Determinants of delivery of health services by community health workers a case of embu district

Public Policy and Administration Research www.iiste.org

ISSN 2224-5731(Paper) ISSN 2225-0972(Online)

Vol.3, No.8, 2013

1

Determinants of Delivery of Health Services by Community Health Workers: A case of Embu District

Judith Museve 1* , Josphat Mutua2

1. School of Health Sciences, Mount Kenya University, P.O.Box 547, 01000, Thika 2. Ministry of Public Health & Sanitation, P.O. Box, 30016, Nairobi

*[email protected]

Abstract

The introduction of the community strategy involving Community Health Workers (CHWs) was to pacify the growing demand for quality healthcare by the general population. However there have been concerns on whether the involvement of the CHWs leads to better delivery of health care services. There is confusion about their services, sustainability and resources. Also, the community’s role in the implementation of the strategy lacks clarity. This study sought to identifying demographic, programmatic and community factors that influence the delivery of health services by CHWs in Embu district.

A cross sectional study design was used, where both quantitative and qualitative data was collected. A structured questionnaire, key informant interviews (KII) and focused group discussions (FGD) were used. The study sample comprised of 137 trained CHWs. The quantitative data was processed and analyzed using the statistical package for social scientists (SPSS, version 17). Odds ratio with 95% confidence intervals was used to show associations and p-value <0.05 was considered a statistically significant level of precision. Qualitative data was analyzed manually by themes, subthemes and reviewed to conform to thematic objectives.

The study found out that Community Health Workers offer vital health services at the community level in Embu district. In terms of delivery of services at level one, referral services, barazas and health education goals were above average in Embu community units; however the score on visiting selected household per month is low. The main factors significantly influencing the delivery of health services at level one is the availability of income and source of income, supplies, period of refresher course, days for giving the services, feedback information and knowledge on disease signs, symptoms and its management. Older and young CHWs were found to be more active than middle aged ones. Male CHWs were more active in the delivery of level one health services than their female counterparts. On average, CHWs in the District work for eight days in a month. Despite supervision being not significant, all none supervised CHWs scored poorly in the delivery of health services. Community Health Committees were ranked the best supervisors. Factors enhancing service delivery are appreciation by community and motivation. Insecurity, lack of motivation, cultural norms, traditional practices, customs and beliefs hinder the delivery of services.

The study recommends the development of a clear contextualized guideline on the qualification, scope, mandate, number of working days per month with commensurate reward and motivation of CHWs in order to improve delivery of level one health services. The strategy to be adopted should be sustainable at community level.

Key words: Delivery, health service, community health workers

1. Introduction

The Community Strategy, which uses community members to render certain basic health services to their communities is a concept used to scale up delivery of health services worldwide. It uses CHWs and community resource persons. There are innumerable experiences on the same throughout the world (WHO, 2007). Quite a number of countries in Africa and South Asia are currently investing in community health workers (CHWs) as a major part of the community strategy to reach the wider population. It is expected that CHWs preferentially reach the poor who are less likely to use health facilities. In Sub-Saharan Africa, Kenya, Uganda, Ghana and South Africa are implementing national programs for CHWs (WHO, 2010). In Kenya, there are six levels of health services delivery; and the community units commonly referred to as level one health service is designed to provide basic community health service which is promotive, preventive and simple curative services. A community unit has 5000 people and is served by 50 CHWs who are volunteers and the community’s own resource persons. The CHWs are trained and supervised by community health extension workers (CHEWS) who are based at level 2 and 3 but are assigned duties to sub-locations to support the CHWs(MOH, 2007). CHWs are particularly important in areas where the use of facility-based services is low. They can increase access to and use of health services, and have played a part in primary health care in tuberculosis, immunization and family



Vol.3, No.8, 2013

2

planning programs. Embu district in Kenya was among the first areas to train community health workers on the Community Strategy in 2007. The district has four administrative Divisions and six functional community units. The residents have no distinct cultural identity; they practice agriculture majoring mainly on food crops like maize, beans, bananas; and also grow tea and coffee.

There is limited research on the quantitative links between the role of CHWs and delivery of services (UNDP, 2005). Therefore there is need to assess the evidence that CHWs can perform the necessary tasks and describe the determinants of delivery of health care services at level one.

1.0 Problem statement

Africa; Kenya included, faces an acute shortage of health workers (WHO, 2006). In the past decade, Kenya’s second National Health Sector Strategic Plan (NHSSP II – 2005-2010) adopted Community Health Strategy as means of addressing the health care needs and challenges at level one. Over the last four years, Embu District has been implementing the community health strategy. Since then, there are six functional community units. Despite this, there has been no change in the morbidity and mortality patterns in the established community units with Malaria (32%), Pneumonia (28%) and Diarrhoea (16%) in children under five years and ameobiasis among the adults still high (DMOH 2009). Elsewhere there has been concern whether the involvement of the community’s own resource persons leads to effective delivery of health care services at level one (Sagini, 2010). Furthermore, there reigns confusion about the sustainability and success of level one workforce, services and resources. The contribution of community health workers in level one health service delivery remains unclear and undocumented in the district (DMOH, 2009).The community’s role in the implementation of the community strategy is not clear (UNICEF 2010). However since its adoption, the roll out of community based health services has taken a different dimension, acceptance and accessibility at different communities (UNICEF, 2008).

