High common childhood illnesses while treatment seeking behavior of mothers’/care givers’ was low at rural northwest Ethiopia

Background: In most developing countries, childhood illness and deaths are among the most challenging health issues. Most of these deaths were due to preventable causes including acute respiratory infections (ARI), diarrhea, and febrile illnesses. Thus, identification of determinants of these childhood disease illness would help to guide strategic planning, and prioritize interventions. Method: Community based cross-sectional study was conducted. Two stage cluster sampling technique was used to select kebeles and then households. A total of 2,158 mothers/care givers with at least one under-five child were included. Mothers/care-givers were interviewed at their homes using a pretested and structured questionnaire. Adjusted Odds Ratio (AOR) with a 95% Confidence Interval (CI) for variables with P-value <0.05 in the multivariable model were reported to show significance and direction of association<0.05 in the multivariable model were reported to show significance and direction of association. Result: The overall prevalence of childhood illnesses was 16.5% (95% CI: 15.0, 18.2). The treatment seeking behavior of the mothers/caregivers was 22.8% (95% CI: 18.7, 27.4). Of the illnesses, 6.5% (95% CI: 5.5, 6.72) were diarrhea, 7.5% (95% CI: 6.5, 8.76) were ARI, and 12.0% (95% CI: 10.66, 13.40) were fever. Mothers’ age 40 years and above (AOR=0.35; 95%CI: 0.16, 0.76), mothers’ occupation (AOR=2.28; 95%CI: 1.84, 4.39), distance from nearest health center (AOR=1.02; 95%CI: 1.01, 1.03), and presence of three or more under-five children in the house (AOR=2.59; 95%CI: 1.03, 5.76) were predictors of childhood illness. Conclusion: In this study, the common childhood illness remains high while treatment seeking behavior of the mothers’/care givers’ was so low. Mothers’ age and occupation, distance from nearest health center, and presence of three or more under-five children in the house were independent determinants of childhood illness. Thus, addressing significant determinants and enhancing mothers’/care givers health seeking behavior would help in reducing morbidity and mortality of under-five children with common childhood illness.

Interval (CI) for variables with P-value <0.05 in the multivariable model were reported to show significance and direction of association<0.05 in the multivariable model were reported to show significance and direction of association. Result: The overall prevalence of childhood illnesses was 16.5% (95% CI: 15.0, 18.2). The treatment seeking behavior of the mothers/caregivers was 22.8% (95% CI: 18.7, 27.4). Of the illnesses, 6.5% (95% CI: 5.5, 6.72) were diarrhea, 7.5% (95% CI: 6.5, In this study, the common childhood illness remains high while treatment seeking behavior of the mothers'/care givers' was so low. Mothers' age and occupation, distance from nearest health center, and presence of three or more under-five children in the house were independent determinants of childhood illness. Thus, addressing significant determinants and enhancing mothers'/care givers health seeking behavior would help in reducing morbidity and mortality of under-five children with common childhood illness. Key word: Childhood illness, determinants, rural, health seeking behavior, Ethiopia Background Childhood illnesses including Acute Respiratory Infection (ARI), diarrhea, and febrile illnesses are 3 among the most serious health issues challenging most developing countries. Globally, an estimated 5.9 million children were died before celebrating their 5 th birth day in 2015 (1). About 75% of underfive deaths were caused by diarrhea, fever, and acute respiratory infection combined (1,2) and resulted to the death of 15,000 children daily from these preventable and treatable diseases (3).
Sub-Saharan Africa still experiences remarkably high child mortality rates (4). In 2017, the region had an average under-five mortality rate of 76 deaths per 1,000 live births. About 1 in 13 children dying before his or her fifth birthday which is nearly 14 times higher than the average ratio of 1 in 185 in high-income countries and 20 times higher than the ratio of 1 in 263 in the region of Australia and New Zealand (2).
In Ethiopia, although significant achievements have been made in reducing childhood mortality at the national level, under-five mortality is high in the Amhara region reported as 85 deaths/1,000 live births (5). Acute respiratory infections, diarrhea and malaria were the leading causes of death for children under-five years of age (1,2,6,7). Diarrhea contributes to more than one in every ten (13%) child deaths (8). Furthermore, according to the 2016 Ethiopia Demographic and Health Survey (EDHS) report, about 12%, 14%, and 7% of under-five children had diarrhea, fever, and acute respiratory illnesses episode 2 weeks preceding the survey (6) In developing countries age of child, sex of child, parental educational status, toilet facility, water source, household income, and distance from health facility, maternal characteristics, and number of children per household were significant factors of childhood illness (9-13).
Health-seeking behavior is a function not only of the access and availability of health facilities, and other sources of healthcare but also motivation and ability of individuals to seek medical treatment (20). The World Health Organization (WHO) estimates that seeking prompt and appropriate care could reduce child deaths due to acute respiratory infections by 20% (21). However, significant numbers of children continue to die without appropriate treatment and ever reaching health facility or due to 4 delays in seeking care in developing countries (22,23). In Ethiopia, health care seeking behavior is poor and only a small proportion of children receive appropriate treatment. Nationally, only 33.3% of under-five children with symptom of acute respiratory infection, 35% with fever and 44% with diarrhea were taken to health facility during 5years preceding 2016 (24). Although different health system initiatives and strategies have been designed and attempted to reduce child mortality in Ethiopia, childhood morbidity and mortality still remain a challenge.
Consequently, findings of this study would be useful used for planners, policy makers, decision makers, and as a baseline for researchers for further interventions addressing child morbidity and mortality. Therefore, this study aimed to determine the prevalence of childhood illness and its associated factors in north west Ethiopia.

