Demand of Governmental Employees for Social Health Insurance and Associated Factors in Takusa woreda, Amhara National Regional State, Ethiopia

Objective: The main aim of this study was to assess the demand of governmental employees for social health insurance and associated factors in Takusa district, Northwest Ethiopia. Methods: Institution-based cross-sectional triangulated with the phenomenological study was conducted. A simple random and a purposive sampling technique were used to recruit 324 employees. Binary logistic regression was performed, and P < 0.05 was considered statistically signicant. In the qualitative study, data were analyzed using thematic analysis. Results: 61.1% of employees had a good demand for social health insurance. Employees who had 5-6 family sizes were 5.5 times more likely to have a demand for SHI than counterparts. Also, employees who live in a rural area were 3.3 times more likely to have SHI demand than urban residents. In the qualitative analysis, lack of awareness, fear of poor-quality health service provision, the unpreparedness of health institutions, lack of inputs, poverty etc. were the main constraints for the poor demand for SHI. Conclusion: In general, the higher demand for SHI in this study was more than half of the Ethiopians but less than abroad. Therefore, the concerned body shall intervene on the factors by implementing a SHI program cooperatively.


Source and study Populations
The source population was all Takusa Woreda governmental employees while the study population was employees who stayed a minimum of 6 months on employment in any place but currently employed in Takusa woreda.

Sample size and sampling technique
The sample size calculated based on the single population proportion formula by assuming 95% con dence interval (CI), 5% marginal error (d), 2,024 total population of Takusa woreda civil servants=N), 5% non-response rate, 54% single population proportion (P) for the demand of SHI in Wolita Sodo town [24]. Based on the above assumptions, the nal sample size to be enrolled in this study was 337 governmental employees. In the case of qualitative study design, FGD with key pling technique was used for quantitative and qualitative study design, respectively.

Data collection procedures
In the case of quantitative study design, all governmental employees' list (frame) was obtained from Takusa woreda Finance O ce. Here, their name was ordered alphabetically and selected randomly by computer among all sectors. Then by providing employees' address, working institution, data collectors were departed into each institution and introduced themselves including the purpose of the study to the selected study participants. Then after, Amharic version self-administered structured questionnaires were given to the employees at their institutions during working day time and returned back in the same day.

Data quality Control
One day of training was given for 2 data collectors (a Woreda health extension o cer and Delgi health center health extension o cer) and a supervisor (Woreda health insurance focal person) who work in that woreda.
Then, data collectors had conducted a pre-test assessment by taking 5% of randomly selected sampling units from other similar populations, Alefa woreda. The questionnaires which were used for data collection were adapted from the different literature review and other similar studies. Based on the pretest result, the questionnaires were modi ed into a local context. The supervisor monitors check the consistency of questions and gave comments until the end of data collection. The principal investigator assured the data quality through checking completeness and consistency of lled questionnaire, data clearing, and analysis of the data through proper data analysis methods. The data were interred into EPI- INFO 7,SPSS v 20, and open code 3.6 twice.

Data processing and analysis
In EPI INFO version 7, hard copy data were entered and then converted to SPSS version 20 program. Again, the existence of the variables, frequencies, outliers, and recoding of the variables was done before any special research was done, verifying the statement. Descriptive data analysis was then conducted, such as text, charts, diagrams, frequencies, mean, median, quartile, and percentages. For all variables, bivariate logistic regression was carried out to nd candidate variables with the enter method for multiple logistic regressions.
Those 9 candidate variables that in the bivariable analysis had a p-value < 0.2 were entered into the multivariable analysis. Finally, the P-value < 0.05 result was found to be statistically important in the Multivariate logistic regression model. To see the power of correlation for variables, the Adjusted Odds ratio was used for 95 percent CI. Bootstrap has been used to see whether or not these observations are in line with another research. Qualitative data were transcribed, interpreted, converted to open-source platform version 3.6, read frequently, coded, and a thematically classi ed. Finally, through thematic study, it was interpreted.

