	783,518	16.3	20,498	16.4	20,499	16.4
66–70	14,794	11.7	649,642	13.5	14,681	11.7	14,681	11.7
71–75	11,434	9.1	563,892	11.8	11,355	9.1	11,355	9.1
76–80	10,062	8	490,882	10.2	9,994	8	9,994	8
81–85	8,346	6.6	329,405	6.9	8,298	6.6	8,298	6.6
86+	6,682	5.3	200,118	4.2	6,640	5.3	6,639	5.3
Type of health care insurance
NHI	120,651	95.5	4,566,313	95.2	119,681	95.5	119,681	95.5
MA	5,672	4.5	231,205	4.8	5,639	4.5	5,639	4.5
Region
Seoul	17,166	13.6	849,805	17.7	17,155	13.7	17,155	13.7
Incheon	7,195	5.7	278,170	5.8	7,189	5.7	7,189	5.7
Gyeonggi	27,123	21.5	1,171,421	24.4	27,109	21.6	27,109	21.6
Gangwon	3,279	2.6	178,995	3.7	3,278	2.6	3,278	2.6
Sejong	559	0.4	24,819	0.5	558	0.4	557	0.4
Daejeon	3,427	2.7	132,846	2.8	3,424	2.7	3,424	2.7
Chungbuk	4,337	3.4	165,193	3.4	4,336	3.5	4,336	3.5
Chungnam	5,378	4.3	219,781	4.6	5,378	4.3	5,379	4.3
Gwangju	5,377	4.3	107,898	2.2	5,372	4.3	5,372	4.3
Jeonbuk	4,560	3.6	193,938	4	4,560	3.6	4,560	3.6
Jeonnam	3,227	2.6	211,365	4.4	3,226	2.6	3,226	2.6
Daegu	13,958	11	209,082	4.4	13,957	11.1	13,955	11.1
Ulsan	2,393	1.9	93,135	1.9	2,391	1.9	2,391	1.9
Busan	7,830	6.2	316,124	6.6	7,825	6.2	7,825	6.2
Gyeongbuk	11,727	9.3	266,917	5.6	11,725	9.4	11,727	9.4
Gyeongnam	7,004	5.5	294,720	6.1	7,003	5.6	7,004	5.6
Jeju	834	0.7	56,419	1.2	834	0.7	833	0.7
—	949	0.8	26,890	0.6	0	0	0	0
CCI
0	84,835	67.2	3,090,278	64.4	84,216	67.2	84,216	67.2
1	27,575	21.8	1,091,047	22.7	27,350	21.8	27,349	21.8
2	9,655	7.6	408,697	8.5	9,583	7.6	9,585	7.6
3	4,204	3.3	202,924	4.2	4,171	3.3	4,170	3.3
—	54	0	4,572	0.1	0	0	0	0

CCI, Charlson Comorbidity Index; MA, Medical Aid; NHI, National Health Insurance.

Table 2.

