INTRODUCTION

Type 2 diabetes mellitus (T2DM) is a chronic metabolic disorder characterised by insulin resistance and insufficient pancreatic insulin production resulting in abnormally high blood glucose levels that persists over time (Ministry of Health Malaysia 2020). The T2DM is commonly recognised as the world’s fastest-growing chronic disease. Globally, diabetes is estimated to impact approximately 463 million individuals, with a projected increase to 700 million by 2045 (Saeedi et al. 2019). In Malaysia, according to the 2019 National Health and Morbidity Survey, the prevalence of diabetes among Malaysian adults aged 18 and above have increased from 11.2% in 2011 to 18.3% in 2019 (Institute for Public Health 2020). 

Over time, diabetic patients may develop long-term macrovascular and microvascular complications such as retinopathy, nephropathy, sexual dysfunction, and diabetic neuropathy (Fowler 2008; Gregg & Bracco 2019). Additionally, uncontrolled T2DM increases the risk of developing conditions predisposing patients to hospitalisation, such as cardiovascular diseases, stroke, nephropathy, infections, and lower-extremity amputations (Khalid et al. 2014). Diabetes was reported to directly result in 1.6 million deaths worldwide in 2016. In Malaysia, approximately 35% of the total mortality was related to diabetes (World Health Organisation 2018). The rising number of diabetic patients places a significant strain on the healthcare system and society. 

The mainstay of T2DM management involves multifactorial strategies, including diet control, exercise, glucose monitoring, weight control, and pharmacological intervention (Glasgow & Strycker 2000). In the most newly diagnosed T2DM cases, the patients are advised for lifestyle modification. If the blood glucose level remains suboptimal, the patients will be prescribed with oral anti-diabetic agents in order to achieve normoglycemia and delay the progression of long-term diabetes complications (Ministry of Health Malaysia 2020). However, as T2DM is a progressive disease, many diabetic patients may eventually require insulin to ensure adequate glycemic control (American Diabetes Association 2018). 

Insulin is a glucose-lowering hormone produced by the pancreas. It is responsible for regulating the body’s carbohydrate, fat, and protein metabolism (American Diabetes Association 2004). As there is presently no cure for diabetes, certain T2DM patients may require insulin in conjunction with oral hypoglycemic medications if the previous treatment did not effectively control the glycemic level. There is a range of insulin injections available. They vary in terms of the onset and the duration of its effects in the body (Ministry of Health Malaysia 2020). In Malaysia, the most often used forms of insulin are short-acting, intermediate-acting, and premixed insulin. Additionally, analogue insulin is available in rapid-acting, long-acting, and premixed forms. At present, most insulin supplied to patients is available to be used with insulin pens. 

Long-term use of pharmacotherapy is common in the treatment of chronic conditions. While these medications are efficient at combating disease, their full benefits are often not achieved if patients do not take the drugs as prescribed. Unfortunately, poor adherence to diabetes medications, particularly insulin, is still a concern. Furthermore, a systematic review revealed glycemic control was deteriorated among T2DM patients during lockdown period of COVID-19 pandemic (Eberle & Stichling 2021). 

Identifying the level of insulin adherence is crucial for improving glycemic control. There are various methods for determining drug adherence including patients’ self-reports, pill counts, electronic medicine monitoring, therapeutic effect measuring, and pharmacy refill data (Farley et al. 2003). As there is no gold standard to measure drug adherence, reviewing medication refills from pharmacy claims data can be beneficial. Refill adherence can be determined based on the concept that patients cannot comply if they do not have enough drugs (Hawkshead & Krousel-Wood 2007).

In Malaysia, after consultations from public health clinics, most of the chronic illness patients will receive a prescription that will last between one and six months, depending on their condition. Since 2011, the Quality Use of Medicine (QUM) Pharmacy Practice Guideline has mandated drugs for prescriptions that last longer than one month to be given monthly (Bahri & Ibrahim 2014). As a result, patients will be required to get their medications, including insulin, at the health facility once a month. The Pharmacy Information System (PhIS) is a comprehensive and integrated system managing pharmacy-related services in most public healthcare settings in Malaysia. Prescription refill records from this centralised system can be a valuable data source to measure patient adherence to chronic medications (Steiner et al. 1988). 

Additionally, following to the implementation of the first phase of the movement control order in Malaysia on 18 March 2020 in response to the COVID-19 pandemic, patients’ access to healthcare facilities for routine care and medication management may have been limited during the lockdown period due to rigorous border control and quarantine efforts (Desvars-Larrive et al. 2020). Despite its evident importance, the literature on patients’ insulin refill adherence during the COVID-19 pandemic is relatively scarce. Therefore, the current study aimed to assess insulin refill adherence among diabetes patients in health clinics in Perlis before and during the COVID-19 pandemic and explore factors related to nonadherence. 