Socio-demographic characteristics
Of the 337 total study participants, 324 (96%) took part in this study, 79% urban and 21% rural. All respondents fell between 18 and 62 years of age, with a mean age of 28.12 years (Table 1). . With a minimum of 500 and a limit of 20,000ETB, the average monthly net employee income was 2,477.17 ( Figure 1).

Health and health related characteristics
Most, 77.5%, workers were able to get a health center, 43.2% were able to get a private clinic, 37% were able to get to the hospital, and 21.6% were able to get a pharmacy nearest health post. Generally speaking, 96% of them were available for health insurance, while just 4% were not.
More than half, 54.9% of study units or their families have been ill for the last 12 months, and the remainder has been well. Out of 180 family sick stories, 44.1% had been ill for less than 3 days, 26.3% had been ill for 1 week, and 19% had been ill for more than 3 weeks. Among the sick respondents, 53.3% had to get outpatient health care from the health department, 21.4% had to get help from the private clinic, and 20.5% from the doctor.
In sick respondents, 49.2% had paid less than 83 ETB, and 22.3% had paid more than 360ETB. Overall, the total amount charged for outpatient sickness in the last 12 months was 548.41, ranging from 0 (maybe workers, exempted facilities, etc.) to 11,000. In all the sick 180 participants, only 24.4% were admitted.
Nearly three-fourths, 71.6 %, reimbursement of medical bills was out of pocket, and others, 73.1% of the state that they were unable to meet all their manufacturing expenses on their own. Furthermore, 67.6% of workers were unhappy with the consistency of health care delivery (Table 3).

Source of information
The awareness-building exercise, accompanied by TV, was the primary source of knowledge for SHI ( g 2).

Demand of Social Health Insurance
The 61.1% of respondents had strong demand, equivalent to two-thirds, while the remaining 38.9% had poor demanded ( Figure 3). The mean monthly payment was 2.23 out of 208 research units that able to pay, varying from 0.01% to 8%. 34.9% were afraid of low-quality health care provision, 27.4% were due to insecurity, among those who were not able to participate and pay for SHI membership. The remaining 16.4% denied should get free health care because of health workers it may be in uenced by corruption; payment of imbalance and service received a lack of understanding.

Factors for Demand of SHI
In the binary logistic regression model, nine variables were eligible to be interred into multiple logistic regression (p<0.2). But in the multiple regression model, only three variables which are rural residence (p=0.039), family size (p=0.04), and cost of outpatient treatment during illness in the last 12 months preceding the data collection period (p=0.006), were found to be associated with demand of SHI among employees (p<0.05).
Employees who had 5-6 family sizes were 5.5 times more likely to had a demand for SHI as compared to those who had 1-2 family size (AOR=5.50, 95%CI=1.01-30.08). It was also found that for rural residents, the likelihood of SHI demand was 3.3 times higher than for urban residents (AOR= 3.29, 95 percent CI= 1.06-10.20). In comparison, workers who charged more than 360 ETB for outpatient care during sickness in the last 12 months were 5.2 times more likely than those who paid less than 83ETB (AOR= 5.22, 95% CI= 1.61-16.95) to have a demand for SHI compared to those who paid less than 83ETB (Table 5). Twelve workers, six men, and six women were interested in 2 FGDs in total. Of these, 8 were professional health professionals. Nearly all FGD respondents were well educated about SHI. SHI was described by the head of the Delgi Health Center as "Like rich people purchasing a car and insuring it, SHI was a way for members to pay based on their income to become a member." A district hospital manager stated the condition as "…All paid members may not get health services by chance, but like other organizations such as 'Eddire' by understanding this it was an opportunity in which one would help others." Health institution that would provide SHI service In the thematic analysis, half of the respondents agreed as any accredited health institution which was voluntary to provide the service would be delegated. The head of Delgi health center said that "… Gambi Hospital can give the service in collaboration with Felegehiwot referral Hospital. Since as they are quali ed and present inside Ethiopia…" On the other hand, a 25 years old female stated that "… start the service by governmental health institutions followed by private institutions based on the standards. If any institutions did not ful ll the standard it would be better to upgrade it through time …" Contrarily, half of them respond as only governmental health institutions, excluding health post would provide the service. A small enterprise o cer complains as "private health institutions were pro t nders, they might need to collect over income… if one person was ill and goes there, the customer would pay 70-100 birr for a card only. If it was for Laboratory, drugs, and other services, we can estimate how much he would pay."