In addition socio-cultural issues such recognition of community health services and service provider, cultural diversity/ dynamics in urban setting, perception of level one health services and client-provider relation have not been fully explored (Haines et al., 2007). This research therefore sought to identify the determinants of delivery of health services at level one by CHWs, with the view to give recommendations that can improve health service delivery at that level.

2. Literature review

2.1Demographic factors and delivery of health services

Provider characteristics such as Age, gender, marital status and education, may influence the perception and delivery of health services by an individual (Simkhada etal., 2007). The association between age of the provider and delivery of health care service has been inconsistent across studies (Babalola and Fatusi, 2009). According to study by Pallikadavath in Asia and Middle East, the age of a provider was found to be significant in Nepal and India, but not in Jordan (Pallikadavath et al., 2004).

The presence of male practitioners in obstetric and gynaecological care has been shown to be an important reason for low use of these services by Asian women in western societies (Whiteford and Szelag, 2000). In their study conducted in the United Kingdom in 1998, Goddard and Smith concluded that the clustering of patients of the same ethnic origin in practices staffed by people with the same gender and cultural background was one reason for the high registration and consultation rates with general medical practitioners in predominantly South Asian communities. However, a study by Haines reported there was no evidence on the relative effectiveness of men versus women as community health workers (Haines et al., 2007).

A study by Simkhada, found that the marital status of a health care provider influenced the delivery of health care services (Simkhada et al., 2007). Yet another study by Kipkorir found out that marriage is highly regarded in the Africa cultural context and marital status determines perceptions of personalities (Kipkorir, 2008). Other studies however, found no association between the healthcare provider marital status and delivery of health care services (Mekonnen and Mekonnen, 2003; Gyimah et al., 2006)

Haines et al (2007) found no association between the education level of the active and inactive CHWs. Yet another study done in Bosomwe district in Ghana on factors influencing the delivery of intermittent preventive treatment (IPT) of malaria in pregnancy, found no association between education level and delivery in health care services (Antwi, 2009).However, in his study Ouma found out that low level of education was associated



Vol.3, No.8, 2013

3

with low delivery of health care services (Ouma et al., 2005). In his study, Elijah found that delivery of health medications increased with higher levels of formal education (Elijah et al., 2002).

In a study carried out in Bangladesh, Rahman found out that the services offered by the CHWs were influenced by the cluster they came from and their residence; that is whether rented or personal (Rahman et al., 2010).

2.2 Programmatic factors and delivery of health services

Timely and accurate information forms the basis for management to plan and for care providers to take appropriate action (Singh, 1992). A study by Mensah found out that very little is known about how health workers value investments in health information technology and its impact on delivery of health services (Mensah, 2007). Delivery of health care services is closely related to financial resources. Incentives such as provider payment mechanisms were found to successfully influence the availability and make-up of services offered, thus promoting health care provision to population groups with low health care access. Differential payment can be done to foster provision in specific geographical areas and among priority populations (Lopez et al., 2004).

Lewis in his study in 2006 concluded that there is need to examine the effectiveness of governance and specifically the efficiency of its role in delivery of health care services since poor governance can undermine health care delivery (Lewis, 2006).

In his study, Haines states that the availability of drugs and cost of travel may influence delivery of health care services although very few studies have assessed the impact of unavailability of drugs and the time used to go and collect them by community health workers (Haines et al., 2007). The distance covered by CHW to offer health services and the availability of transport options can have a significant impact on appropriate and timely delivery of health services (Furuta and Salway, 2006).

The level of training of provider has a big influence on the delivery of service (Brabin et al., 2009). A study done in rural western Kenya to assess the effect of health care workers training on the use of IPT for malaria in pregnancy showed that there was an increase in coverage for IPT 1 from 19% in 2002 to 61% 2005 after health care workers were retrained (Ouma et al., 2005). However, a survey conducted in three health centers in Kampala showed no effect on malaria guidelines and treatment even after the training of health workers (Nankwanga and Gorette, 2008). In his study report, Kaseje points out that although the training of CHWs is a key and major activity in Kenya , it still was not clear on fundamental issues such as duration of training, the content of training, the designated trainers and the role of the community in the training of the CHWs (Kaseje, 2006).