MATERIALS and METHODS

This cross-sectional study was conducted from 1st January 2022 to 30th June 2022. This study was registered with the National Medical Research Register (NMRR ID-22-00227-KJZ) with the Medical Research and Ethics Committee approval. Permission to carry out this study and access to the PhIS data was also obtained from the Kangar District Health Officer before conducting this study.

Five health clinics were selected based on different geographical areas in Perlis, i.e. Padang Besar Health Clinic in the north, Kuala Perlis Health Clinic in the west, Arau Health Clinic in the east, Kuala Sanglang Health Clinic in the south, and UTC Health Clinic in the central area of Perlis. The current study involved T2DM patients who had been prescribed with insulin therapy from primary health care centres in Perlis. The inclusion criteria were T2DM patients aged 18 years old and above using insulin therapy during the period from 1st September 2019 to 30th September 2020. Pregnant patients with gestational diabetes mellitus were excluded from this study. 

The sample size estimation for the first specific objective was calculated using population proportion formulae (Lemeshow et al. 1990). According to a previous study, the prevalence of insulin adherence was 57.0% (Chew et al. 2015). If Type 1 error probability was set at 5% with a statistical power of 95%, a total of 318 samples was needed. The final sample size was 398 after considering the 20% dropout rate. A list of eligible patients was generated from the electronic system (PhIS). Then, simple random sampling method was used for participant selection in this study. Random numbers were generated from Microsoft Excel to create a sampling frame for participant selection.

A standardised data collection form was formulated to assess patients’ demographic data (age, gender, and ethnic), comorbidities, insulin-related information (types of insulin used, insulin regimen, total daily insulin dose, and the number of daily injections), and insulin refill records before (September 2019 to February 2020) and during (April to September 2022) the COVID-19 pandemic. Adherence to insulin refill were measured by medication possession ratio (MPR) (Stolpe et al. 2016) based on prescription data obtained from the PhIS. Patients were considered adherent at an MPR≥80% (Stolpe et al. 2016). 

MPR =            x 100%

The distribution of continuous variables (MPR score, age, and total daily insulin dose) were assessed by using skewness and kurtosis (Kim 2013). Then, student's t-test, Pearson’s Correlation, and One-way ANOVA were performed in univariable analysis. All variables with p<0.25 from the univariable analysis were included for downstream analysis (Zhang 2016). A general linear model (multiple linear regression) was used to identify factors associated with adherence to insulin refills by adjusting the confounding variables. A p-value of less than 0.05 was considered statistically significant. All analyses were performed in SPSS software version 21.0 (IBM Corp, Armonk, NY). 

RESULTS 

A total of 426 patients were included in the final analysis. About two-thirds of the patients were between the ages of 51 and 70. The majority of this study’s patients were female (63.1%) and Malay (94.3%). Premixed (41.8%) and basal-bolus (44.1%) insulin were prescribed for most patients. Half of the patients had a total daily insulin dose of less than 50 units, while 41.5% of patients used insulin doses between 51 and 100 units. Analogue insulin was utilised by only six patients. A considerable proportion of patients were also diagnosed with hypertension (83.6%) and dyslipidemia (92.5%). The sociodemographic characteristics and insulin-related information of patients were shown in Table 1. 

Insulin refill adherence was also found to be significantly poorer during the COVID-19 pandemic (mean=59.24, SD=28.97) than before the pandemic (mean=68.31, SD=31.27) (p<0.001) (Table 2). About 40% (n=170) of participants had good adherence to their insulin refills (MPR≥80%) prior to the COVID-19 pandemic. However, only 122 individuals (28.6%) adhered well to their insulin refill schedule during the COVID-19 pandemic (Table 3).

Table 4 showed the association between sociodemographic data, comorbidities, and insulin-related information towards insulin refill adherence. Univariate analysis revealed three factors (insulin regimen used, total daily insulin, and hypertension) were significantly associated with insulin refill adherence. Only total daily insulin dose (adjusted β=-0.129, p=0.012) and not having hypertension (adjusted β=-7.359, p=0.043) were significantly associated with insulin refill adherence in the general linear model after adjusting for confounding variables (Table 5).  

DISCUSSION

Globally, poor adherence to insulin regimens is common among diabetic patients (Chew et al. 2015). Nonadherence to insulin was found in most countries, including Japan (44%), the United States (42%), the United Kingdom (41%), Germany (40%), China (33%), Turkey (24%), Spain (23%) and France (19%) (Doggrell & Chan 2015). Several studies conducted in Malaysia among diabetic patients found about 41% to 53% of patients were nonadherent (Ahmad et al. 2013; Chew et al. 2015; Chua & Chan 2011). The current study examining patients’ refill records revealed that 60.1% of patients had a suboptimal level of insulin refills during the study period.