Study design, setting and source population
Community based cross-sectional study was conducted from March 3-18, 2016 in North Gondar zone, which is located in Amhara region in the northwest part of Ethiopia (18). The zone has 24 districts.
According to central statistical agency (CSA) 2017, the estimated population in the North Gondar zone was 3,654,920. Of which, 858,906 were women in the age ranges between 15-49 years old (reference). The estimated under-five children were 493,414. As of 2016, the zone has 9 government hospitals, 126 health centers and 563 health posts providing health care services. Mothers/care givers with child age between one and 59 months were included in the study.

Data collection procedures and variables
Two stage stratified cluster sampling technique was used to select kebeles and then households. A pretested structured interviewer administered questionnaire was used to collect the data. The questionnaire was translated to Amharic and back translated to English to check the consistency.
Mothers/care-givers were interviewed at their homes. In case of more than one child of one and 59 months in the household, data was collected from the last child. Fifty four data collectors and five supervisors, all of them with at least first degree in health sciences were involved in the data collection process. Two days training on the questionnaire, way of interviewing, and ethical issues 5 was given prior to the data collection.
The occurrence of childhood illness during one and 59 months of life and seeking of health care of mothers while the child was sick were the outcome variables. The independent variables were classified as socio-demographic factors including age of mother, marital status, maternal education, maternal occupation, occupation of father, father education, religion, number of sick children, and wealth index; environmental factors were source of water, cook in the house, use of wood, types of toilet facility, and types of cooking fuel; clinical factors including birth weight, pregnancy status, place of last delivery, PNC attendance, home delivery practice, health education by health post, pre-lacteal feeding, initiation of breast feeding, bottle feeding, and pregnancy complication.
In this study, common childhood illnesses include ARI, diarrheal diseases, febrile illnesses, and pain in the ear. Presence of childhood illness was defined as the presence of at least one of the following (cough, diarrhea, fever, or ear pain) two weeks preceding the survey. ARI was defined as the presence of cough and difficulty of breathing which is reported by mothers or care givers within two weeks preceding the survey. Diarrhea was defined as the presence of three or more loose/watery stools per day, or blood in stool as perceived and reported by mothers or care givers within two weeks preceding the survey. Fever was defined as subjective feeling of elevated body temperature reported by mothers or care givers within two weeks preceding the survey. Health care seeking behavior was defined us mothers/caregivers response for signs and symptoms of illnesses to reduce severity and complication after recognizing the child's illness.

