The nonadherence to prescriptions among type 2 diabetes patients, and its determining factors

The nonadherence to prescriptions


INTRODUCTION
Drug adherence is the degree to which a patient follows and strictly adheres to medical recommendations, such as medication schedule, dose, and frequency [1].When a patient takes prescribed medications at the doses and times indicated by a healthcare professional and with the patient's agreement, they are considered compliant [2,3].
Patients are typically required to strictly adhere to prescribed medical treatment to obtain the desired medical outcomes.From the perspective of disease treatment, the patient's inability to completely comply to the prescribed pharmaceutical regimen may certainly have severe and harmful effects.World Health Organization (WHO) reports that roughly 50.0% of patients do not take their recommended drugs [4].This percentage can be significantly higher in underdeveloped nations due to a variety of reasons, including patient literacy, the prescription of complex medications, and the duration of the sickness [4].For the latter, there is evidence that individuals with chronic illnesses struggle to adhere to their prescribed treatment regimen [5].For example, it was observed that the adherence rate for medications intended to be taken over an extended period fell by almost 50.0%[6].According to [7], disease duration is a significant factor that can influence drug adherence.It was reported that the nonadherence rate was 53.0% for patients with a disease duration of more than one year and 1.0% for patients with a sickness duration of less than one month [7].
Type 2 diabetes mellitus (T2DM), a chronic metabolic condition characterized by high glucose levels in the blood, is projected to affect 642 million patients worldwide by 2040 [8,9].Although the number of diabetes patients in Kazakhstan, a developing landlocked country in central Asia, is unclear, it is believed to be comparable to the global incidence of 12.5% among individuals over 50 [10].
The rate and prevalence of nonadherence to antidiabetic medication in Kazakhstan are unknown and have not been studied.Thus, the purpose of the present study is to assess the rate and prevalence of nonadherence to antidiabetic therapies among T2DM patients, as well as the factors that contribute to this phenomenon.Determining the prevalence of adherence and the factors that influence it can assist practitioners provide patients with better healthcare and more successful treatment regimens.

STUDY DESIGN
After reviewing relevant literature, a questionnaire with three parts (general background, KAP section, and reasons for non-adherence) was developed.The questionnaire was piloted with 22 randomly selected individuals to validate it.Eligible patients were interviewed in person, but due to the pandemic, some patients were interviewed over the phone.Patients' adherence was evaluated using both direct questions about adherence and a summary of their responses to nonadherence statements.

Statistical Evaluation
StataCorp Stata 14.2 software was used for statistical analysis in this study.Initially, descriptive analysis of the given responses was performed, with means, medians, standard deviations, and frequencies calculated.The strength of the association variables was assessed using simple logistic regression.To control for the presence of confounders, multivariate analysis was also performed using logistic regression.The p>|z| value was checked in each analysis to determine the statistical significance of the results.

Patient Qualifying Requirements
The inclusion criteria for the trial were consenting individuals older than 18 years with confirmed T2DM.Patients in significant distress, newly diagnosed individuals, and patients with apparent psychiatric disorders were excluded from the study.

Dependent variables
Dependent variables are non-adherence to prescriptions.

Independent variables
Independent variables are socio-demographic factors, medication-related factors, morbidity-related factors, and patient-provider relationship.

Awareness of Diabetes
Patients' awareness of the condition and treatment was classified as "full knowledge" when they provided all correct responses, and "minimal knowledge" when they provided only erroneous responses or expressly said they did not know."Relatively adequate" and "relatively insufficient" refer to patients who provided more (but not all) or fewer correct answers alongside the wrong ones.
According to patient responses, the explicit rate of adherence to anti-diabetic medicines is 56.6% for full adherence, 10.3% for non-adherence, and 33.0% for partial adherence.Responses such as "adherent but not as prescribed" and "adherent but occasionally skip or forget" were used to measure partial adherence (Table 2).

Attitudes & Habits of Respondents Towards Antidiabetic Medications
Among respondents with comorbidities, hypertension is the most prevalent (60.3%).Thus, nearly three-quarters of respondents (73.3%) reported using prescriptions other than anti-diabetic agents, including 36.6% who take vitamins and 27.3% who use herbal products for therapeutic purposes (Table 3).

