Factors affecting healthcare-seeking for children below five years with symptoms of acute respiratory tract infection in Ethiopia: a Cross-sectional study based on the 2016 Demographic and Health Survey

(1)

1 | P a g e

Factors affecting healthcare-seeking for children below five

years with symptoms of acute respiratory tract infection in

Ethiopia: a Cross-sectional study based on the 2016

Demographic and Health Survey

Meron Tilahun

Master Degree Project in International Heath, 30 credits. Spring 2018

International Maternal and Child Health

Department of Women’s and Children’s Health

Supervisor : Andreas Mårtensson

Word count – 10,100

(2)

2 | P a g e

Abstract

Background

Acute respiratory tract infection is the main cause of death among children below the age of five in Ethiopia. Although the primary health care strategy have shown improvement in access, health care seeking remains low. This study investigates factors related to the healthcare-seeking behavior of mothers for children with symptoms of acute respiratory tract infections in Ethiopia.

Methods

The study was based on secondary analysis of cross-sectional population-based data from the 2016 Ethiopia Demographic and Health Survey (EDHS). A total of 432 children who had symptoms of acute respiratory tract infections in the two weeks prior to survey were analyzed. Logistic regression models were used to identify associated factors.

Results

More than two-third of the children who were reported to be having symptoms of acute respiratory tract infection in the two weeks preceding survey did not receive healthcare. One-thirds of children received care from public health care providers. Among predisposing factors, having a female child lowered the odds of seeking health care while maternal education increased the odds of seeking health care. Among enabling factors, as compared to male headed household, female headed households were 1.9 times more likely to seek health care for sick child. Mother’s in rural areas were less likely to seek care from health care facilities.

Conclusion

(3)

3 | P a g e

List of abbreviations

ARI Acute Respiratory tract Infection

ALRTI Acute Lower Respiratory Tract Infection aOR Adjusted Odds Ratio

AURTI Acute Upper Respiratory Tract Infection cOR Crude Odds Ratio

CI Confidence Interval

DHS Demographic and Health Surveys GBD Global Burden of Disease

GDP Gross Domestic Product

HSDP Health Sector Development Program

HibCV Hemophilius influenzae type b Conjugate Vaccine ICCM Integrated Community Case Management

IMCI Integrated Management of Childhood Illness LMIC Low and Middle-Income Countries

PCV Pneumococcal Conjugate Vaccines PHC Population and Housing Census

(4)

4 | P a g e

List of tables

Table 1: Description of variables used in this study with their code and categorization

Table 2: Absolute and percentage frequency distribution of participants’ socio-demographic characteristics based on Ethiopia DHS data, 2016 (n=432)

(5)

5 | P a g e

List of figures

Figure 1. Conceptual framework for healthcare-seeking : adapted and modified from Andersen Behavioral Model (47)

Figure 2 .Map of Ethiopia (Adopted from Ethiopia Atlas of Key Demographic and Health Indicators, 2016)

(6)

6 | P a g e

Globally, the death of children below the age of five declined from 12 million in 1990 to 5.6 million in 2016 (1). However, with the advances, several countries continue to register a higher rate of child deaths, particularly in sub-Saharan Africa (1). Most of these deaths resulted from preventable and treatable infectious diseases (1). Acute respiratory tract infections (ARIs), diarrheal diseases and malaria remain the leading causes of death among children below five years – accounting for more than a third of global deaths and about 40% of deaths in sub-Saharan Africa (1,2). According to the Global Burden of Disease (GBD) 2013 report, acute lower respiratory tract infection caused more than 2 million deaths in children younger than five years (3).

The respiratory tract infections among children also impose an intermittent and large burden on the infrastructure of healthcare. Severe forms of ARI are the most common reason for hospital admissions in children less than five years accounting for 60% of all hospitalizations and for 30% in-hospital deaths in children aged 0–4 years in 2010 (4). According to the World Health Organization 2016 report, ARI was the second largest cause of child death in the world and the top killer in sub-Saharan Africa (5). Data from national Demographic and Health Survey 2011 reported the prevalence of symptoms of ARI in Ethiopia to be 7% (6) .

1.2 Classification of ARI

(9)

9 | P a g e

the respiratory tract but can also extend to other systems causing common childhood infections including diphtheria, measles and pertussis(8).

1.2.1 Causes and symptoms of ARI

ARIs are commonly caused by bacteria and viruses (8). The most frequently detected viral causes of ARI include respiratory syncytial viruses (RSVs), parainfluenza viruses, influenza virus A and B, parainfluenza virus (PIV), human metapneumovirus (HMPV) and adenovirus (AV) (8,9). However, presence respiratory viral pathogen may reduce the body’s defense mechanism and antibacterial activities of the lungs which eventually leads to secondary bacterial infections (10). Coinfections with both causative agents is also common particularly in children younger than five years. The most frequently isolated ARI causing bacterial agents include Streptococcus pneumonia and Haemophilus influenzae type b (Hib). Other bacterial agents such as Mycoplasma pneumoniae, Chlamydophila pneumoniae, and Klebsiella pneumaniae are less frequently reported causes of ARI (10).

In children, the upper respiratory tract infections are mostly accompanied by cough, nasal congestion, sore throat and headache with fever. This infections can spread through direct hand contact with possibly infectious secretions or by respiratory droplets that were coughed or sneezed out by an infected person. Onset of the symptom usually appear one to three days after exposure to the causative agent and last up to 10 days, but may persist for longer periods (10,11). While the development of lower respiratory tract infections among children leads to shortness of breath, weakness, fever, cough and fatigue. The respiratory rate and lower chest wall indrawing are commonly used clinical signs to identify and manage potentially serious childhood lower respiratory illnesses (10).

The Demographic and Health Survey (DHS) defines childhood ARI symptoms to include any of the following conditions: cough accompanied by short, rapid breathing, and/or chest related difficult breathing reported in the two weeks prior to survey (12). These ARI symptoms are considered as a proxy for pneumonia which is one of the major cause of death in young children of Ethiopia.

1.2.2 Risk factors

(10)

10 | P a g e

pollution, overcrowding and malnutrition as major factors (13-23). A prospective cohort study on children younger than five years found increased risk of ARI for children with low birth weight (<2500g). The risk was six times higher compared to a normal birth weight infant (13). Furthermore, low birth weight babies were shown to have two times higher risk of ARI-related hospital admissions (13).

Malnutrition was found to increase vulnerability of children to ARI episodes (15,16). A study in Nigeria reported that malnourished children have three times higher risk of developing ARI than the well-nourished ones (16). The increased risk is largely due to impaired immune function that resulted from poor nutrition (15,16). Furthermore, the risk of dying was shown to be higher when children concurrently develop both conditions (18).

Exposure to indoor air pollution was described as one of the modifiable risk factor for childhood ARI (17,19,20).This condition increase the risk of acquiring ALRTI by 78% and was the cause for around one million ARI-related deaths worldwide (19). Indoor air pollution (IAP) tends to be high in low income settings due to the frequent use of solid fuels for household energy needs. IAP may increase the incidence of ARI through affecting the respiratory defense mechanism against pathogens (20).

