Factors Associated with Self-Care Practice among Adult Diabetes Patients

Study Design, Setting and Populations

The hospital-based cross-sectional study design was conducted among adults with diabetes mellitus visiting Dawro Tercha zonal diabetes care center from June to August 2020. Tercha zonal diabetes care center is the largest zonal referral diabetes care center in Southern Nation, Nationalities, and Peoples of Ethiopia. It was found in Dawro Zone which is located 490 km away from Addis Ababa capital city of Ethiopia. The inclusion criteria of this study were adults with DM (18 years and above) and attending Tercha referral and zonal diabetes-care center for follow-up. All patients who were critically ill and those have mental and visual problems, and gestational diabetes was excluded from the studies.

Variables

The self-care practice among adults diagnosed with Diabetes Mellites (DM) was the dependent variable. But, demographic factors such as age group, sex, level of education, marital status, Occupation, and clinical factors such as the family history of diabetes, duration of diabetes, type of diabetes, patient’s body weight, presence of diabetes complications, and presence of glucometer, history of smoking, social support, and were treated as independent factors.

**Sample Size Determination**

The sample size was calculated by the using single population proportion formula,

$$n = \frac{(Z \alpha / 2)^2 p(1-p)}{d^2}$$

where $n =$ minimum sample size required for the study, $Z =$ standard normal distribution ($Z = 1.96$) with CI of 95% and $d =$ is a tolerable margin of error ($d = 0.05$), $P =$ estimate of our target population having poor diabetes self-care practice of 38.1% which was taken from related studies conducted in Harar and Dire Dawa hospital, eastern Ethiopia [21].

$$n = \frac{(Z \alpha / 2)^2 p(1-p)}{d^2} = 362$$

During the time of data collection, a record from the patient profile indicates that a total of 621 adult patients with diabetes mellites were being followed up at Tercha diabetes care center. Since the total number of the study population is less than ten thousand, we used an adjustment for finite population correction formula ($Nc=n/ (1+n/N)$). Where, $n=$ sample size calculated, $Nc=$ sample size after use of correction formula, and $N=$ number of the source population. ($Nc=362/ (1+362/621) = 229$. Assuming a 5% nonresponse rate: $(0.05) \times (229) = 240$ was calculated for the study. Using systematic random sampling techniques, every third patient who fulfils the inclusion criteria was invited to participate in the study.

**Data Collection Tools, Procedure and Quality Assurance**

The data collection tool was developed based on previous similar literature [16, 18, 21]. Initially, the questionnaire was prepared in English Language and translated into Amharic. A forward-backward translation approach was applied in accordance to the World Health Organization’s procedures [22]. Before initial data collection, we mended the questionnaires based on expert opinion. Then, the content validity index of the study tool was calculated and rated at 0.86. Based on the result of the pre-test, ambiguous questions were modified for clarity and consistency. The Amharic version was pilot tested in a similar hospital but outside of the study area with 25 adults with diabetes. Then, the structured questionnaires had three main categories:

• Socio-demographic questionnaires (age, sex, place of residence, income, occupation, educational status, and marital status) were measured using nine questionnaires.
• Diabetes-related clinical characteristics (presence of diabetic-related complications, body weight, types of DM, Family history of DM, history of smoking, social support, and presence of glucometer,) were measured using ten questionnaires.

- A Self-care activities questionnaire that is concerned with diet, physical activities, SMBG, foot care, and medication adherence were measured by five items, and had Yes and No responses for each item.

Three nurses with BSc degrees and two nurses with MSc were recruited for data collection and supervision respectively. The data collectors were provided with one day of intensive training about the data collection procedure, (how to administer the self-administered questionnaire, informed consent, keep confidentiality, and respect the right of the participants). The completeness and consistency of the collected data were checked daily and closely supervised by the principal investigator. Since data were collected during the COVID-19 pandemic, the World Health Organization (WHO) Safety guidelines and protocols were strictly followed at all times.

**Operational Definitions**

In this study, the proportion of participants with good and poor self-care practice was computed. The sum of the variable between diet, exercise, SMBG, Foot care, and medication adherence was calculated. Then we used the mean to dichotomize the self-care score. Those who scored above the mean in the overall self-care score were categorized as having good self-care practice and those who scored below the mean score were categorized as having poor self-care practice.

**Data Entry, Processing and Analysis**

The data were verified, coded, and entered into Epi data 4.6 software and then exported to Statistical Package for Social Sciences, (SPSS) version 25.0 Software for analysis. Descriptive statistics such as frequency, percentages, mean, standard deviation ($M+SD$), and ranges were used for the interpretation of outcome variables. A cross-tabulation was computed for the cross-comparison of dependent and independent variables needed for graph and logistic regression. A binary logistic regression and multivariate logistic regression analysis were carried out to determine the association between dependent and independent variables. The variable in bivariate analysis with p-value < 0.25 was entered into multivariate logistic regression. A statistically significant association was declared at a P-value less than 0.05 in multivariate logistic regression analysis. The strength of the association of factors with knowledge and practice was demonstrated by computing the odds ratio (OR) and the adjusted odds ratio (AOR) with a 95% confidence interval (CI).