General Characteristics of the Study Participants with Diabetes Mellitus

VARIABLES	BEFORE MATCHING ( N = 2,510,701)				AFTER MATCHING ( N = 154,212)
	CASE ( N = 77,699)		CONTROL ( N = 2,433,002)		CASE ( N = 77,106)		CONTROL ( N = 77,106)
	N	%	N	%	N	%	N	%
Sex
Male	42,284	54.4	1,351,931	55.6	41,944	54.4	41,944	54.4
Female	35,415	45.6	1,081,071	44.4	35,162	45.6	35,162	45.6
Age
<30	718	1	21,977	1	713	1	713	1
31–35	662	0.9	17,830	0.7	651	0.8	651	0.8
36–40	1,734	2.2	44,445	1.8	1,717	2.2	1,716	2.2
40–45	3,348	4.3	83,085	3.4	3,325	4.3	3,325	4.3
46–50	6,207	8	156,419	6.4	6,163	8	6,163	8
51–55	9,519	12.3	242,501	10	9,430	12.2	9,430	12.2
56–60	12,160	15.7	345,527	14.2	12,056	15.6	12,056	15.6
61–65	12,611	16.2	396,083	16.3	12,517	16.2	12,518	16.2
66–70	9,633	12.4	346,602	14.2	9,555	12.4	9,555	12.4
71–75	7,504	9.7	298,931	12.3	7,451	9.7	7,451	9.7
76–80	6,789	8.7	258,867	10.6	6,757	8.8	6,757	8.8
81–85	4,580	5.9	154,280	6.3	4,546	5.9	4,546	5.9
86+	2,234	2.9	66,455	2.7	2,225	2.9	2,225	2.9
Type of health care insurance
NHI	73,048	94	2,261,232	92.9	72,482	94	72,482	94
MA	4,651	6	171,770	7.1	4,624	6	4,624	6
Region
Seoul	10,599	13.6	409,886	16.8	10,590	13.7	10,590	13.8
Incheon	4,082	5.3	142,736	5.9	4,078	5.3	4,079	5.3
Gyeonggi	17,546	22.6	583,113	24	17,529	22.7	17,529	22.7
Gangwon	1,766	2.3	90,311	3.7	1,763	2.3	1,763	2.3
Sejong	297	0.4	11,188	0.5	297	0.4	297	0.4
Daejeon	2,071	2.7	67,422	2.8	2,070	2.7	2,070	2.7
Chungbuk	2,350	3	83,045	3.4	2,350	3	2,350	3
Chungnam	2,801	3.6	112,074	4.6	2,800	3.6	2,800	3.6
Gwangju	3,085	4	60,599	2.5	3,085	4	3,085	4
Jeonbuk	2,740	3.5	101,026	4.2	2,739	3.6	2,739	3.6
Jeonnam	2,504	3.2	111,065	4.6	2,504	3.2	2,503	3.2
Daegu	7,803	10	108,529	4.5	7,803	10.1	7,803	10.1
Ulsan	1,927	2.5	46,064	1.9	1,926	2.5	1,926	2.5
Busan	4,308	5.5	169,630	7	4,305	5.6	4,305	5.6
Gyeongbuk	8,137	10.5	138,828	5.7	8,135	10.6	8,135	10.5
Gyeongnam	4,635	6	154,458	6.3	4,635	6	4,635	6
Jeju	497	0.6	27,007	1.1	497	0.6	497	0.6
—	551	0.7	16,021	0.7	0	0	0	0
CCI
0	53,500	68.9	1,557,547	64	53,123	68.9	53,123	68.9
1	15,952	20.5	543,049	22.3	15,839	20.5	15,839	20.5
2	5,718	7.4	219,631	9	5,667	7.3	5,667	7.4
3	2,487	3.2	110,518	4.5	2,477	3.2	2,477	3.2
—	42	0.1	2,257	0.1	0	0	0	0

The effect of the temporary introduction of audio-only telehealth services on patterns of health care utilization is shown using a DID estimate in Table 3. With regard to the number of outpatient visits, introduction of the audio-only telehealth services was associated with a 0.07 (p = 0.0027) increase in the number of visits in patients with hypertension and a 0.32 (p < 0.0001) increase in patients with diabetes. The use of telehealth services was also associated with a decrease in hospitalizations (−0.2%, p = 0.0007) and emergency department visits only in patients with hypertension (−0.11%, p = 0.0016) as those diagnosed with diabetes showed statistically insignificant trends.

Table 3.

Difference in Difference Analysis of Introducing Audio-Based Telehealth on Patterns of Health Care Utilization

DISEASE	GROUP	OUTPATIENT VISITS ( N )			DID ESTIMATE	p	HOSPITALIZATIONS (%)			DID ESTIMATE	p	EMERGENCY DEPARTMENT VISITS (%)			DID ESTIMATE	p
DISEASE	GROUP	BEFORE	AFTER	DIFFERENCE	DID ESTIMATE	p	BEFORE	AFTER	DIFFERENCE	DID ESTIMATE	p	BEFORE	AFTER	DIFFERENCE	DID ESTIMATE	p
Hypertension	Control	11.2	7.7	−3.4	0.07	0.0027	1.1	1.7	0.6	−0.2	0.0007	0.3	0.6	0.24	−0.11	0.0016
Hypertension	Case	12.2	8.8	−3.4	0.07	0.0027	1	1.4	0.4	−0.2	0.0007	0.3	0.4	0.13	−0.11	0.0016
Diabetes	Control	12	8.2	−3.8	0.32	<0.0001	4.3	4.4	0.1	0.0	0.8750	1.1	1.2	0.1	0.0	0.9860
Diabetes	Case	13.1	9.6	−3.5	0.32	<0.0001	4	4.1	0.1	0.0	0.8750	1	1.2	0.1	0.0	0.9860

DID, difference-in-difference.

Likewise, the effect of telehealth services on patterns of medication prescription is demonstrated in Table 4. In terms of MPR, policy implementation resulted in an overall increase in MPR in both patients with hypertension (3.0%, p < 0.0001) and diabetes (3.4%, p < 0.0001). Similarly, policy implementation was related to an escalation in the percentage of hypertension (3.1%, p < 0.0001) and diabetes (1.7%, p < 0.0001) patients who received appropriate medication prescription.