Preparedness of health institutions
Regarding the preparedness of health institutions, the majority of FGD study groups replied as almost all of them were as it was. Health institutions that would take contract agreement were not selected; lack of additional manpower, classrooms, supplies; lack of discussion forum and promotion; lack of infrastructure; unpreparedness of membership les were some of the indicators. A Nurse employee said that "… in our woreda among ve health centers, 2 of them hadn't, even once, the medical laboratory professional." A 25 year's old employee responds as "...Still, those health institutions which will provide the service are not identi ed. It indicates as there was no preparedness. The manpower, classes, and other inputs are as previous." A 32 years old nurse said that "employers hadn't any idea about SHI…they didn't know how much percent they would pay for their employee, how would pay it…"

Scope of SHI program
Thematic analysis showed that the scope of SHI services would be all health services, except beauty, at any time and place up to abroad referral system including both medical & non-medical related costs. A Small enterprise o cer, 42 years old, said that "…SHI must provide all health services at any time and any place, except those done for beauty. There must be a referral system from Woreda up to specialized Federal Hospitals, even outside of the country. Unless it incorporates complicated services why it needs to be a member of SHI because he can get simple health services by himself…" Majority of them also agreed as all employee's would-be member of SHI.
A 28 years old district hospital manager said that "…when I see mine, it would be 117.33ETB per month. I should save it… since I am health professionals, had better awareness about the disease, and I can care for myself from the disease." He also stated his fear as "…If members paid consecutively and unfortunately, they did not get health service by SHIA due to being healthy, they might think paying without bene t…". They also mentioned that members might be enforced by health professionals to give referral papers for them whether the case was beyond the scope of contracted health institutions or not.

Strength and weakness of SHI scheme
According to Delgi health center head, SHI had holistic pro ts. He stated it as "…when health institutions were strengthened, supplies would be accessible and then accessible quality service would be given to members. It also reduces those ill persons that stayed at home due to shortage of money…" A 32 years old environmental health o cer expressed the strength of SHI as "… civil servants' health-seeking behaviors would be increased. When they were felt ill, they would come to get treatment. SHI was used as retires. In the retires program, some percent of the money was collected throughout their active working time and it would be used at the time of passive employment. SHI was just like this; the income was collected during the healthy time and used when they were ill". The majority of them agreed as the SHI program provides both prevention and treatment services. They state it as, when members feel simply ill, they would come and got health service, pieces of advice about health education, because they had no nancial problems. So, this can be considered as preventive as well as curative. On the other hand, participants respond as SHI had its weakness. A 25 female raised the issue in that "… membership payment, and its bene t was not balanced. Even after paying this amount of money, complicated health services were not provided…" year's old female said that "…the idea should not be declared from administrative bodies only and forwarded to members. Rather, it was better to discuss with employees and customize with our country context…" She also said that "….as membership monthly payment was too much it was better to create a discussion forum and then after arriving on consensus to increase by starting from smallest monthly payment system…" This study revealed that the following indicators were enabling factors for good demand of SHI in the future.