Human resource is the one of the most important components that determines the performance of public health system (WHO, 2006). There is contradictory evidence on the contributions of different categories of health workers and the role of health workers relative to other health system inputs in increasing delivery of essential services, particularly in developing countries (Kruk et al., 2009). The role of CHWs include home visits, environmental sanitation, provision of water supply, first aid, treatment of minor and common illnesses, nutrition counseling, health education and promotion, surveillance, maternal child health, family planning, referrals, record keeping and data collection among others (Ofuso –Amaah, 2003).

Study findings show that wide differences in social status between the practitioner and patient may inhibit service delivery (Shah, 2007). Few comprehensive studies have been completed to analyze the patient – provider relationship and its effect on health care delivery (Turin, 2010).

2.3 Knowledge, attitude, culture and practice

Knowledge may provide consumers with a basis for evaluating whether they or their dependant require treatment. Evaluating and making the best use of information on good and bad health providers requires a measure of sophistication in the target group. There is need to assess the level of knowledge on mitigation measure and knowledge on impact of diseases of CHW in relation to the way level one health service is delivered (Haines et al., 2007).

Attitudes about medications, illness and health service by provider may interfere with the delivery of health service (Deventer and Radebe, 2009). Stigma towards certain conditions such as HIV/AIDS has effects on the delivery of health services (Turin, 2010). The tendency of patients to doubt or question advice offered by medical practitioners may also contribute to delivery of health services (Frank, 2009). Another study by Simkhada reports that the traditional perception of events may tie providers to encourage the use of alternative options and only go for formal systems when the traditional option fails (Simkhada et al., 2007). Lack of



Vol.3, No.8, 2013

4

satisfaction with a service rendered could be a major demotivating factor to a service provider (Aldana et al., 2001). Other studies reported the negative attitude of health care workers and the poor relations between healthcare workers and women as major barriers to service provision (Mathole et al, 2004; Simkhada et al., 2007).

Many cultural or social factors may impede the delivery of health services. A study by Addai reported that the cultural perspective on the delivery of health services suggests that medical need is not only determined by the presence of physical disease but also by the cultural perception of illness. (Addai,2000). In most African rural communities, health services coexist with indigenous health care services therefore clients must choose between the options (Addai, 2000). The use of traditional medicines and traditional doctors is not included in healthcare delivery data in Kenya yet there are strong cultural associations with traditional forms of healthcare (Turin, 2010). In communities where women are not expected to mix freely with men, delivery of health services by opposite sex may be impeded (Gabrysh and Campbell, 2009).

2.4 Community factors and health service delivery

A study by Duong reported that power hierarchy at home plays a central role in determining the delivery and utilization of health services (Duong, 2005). Few studies have looked at how family support and the provider’s position within the household influence the delivery of health services (Furuta and Salway, 2006). This is supported by a study by Dudgeon which reports that the influence of intentions and practices of close relatives’ especially old people on the delivery of health service have not been thoroughly investigated (Dudgeon and Inhorn, 2004).

The influence on individual health behavior extends to beliefs and practices of others in the community (Mayhem, 2008). The role of the community factors on decisions to deliver health services has been largely ignored (Cheboi, 2011).These community beliefs and norms are reflected in an individual’s health decisions whereby one’s behavior towards delivery of health services is influenced by how a person thinks the community views his or her actions (Stephenson et al., 2005). Another study by Wamai, reports that the delivery of health care may be characterized by the type of care, purpose and acceptance of the services (Wamai, 2009). A number of studies have reported the importance of community support. One study showed that social support is strong predictor of delivery of any services (Park et al., 2010).Yet another study reported that the widely publicized views of politicians, religious groups and other opinion leaders on use of health services play an important role in skepticism towards delivery and reception of health services (Frank. 2009). A study in Pakistan found that the resistance by a husband and cultural unacceptability of a health service were more important determinants than fears of further worsening of disease status (Sathar et al., 2001). Lifestyle is a motivator of delivery of health service but few studies have looked at the effect of community lifestyle on delivery of health services (Shah, 2007).

Health communication is a dynamic process that at any point in time has a status that may or may not be appropriate for specific population groups it is meant to inform. However, measuring the status of health communication for a given population is a complex and ill-defined process (Calderon et al., 2004).

The issue of personal safety and security is a prerequisite for the initiation and continuation of delivery of health services therefore there is need to assess its role in the delivery of level one health service (Sibhatu, 2008).