Table 4.

Difference in Difference Analysis of Introducing Audio-Based Telehealth on Medication Prescription

DISEASE	GROUP	MEDICATION POSSESSION RATIO (%)			DID ESTIMATE	p	PROPORTION OF PATIENTS WHO RECEIVED APPROPRIATE MEDICATION PRESCRIPTION (%)			DID ESTIMATE	p
DISEASE	GROUP	BEFORE	AFTER	DIFFERENCE	DID ESTIMATE	p	BEFORE	AFTER	DIFFERENCE	DID ESTIMATE	p
Hypertension	Control	86	84.8	−1.1	3.0	<0.0001	73.4	71	−2.4	3.1	<0.0001
Hypertension	Case	88.1	89.9	1.8	3.0	<0.0001	76.8	77.4	0.7	3.1	<0.0001
Diabetes	Control	84.2	84	−0.3	3.4	<0.0001	68.4	67.1	−1.3	1.7	<0.0001
Diabetes	Case	86.2	89.3	3.2	3.4	<0.0001	72.1	72.5	0.4	1.7	<0.0001

Discussion

This study investigated the effect of the implementation of a nationwide audio-based telehealth services on health care utilization and continuity of medication prescription in patients diagnosed with hypertension or diabetes using a DID estimate. The results revealed that individuals diagnosed with hypertension or diabetes, who utilized both in-person and telehealth services, show an increase in outpatient visits after policy implementation than those who continued to receive only in-person services. These individuals were also more likely to report an increase in MPR and receive appropriate medication prescription. In the case of hospitalizations and emergency department visits, only patients with hypertension showed a decrease.

Concerning patterns of health care utilization, the findings present that telehealth services have led to an increase in outpatient visits. A decrease in hospitalizations and emergency department visits was also found, although the reductions were only significant in patients with hypertension. Such tendencies suggest that the availability of telehealth service may exert a positive effect on the utilization of health care services because continuous primary care is important in managing chronic diseases and preventing its related complications. As individuals are required to visit health care institutions over long periods to manage chronic conditions so that they can be monitored and equipped with required self-management skills, primary care is a well-suited method to provide continuity and coordination, improve health outcomes, and promote cost-effectiveness.^24,25

Considering that the Korean health care system does not have a strong referral system, allowing patients to visit all levels of health care facilities easily, an increase in outpatient visits and a decrease in hospitalizations and emergency department visits may infer a more efficient use of health care resources.^26,27 The fact that hypertension and diabetes are ACSC, which means that timely and adequate treatment delivered in an outpatient setting can potentially circumvent the requirement for hospitalizations, further implies that telehealth services can be an efficient source of medium to manage chronic diseases.²⁸ Emergency department visit is also an indicator of restricted access to primary care, in which inadequate care can lead to less favorable clinical outcomes, while inducing higher health care spending.²⁹

The results also reveal that telehealth utilization, which involved prescriptions, correlated with rises in MPR and the proportion of patients who received appropriate medication prescriptions. The findings are noteworthy because a higher MPR, in specific an MPR of 80% or above, is reported as being adequate to effectively manage chronic diseases.¹⁵ In addition, appropriate medication prescription, referring to an MPR of 80–110%, is an important indicator as it infers a sufficient, but not an oversupply of prescriptions.

Specifically, low adherence to antihypertensive drugs has been cited as a major contributing factor to the suboptimal control of chronic diseases and increasing health care costs.^30,31 Appropriate adherence to prescribed antidiabetic medicine has also been reported as an essential factor in managing hemoglobin A1c levels.³² Since outpatients with chronic diseases often take medications for extended periods, while prescriptions for these drugs generally remain stable unless the patient experiences an unstable condition or adverse event, the use of telehealth may have been particularly feasible in managing hypertension and diabetes when access to health care institutions became difficult.³³

The COVID-19 outbreak has created an unprecedented opportunity for telehealth to be reimbursed and plays a larger role in the health care system.³⁴ This study demonstrates that providing an even audio-based telehealth service may provide benefits in improving patterns of health care utilization and medication prescription in patients with hypertension or diabetes, who require continuous management. The utilization of telehealth may decrease as the pandemic evens out, with the possible preference of in-person consultations over audio visits by health care providers and patients and the cessation of health care policies that promoted telehealth services. Yet the findings highlight potentially important inferences on the benefits of telehealth services in the future by suggesting that audio-based visits can enhance treatment adherence and lead to improved health outcomes.