Furthermore, their insulin refill adherence rate was much lower during the initial period of the COVID-19 pandemic in Malaysia. Several episodes of lockdowns and movement restrictions during the initial phase of the pandemic were major problems for patients with chronic diseases requiring follow-ups and prescription refills (Kretchy et al. 2021). Nonadherence to prescribed regimens is associated with suboptimal treatment outcome, higher risk of hypo- or hyperglycemia, and significantly increased mortality rates (Polonsky & Henry 2016). The pandemic has highlighted the urgent need for medical compliance to be reconceptualised.

Consistent with earlier research findings, gender and race have not been seen to be associated with the level of medication adherence (Ahmad et al. 2013; Mariye et al. 2019; Peyrot et al. 2010). Most studies found that younger adults were more likely to be nonadherent to their medication (Ahmad et al. 2013; Chew et al. 2015). Some patients intentionally skipped taking their insulin doses for weight control (Peyrot et al. 2010). Interestingly, this study found no significant difference in patients’ age. 

Insulin nonadherence normally occurs when the treatment regimen is complicated, whether in terms of administration method, scheduling, or storage (Jimmy & Jose 2011). Those with complex insulin regimes, particularly with a greater number of insulin injections per day, had a higher rate of insulin injection omission (Peyrot et al. 2010). Nevertheless, this study did not discover that the types of insulin regimens significantly affected the degree of patient adherence. On the other hand, this study found that patients with greater daily insulin doses had poor refill adherence levels. A larger daily insulin dose may result in a higher rate of minor hypoglycemia and, therefore, a higher likelihood of skipping an insulin dose (Yavuz et al. 2015). Consequently, these patients may have excess insulin supply, resulting in a lower rate of insulin refills. 

Previous literature conducted among T2DM patients attending public health clinics in Malaysia found that patients’ medication adherence was lower if they had concurrent comorbidities such as hypertension and dyslipidemia due to complex medication regimens (Ahmad et al. 2013). Interestingly, this study found that T2DM patients with hypertension had higher adherence to insulin refills. These patients could be more health conscious of adhering to therapeutic recommendations and follow-up appointments closely after being diagnosed with additional comorbidities (Chew et al. 2015).

In most countries where diabetes treatment is often from out-of-pocket expense, and treatment involving insulin is costly, financial factors are often associated with nonadherence to insulin (Bermeo-Cabrera et al. 2018; Sharma et al. 2020). However, the financial factor was not examined in this study. In Malaysia, the vision of health is to provide a healthcare system that is equitable and affordable to the population (World Health Organisation 2012). The Ministry of Health is the key healthcare provider in Malaysia, with an extensive network of primary care clinics and large public hospitals in every state. Every citizen of Malaysia is eligible for public healthcare services that are heavily subsidised by the government (Tumin et al. 2016). 

There were a few limitations in the study. This study only obtained data from the pharmacy database (PhIS). The authors may lose information that was unavailable in the system, such as marital status, education levels, monthly household incomes, and duration of diabetes, which could influence their insulin refill adherence. However, the data was more reliable compared to patients’ self-reporting adherence measurement, as the latter may be subjected to reporting bias. Some patients may overreport their adherence for various reasons. 

In early 2022, the Pharmaceutical Services Divisions of Perlis State Health Department introduced a new Pharmacy Value-Added Services (VAS), namely UM3 (Pembekalan Ubat Susulan Melalui Pihak Ketiga). This new service was introduced through the concept of delighting the customer, ensuring continual access to follow-up medication supplies, and preventing congestion in the pharmacy waiting areas. Refill medications, including insulin, can be delivered to a location of the patient’s choice via collaboration with a chosen delivery company to ensure continuity of medicine for all patients with the new norm.  

The current study contributed to the existing literature on patients’ insulin refill adherence, allowing future research to expand based on the findings, including insulin-taking behaviour and the underlying factors contributing to nonadherence. Further studies investigating the true factors associated with insulin nonadherence should be explored either through qualitative in-depth interviews or a case-control study.

CONCLUSION

In conclusion, this study highlighted that overall insulin refill adherence among patients in public health clinics in Perlis was low, particularly during the COVID-19 pandemic. Poor insulin adherence is a significant issue that can lead to suboptimal glucose control and increased utilisation of healthcare resources. Interventions should focus on T2DM patients in health clinics, particularly those who require high total daily insulin doses and those without hypertension, to improve medication adherence and treatment outcomes. 

ACKNOWLEDGEMENT 

The authors would like to thank the Director General of Health for the permission to publish this study.