Influencing Factors for Nonadherence
Statements were formulated so that "strongly disagree" range of replies showed that the statement in question is not a reason for a certain patient's non-adherence, whereas "strongly agree" range might be a reason for a patient to omit drugs.Medication cost was the lone exception, as many patients get their medications through the hospital without having to purchase them, and hence tended to respond, "not sure."The healthcare system, notably delays in medicine availability from the hospital, was also indicated as a contributing factor to inadequate adherence.However, this delay appeared to depend on when the survey was conducted.For instance, respondents surveyed prior to the New Year's holiday tended to report no problems with hospital drug discharge delays, whereas respondents surveyed after the holiday tended to complain about the lengthy absence of their medications.Overall, the expense of medications appears to be the most influential factor in nonadherence (Table 4).

Bivariable Tests
Using StataCorp Stata 14.2 to evaluate various determinants of adherence demonstrates that various factors behave differently in relation to medical adherence.For instance, study of age (split into decades) reveals a statistically significant correlation between age groups "60-69" and "70-99" and commitment to religious practices (odds ratio [OR] are 0.23 and 0.19, respectively).The link between adherence and gender and between adherence and education level is not statistically significant.There was no statistically significant correlation between adherence and smoking, however drinking alcohol one-two times per week raises the likelihood of non-adherence by 5.09.The chances ratio for having a comorbidity and being non-adherent is 0.43, whereas OR for using non-diabetic medication and being non-adherent is 0.33.The only statistically significant link between employment levels and non-adherence is between full employment and non-adherence (OR=2.12).Table 5 summarizes the bivariate analysis.Significantly, the initial classification of adherence as "complete," "partial," and "minimum" was revised to "adherent" and "non-adherent."The last category included both minimal and partial compliance.Importantly, the Reliability of the scale score utilizing the internal-consistency reliability assessment based on Cronbach's alpha is 0.8488, indicating that the obtained values are reliable.

Logistic Regression Analysis
Table 6 illustrates the analysis of the logistic regression that examines factors associated with anti-diabetic drug adherence or non-adherence.The higher the patient's agreement, the more likely it is that he or she will not adhere to the recommended treatment.It should be noted, however, that the response rate was so low in many instances, notably in the "agree" spectrum, that OR estimates were rather high.Many responses came within the range of "disagree," with "strongly disagree" earning the most votes.