This study has some limitations. First, this study analyzed only the short-term impact of the policy on telehealth and does not include a longer time span. Second, the investigated type of telehealth service was limited to audio-only technology and does not extend to video- or app-based technology. Third, hypertension and diabetes were identified based on ICD-10 codes as this study used claims data. Hence, not all individuals with blood pressure or glycemia-related abnormalities may have been identified. However, despite the limitations stated above, this study is unique in that it is the first to examine the effect of utilizing telehealth services on patterns of health care utilization and medication prescription using large, nationwide data in Korea during the COVID-19 pandemic.

Conclusions

The introduction of audio-based telehealth consultation and prescription services was associated with an improvement in patterns of health care utilization and medication prescription in patients with hypertension or diabetes. Specifically, policy implementation resulted in an increase in outpatient visits, MPR, and the proportion of individuals receiving appropriate medication prescription. A decrease in hospitalizations and emergency department visits was also found in patients with hypertension. The findings together reveal the potential benefit of telehealth on enhancing treatment adherence and improving health outcomes in patients with chronic diseases.

Footnotes

Acknowledgments

We would like to thank the Health Insurance Review & Assessment of the Republic of Korea for providing data. The institutions played no role in the study design, analysis, and interpretation.

Authors' Contributions

Conception and design: M.C., W.K., M.K., R.Y., Y.H., and J.S.; data collection and analysis: M.C.; data interpretation: M.C., W.K., M.K., R.Y., Y.H., D.W.L., and J.S.; article writing: M.C. and W.K.; and article approval: M.C., W.K., M.K., R.Y., Y.H., D.W.L., and J.S.

Disclosure Statement

No competing financial interests exist.

Funding Information

No funding was received for this article.

References

Rutledge

, Kott

, Schweickert

, et al. Telehealth and eHealth in nurse practitioner training: Current perspectives. Adv Med Educ Pract, 2017; 8:399–409; doi: 10.2147/AMEP.S116071

Balestra

. Telehealth and legal implications for nurse practitioners. J Nurse Pract, 2018; 14(1):33–39; doi: 10.1016/j.nurpra.2017.10.003

Gilman

, Stensland

. Telehealth and Medicare: Payment policy, current use, and prospects for growth. Medicare Medicaid Res Rev, 2013; 3(4):mmrr.003.04.a04; doi: 10.5600/mmrr.003.04.a04

Gajarawala

, Pelkowski

. Telehealth benefits and barriers. J Nurse Pract, 2021; 17(2):218–221; doi: 10.1016/j.nurpra.2020.09.013

Yang

, Kozhimannil

. Medication abortion through telemedicine: Implications of a ruling by the Iowa Supreme Court. Obstet Gynecol, 2016; 127(2):313–316; doi: 10.1097/AOG.0000000000001251

De Vos

. The effect of COVID-19 and subsequent social distancing on travel behavior. Transp Res Interdiscip Perspect, 2020; 5:100121; doi: 10.1016/j.trip.2020.100121

Hou

, Lai

, Ben-Ezra

, et al. Regularizing daily routines for mental health during and after the COVID-19 pandemic. J Glob Health, 2020; 10(2):020315; doi: 10.7189/jogh.10.020315

Kim

. Changes and implications of health insurance expenses before and after the COVID-19. National Assembly Budget Office, 2020; 26:1–4.

, Huh

, Kang

, et al. Effect of underlying comorbidities on the infection and severity of COVID-19 in Korea: A Nationwide Case-Control Study. J Korean Med Sci, 2020; 35(25):e237; doi: 10.3346/jkms.2020.35.e237

10.

Kuecker

, Kashyap

, Seckel

. Implementation of a protocol to manage patients at risk for hospitalization due to an ambulatory care sensitive condition. Fed Pract, 2020; 37(8):380–383; doi: 10.12788/fp.0030

11.

Loyd

, Blue

, Turner

, et al. National norms for hospitalizations due to ambulatory care sensitive conditions among adults in the US. J Gen Intern Med, 2023; 38(13):2953–2959; doi: 10.1007/s11606-023-08161-z

12.

Bosworth

, Granger

, Mendys

, et al. Medication adherence: A call for action. Am Heart J, 2011; 162(3):412–424; doi: 10.1016/j.ahj.2011.06.007

13.

Lam

, Fresco

. Medication adherence measures: An overview. Biomed Res Int, 2015; 2015:217047; doi: 10.1155/2015/217047

14.

Ali

, Kerr

, Lee

, et al. What is the statin possession ratio cut-off which identifies those at increased risk following acute coronary syndrome. Heart Lung Circ, 2023; 32(4):487–496; doi: 10.1016/j.hlc.2022.11.012

15.