2.5 Summary of literature review

There are limited studies evaluating demographic characteristics of the level one health service providers such as age but not by cohorts, gender and marital status. Several studies have examined the role of education status, residency, source of income, knowledge of the health provider, attitude and practice but these studies were limited to quantitative research and non on qualitative research design. On programmatic factors, there is so much literature on the cost of financing but not on community based health care financing, quality of services, governance and accessibility of drugs, however the findings are inconclusive and inconsistent. Studies on the role of supervision, monitoring and evaluation, communication, patient-provider relationship and area covered by community health worker are limited. Studies on the role of community factors in relation to service delivery have been done but the results are inconsistent across studies. The roles of alternative medicine, beliefs and norms have not been fully explored. 3. Methodology



Vol.3, No.8, 2013

5

3.1 Study Design

An explorative, cross-sectional study was conducted to assess the level and factors influencing the delivery of health services at level one by community health workers. Both quantitative and qualitative data was collected. Study Population was the active community health workers in Embu district. The population of community health workers in the district was 207. Sample Size determining the population proportion was obtained using Fisher et al, 1993 formula:

n = Z2pq

d2 ..............................equation 1

Where:

Z = 1.96 (if the population is approximately normal when 95% bounds on the values in the

population)

p = proportion of CHWs expected to deliver health service, 50% (proportion unknown)

1- p = proportion of CHWs expected not to deliver health services

d = 0.05 (degree of accuracy at 95% confidence level)

Hence n = 1.962 *0.5(1 – 0.5) = 384.16 = 384

0.052

nf = n

1+ n

N …………………equation 2

In determining the sample size for this research, a sample size of 384 was needed (rounded from 384.16). Using the finite correction factor in above equation (equation 2), with N = 207, 1.96 (95% confidence), leads to

nf = 384 = 132.4 = 133

1 + 384

207

A 5% attrition rate was added; hence final sample size was 140 (rounded from 139.65)

Systematic sampling method was used to identify the respondents. A register of the CHWs was obtained from the DMOH’s office in Embu. Respondents were equally distributed per the six community units.

140 = 23 respondents per community unit (2community units had 24 respondents) 6

3.2Sampling procedure

K th = Number of active CHWs in the area

Number of desired sample size

= 207 = 1.5 = 2

140

A table of random numbers was used to identify the first respondent and thereafter every 2nd CHW from the register was interviewed until the 23rd respondent per community unit and 24th respondent in the last two units. Due to the complex and dynamic nature of interaction and development of indicators that measure overall rate of delivery of level one health services; this study used four key services of level one to rate delivery of the services; number of referred clients, number of health education forums conducted, number of barazas addressed and number of households visited. A code of one was allocated for every service offered above given target



Vol.3, No.8, 2013

6

(yes=1) and zero for services delivered below given targets (no=0). A dichotomous outcome (delivery of level one health services) was done by scoring four target variables where one(1) meant yes for achieving the set targets per month and zero(0) meant number not achieving the set targets. The overall results were computed for all the questionnaires and the aggregate “yes” meant “delivery of service” while for “no” meant “no service delivery” by CHWs.

3.3 Data analysis

The quantitative data was cleaned, coded, entered and analyzed using the Statistical Package for Social Scientists (SPSS version 17). Association between socio-demographic factors and service delivery, programmatic factors and delivery of health service, knowledge, and practice and health service delivery were determined. A contingence table for testing dependence through chi-squire test was developed. Level of health service delivery was defined as those who achieved the set targets per month while those who did not achieve the said targets were considered ‘No service delivery’. The analysis sought to measure the associations between variables, and odds ratio was used to show the strength of the associations. A value of p <0.05 and OR>1 were taken to be significant.

4. Findings

4.0 Introduction

This chapter presents the findings of the study on the determinants of health services at the level one in Embu district. A total of 137 CHWs against a target of 140 (CHW) participated in the study. Out of this, 79% of them were females while the rest were males. This was a response rate of about 97%. Qualitative data was obtained from key informants namely; the community health extension worker, area assistant chief, a church elder and some selected CHWs in a focus group discussion. Data was treated by use of inferential comparison and reviewed to confirm its reflection of the objectives.

4.1 Rates of delivery of health services at level one

Majority (2/3) of the community units had overall good delivery of health services. A half of the community units achieved the desired number of barazas, 2/3 of them had met the threshold for health education forums while ½ of the units met the threshold for client referral. Household visits scored poorly in all community units. Overall, the 3 key targets of health services delivery at level one rated above average (56% of barazas, 62% of health education and 56% of referral of patients). Conversely, target household visits rated way below average at 16%. Sixty percent (60%) of the CHWs met the criteria and threshold of delivery of level one health services. Delivery of level one health services was based on the rates of four level one health services. Majority (60%) of the CHWs scored above average in three of these four targets (number of barazas, health education and referral of patients). However 84% of these CHWs rated poorly in the number of household visited per month

4.2 Association between demographic factors and health services delivery

Table 1 below shows the association between demographic factors and health service delivery at level one. The median age of the respondents was 39years with majority of them being aged 26 – 45 years. From the results, only a source of income was significantly associated with the delivery of health services at level one (p=0.036). CHWs that were most likely to deliver health services are those who were salaried, had family support or were farmers. Self-employed CHWs were the least likely to deliver health services.

Although gender was not significantly associated with delivery of health care services, 72 % of the male CHWs met the threshold for delivery of level one health services compared to 58% of the females. CHWs who had completed secondary school education were the most likely to either deliver or not deliver health services.