Data Processing and Analysis
The data were entered and cleaned using Epi-Info version 7 and analyzed using STATA version14.
Data cleaning, coding and recoding was made. Descriptive statistics in the form of means for continuous variables and percentages for categorical variable were made. Then the findings were presented by tables and texts. Variables with 0.2 p-values ≤0.2 in the bi-variable analysis were fitted in the multivariable model. Adjusted Odds Ratio (AOR) with a 95% Confidence Interval (CI) and pvalue <0.05 in the multivariable model were used to declare significant association with childhood illness. Goodness of fit was checked using Hosmer and Lemeshow test with p-value of 0.37.

Factors affecting childhood illnesses
In the bi-variate analysis age of mother, bottle feeding, distance in kilometer from the health center , occupation of mother, wealth index, source of drinking water, cook in the house hold, and number of children in the household were significant determinates of childhood illness. However, in the multivariable analysis, age of mother, distance in kilometer from the nearest the health center, occupation of mother, and number of underfive children in the house were significant factors of childhood illness/ post neonatal illness (Table 5).

Discussion
In our study, the overall prevalence of childhood illnesses among under-five children preceding two weeks of the survey was 16.5%. Of the illnesses, 6.5% were diarrhea, 7.5% were ARI, and 12.0% were fever.. When we compared this finding with the reports from Ethiopian Demographic and Health Survey, the prevalence of diarrhea and fever were higher than the national figures (%?), however the two weeks prevalence of ARI was comparable among children of under five years of age northwest Ethiopia (24). Similarly, the result of this study noted lower prevalence of childhood illness compared to some studies conducted elsewhere (9,11,13,16). There could be several explanations for the differences in prevalence of the illnesses with this findings and across literatures. For instance, the study in Tanzania included three remote districts located 200-400 kilometers from the capital of the country.
All the three districts were predominantly rural and impoverished with poor transportation, infrastructure, and subsistence agriculture driven economy there by limited availability and poor access to health care services. Likewise, the study done in Kenya was from two slum areas where the majority of the residents seek health care outside of health institutions. They mainly visited local drug shops, faith based institutions, and to some extent private clinics. And evidences showed a strong relationship between increased occurrence of childhood morbidity and health seeking behavior outside of health institutions (25)(26)(27).
Furthermore, our finding noted lower prevalence of childhood illnesses, particularly childhood diarrhea, compared to different studies conducted at different regions and districts in Ethiopia (28)(29)(30)(31)(32)(33)(34)(35)(36)(37)(38). Sample size, year of study, and study setting were some of the factors for the differences in prevalence of childhood illnesses across literatures. For example, one of the studies cited above was conducted in three small kebeles and used only 405 care givers in the study.
The treatment seeking behavior of the mothers/caregivers for any of the childhood illness was 54%.
Out of 370 children of under five years of age who had one or more childhood illnesses, mothers/caregivers sought care at health facilities for 63 (45%), 97 (37.60%), and 17 (10.33%)of children for diarrhea, fever, and ARI, respectively. Treatment seeking behavior from health care facilities in this study was consistent for diarrhea and fever but lower for ARI compared to the EDHS 2016 report. The lower care seeking behavior for ARI could be the mothers' perception of mild illness of cough as common cold (39). This study has also noted that the treatment seeking behavior of the mothers/caregivers was lower compared to similar studies done elsewhere (26,(40)(41)(42)(43). There could be several explanations for the differences in treatment seeking behavior of the mothers/caregivers for common child hood illness. For instance, visiting traditional healer first, financial constraint, perception that illness was not serious and the expectation that illness would recover soon were some of the possible reasons (25,40,44). However, treatment seeking behavior of the mothers/caregivers in our study was higher compared to similar studies conducted in other parts of the country elsewhere (7,35,45,46). The possible reason is that some of the above studies were conducted in relatively urban areas where residents had better socioeconomic status, and good awareness and attitude to seek modern care for their children (46,47).