On the other hand, exclusive breastfeeding for six months was shown to reduce the risk of childhood ALRTI (21,22). A population based cohort study on children 0-4 years reported that breastfeeding for six month or longer decrease the risk of ALRTI by 30% (21). Similarly, children exclusively breastfed for six month had fewer ARI episodes and hospital admissions than their counterparts (22). It has also been shown that suboptimal breastfeeding was the cause for about 44% of infection related deaths in the first year of life including those with caused by ALRTI (23).

1.2.3 Prevention and Treatment

(11)

11 | P a g e

The extensive application of vaccines against Hib, influenza, measles, pertussis, pneumococcus, and diphtheria can significantly reduce ARIs in children (26). The two vaccines included in national immunization programs of all countries, Bordetella pertussis and measles vaccines, and two recent vaccines, H.influenzae type b conjugate vaccine (HibCV) and pneumococcal conjugate vaccines (PCVs) significantly contributed to the prevention of childhood deaths due to acute respiratory episodes (26).

The HibCV reduces the risk of meningitis and pneumonia in young children. The vaccine is to be administered in a three-dose regimen before 14 months of age. The first dose is given at two months, followed by a 0.5 ml dose at four and six months. A booster dose of 0.5 ml is also recommended at 12-15 months of age (27). Using a quasi-randomized approach, a study in Bangladesh reported that HibCV resulted in 34-44% reduction in the risk of radiologically apparent pneumonia following at least 2 doses of the vaccine (28). Similarly, a case-control study in Pakistan estimated 62% reduction in radiologically confirmed pneumonia with three doses of HibCV (29).

PCVs, especially PCV13 provides immunity against 13 serotypes of S.pneumoniae species. PCV 13 immunization series consists of a 0.5 ml intramuscular injection administered to children at 2 months, 4 months, 6 months, and 12-15 months of age (30).

Appropriate therapy for ARI relies on accurate assessment of disease severity using clinical signs outlined in Integrated Management of Childhood Illness (IMCI) guidelines (31). For children aged 2-59 months presenting with fast breathing and no chest indrawing or danger sign, the guideline recommends treating a child as an outpatient with oral amoxicillin for three days. While children with chest indrawing (with or with-out fast breathing) should be initially treated with oral amoxicillin 40mg/kg twice daily for five days. Oxygen therapy can also be added if necessary (31).

1.3. ARI in Ethiopia

(12)

12 | P a g e

1.4 Health care delivery in Ethiopia

Health care delivery in Ethiopia uses a three-tiered public health delivery system (34). At the primary care level, a primary health care unit (PHCU) comprises primary hospital, health centers, and health posts. Health centers provide basic preventive and curative services and serves as a referral center and practical training center for community health workers. Secondary care includes general hospitals that provides inpatient and outpatient care and serves as a referral center for PHCU. Tertiary care consists of specialized hospitals that serve as a referral from general hospitals. Private for profit and non-governmental organizations also play a vital role in increasing health service coverage and utilization (34).

Following the principles of Alma Alta, the public health care system in Ethiopia adopted the primary care strategy (35). At the community level, the Community Health Workers (CHWs) program and Women’s Development Army (WDA) have been established. This programs were devised to reduce inequalities in access to essential health care services, particularly for underserved households and communities (35). The CHW structure includes Health Extension Workers (HEW) and a recently developed nationwide Women’s Development Army (WDA). Two HEWs are assigned to oversee one health post in Ethiopia. The WDA complement the service provided by CHWs who are each assigned 25 people (35). CHWs play an important role in disease prevention, health promotion and selected curative services at the community and household level. Under the Devolution Plan of the Government of Ethiopia, the regional governments have been given comprehensive administrative and financial autonomy in almost all sectors, including health (35). Thus the nine regions in the country are responsible for developing their own strategies, programs and interventions based on their locally generated data and identified needs. According to UNICEF 2006/2016 budget brief, the government of Ethiopia spent 1.6% of its GDP on health care on 2015/2016 fiscal year (36). This proportion of total health expenditure out of GDP is very low compared to other countries in sub-Sahara Africa and has resulted in significant personal out of pocket spending (36).

(13)

13 | P a g e

but also many informal providers including traditional or spiritual healers, traditional birth attendants, and herbalists. Non-governmental organizations (NGOs) are also active in the health and social sector. In urban parts of the country, some public–private partnership initiatives exist through franchising of private health outlets. These have been successful to a large extent in raising the level of awareness of positive health behavior among the people. For instance, the increasing contraceptive prevalence rate is due to the efforts of NGO sector and the CHWs (35,12). However, primary health care activities have not brought about expected improvements in health practices, especially of rural population groups. In rural Ethiopia, more than 90% of births are delivered at home (12). For over half the population (84%) living in the rural part of the country, poverty coupled with illiteracy, the low status of women and inadequate water and sanitation facilities have had a deep impact on health indicators (12). Beside limited knowledge of illness and wellness, cultural prescriptions, perceptions of a health service and provider and social barriers, cost has been a major barrier to the provision of an effective health service. This has affected the physical and financial accessibility of the health services.

1.5 Healthcare-seeking in Ethiopia

Healthcare-seeking behavior is an action undertaken by individuals to overcome perceived health problem or illness (38).

(14)

14 | P a g e

The health-seeking behavior of mothers’ of children with ARIs is considerably low in Ethiopia (6,40). In order to improve it, programs should be initiated that majorly target the mothers with less education, particularly those living in rural areas of the country. Early marriages are prevalent in Ethiopia denying the young mothers with sufficient maturity to care for the child (6). The healthcare-seeking behavior in Ethiopia is influenced by environmental factors, the government policies and individual beliefs and culture (6).

1.6 Factors affecting Healthcare-seeking for ARI

Child morbidity and ill-health complications could be prevented by appropriate medical care seeking. Appropriate health-seeking behavior has the potential of improving the survival rates of children (39). Health seeking-behavior can be influenced by, the caretaker’s age, wealth index, rural or urban setting, and education, the symptom types, family size, the perception of illness severity, beliefs, and the ability to recognize the signs and symptoms of ARIs (40-45).

The caretaker’s age affects the health-seeking behavior. Younger mothers in the rural areas between the ages of 15 and 34 have a higher health-seeking behavior than older mothers above the age of 35 (41). This can be attributed to the ease and ability of the younger mothers to have access to mass media due to higher levels of education. Wealth index, especially in rural residence, also affects health-seeking behavior (41,44). Mothers or families with higher wealth index in the rural areas can afford the cost of transport to the health facilities. It is, therefore, prudent that the government provides a cheaper and faster means of transport to the health facilities or bring the health centers closer to the people. The inability of the caregiver to recognize the potentially fatal conditions and seek numerous care practices causes a significant delay in receiving the relevant health care (40). The delay can result in health complications and cause the health care service ineffective.