DISCUSSION
Medical non-adherence is one of the most significant challenges in public health.Not only does it have negative consequences on people' health, but it also increases the cost of health care.This is particularly more damaging in the case of chronic medical diseases, as patients' adherence to therapy tends to diminish dramatically over very brief intervals [11].Consequently, addressing this issue and contributing to the resolution of medical non-adherence is an important objective.
It is well-established that patients' adherence to antidiabetic medicine is strongly correlated with decreased incidences of diabetic complications [12].Simultaneously, poor, or non-adherence to medication is a leading cause of uncontrolled diabetes and may lead to the complications associated with it [13].Considering this, examining the rates of non-adherence and the cause for this feature is a vital step in the prevention of complex diabetes.In this study, an effort was made to shed light on potential variables for medical nonadherence as well as general non-adherence rates.
The reported medication adherence rate was 56.6%, with the remainder being either non-or partially-adherent to antidiabetics.According to our knowledge, no comparable research has been conducted in the Republic of Kazakhstan, hence no comparative studies are available.In general, the unsatisfactory rate of adherence suggests that there may be a multitude of circumstances that limit appropriate drug compliance.These may include patients' sociodemographic characteristics, illiteracy of the significance of controlling diabetes, and the patient-provider relationship.In the present study, the authors attempted to evaluate all these factors.Even though, at the time of the report, the desired 422 responses had not been acquired, 280 responses had been received, which, with a confidence level of 90% and an error margin of 5%, was adequate to make reasonable conclusions.
Among the sociodemographic variables, age group, work level, and alcohol consumption may be associated with better or worse medication adherence.Consequently, those over the age of 60 are less likely to be non-adherent, those who are fully employed have lower rates of adherence, and those who use alcohol one-two times per week are at a greater risk of nonadherence.Notably, neither education levels nor gender revealed such relationships.
In addition to non-antidiabetic medicines, the comorbidity status of patients was also of relevance.Both variables had stronger relationships with adherence to anti-diabetic treatments compared to those with no comorbidities or no additional medicines.There was no statistically significant link between diabetes in the family and rates of nonadherence.
Table 5 presents ORs for antidiabetic therapy adherence for each of the questionnaire items (see Appendix A).Each response other than "strongly disagree" is associated with ostensibly high ORs.These tendencies may be explained by the relatively low response rates for the offered answer choices.As anticipated, self-identified adherent respondents tend to respond with "strongly disagree" to all issues.Overall, the statistics indicate that 17.7% of surveyed patients believe their medications decrease their well-being, 29.9% believe their medications are too expensive, 19.9% do not believe their medications have been correctly prescribed, and 32.7% of patients report having difficulties acquiring medications on time due to hospital-specific factors.Other factors are seen less frequently and are included in Table 3.
Since such analysis of non-adherence is rare and might arguably be the only one to the moment of writing in the Republic of Kazakhstan, it may be useful for medical practitioners to better comprehend their patients and develop more efficient and successful therapy strategies.Among the study's other merits is the relatively large sample size, which provides for a more accurate depiction of how a typical patient with the given condition may present.Another strength of the study is that it addresses various medical diseases and attitudes about them, as well as a wide range of causes for nonadherence to non-diabetic medications.Kazakhstan's polyclinics and hospitals vary greatly from one another, making it impossible for the study to accurately represent the condition throughout the country.Another potential issue with the study may be that it does not reflect the bilingualism prevalent in the Republic of Kazakhstan, so overlooking a potentially significant factor of non-adherence.In addition, the constructed study questionnaire does not fully reflect the diversity of possible medications available to patients; in some instances, patients purchase medications with the same active ingredient but under a different brand name, which makes their responses to surveys more complicated.In addition to this, the study omits the odd discovery that the same active ingredient sold under different brand names generates varied reactions in individuals.Due to the nature of the study, it may be susceptible to recollection bias on the part of patients as well as a tendency to provide less-than-honest responses.Lastly, the study is unable to track how patients feel about various drugs in particular: in many circumstances, patients are adherent to some treatments while being resentful of other medication options; yet the study generalizes the attitude towards all the available medications.
Because the socioeconomic position in Kazakhstan is not uniformly distributed, the current study could be broadened by analyzing more cities and hospitals within cities. Another possible route for the study would be a targeted review of the most used pharmaceuticals in Kazakhstan to forecast if a particular drug under a particular brand name will be effective or ineffective.Finally, it remains to be investigated whether the current worldwide situation regarding COVID-19 may have an impact on patient adherence rates.

Study Limitations
The current study was conducted in Astana, the capital city of Kazakhstan, which does not necessarily reflect the general practice amongst diabetic patients in the entire country.The second drawback is that we did not investigate the antidiabetic medications used, or their side effects as factors for nonadherence, which are beyond the scope of this study.

CONCLUSIONS
In conclusion, this is the first study in the country to investigate medication adherence rate to antidiabetic drugs.The study shows there is a very low rate of adherence to antidiabetic medications amongst the studied participants, comparable to WHO predicted rate for developing countries.There are several factors that appear to negatively influence medication adherence, including older age, full-time employment, and alcohol consumptions.

Medication adherence
The following table presents with statements about possible reasons not to take medicines.Please, fill the table in and, should you have other suggestions, add them in the end of the list:

Table 1 .
Summary of demographic variables analyzed in the study

Table 1 (
continued).Summary of demographic variables analyzed in the study

Table 2 .
Summary of KAP section

Table 3 .
Comorbidities and non-diabetic medication use

Table 4 .
Reasons for non-adherence and associated response rates in percent

Table 5 .
Bivariate analysis of the independent variable and possible effect modifiers

Table 6 .
Odds Ratios for reasons in relationship to non-adherence