Vol.3, No.8, 2013

7

Table 1: Association between demographic factors and delivery of health services at level one

Demographic factors Delivery of level one health services , N=137

Service, n(%)

No service n(%)

df P-value

Gender:

Male Female

21(72.4%) 63((58.3%)

8(27.6%) 45(41.7%)

1.911

1

0.167

Respondents age (years):

15-25 26-35 36-45 46-55 Over 56

5(71.4) 26(56.4) 27(54.0) 20(76.9) 6(75.0)

2(28.6) 20(43.5) 23(46.0) 6(23.1) 2(25.0)

5.177

4

0.27

Source of income:

Salary Farmer Self-employed Casual Family support

2(100.0) 35(66.0) 15(37.5) 19(54.3) 5(71.4)

0(0.0) 18(34.0) 25(62.5) 16(45.7) 2(28.6)

9.973

4

0.036

Education level:

None Primary(incomplete) Primary(complete) Secondary(incomplete) Secondary(complete) Tertiary

1(100) 3(100.0) 11(45.8) 18(72.0) 48(59.3) 5(100)

0(0.0) 0(0.0) 13(54.2) 7(28.0) 33(40.7) 0(0.0)

9.624

5

0.080

Marital status:

Married Single Divorced Separated Widowed

68(62.4) 10(52.6) 1(100.0) 2(66.7) 3(60.0)

41(37.6) 9(47.4) 0(0) 1(33.3) 2(40.0)

1.327

4

0.920

Religion:

Catholics Protestants Muslims Indigenous

29(61.7) 50(60.2) 4(66.7) 1(100)

18(38.3) 33(39.8) 2(33.3) 0(0.0)

0.747

3

0.783

Residency in area(years):

1-2 3-4 >5

1(100.0) 10(50.0) 73(62.9)

0(0.0) 10(50.0) 43(37.1)

1.838

2

0.399

P <0.05

From Figure 1 below, level of income was significantly associated with delivery of health services (p=0.036). The results were computed using Fischer’s exact test due to counts of less than 5. Majority of the respondents earned a monthly income of less than 2,500 Kenyan shillings. These respondents were more than three times likely to deliver health services compared to those earning more than 2,500 Kenyan shillings.

Figure 1: Association between level of income and delivery of health services



Vol.3, No.8, 2013

8

P<0.05 4.3 Association between programmatic factors and delivery of health services

Table 2 below summarizes the association between key programmatic factors and delivery of health services. The results show that most of the CHWs work for 6-10 days per month and they are mainly supervised by the Community Health Extension Worker. The group working for 6-10 days had the lowest rate of service delivery. This group had the largest number of CHWs offering service and those not offering the service. The median number of days the respondents committed to health service work was 8 days per month (IQR 6-10). Only a few CHWs work more than 10 days in a month and the number decreases with increase in the number of days.

The CHWs closer to a health facility were more likely to deliver health services, with stationary being cited as the main supply influencing delivery of health services. All the CHWs utilize feedback reports with a large majority of them having been trained in Infant and Young child feeding. A significant number (48%) of the CHWs had not received any refresher training.

The type of refresher course (p=0.010), period of training (p=0.019) and use of feedback reports (p=0.012) were significantly associated with delivery of health services. CHWs who were trained in Breast feeding, Infant and Young Child feeding and Home-based care were the most likely to deliver health care services. Those who were trained for a longer period of time (> 1week) were the most likely (80%) to deliver health services. Likewise, those who used planning feedback reports were the most likely to deliver the services.

This finding was supported by the views of the FGD discussants who unanimously agreed that training was crucial in their work. A key informant who was an Assistant chief summed it thus; “Since the selection and training of CHWs, my work has been made easy. I am able to monitor what is happening in every village using my phone… Jointly with them, we have increased toilet coverage greatly.” Another discussant in FGD 2 affirmed the importance of training as follows; “Since our last training in 2008, I have never attended any other course on public health. Sometimes I feel that am not sure of what to tell people. I think I have even forgotten some of the issues we trained on”.

CHWs who received drugs as supplies were the most likely to offer service delivery. Although supplies were not significantly associated with service delivery, all the FGD discussants thought they were a major challenge affecting their work. The magnitude of the supplies problem was vividly captured by a discussant who said: “Why has the government taken so long to give us the kits they promised. A CHW without a First-aid kit is like a soldier without a gun”.

The feedback reports were found to be critical in service delivery with majority of the respondents using the reports mainly for planning. This finding was in agreement with the views of both the discussants and key informants as per the following excerpts:



Vol.3, No.8, 2013

9

One Key informant summarized it thus: The CHWs reports are very important to my office. It assists me to identify those births that have not been notified and registered”.

One discussant in FGD 3 summed their view as: We make monthly reports and take them to the Public Health Officer’s office. At times we don’t know what happens after that”.