(15)

15 | P a g e

1.7 Rationale of the study

The continued advancements in medicine and healthcare have significantly improved health. Amidst these advances, the childhood infections continue to impose significant burden due to delays in seeking prompt and appropriate care or not seeking health care at all. A cross-sectional study conducted in six high child mortality countries of sub-Saharan Africa reported that 85% of children with suspected ALRTI were taken to a provider by their caregiver in Tanzania, as opposed to only 30% in Ethiopia where care seeking was found to be the lowest among the six countries (40). The reason could be because of numerous care-seeking practices and the difficulty of the caregiver’s to recognize potential fatal conditions.

ARI continues to be one of the major health issues in Ethiopia, especially when there has been limited evidence regarding the healthcare-seeking behavior and associated factors. Therefore, analysis of care seeking behaviors by caregivers, and child, caregiver and household characteristics associated with care seeking is required to optimize future strategies and achieve sustainable development goal targets on child survival by 2030.

1.8 Theoretical framework

(16)

16 | P a g e

Figure 1. Conceptual framework for healthcare-seeking : adapted and modified from Andersen Behavioral Model (46)

1.9 Aim of study

The aim of the study was to evaluate factors influencing healthcare-seeking for ARI symptoms in children below the age of five years based on data from EDHS, 2016.

Specific objectives

Ⅰ. To determine the prevalence of healthcare-seeking for children below the age of five years with symptoms of ARI in Ethiopia.

Ⅱ. To evaluate socio-demographic (predisposing), enabling and external factors affecting healthcare-seeking behavior of mothers for children with symptoms of ARI.

1.10 Research question

What is the prevalence and the factors influencing healthcare-seeking for children below five years of age with symptoms of ARI in Ethiopia?

2. Methods 2.1 Study setting

(17)

17 | P a g e

the age of 15 and 14.6% under the age of five years . The average household size was four persons with the urban population having a smaller mean household size than the rural population (12). Based on the World Bank 2007 report, the average life expectancy at birth was 58 years (47). The federal government structure is subdivided into nine regional states namely: Tigray, Afar, Amhara, Oromia, Somali, Southern Nation Nationalities and Peoples Region (SNNPR), Benishangul-Gumuz, Gambella, and Harari regions and two city administrative councils: Dire Dawa and Addis Ababa. The regional states as well as the two city administrative councils are further divided into Woredas (districts) and Kebeles (sub-districts) (12). A woreda/district is the basic decentralized administrative unit and has an administrative council . The woredas are further divided into kebeles (villages) organized under peasant associations in rural areas and urban dwellers’ associations in towns (12).

Figure 2 .Map of Ethiopia (Adopted from Ethiopia Atlas of Key Demographic and Health Indicators, 2016)

2.2 Study design

(18)

18 | P a g e

2.3 Study population

This study was based on children below the age of five years who had ARI symptoms (i.e. cough accompanied by short, rapid breathing) in the two weeks preceding the Ethiopia Demographic and Health Survey (EDHS) which was carried out between January – June 2016. The 2016 EDHS included 10641 children of women born in the last five years (12). For this study, a sub-sample of children with a reported history of cough with short, rapid breathing in the two weeks prior to survey were analyzed for being taken to health care facility (Figure 3).

2.4 Sampling

2.4.1 Sampling technique

The sample for the 2016 EDHS was nationally representative covering the whole country. The survey employed a sampling frame from the Ethiopia Population and Housing Census, 2007 (PHC) (12). The sampling frame developed for 2007 census contains details about the census location, type of residence (urban or rural), and estimated number of residential households in survey areas. With the exception of census areas in six zones of the Somali region, each area has accompanying sketch map describing the geographic boundaries. Satellite maps were used for the remaining six zones in Somali region.

(19)

19 | P a g e

The 2016 EDHS was carried out nationally between January and June, 2016. All women age 15-49 and all men age 15-59 who were either permanent residents of the chosen households or visitors who stayed in the household the night before the survey were eligible to participate (12).

From the households included in the survey, a total of 16,583 women were found to be eligible for individual interviews and 15,683 of them were successfully interviewed which translates to a response rate of 95% for the women (12).

2.4.2 Data collection

Data was collected from January 18, 2016 to June 27, 2016 by 36 teams, each consisting of a team supervisor, a field editor, three female interviewers, one male interviewer, two biomarker technicians and one driver (12).

EDHS 2016 used five questionnaires to collect data on key demographic and health indicators at the national level. The questionnaires were originally written in English and then translated to three major local languages: Amarigna, Tigrigna, and Oromiffa. The questionnaires were pretested in selected clusters so as to modify questionnaire translation. Following the pretest, training of field staffs was made with the modified questionnaire. The survey used five types of questionnaires: the Household Questionnaire, the Woman’s Questionnaire, the Man’s Questionnaire, the Biomarker Questionnaire, and the Health Facility Questionnaires (12).

Information on age, sex, marital status, education, and relationship to the head of the household was collected using the Household Questionnaire. The data on age and sex of household members obtained in the Household Questionnaire were used to identify women between 15-49 years and men between 15-59 years eligible for individual interviews. The Household Questionnaire also collected data on the characteristics of the household’s as a dwelling unit, such as source of water, type of toilet facilities , asset ownership, and household construction materials. An additional form to estimate the prevalence of injuries among household members was also included in the Household Questionnaire (12).

(20)

20 | P a g e

diarrhea, fever, and cough as well as information on treatment of childhood diseases was collected using this questionnaire (12).

The 2016 EDHS also included a Health Facility Questionnaire. This questionnaire was used to document vaccination information for all children without a vaccination card which was identified through the Woman’s Questionnaire.

2.5 Variables

2.5.1 Outcome variable

The main outcome variable was whether or not mothers had sought health care from a government health facilities and private hospitals/clinics for their child with symptoms of ARI. The responses were categorized into “yes” if care was sought from government hospitals, health centers, health post and private hospitals/clinics or “no” if care was not sought from any government or private health facilities for childhood ARI symptoms (Table 1).

2.5.2 Predictor variables

This study contained sixteen independent variables which were analyzed for their association with healthcare-seeking for children with ARI symptoms. These variables were identified and included in this study based on previous literatures conducted in similar topic. The predictor variables which were assessed for their association with healthcare-seeking have been divided into variables related to the predisposing factors, enabling factors, and external factors based on the theoretical framework for this study (46).

Predisposing factors Child Characteristics

 Child age in months: Categorized as 0-11 months, 12-23 months, 24-35 months, 36-47 months, 48-59 months.

 Sex of child: Categorized as male or female. Maternal Characteristics

 Maternal age : This was maternal age in completed years grouped into three categories, 15-25 years, 26-36 years, 37-49 years.

(21)

21 | P a g e

 Current Marital status : This was categorized into three categories single, married, widowed/divorced/separated.

 Maternal occupation: Categorized into not working, professional/technical/managerial, agricultural-self-employed, unskilled and other.