Stroupe

, Teal

, Tu

, et al. Association of refill adherence and health care use among adults with hypertension in and urban health care system. Pharmacotherapy, 2006; 26(6):779–789; doi: 10.1592/phco.26.6.779

16.

Stroupe

, Teal

, Weiner

, et al. Health care and medication costs and use among older adults with heart failure. Am J Med, 2004; 116(7):443–450; doi: 10.1016/j.amjmed.2003.11.016

17.

Hyun

, Kang

, Lee

. Does long-term care insurance affect the length of stay in hospitals for the elderly in Korea?: A difference-in-difference method. BMC Health Serv Res, 2014; 14(1):1–7; doi: 10.1186/s12913-014-0630-1

18.

Dimick

, Ryan

. Methods for evaluating changes in health care policy: The difference-in-differences approach. JAMA, 2014; 312(22):2401–2402; doi: 10.1001/jama.2014.16153

19.

Stuart

, Huskamp

, Duckworth

, et al. Using propensity scores in difference-in-differences models to estimate the effects of a policy change. Health Serv Outcomes Res Methodol, 2014; 14(4):166–182; doi: 10.1007/s10742-014-0123-z

20.

Austin

. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Mutlivariate Behav Res, 2011; 46(3):399–424; doi: 10.1080/00273171.2011.568786

21.

D'Hoore

, Bouckaert

, Tilquin

. Practical considerations on the use of the Charlson comorbidity index with administrative data bases. J Clin Epidemiol, 1996; 49(12):1429–1433; doi: 10.1016/s0895-4356(96)00271-5

22.

Austin

. Optimal caliper widths for propensity-score matching when estimating differences in means and differences in proportions in observational studies. Pharm Stat, 2011; 10(2):150–161; doi: 10.1002/pst.433

23.

Austin

. A comparison of 12 algorithms for matching on the propensity score. Stat Med, 2014; 33(6):1057–1069; doi: 10.1002/sim.6004

24.

Rijken

, Bekkema

, Boeckxstaens

, et al. Chronic Disease Management Programmes: An adequate response to patients' needs?. Health Expect, 2014; 17(5):608–621; doi: 10.1111/j.1369-7625.2012.00786.x

25.

Lall

, Engel

, Srinivasan

, et al. Improving primary care for diabetes and hypertension: Findings from implementation research in rural South India. BMJ Open, 2020; 10(12):e040271; doi: 10.1136/bmjopen-2020-040271

26.

Kim

, Cho

, Kim

, et al. Primary care patients' preference for hospitals over clinics in Korea. Int J Environ Res Public Health, 2018; 15(6):1119; doi: 10.3390/ijerph15061119

27.

Song

. The South Korean health care system. Japan Med Assoc J, 2009; 52(3):206–209;

28.

Rocha

JVM

, Santana

, Tello

. Hospitalization for ambulatory care sensitive conditions: What conditions make inter-country comparisons possible?. Health Policy Open, 2021; 2:100030; doi: 10.1016/j.hpopen.2021.100030

29.

Lee

, Sung

, Lim

, et al. Emergency department visits can be reduced by having a regular doctor for adults with diabetes mellitus: Secondary analysis of 2013 Korea Health Panel Data. J Korean Med Sci, 2017; 32(12):1921–1930; doi: 10.3346/jkms.2017.32.12.1921

30.

Mann

, Allegrante

, Natarajan

, et al. Predictors of adherence to statins for primary prevention. Cardiovasc Drugs Ther, 2007; 21(4):311–316; doi: 10.1007/s10557-007-6040-4

31.

Zullig

, Gellad

, Moaddeb

, et al. Improving diabetes medication adherence: Successful, scalable interventions. Patient Prefer Adherence, 2015; 9:139–149; doi: 10.2147/PPA.S69651

32.

Bailey

, Kodack

. Patient adherence to medication requirements for therapy of type 2 diabetes. Int J Clin Pract, 2011; 65(3):314–322; doi: 10.1111/j.1742-1241.2010.02544.x

33.

Kurata

, Onuki

, Yoshizumi

, et al. Proportion of Japanese outpatients filling prescriptions for long-term medication regimens. Patient Prefer Adherence, 2019; 13:667–673; doi: 10.2147/PPA.S191512

34.

Smith

, Thomas

, Snoswell

, et al. Telehealth for global emergencies: Implications for coronavirus disease 2019 (COVID-19). J Telemed Telecare, 2020; 26(5):309–313; doi: 10.1177/1357633X20916567