Table 2: Association between programmatic factors and delivery of health services

Programmatic factor Delivery of level one health services (N=137)

Service (%) No service (%)

df p-value

Supplies: Drugs Stationary Means of transport Identification badges None

3(75.0) 46(59.7) 8(66.7) 8(61.5) 19(61.3)

1(25.0) 31(40.3) 4(33.3) 5(38.5) 12(38.7)

0.541

4

0.997

Type of refresher training:

Home-based care PMTCT Disability Reproductive health Infant & young child feeding Breast feeding None

5(71.4) 3(60.0) 2(33.3) 2(50.0) 23(71.9) 16(94.1) 34(51.5)

2(28.6) 2(40.0) 4(66.7) 2(50.0) 9(28.2) 1(5.9) 32(48.5)

15.480

6

0.010

Period of refresher training:

< 1week 1 week >1 week None

45(73.8) 0(0.0) 4(80.0) 35(53.0)

16(26.2) 5(100.0) 1(20.0) 31(47.0)

8.582

3

0.019

Working days per month:

< 5 6 - 10 11 – 15 16 -20 21 -25 26 -30

10(47.6) 53(63.9) 3(27.3) 13(76.5) 1(100) 4(100)

11(52.4) 30(36.1) 8(72.7) 4(23.5) 0(0.0) 0(0.0)

1.838

5

0.399

Supervision:

Community Health Committee Community Health Extension worker NGO representative None

4(66.7) 76(63.3) 4(50.0) 0(0.0)

2(33.3) 44(36.7) 4(50.0) 3(100.0)

5.465

3

0.141

Proximity to Health facility:

< 5 km 5 – 10 km >10 km

72(63.7) 11(50.0) 1(50.0)

41(36.3) 11(50.0) 1(50.0)

1.570

2

4.56

Use of feedback report:

Planning Address gaps All the above None

25(78.1) 11(52.4) 40(48.2) 0(0.0)

7(21.9) 10(47.6) 43(51.8) 1((100)

8.463

3

0.012

*Fischer exact test was applied where parameters had options with less than 5 counts.



Vol.3, No.8, 2013

10

Figure 2 below shows the association between the number of working days per month and health service delivery. Generally, the rate of health service delivery increases with increase in the number of working days.

Figure 2: Association between number of working days and delivery of health services

4.4 Association between Knowledge, attitude, practice and delivery of health services

4.4.1 Knowledge on disease conditions and delivery of health services

Table 3 below shows the results of the assessment of the knowledge of CHWs on disease symptoms and signs; and management of common disease conditions. There was a statistical significance between the knowledge on diarrhoea signs and symptoms and service delivery (p=0.016). Majority of the respondents correctly identified the signs and symptoms of diarrhoea.

Table 3: Association between Knowledge of signs and symptoms of diarrhoea and delivery of health services

Knowledge factor Delivery of health services

df

p-value Service No service

Diarrhoea Signs and symptoms:

1 loose motion per day

4 loose motions per day

Total

12(85.7%)

64(52.0%)

76(55.5%)

2(14.3%)

59(48.0%)

61(45.5%)

5.77

1

0.016

4.4.2 Association between practice factors and health service delivery

4.4.2.1 Channels of communication

The channel of communication was significantly associated with the delivery of services (p=0.044). All the respondents who used baraza and church as the only channel of communication delivered the health services. The CHWs who utilized all forms of channels of communication had the highest likely hood of delivering health services.



Vol.3, No.8, 2013

11

Figure 3: shows the association between the channels of communication and the delivery of services.

4.4.2.2 Supervision of CHWs and delivery of health services

Figure 4 below shows the association between supervision of CHWs and the delivery of health services. From the results, although programmatic factors were not significantly associated with delivery of services, supervision was observed to be critical. All CHWs who were not supervised did not deliver health services. Although community health committees were ranked as the best supervisors, majority of the CHWs are supervised by CHEWs .

Figure 4: Association between supervision and delivery of services

The FGD discussants agreed with this result and acknowledged the challenge of supervision. One key informant summed it thus: “We are only two Public Health Officers in my division which has 2 community units. The official responsibilities are too much for the two of us to also monitor the work of the CHWs, especially considering the terrain of this place”.

Although community health committees had been cited as the best supervisors, one discussant in FGD two gave a conflicting opinion thus: “We are told to work closely and report to community health committees. But some of them don’t know our work. They were not trained so it’s probably not their fault. They have not been sensitized on their roles”.



Vol.3, No.8, 2013

12

4.4.3 Traditional practices, customs and beliefs

The community still subscribes to the use of traditional and other non-conventional medicines in the management of common illnesses. This can be attributed to cultural values, cost and accessibility of medical services. For instance, the community advocates for the use of herbs (mwarobaini) for treatment of malaria; and ovulectomy for the management of upper respiratory tract infections and coughs. Ash from specific plants is used in the treatment of diarrhoea and removal of milk teeth among children. One key informant confirmed the use of ovulectomy as follows: “Ovulectomy is a deep rooted cultural practice here. Even one of our former senior politicians had his uvula cut in Manyatta division in Embu district”.