Enabling factors

 Household Wealth index : This index was created from data collected on individual household’s possession of selected assets like television and bicycle, materials used for household construction and type of water source and sanitation facilities. Through principal component analysis of the collected information, households were categorized into five wealth quantiles as poorest, poorer, middle, richer and richest categories.

 Number of dejure household members: This was the number of household members that usually live in the household. It was categorized into three groups as 1-5 members, 6-10 members, and 11-16 members.

 Number of children five and under in the household: This measures the number of children who were residents in the household and aged 0-59 months. Visiting children were excluded. It was categorized into two groups as 0-2 children, 3-6 children. Up to two children in the household was used as the reference category.

 Sex of the Household Head : Categorized as male or female. Male was used as the reference category.

 Respondent’s perception of distance to health facility: The survey included questionnaire to assess if distance to health care facility was a problem from getting a medical advice or treatment. The responses were placed into three categories: those who experience difficulty in reaching medical service were categorized as having ‘big problem’ and those with no such problem were categorized as having ‘no problem’ and ‘not a big problem’ to reach health service. The last category was merged into no problem.

(22)

22 | P a g e

and ‘1’ for the rest of the responses. Then a numerical variable containing the sum scores (0-3) was created. Finally, those having exposure for at least one media service (i.e. more than zero value in the sum scores) were categorized as ‘yes’ while those with zero score for any media were categorized as ‘no’.

External factors

 Place of residence: Categorized as urban and rural.

2.6 Statistical analysis 2.6.1 Data management

For this study, Child Recode file and Household Recode file from EDHS 2016 were used. The datasets were downloaded from DHS website and saved in .sav file format. The Child Recode file was used as a base and merged with Household Recode file using R commander Version 2.3-1. The DHS VI recode manual was used to identify and describe the variables used for this study. The working dataset was limited to children aged 0-59 months with reported symptoms of cough accompanied by short, rapid breath in the two weeks prior to survey. The DHS questionnaires obtained data on the respiratory symptoms of children aged 0-59 months. The responses were obtained by asking mother’s weather their children had cough in the two weeks preceding survey. For children who had a cough the mothers were further asked if the child also suffered from short, rapid breaths related with a chest or blocked rainy nose. Children having all the above three symptoms were identified from the dataset and those with ‘no’, ‘don’t know’ and missing responses on each of the symptoms were excluded. Additionally, those who had cough a day prior to survey were not included. The dataset contained twenty two variables relevant to the study and all the other variables were deleted. These variables were re-coded and categorized as described in Table 1. Data was analyzed using R statistical software Version 3.2.2 with R commander statistical package (Rcmdr) Version 2.3-1.

2.6.2 Statistical methods

(23)

23 | P a g e

performed for each of the explanatory variables and the outcome variable. Variables excluding the null value of one in the bivariate analysis were considered for multivariate analysis. The crude and adjusted odds ratio from the regression analysis were reported with corresponding 95% confidence intervals (Table 3).

2.6.3 Missing Values

The outcome variable had four observations with NA’s (not applicable) that were excluded. In addition, pairwise checks among predictor variables was performed by means of Chi-squared test to determine multicollinearity. The variable ‘Household has electricity’ was excluded from analysis due to its multicollinearity with the variable ‘Exposure to media’ (p<0.001). In addition, the variable region was excluded from the regression analysis due to multicollinearity with place of residence.

3. Ethical considerations

Ethical approval for EDHS 2016 was obtained from the National Research Ethics Review Committee (NRERC) at the Ministry of Science and Technology in Ethiopia and the Institutional Review Board of ICF International.Approval was also granted by the DHS program to use the dataset for the present study. Prior to each interview, well informed written consent was obtained from survey participants. In addition, an informed consent was obtained from parent or caregiver prior to participation by a child. Participants were informed about the purpose of the study, expected duration and the potential risk and benefits of participation. Participation was entirely voluntary. The respondents were fully aware that they had the right not to participate or to withdraw from the study at any time during the survey. Confidentiality was maintained at all steps of data collection and processing. Survey respondents were identified only by an assigned code to ensure their privacy and well-being.

4. Results

(24)

24 | P a g e

Figure 3. Flow diagram of participants included in the study from EDHS, 2016

4.1 Characteristics of study participants

There was a balanced distribution of children with regards to sex. As shown in table 1, slightly more than one third of the children were aged between 36-47 month. Among the mothers, half of them were in the age group of 26-36 years and more than one third were aged 15-25 years. About 66% of the mothers had no or little formal education followed by primary education (27%) and the least had secondary and higher educational status (7%). About 69% of the mothers in the sample had no exposure to media that included newspaper, radio and television. The majority of mothers were currently married (95%) and the rest were either single, widowed, separated or divorced. Regarding occupation, more than half of the mothers were not employed (55%), about 19% were employed in either professional, managerial, and technical jobs and around 22% were engaged in agricultural sector. The sample consisted 42% Muslim participants and those belonging to Traditional faiths followed by 16% Protestant respondents. The rest belong to Orthodox and Catholic faiths.

(25)

25 | P a g e

had one to five members while 4% had more than eleven members living in the same household. Only 17% of the households had three to six children who were residents in the household during survey while 83% had a maximum of two children in the household. A higher percentage of mothers (65%) had experienced difficulty in getting to health facility. As shown in table 1, males headed about 85% of the sampled households. Over 80% of the households had no electricity. The majority of households reside in Oromia region (27%) followed by Tigray (17%) and Amhara region (15%). Harari region had the least number of respondents (1%) during the survey. A majority of the respondents were living in rural areas (87%).

Table 2: Absolute and percentage frequency distribution of participants’ socio-demographic characteristics based on Ethiopia DHS data, 2016 (n=432)

Characteristics Total population

n %

Predisposing factors

Child’s age in months

0-11 12-23 24-35 36-47 48-60 48 71 65 130 115 11.19 16.55 15.15 30.30 26.81 Sex of child Male Female 220 212 50.93 49.07

Maternal age in years

(26)

26 | P a g e

No education Primary

Secondary and higher 287 115 30 66.44 26.62 6.94 Exposure to media

(newspaper, radio, television) No Yes 296 136 68.52 31.48 Religion

Orthodox and Catholic Protestant

Muslim and Traditional 180 69 183 41.66 15.97 42.37

Current marital status

Single Married Widowed/separated/divorced 3 412 17 0.69 95.37 3.94 Maternal occupation Not working Professional/technical/managerial Agricultural-employee Others 237 80 97 18 54.86 18.52 22.45 4.17 Enabling factors

Household wealth index

(27)

27 | P a g e

Richest 58 13.43

Number of dejure household members

1-5 6-10 11-16 194 221 17 44.91 51.16 3.93

Number of children five and under in household 0-2 3-6 360 72 83.33 16.67

Sex of household head

Male Female 367 65 84.95 15.05

Mother’s perception of distance to health facility No problem Big problem 150 282 34.72 65.28

Household has electricity

(28)

28 | P a g e Gambela Harari Addis Ababa Dire Dawa 13 4 12 16 3.01 0.93 2.78 3.70 Place of residence Urban Rural 57 375 13.19 86.81

4.2 Prevalence of healthcare-seeking for symptoms of ARI

One of the objectives of this study was to determine the prevalence of care-seeking among mothers of children below five years with ARI symptoms in the two weeks preceding the survey. Of the 432 children included in the study, healthcare was sought for 158 (36.5%) children with ARI symptoms that occurred during the two weeks prior to survey. Among those who seek health care for ARI symptoms, the proportion of children who seek care from government facilities was 31.0% and from private facilities was 5.5% (Data not shown). The result that majority consulted public health facilities might be because of the provision of health services at subsidized cost or the relative accessibility of public facilities as compared to private facilities (35).