Different category of factors including maternal socio-demographic characteristics, child related and household environmental factors were significantly associated with childhood illness. Accordingly, maternal age, number of children in the household, mother occupation, and distance from the nearest health center were the independent determinants of childhood illness.
In this study, it was noted a child lived in a family with three and/or more under-five children was more likely to have childhood illness than a household with only one child. This finding was supported with similar studies conducted elsewhere (36,(48)(49)(50). This could be explained by the fact that when the number of children in the household increases, it is expected that children could be more vulnerable to contamination because the quality of care and attention from parents decreases as mothers become incapable of caring for children. Furthermore, children who get the disease may easily transmit the disease to others who live in the same area.
Moreover, the odds of developing childhood illness among children of semi-skilled/skilled mother occupational status were 2.3 times compared with unskilled mothers. This finding was in agreement with previous studies (37,(50)(51)(52). This can be justified in that maternal occupation and income are factors that provide information about the level of autonomy of the woman that could empower herself to take care of her child (53,54). However, this finding was in contrast with a study done in Ethiopia to determine risks of underfive diarrheal and fever morbidity (55). The possible reason is that mothers with skilled occupation were working at outside of their home and children of those mothers were not received the quality care by the servants/ care giver than were supposed to get from their mothers.
Occurrence of childhood illness was inversely associated with distance to health facility. Children who lived near to health facility received more care at the facility than those who lived far from a health facility (56,57). These finding was also consistent with other data from previous reports (9,28,58).
This can be explained by the fact that children living in families located far from health facilities are denied of the availability and get access to the preventive and treatment services for common health problems which in turn led to occurrence increased childhood morbidity.
Finally, this study was not without its limitation. Since data collections were based on mothers/caregivers responses that there could be a room for recall bias. Besides, had limitations about the actual symptoms of illnesses as defined in health care setting and also had gaps with respect to the actual and reported care seeking behavior of mothers/caregivers for childhood illness.
This study was not free of limitations. The morbidity data collected were subjective in the sense that morbidity data were based on the caregivers' perception of illness without validation by medical personnel. Moreover, this study used a recall period for six weeks and data was based on selfreported treatment seeking patterns, thus susceptible to recall bias and social desirability bias.

Conclusion
In this study, common childhood illnesses remains high while treatment seeking behavior of mothers'/care givers' was so low. Mothers' age and occupation, distance from nearest health center), and presence of three or more under five children in the house) were independent determinants of childhood illness. Thus, addressing significant determinants and enhancing mothers'/care givers health seeking behavior would help in reducing morbidity and mortality of under-five children with common childhood illness.

List Of Abbreviations
The following are some of the key abbreviations used in this study. ARI-acute respiratory infections, PNC-post natal care, EDHS-Ethiopian demographic and health survey, U5-under five, WHO-world health Organization, AOR-adjusted odds ratio, CI-confidence interval, SD-standard deviation, EMDHS -Ethiopia Mini Demographic and Health Survey

Ethical consideration
Ethical approval was obtained from the University of Gondar Institutional Review Board (IRB).
Permission letter was obtained from all local administrators and health managers. An informed consent was obtained from mothers/care givers. For mothers/care givers less than 18 years of old, assent was obtained. All study participants were informed that the participation was voluntarily. The potential benefits, harms, the confidentiality, and the possibility of withdrawing from the interview at any time of the interview and after were also informed.

Consent for publication
Not applicable

Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
The authors declare they have no competing conflicts of interest.

Funding
The study has no funding source.