It is a custom of the community to circumcise girls. Female circumcision is still common with any woman who has not undergone the rite being socially discriminated among community members. One elderly male discussant commented as follows; “An uncircumcised girl cannot address the village elders, it’s an abomination. Hence, this young girls walking around the village as health educators are only wasting their time”.

The community deeply believes and subscribes to witchcraft and religious fanaticism. The people seek the services of medicine men and religious prayers in the management of common ailments. A discussant in FGD one summarized it as follows: “My parents treated me using medicine from the local medicine man and I grew up as a healthy person. Why shouldn’t I do the same to my children and relatives? Nobody in this village has ever complained or failed to respond to this medicine. It’s our way of life; they are wasting their time”.

Some young mothers seek the opinion of their mother-in-laws (especially if the child is named after her) or husband before taking a child for treatment.

4.5 Community based factors affecting delivery of health services

Table 4 below illustrates the association between community based factors and delivery of health services. The results show that a large majority of the CHWs (96%) were appreciated by the community. Out of these, 53% of them achieved the required threshold for delivery of health services.

Insecurity and lack of motivation are the key challenges in the delivery of health services, with those who cited them as a challenge being the least likely to deliver health services. Apparently, very few of the CHWs viewed culture as a challenge to their work. From the FGD discussants, it emerged that male chauvinism was common in the community with the entry of females into the community services being accepted skeptically. From most of the FGDs , it seemed that age and gender were critical factors in the delivery of the health services. One discussant in FGD two summed it thus: “We are not well received in the community. It appears there are gender preferences with the males accepting male CHWs more and the trend is the same for females… Young people as health educators are only wasting their time since elderly people accept information from people of a certain age and not just everybody”.

Appreciation by the community was significantly associated with delivery of services (p=0.025). Community appreciation impacted positively on the delivery of health services.



Vol.3, No.8, 2013

13

Table 4: Association between community based factors and delivery of level one health services

Community factors Delivery of level one health services (N=137)

df

p-value Service

n(%)

No service n(%)

Source of support:

Spouse Family Community Provincial administration None

24(68.6) 15(51.7) 24(63.2) 4(80.0) 17(56.6)

11(31.4) 14(48.3) 14(36.8) 1(20.0) 13(43.4)

2.965

4

0.564

Type of support:

Transport Encouragement None

13(72.2) 54(60.7) 17(56.7)

5(27.8) 35(39.3) 13(43.3)

1.191

2

0.551

Community appreciation:

Yes No

70(53.4) 6(100.0)

61(46.6) 0(0.0)

5.04

1

0.025

Challenges:

Insecurity Terrain Lack of motivation Family issues Long distance Cultural issues

8(47.0) 4(66.7) 48(57.1) 4(66.7) 17(80.9) 3(100)

9(53.0) 2(33.3) 36(42.9) 2(33.3) 4(19.1) 0(0.0)

7.525

5

0.184

Figure 5 below shows the relation between the type of support and delivery of health service. The findings indicate that those who were supported by the provincial administration, spouse and community were the most likely to deliver health services compared to those with general family and no support.

Figure 5: Association between type of support and delivery of health services

4.6 Multivariate analysis of key determinants of health service delivery

Table 5 below shows the results of a multivariate analysis of the key variables associated with service delivery using multinomial logistic regression.



Vol.3, No.8, 2013

14

With adjustment for no refresher course, the results show that the CHWs who attended a refresher course pertaining to importance of breastfeeding were fifteen times more likely to deliver health services at level one (p=0.01, 0R= 15.057) compared to those who had training on infant and child feeding. There was no significant association between those who attended HBC, PMTCT and reproductive health, and delivery of health services at level one. Apparently, CHWs who had a refresher course lasting less than one week, were more likely to deliver health services (p=0.026, OR=0.424). When adjusted for no feedback report, the combined use of feedback reports to plan and address gaps was highly associated with delivery of health services at level one (p=0.001, OR=3.609)