4.3 Factors influencing healthcare-seeking for symptoms of ARI 4.3.1 Bivariate analysis

(29)

29 | P a g e

wealth index, number of children below five years and number of dejure household members showed no significant association with healthcare-seeking practice on crude analysis (Table 3).

4.3.2 Multivariate analysis

For multivariate analysis six variables fulfilled the criteria (confidence interval without the null value of one) and were included in the final model. The variables were sex of child, maternal education, exposure to media, sex of household head, mother’s perception of distance to facility, and place of residence (significant values highlighted in bold in Table 3).

(30)

30 | P a g e

Table 3 Crude and Adjusted odds ratios with 95% confidence intervals from logistic regression model for healthcare-seeking in relation to predisposing, enabling, and external factors in 0-59 month children who had ARI symptoms in the two weeks preceding survey using Ethiopia DHS, 2016 (n=432)

Predictor variables Crude analysis Adjusted analysis

OR (95%CI) OR (95%CI)

Predisposing factors

Child’s age in months

0-11 1.47 (0.73 – 2.92) 12-23 2.18 (1.00 – 3.98) 24-35 1.59 (0.85 – 2.96) - - 36-47 Reference 48-59 0.98 (0.56 – 1.69) Sex of child

Male Reference Reference

Female 0.65 (0.42 – 0.72) 0.64 (0.42 - 0.97)

Maternal age in years

15-25 1.54 (1.00 – 2.38)

26-36 Reference - -

37-49 0.91 (0.49 - 1.62)

Maternal education

No education Reference Reference

Primary 1.77 (1.13 - 2.76) 1.58 (1.29 - 2.51)

Secondary and higher 3.33 (1.55 - 7.39) 1.92 (1.80 - 4.66) Religion

Orthodox and Catholic Protestant Muslim and Traditional

1.07 (0.69 – 1.65) 1.46 (0.82 – 2.57)

Reference

- -

Exposure to media (newspaper, radio, television)

No Reference Reference

Yes 1.82 (1.20 - 2.77) 1.41 (0.88 - 2.25)

Current marital status

Single 3.60 (0.34 -77.97)

Married Reference

Widowed/separated/divorced 2.02 (0.75 – 5.50)

Maternal occupation

Not working Reference

Professional/technical/managerial 1.29 (0.77 - 2.16)

Agricultural-employee 0.70 (0.42 - 1.17) - -

(31)

31 | P a g e

Enabling factors

Household wealth index

Poorest Reference

Poorer 1.07 (0.60 - 1.89)

Middle 1.42 (0.77 - 2.57) - -

Richer 1.87 (1.00 - 3.43)

Richest 2.99 (0.30 - 5.71)

Number of dejure household members

1-5 1.64 (1.00 – 2.46)

6-10 Reference

11-16 1.20 (0.40 - 3.29) - -

Number of children five and under in household

0-2 Reference

3-6 0.72 (0.41 - 1.23) - -

Sex of household head

Male Reference Reference

Female 2.31 (1.35 – 3.96) 1.94 (1.10 - 3.42)

Mother’s perception of distance to health facility

No problem 1.62 (1.37 -2.43) 1.13 (0.71 - 1.77)

Big problem Reference Reference

External factors

Place of residence

Urban 2.99 (1.69 – 5.35) 1.73 (1.59 - 3.40)

Rural Reference

Numbers in bold – Significant at 95% CI , OR -odds ratio, CI- Confidence Interval

5. Discussion 5.1 Main findings

(32)

32 | P a g e

mothers/caregivers about severity of childhood respiratory symptoms. Further, poor perception towards health facilities and cultural beliefs and practices may also influence acceptability and utilization of health care services.

Here it is important to note that, ideally healthcare should be sought for all children with ARI symptoms although the health care infrastructure and structural factors hinder the quality and quantity of health care service. Prompt healthcare seeking for ARI symptoms is crucial considering the high burden of child death in Ethiopia due to ARI, particularly pneumonia (63).

5.2 Study findings in relation to other studies

The study result identified significant factors associated with healthcare-seeking for children with symptoms of ARI including: sex of child, maternal education level, sex of household head, and place of residence. Each factor is considered separately based on the theoretical framework of this study and discussed in the context of Ethiopia.

5.2.1 Predisposing factors Sex of child

(33)

33 | P a g e

Maternal education

Mother’s level of education was a significant predictor of healthcare-seeking for children after controlling for other demographic, socioeconomic, and external factors.The odds of seeking health services for sick children significantly increased with maternal level of education. This is in line with what has been reported in Ethiopia (41) and Nigeria (44). This may be due to the fact that when the primary caregivers have the relevant educational level, they can have the access to information on the signs and symptoms of the ARIs. The caregivers would also have the necessary information and means to seek medical care for the children for early diagnosis, treatment, and prevention of the diseases. In addition, the level of education influences the ability of the caregiver to seek medications such as vaccinations and early hospital visits in case of sickness (53). This result however disagrees with a recent study conducted in India (54). This study used demographic and health survey data to examine the effect of social and economic factors on the decision of parents to seek care for their children and on their choice of providers. But the study found no significant association between healthcare-seeking and mothers level of education. The fact that maternal level of education in the current study remained statistically robust and that other measures such as household wealth and employment status did not mediate the effects of this variable means that being wealthy or employed among Ethiopian women do not necessarily affect their decision making regarding seeking professional care from health facilities. Rather being educated had an independent net effect to seek care for sick children.

5.2.2 Enabling factors Sex of household head

(34)

34 | P a g e

women and their children. In this regard, providing formal education could bring social and economic liberty which could enhance their chance of accessing professional care from health facilities (53).