Discussion
In this study, the demand of employees for SHI was 61.1%. It revealed that family size, rural residence, and cost of treatment were positively associated with the demand for SHI. The demand for this study, 61.1% (95%CI = 55.6-66.7%), was greater than that of Jimma, 51.5%, and Wolita Sodo, 54% [18][19]. It might be due to employees in this study were nancially secured; had greater con dence in the implementation and bene t of the newly proposed program; had high intension to help risk groups, quality health service provision, had relatively better awareness than the Jimma community, and Wolita Sodo teachers. District in Ethiopia (80%) [18][19][20][21][22]. This might be due to the HI program was in the infancy stage in Ethiopia in general, it had already started in other countries like Fogra. As a result, its bene t would be well known, and uptake of health insurance would be increased. The other explanation might be due to variation of socio-demographic and economic factors.
But it was greater than that of Nigeria (52.5%) [23] and Jimma (51.5%) [19]. It might be due to Takusa woreda employees were relatively nancially well secured than Nigeria (literature shows that Populations who live below the poverty line in Ethiopia (27.8%) was less than that of Nigerian (70.2%) and Jimma. The other reason might be due to variation of study units, household heads in Jimma, and Secretariat in Nigeria. Surprisingly this nding was in line with that of China (70%) [24] and Wolita Sodo (71.3%) [18].
The willingness to join SHI in this study was also supported by a qualitative study. More than half of them agreed to join SHI by putting preconditions. A 25 years old female stated as "I am voluntary to join SHI however, the idea must not be decided by the authoritative body only …Rather it would be better to have discussion forum between employees and authorized body and able to amend accordingly". In this study, Willingness to pay for SHI, 64.2% (95%CI = 59.3-69.4%), was greater than that of Lose Angeles,54% [25]. It may be due to Takusa woreda civil servants were almost sedentary while Mexican were immigrants who live in Lose Angeles. Also, it was more than done in Namibia, 42.25% [21]. This might be due to the implemented health insurance program in Namibia did not cover all components of health services, and members might be dissatis ed. Again, it was greater than that of Wolita Sodo in 2012, which was 53.1%. It might be due to net income and awareness of this study participants were better than Wolita Sodo teachers [18]. But it was less than a study done in Fogera Woreda in 2013 on community. The possible justi cation for this community might be their wealthiness through rice production. Its bene t might be the well-known quality of health service might be improved; due to environmental conditions, they might be more vulnerable to diseases like malaria.
Though 64.2% were willing to pay, only 61.1% had good demand for SHI. The discrepancy (3.09%) might be due to the participants could pay, but they were not voluntary to join the SHI scheme. They might be health practitioners, Bank professionals, polices, etc. who were not paid for medical bills in the former health service provision system and others. This was supported by FGD results. A 28 years old district hospital manager said that "I will not be members of SHI. Because I am health service providers, which make me more vulnerable to occupationally related diseases. So, I should get the service freely… on the other hand, as a health professional, I can promote my health better than non-health professionals".
Increment of SHI demand as FS increases in this study was supported by the study conducted in Fogera and Ghana [22,26]. It might be due to as the size of the household increases, the nancial burden faced by households during illness to get health care services increases and health insurance participation would rise. However, in Zimbabwe family size was associated negatively [27]. It might be due to an increment of payments for CBHI membership, as the family size rises, which leads to unaffordable, and hence a family might decide to withdraw from the program. This may be seen when there were no chronic illnesses in the family and yet family income is low.
It was also found that participants who live in rural areas were 3.3 times more likely to had SHI demand compared to urban residents. This might be rural employees were not well resourced nancially as compared to urban; they might mostly vulnerable to diseases because they were far from health information to had a self-car, inaccessible for health services. Culture and belief-systems of the rural area might be increased the probability of their illness. Due to these, their intention to enroll in the SHI program might be increased. Moreover, the increment of uptake of SHI, as the cost of health service payment increased, in this study was supported by research done in Pakistan [28]. It might be due to as people became ill and their medical bills increased, their demand for SHI might be raised.
The main reasons for the poor demand for SHI in this study were similar to that of Jimma, Wolita Sodo, and Fogera [18][19]22]. These were: poor quality of health services; Poverty; fear of poor implementation; feeling of being healthy; the pocket payment was better and restriction of health service provision. This was supported by a qualitative study. It revealed that inputs like water, light, equipment, drugs, manpower were scarce; health professionals were not responsible and accountable for their work.
The majority of FGD participants agreed that preparedness of health institution which would provide SHI services was as it was. A Nurse employee said that "… in our woreda among ve health centers, 2 of them hadn't, even once, the medical laboratory professional." Again a 25 year's old employee responds as "...Still, those health institutions which would provide the service were not identi ed… there is no any preparedness…manpower, classes and other inputs are as previous." A 28 years old district hospital manager said that "…If members paid consecutively and unfortunately, they did not get health service by SHIA due to being healthy, they might think paying without bene t…". They also stated as members might enforce health professionals to give referral papers for them whether the case was beyond the scope of contracted health institutions or not.
The other main factor which would reduce the uptake of SHI was the scope of health services provided by SHIA. Thematic analysis showed that the scope of SHI services would be all health services, except beauty, at any time and place up to abroad referral system including both medical & non-medical related costs. A small enterprise o cer said that "unless it incorporates complicated services why it needs to be a member of SHI because anyone can get simple services by themselves." Type of health institution that provides SHI service were another fear employee for the quality of SHI service provision. Half of the respondents agreed as any accredited health institution which was voluntary to provide the service would deliver the service. Contrarily, half of them respond as only governmental health institutions, excluding health post would provide the service. A Small enterprise o cer complains as "private health institutions were pro t nders, they may need to collect over income… if one person was ill and gone to their customers would pay 70-100 birr for a card only. If it was for Laboratory, drugs, and other services, we can estimate how much he would pay." In general, a qualitative study revealed that the main constraints for successful implementation of SHI would be SHIA related factors, SHI provider health institutions, employees' related factors, and other factors. The quantitative study was triangulated with a qualitative study. The main limitations for this study were recall bias on money paid for health care and illness episode in the family, and preparedness of health institutions were assessed by FGD instead of facility auditing.