Socio-Demographic Characteristics of Study Participants

In total, 213 diabetic patients participated in this study with a response rate of (89%). Table 1 shows the socio- demographic characteristics of participants. The mean age of participants was 44.4 + 14.2 years. The male population outweighed by 114(53.5%). A significant number of the respondents 169 (79.3%) were married and 112(53.1%) of them were rural residents. The majority 94(44.1%) of the respondents had attended primary education and 88 (57.7%) were civil servants. Nearly two-thirds of 145(68.1%) of the participants had low income.

Clinical Characteristics of Study Participants

Table 2, presents the diabetes-related clinical characteristics of the respondents. About 77 (36.2%) of respondents’ duration of diabetes was 10 years and above. More than half 134 (62.9 %) of respondents had no family history of diabetes. The mean of respondents’ body weight was 64.9 + 12.6 Kg. Nearly, three-fifth 126(59.2%) of participants had type-II DM and about 131(61.5%) developed diabetes-related complications. Approximately three-fourths 159(74.6%) of the participants had no glucometer in their home. A vast majority 201(94.4) of the respondents had social support.

Key: DM= Diabetes mellites, Kg = Kilogram, M+ SD= Mean and standard deviation Table 2: Diabetes-related clinical characteristics of study participants.

The Level of Self-Care Practice among Study Participants

Concerning dietary parameters, the majority of study participants 154(72.3%) have no healthful eating plan every week and a significant number 172(80.8%) of them did not perform a low level of physical exercise. About 164(77.0 %) did not test their blood sugar every week and 155(72.8%) of them did not check their foot on weekly bases and 137(63.3%) of them were not adherent to their antidiabetic medications

Click to enlarge

Regarding overall status of self-care practice of the respondents, only 84 (39.4%) diabetic patients had a good level of self-care practices and from the total of respondents,

129 (60.6%) had poor levels of self-care activities (Figure 1).

Factors Associated with Self-Care Practice among Adults with Diabetes Patient

As presented in Table 3, seven variables (age, body weight, educational level, place of residence, having diabetes complications, type of DM, and having private glucometer) were used as candidate variables (at P-value <0.25) and entered together into a multivariate logistic regression. A P-value < 0.05 in multivariate analysis was taken as a cut-point value to be statistically significant. Accordingly, bivariate logistic regression analysis revealed, attending primary education [COR=7.971, (95% CI: 1.020-62.264), P=0.048], Secondary education [COR= 18.452, (95%CI: 2.327-146.326), P= 0.006] and college and above education [COR=100.571, (95% CI: 11.547- 875.969), P <0.001] were found more likely to have good self-care practices compared to those who were illiterates respectively. However, being an urban residence[COR=0.431,(95% CI:0.244-0.760), P=0.004], having DM complication [COR= 0.531, (95 %CI: 0.302 – 0.934), P=0.028], Having Type-II DM [COR=0.536,(95% CI:0.306- 0.938), P= 0.029] and absence of private glucometer [COR=0. 387, (95% CI:0.200– 10.750), P=0.005] were found less likely to have good self-care practice compared to those who were rural residents and did not develop DM related-complication, with type-I DM and had private glucometers respectively.

After controlling some confounders, the multivariate logistic regression analysis revealed that attending secondary education P =0.01 and college and above P <0.001 were found more likely to have good self-care practice compared to those who have no formal education respectively. However, the odds of participants who did not develop DM complications P=0.013 were found 0.412 times less likely to have good self-care practice compared to those who developed DM complications.

Keys: AOR= Adjusted Odd Ratio, COR= Crude Odd Ratio, DM= Diabetes Mellitus, CI= confidence Intervals, Ref = Reference, *Statistically significant at (p<0.05), **Statistically significant at (p<0.01) Table 3: Bivariate and multivariate analysis to identify factors associated with self-care practice.

Ethical Approval and Consent to Participation

Before the start of the study, ethical clearance was obtained from the Institutional Research Ethical Review

Board (IRERB) of the College of health science, Addis Ababa University, (with protocol number: 037/20/SNM, meeting number: 09/2012E.C), in accordance with the declaration of Helsinki. After ethical clearance was obtained, a letter of support was written from Addis Ababa University, School of Nursing and Midwifery to Tercha Referral Hospital. Then, permission was obtained from the diabetic care- center of Tercha Referral Hospital. Study participants were asked for their willingness to participate in the study. All the reasons why the participants were selected and why the research was being conducted were explained verbally to the study participants. The participants were fully explained that they have the right not to participate, to discontinue at any time in between, or have a right not to answer any questions they were not willing to answer. Both verbal and written informed consent was obtained from the study participants. For participants who cannot read and write, verbal consent was obtained from their parents and legal guardians in accordance with the FDRE National Research Ethics Review Guideline. This procedure was discussed with Tercha Hospital administrators and approved by the Institutional Research Board of Addis Ababa University, College of Health Science. Participants were assured of confidentiality and anonymity of their responses. Identification number and codes were used in the survey questionnaires. The collected data were stored and kept on a password-protected computer. Only the principal investigator had access to the computerized data.

Acknowledgments

The authors would like to acknowledge data collectors, supervisors, and the study participants for their valuable contribution to the study. We would also like to extend our gratitude to the medical directors and managers of Tercha Referral Hospital for their unreserved support.

Authors’ Contribution: BAA, ZAM, and BDT: Conceived a data, designed the study, performed analysis, interpreted a data, drafted and finally approved the manuscript for publication. GM, GNB, TG, YLL, AA, MGF, and BB have assisted in designing the study, conceptualization, analysing and interpreting a data, and involved in review process of the finial manuscript.