5.2.3 External factors Place of residence

The finding of this study showed that mothers who live in urban areas were more likely to seek care from the health facilities as compared to the rural mothers even after controlling for mother’s perception of distance to health care facility (aOR= 1.73; 95% CI: 1.59-3.40). This finding was consistent with a study conducted by Astale and Chenaut et al (41) which found statistically significant differences in care-seeking between urban and rural areas. Another community based cross-sectional study in Northwest Ethiopia reported similar findings (56). According to this community based study, the mothers dwelling in the urban areas were 11.5 times more likely to seek medical care than their rural counterparts (56). The study based this finding on the accessibility of the healthcare centers and hospitals to the urban mothers than for the rural mothers. From the findings shown in similar setting, it can be concluded that there is a significant disparity of the health-seeking behavior of the mothers in the urban and rural dwelling of Ethiopia. This disparity can be attributed to the accessibility of healthcare and information for childhood illness in the urban setting than in the rural areas. The results of the current study is also consistent with previous cross-sectional study conducted in six high pneumonia mortality countries in sub-Saharan Africa (40). The urban mothers seek medical attention for their children on the first day of the onset of the illness more often than the rural mothers. The delay in seeking medication in most of the rural areas can be attributed to the wealth index, socio-cultural factors, and seeking alternative home remedies. The poverty in most of the rural areas makes seeking healthcare a choice. It is, therefore, prudent that the government in partnership with other private organizations works on strengthening the accessibility of the healthcare services in rural areas and increase awareness on the need and importance of prompt care-seeking.

5.3 Non-significant variables

(35)

35 | P a g e

size used for the current study. On the contrary, other studies conducted in sub-Saharan Africa (41), Nepal (42), and India (58) found age of the child to have a significant effect on care-seeking for symptoms of ARI. The studies reported that the odds of seeking health care was higher for younger children than for older ones.

This study found no significant association between health care seeking for ARI symptoms and mother’s age. This result is in agreement with those studies conducted in other low income settings (44,58). Contrary to this, a study by Astale and Chenaut et al (41) demonstrated that care seeking is highly prevalent among younger mothers between the ages of 15 and 34 than in older mothers above the age of 35. However, this is also dependent on whether they are in a rural or urban setting. Although it is not certain how young age could be an advantage in this regard, it can be argued that younger mothers could have more exposure to the various forms of mass media than the older women which enhances their awareness and health-seeking behavior once the children experience any form of health complications. A study in Indonesia suggested that cultural factors and beliefs might affect the health-seeking behavior of older women especially in the rural settings. In rural areas older caregivers prefer to treat childhood illnesses with medicines available at home or herbal medicines or home remedies before visiting a health facility (59).

Several studies from LMIC reported that children from the wealthier households are more likely to be taken to health facilities than children from the poorer households (39,40,44). In this study, also children from the richest households were more likely to be taken to care than those from the poorest category, although this finding was statistically insignificant. Moreover, the results of the present study also showed that only a small percentage of the participants are in the higher and highest wealth quintiles and most of the population in the sample are in the lowest wealth quintile. This could have affected the significance of the results and could also be associated to the fact that most of participants in this study live in rural areas (87%). This shows unequal distribution of wealth in the regions and especially in rural areas. Further study that examine the impact of household wealth on healthcare-seeking may be required.

(36)

36 | P a g e

This result might be due to the wider gap in number of under five children observed in the households included in the present study.

The number of household member has been suggested to have an influence on healthcare-seeking behavior (41). Previous research in Ethiopia indicates that smaller family size allowed the parents to invest adequate time and money in the sickening child. Based on a study from Astale and Chenaut et al, a smaller family size was a significant factor to seek health care in both urban and rural family settings (41). The average family size in their study was 3.7 people in urban households while in rural households, the average family size was 4.9 people (41). The smaller family size enables the family to comfortably seek health care for all the family members appropriately. This help in early diagnosis, treatment, and prevention of the respiratory infections. In this study number of household members did not influence care seeking for ARI symptoms. This finding agrees with a study by Kolola et al that reported no significant association between household size and healthcare-seeking (45).

Other studies have suggested that working mothers seek health service for their children more than non-working mothers (44,56). This may be due to the fact that working mothers may have more financial ability to seek health care for their children than unemployed mothers. The current study found no significant association between maternal occupation and healthcare-seeking for children with ARI symptoms.

The present study reported difference in healthcare seeking for mothers who experience difficulty in getting to health care facility versus mothers who do not experience difficulty in reaching health facilities .The odds of seeking care for mothers who reported no problem for distance to health facility is 1.7 times higher than those experience problem in reaching health care facilities. Previous studies have also supported this result (44,57,58). However, when the variable was controlled for other predisposing and enabling factors the effect is attenuated. This may be partly explained by the small sample size used for analysis that limit any significant finding.

5.4 Study results in relation to conceptual framework

(37)

37 | P a g e

the need to promote female school enrolment, health extension programs and establishment of more primary health care centers in rural areas.

5.5 Strength and Limitations

This study was limited through excluding care sought from informal sources such as local pharmacy shops, traditional practitioners and faith healers. Having this additional information would allow to determine the primary source of care and associated factors related with the choice. Other potential limitations for this study include recall bias and non-availability of some variables in survey dataset. In addition, the final sample used for the analyses included only those respondents’ who reported their children to be having ARI symptoms during previous two weeks. This may lead to recall bias that could affect the accuracy of the study. Moreover, mother’s report about child’s ARI symptoms may not have been correct due to recall bias and there might also be misconceptions about childhood ARI symptoms as a result of low educational level of respondent’s. This may have led to under- or over-reporting of childhood ARI symptoms. The study also included children between the age of 0 to 59 months where identifying sign and symptoms is rather difficult in the very young once. Further to this, survey data did not have any means to justify the reports of ARI symptoms given by mothers. Data about the presence of health facilities and their accessibility which influence care-seeking were not included in the analysis. However, place of residence and mother’s perception of distance to health care facilities were used as proxy for these explanatory variables. Furthermore, information about perceived severity of illness and health beliefs which may affect use of health services was not included . The current study did not take into account the impact of seasonal changes which may play a role in the prevalence of childhood ARI symptoms and care seeking behavior. Additionally, cultural and social norms of the study population were not included in the analysis of the study. Despite the above mentioned limitations, the study used a population based cross-sectional data that included all the regions of the country. The other strength was the use of a conceptual framework based on Andersen’s health behavior model which is a widely used method for evaluating potential factors associated with health services utilization.

5.5.1 Internal and External validity

(38)

38 | P a g e

was also made to include variables that potentially influence care seeking behavior. However, the fact that the factors related to healthcare-seeking are numerous and the effects of uncontrolled events might affect internal validity regardless of use of a large scale demographic data and the choice of statistical methods used to detect the associations. The effect of selection bias is minimal for this study as children with the described symptoms of ARI (i.e. cough accompanied with short, rapid breathing) was included in the analysis which could be used to measure care seeking behavior by their caregiver. Consequently, the findings of the study could be generalized to children who have had ARI symptoms or to a population having similar characteristics with the present study.