Conclusions
Generally, when the result of this study was compared with most studies, it was greater than Ethiopians but less than abroad. In this study, family size, rural residence, and the maximum cost of outpatient treatment were found to be positively associated with good demand for SHI. Though the majority of study units had poor awareness, most of them had a good attitude and demand for SHI. It indicates as if an awareness creation session was conducted with different modes and preconditions were ful lled, their intention to uptake SHI would be enhanced. To have good demand for SHI, this study demonstrated to conduct awareness creation session forum; maintain quality of health service, prepare health institutions and use of networked nance system was the predominant enabling factors.
We recommended the social health insurance agencies to conduct awareness creation sessions about health insurance and its bene ts to employees, and to select institution which takes contract; conduct audit and trust assessment regularly, Regional Health Bureau and Ministry of Health to advocate SHI by mass media to enhance its policies regarding medical record system, information utilization to support effective implementation of SHI, and researchers to do further investigation on SHI, especially on quality of health service and preparedness of health institutions.

Declarations Ethical considerations
Ethical clearance was obtained from Review Committee of University of Gondar College of medicine and health science Institute Public Health. Permission Letters was obtained from administration and Health o ce. Informed written consent was obtained from respondents. Data collectors informed the study participants about their right to give, interrupt or refuse to give the data and its con dentiality would be maintained.

Consent for publication
Not applicable Availability of data and materials The original data are available at the corresponding author, so that an interested reader can get the data from the corresponding author with reasonable request.

Competing interests
The authors declared that they have no competing interests regarding the publication of this article Funding Not applicable Authors' contributions TG: Conceptualization of the idea of the study. TG, MY, and AF were participated in the data collection, data curation, analysis, and interpretation of the data. MG and GA were participated in data collection and analysis. All authors were actively participated for the manuscript preparation and nalization of this article. All authors read and approved the nal manuscript. Table 2   Table 2 was not provided with this version Source of information for SHI among employees in Takusa Woreda, 2015 (n=216)