5.6 Public health relevance and recommendations

Although there have been progress regarding child survival globally, about one child out of 13 dies before reaching the age of five in sub-Saharan Africa compared with 1 in 189 for those who are born in high- income countries. The Sustainable Development Goal on child survival targets to reduce the number of deaths of children below the age of five by as low as 25 deaths per 1000 live birth between 2017 and 2030, yet current estimates suggest that more than three quarters of all countries in sub-Saharan Africa will not achieve the under-five mortality target, and 13 countries in the same region will not reach the target until after 2050 (60).The government of Ethiopia aimed to reduce the number of under five deaths to at least 29 deaths per 1000 live births by 2019/2020 (35). To achieve this target a national new born and child survival strategic plan was devised to assist in case management of childhood infections and reduce inequity in access to and utilization of newborn and child health services (35). In this aspect assessment of care seeking and associated factors is an important step towards achieving the goal. Such research will have an impact on the direction or implementation of currently launched health reforms in Ethiopia. It will rationally inform the decision makers in government and private sector about the re-structuring of the health care administration and re-designing the interventions. Such research could also be instrumental in identifying the possible ways of partnership and collaboration to strengthen the entire health system.

(39)

39 | P a g e

towards improving child survival and achieving development goals. Although the government of Ethiopia employed a primary health care approach to deliver maternal and child health care services, this need to be further strengthened to bridge disparities in access to life saving interventions including diagnosis to suspected pneumonia and treatment to other childhood illness. Further to this, the demand for health care services could further be increased through promoting women’s education. Educating women present an opportunity to increase their understanding of health priorities and medical care needs.

Study finding showed that women empowerment is an important determinant of healthcare-seeking as that should be given attention in countries like Ethiopia. It has been argued that financially empowered women can play a more active role in household decision-making and have greater potential to increase spending on education and health (61,62). Women empowerment increase the freedom of choice and action to shape women’s lives and consequently not only contributes to individual woman, but to the family, society and the country as a whole. Women empowerment is considered as necessary determinant for health, although it is not a sufficient condition. Empowering women has numerous positive impacts that promote and ensure increased access to knowledge, social and economic resources, and greater involvement and participation in individual and household decision making processes (61,62).

It is also recommend to strengthening community-based preventive and curative health services by means of Health Extension Workers to address the problem of childhood deaths from preventable and easily treatable diseases.

5.7 Conclusion

(40)

40 | P a g e

improved by strengthening the countries educational system and enhancing their access to education. Good education may have improve mothers decision-making power in the household and their understanding of the importance of early treatment and prevention resulting in an increased likelihood of their healthcare-seeking behavior. Access to quality education also impacts socioeconomic status which would enhance healthcare-seeking behavior for childhood illnesses and general health service utilization. Hence, women’s education positively impacts socio-economic development and national productivity as well as increase the household wealth status of families and communities. Further research is vital to conceptualize the habits and practices of caregivers during childhood sickness.

6. Acknowledgment

I would like to express my sincere gratitude to Swedish Institute (SI) for providing me with a scholarship to purse this master’s program.

I would also like to thank all IMCH teachers, staffs for their support and guidance. A very special gratitude to my supervisor Andreas Mårtensson for the continuous support, insightful comments and encouragement during the thesis writing, and Erik Olsson for the constructive feedback. Thank you Gbemisola Allwell-Brown for your comments and motivation.

I am very grateful to my fellow classmates and peer readers for the inspiration and valuable inputs. I learned a lot from you all !

This journey would not be possible without the continuous support of family and friends. Many Thanks!

(41)

41 | P a g e

1.WHO | Children: reducing mortality [Internet]. WHO. Available from:

http://www.who.int/mediacentre/factsheets/fs178/en/

2. Under-Five Mortality [Internet]. UNICEF DATA. Available from: //data.unicef.org/topic/child-survival/under-five-mortality/

3. GBD 2013 Mortality and Causes of Death Collaborators. Global, regional, and national age-sex specific all-cause and cause-specific mortality for 240 causes of death, 1990–2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet. 2015; 385: 117–171

4.Nair H, Simões EA, Rudan I, Gessner BD, Azziz-Baumgartner E, Zhang JSF, et al. Global and regional burden of hospital admissions for severe acute lower respiratory infections in young children in 2010: a systematic analysis. The Lancet. 2013 Apr;381(9875):1380–90.

5.WHO | Children: reducing mortality. WHO. Available from:

http://www.who.int/mediacentre/factsheets/fs178/en/

6. Agency/Ethiopia CS, International ICF. Ethiopia Demographic and Health Survey 2011. 2012 Available from: http://dhsprogram.com/publications/publication-fr255-dhs-final-reports.cfm 7.SEER Training: Conducting Passages [Internet]. Available from:

https://training.seer.cancer.gov/anatomy/respiratory/passages/

8. Simoes EAF, Cherian T, Chow J, Shahid-Salles SA, Laxminarayan R, John TJ. Acute Respiratory Infections in Children . The International Bank for Reconstruction and Development / The World Bank; 2006. Available from: https://www.ncbi.nlm.nih.gov/books/NBK11786/

9. Richter J, Panayiotou C, Tryfonos C, Koptides D, Koliou M, Kalogirou N, et al. Aetiology of Acute Respiratory Tract Infections in Hospitalised Children in Cyprus. PLOS ONE. 2016 Jan 13;11(1):e0147041.

10.Bhuyan GS, Hossain MA, Sarker SK, Rahat A, Islam MT, Haque TN, et al. Bacterial and viral pathogen spectra of acute respiratory infections in under-5 children in hospital settings in Dhaka city. PLOS ONE. 2017 Mar 27;12(3):e0174488

11. Cotton M, Innes S, Jaspan H, Madide A, Rabie H. Management of upper respiratory tract infections in children. S Afr Fam Pract (2004). 2008;50(2):6–12.

12. CSA/Ethiopia CSA-, ICF. Ethiopia Demographic and Health Sruvey 2016. 2017; Available from: http://dhsprogram.com/publications/publication-FR328-DHS-Final-Reports.cfm

13. Martins ALO, da Silva Fernandes Nascimento D, Schneider IJC, Schuelter-Trevisol F. Incidence of community-acquired infections of lower airways among infants. Revista Paulista de Pediatria (English Edition). 2016 Jun 1;34(2):204–9.

(42)

42 | P a g e

15. Katona P, Katona-Apte J. The Interaction between Nutrition and Infection. Clin Infect Dis. 2008 May 15;46(10):1582–8.

16.Rodríguez L, Cervantes E, Ortiz R. Malnutrition and Gastrointestinal and Respiratory Infections in Children: A Public Health Problem. Int J Environ Res Public Health. 2011 Apr;8(4):1174–205.

17.Ujunwa F, Ezeonu C. Risk Factors for Acute Respiratory Tract Infections in Under-five Children in Enugu Southeast Nigeria. Ann Med Health Sci Res. 2014;4(1):95–9.

18.Chisti MJ, Salam MA, Bardhan PK, Faruque ASG, Shahid ASMSB, Shahunja KM, et al. Treatment Failure and Mortality amongst Children with Severe Acute Malnutrition Presenting with Cough or Respiratory Difficulty and Radiological Pneumonia. PLOS ONE. 2015 Oct 9;10(10):e0140327.

19.WHO | Indoor air pollution from unprocessed solid fuel use and pneumonia risk in children aged under five years: a systematic review and meta-analysis. WHO.Available from:

http://www.who.int/bulletin/volumes/86/5/07-044529/en/

20. Nandasena S, Wickremasinghe AR, Sathiakumar N. Indoor air pollution and respiratory health of children in the developing world. World J Clin Pediatr. 2013 May 8;2(2):6–15.

21.Tromp I, Jong JK, Raat H, Jaddoe V, Franco O, Hofman A, et al. Breastfeeding and the risk of respiratory tract infections after infancy: The Generation R Study. PLOS ONE. 2017 Feb 23;12(2):e0172763.

22. Ladomenou F, Moschandreas J, Kafatos A, Tselentis Y, Galanakis E. Protective effect of exclusive breastfeeding against infections during infancy: a prospective study. Archives of Disease in Childhood. 2010 Aug 1;archdischild169912.

23. WHO | Acute lower respiratory infections in childhood: opportunities for reducing the global burden through nutritional interventions [Internet]. WHO. Available from:

24.WHO | Integrated Management of Childhood Illness (IMCI) [Internet]. WHO. Available from:

http://www.who.int/maternal_child_adolescent/topics/child/imci/en/

25.Pneumonia [Internet]. UNICEF.. Available from:

https://www.unicef.org/health/index_91917.html

26. WHO | Vaccines to prevent pneumonia and improve child survival. WHO. Available from:

(43)

43 | P a g e

28.Baqui AH, El Arifeen S, Saha SK, Persson L, Zaman K, Gessner BD, et al. Effectiveness of Haemophilus influenzae Type B Conjugate Vaccine on Prevention of Pneumonia and Meningitis in Bangladeshi Children: A Case Control Study. The Pediatric Infectious Disease Journal. 2007 Jul;26(7):565–71.

29. Khowaja AR, Mohiuddin S, Cohen AL, Mirza W, Nadeem N, Zuberi T, et al. Effectiveness of Haemophilus influenzae Type b Conjugate Vaccine on Radiologically-Confirmed Pneumonia in Young Children in Pakistan. J Pediatr. 2013 Jul;163(1 Suppl):S79–S85.e1.

30.Vaccine Information Statement | Pneumococcal Conjugate | VIS | CDC.Available from:

https://www.cdc.gov/vaccines/hcp/vis/vis-statements/pcv13.html

31.Revised WHO Classification and Treatment of Pneumonia in Children at Health Facilities [Internet]. World Health Organization; 2014. Available from:

https://www.ncbi.nlm.nih.gov/books/NBK264162/

32.Gordon DM, Frenning S, Draper HR, Kokeb M. Prevalence and Burden of Diseases Presenting to a General Pediatrics Ward in Gondar, Ethiopia. J Trop Pediatr. 2013 Oct 1;59(5):350–7. 33. Bohn JA, Kassaye BM, Record D, Chou BC, Kraft IL, Purdy JC, et al. Demographic and mortality analysis of hospitalized children at a referral hospital in Addis Ababa, Ethiopia. BMC

Pediatr [Internet]. 2016 Oct 21;16. Available from:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5073447/

34. Federal Ministry of Health (2015) Health Sector Transformation Plan 2015/16 - 2019/20. Addis Ababa, Ethiopia.

35. HSTP : Health Sector Transformation Plan : 2015/16 - 2019/20 (2008-2012 EFY) / Addis Ababa, Ethiopia.

36.UNICEF_Ethiopia_2017_--_Health_and_Nutrition_Budget_Brief.pdf.

37.WHO | Global Strategy on Human Resources for Health: Workforce 2030 [Internet]. WHO. Available from: http://www.who.int/hrh/resources/globstrathrh-2030/en/

38.Ward, H., Mertens, T. and Thomas, C. Health seeking behaviour and the control of sexually transmitted disease in Health Policy and planning. 1997; 12:19-28.

39.Wardlaw, Tessa M, Johansson, Emily White, Hodge, Matthew, World Health Organization & UNICEF. (2006( .Pneumonia :the forgotten killer of children .Geneva :World Health Organization .http://www.who.int/iris/handle/10665/43640

(44)

44 | P a g e

41.Astale T, Chenault M. Help-Seeking Behavior for Children with Acute Respiratory Infection in Ethiopia: Results from 2011 Ethiopia Demographic and Health Survey. PLoS One [Internet]. 2015 Nov 11;10(11).

42.Sreeramareddy CT, Shankar RP, Sreekumaran BV, Subba SH, Joshi HS, Ramachandran U. Care seeking behaviour for childhood illness- a questionnaire survey in western Nepal. BMC International Health and Human Rights. 2006 May 23;6:7.

43.Gebretsadik A, Worku A, Berhane Y. Less Than One-Third of Caretakers Sought Formal Health Care Facilities for Common Childhood Illnesses in Ethiopia: Evidence from the 2011 Ethiopian Demographic Health Survey. International Journal of Family Medicine. 2015 . Available from: https://www.hindawi.com/journals/ijfm/2015/516532/

44.Adedokun ST, Adekanmbi VT, Uthman OA, Lilford RJ. Contextual factors associated with health care service utilization for children with acute childhood illnesses in Nigeria. PLOS ONE. 2017 Mar 15;12(3):e0173578

45.Kolola T, Gezahegn T, Addisie M. Health Care Seeking Behavior for Common Childhood Illnesses in Jeldu District, Oromia Regional State, Ethiopia. PLoS One [Internet]. 2016 Oct 14 ;11(10).

46. Andersen RM. Revisiting the Behavioral Model and Access to Medical Care: Does it Matter? Journal of Health and Social Behavior. 1995;36(1):1–10.

47.Life expectancy at birth, total (years) | Data . Available from:

https://data.worldbank.org/indicator/SP.DYN.LE00.IN?locations=ET

48.Pandey A, Sengupta PG, Mondal SK, Gupta DN, Manna B, Ghosh S, et al. Gender Differences in Healthcare-seeking during Common Illnesses in a Rural Community of West Bengal, India. Journal of Health, Population and Nutrition. 2002;20(4):306–11..

49. Lailulo YA, Sathiya Susuman A, Blignaut R. Correlates of gender characteristics, health and empowerment of women in Ethiopia. BMC Womens Health [Internet]. 2015 Dec 7 ;15.

Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4672522/

50. Adeleye OA, Okonkwo CA. Ideal Child Gender Preference in Men's Worldview and Their Knowledge of Related Maternal Mortality Indices in Ekiadolor, Southern Nigeria. Asian Journal of Medical Sciences. 2010;2(3):146–151

51. Hussain R, Fikree FF, Berendes HW. The role of son preference in reproductive behaviour in Pakistan. Bull World Health Organ. 2000;78(3)

Factors affecting healthcare-seeking for children below five years with symptoms of acute respiratory tract infection in Ethiopia: a Cross-sectional study based on the 2016 Demographic and Health Survey