Table 5: Multivariate analysis for key independent variables

Variable Levels Exp(β) 95% CL for Exp(β) p- value

Lower Upper

Type of refresher course

HBC 2.353 0.426 13.002 2.353

PMTCT 1.412 0.221 9.007 1.412

Disabilities 0.471 0.081 2.748 0.471

Reproductive health

0.941 0.125 7.085 0.941

Infant and

child feeding

2.405 0.961 5.971 0.057

Breast feeding

15.057 1.887 20.196 0.010

None Ref - - -

Period of refresher course

None Ref - - -

< I week 0.424 0.199 0.903 0.026

I week 1.298 0.318 5.289 0.716

>1 week 2.513 2.513 2.513 2.513

Use of feedback reports

Planning 2.106 6.071 6.455 0.001

Address gaps 4.223 1.278 1.396 0.001

All the above 3.609 3.609 3.809 0.001

None Ref - - -

No of working Days/month

0-5 1.925 0.633 5.653 0.248

6-10 0.991 0.428 2.282 0.048

Over 11 days Ref - - -

Knowledge of signs& symptoms of diarrhoea

I loose motion 0.283 0.059 1.350 0.113

4 loose motions Ref - - -

5. Discussion

5.1 Overall delivery of level one health services

From the study findings on determinants of delivery of health services at level one, the service is considered good. About 60% of the CHWs met the criteria and threshold of delivery of level one health services. Delivery of was based on the rates of four level one health services. Majority (60%) of the CHWs scored above average in three of these four target (number of barazas, health education and referral of patients). However 84% of these CHWs rated poorly in the number of households visited per month. This can be attributed to the fact that visiting households is an involving personal task, entailing travelling, planning and coordinating personal security, whereas the three other services are group based targets which are delivered at any community function or



Vol.3, No.8, 2013

15

meeting. This is in agreement with the findings of Furuta and Salway (2006) which show that the distance covered by CHW to offer health services and the availability of transport options can have a significant impact on appropriate and timely delivery of health services.

The rates were similar in both urban and rural based community units except in visiting of households where the urban based community units scored low (Dallas 10% and Njukiri 11%). This may be due to accessibility, privacy, diversity and the dynamics of urban life. This finding is in line with that of Shah (2007) who reported that wide differences in social status between practitioner and patient may inhibit service delivery.

5.2 Influence of demographic factors on delivery of health services at level one

This study found out that income and source of income for CHWs had influence on delivery of level one health services. This was cross cutting at all levels of community units. All the CHWs earning a monthly salary met the threshold of delivery of level one health services while 62% of those who were self employed did not meet the threshold of delivery of level one health services. Among those earning a salary, those with less earnings were more active than those with more earnings. This can be attributed to the fact that unemployed CHWs have other competing tasks of maintaining their livelihoods while their employed peers can organize their time better and facilitate their movement within and about the community unit. These findings concur with those of Rahman (2010) in a study on factors affecting the recruitment and retention of community workers in Bangladesh which reported that the services offered by a CHW was influenced by the cluster they come from and the kind of house they live in whether rented or personal. There was no statistical significance between age and delivery of level one health services; however young and elderly CHWs scored highly in the delivery of level one health services than the middle aged. It can be hypothesized that old people have no competing tasks and are self actualized hence are committed to their work while young people are enthusiastic and eager to perform in the first job assignments. The middle aged persons are busy taking care of their young families, high ambitions, leisure and other community demands. However this contradicts with a study by Ndedda (2010) in Busia on social demographic determinants of CHWs performance where CHWs aged 30-40 years were the most active. These findings substantiate those of Babalola and Fatusi(2009) in which age of the service provider was significant in South Asia but not in Middle East. Gender had no statistical significance with delivery of level one health services. Majority of the CHWs were females agreeing with Ndedda (2010) but males participated more than female colleagues in the delivery of level one health services in all except Kavutiri community unit. This is probably because males have fewer tasks in the households and community level than females.

CHW’s level of education had no statistical association with delivery of health services at level one of health care. CHWs with no basic education offered similar services to CHWs with tertiary level of education. The trends were the same in both rural and peri urban based community units. This study concurs with that of Haines et al (2007) which found no significant association between the education status of active and inactive CHWs. Marital status, period of residence and religion had no statistical value on delivery of level one health services. This is in agreement with the report of Mekonnen (2003) and Gyimah et al (2006) which found no association between the health care provider marital status and delivery of health care services.

5.3 Programmatic factors and the delivery of health services

Supplies elicited significant statistics with delivery of level one health services. Seventy five percent (75%) of the CHWs who had received drug supplies performed their role and scored highly in the delivery of level one health services. This agrees with WHO (2011) report which showed that community health workers supplied with injectables enhanced access and uptake of contraceptives at the community level. This was supported by FGD discussants who reiterated that lack of supplies especially drugs was a major challenge. Training in general had no statistical value in the delivery of level one health services but the type and period of training or refresher course had significance. CHWs who were trained in Breast feeding, Infant and Young Child feeding and Home-based care were the most likely to deliver health care services. These are common areas of concern in the community hence have greater demand and interest. Those who had attended refresher course for more than one week were more {2 (100%); 1 (57%)} involved than those who had not in the delivery of level one health services. This may be because they were eager to share their new knowledge with the community and secondly their tools of services were restocked. This is in agreement with the study by Brabin et al (2009) which showed that the level of training of provider has a big influence on the delivery of service. The argument is supported by another study done in Busia by Ouma (2005) which reported that the level of service delivery increased considerably after the retraining of CHWs. This was also supported by all FGD discussants who felt that training gave them the confidence to deliver services and retraining kept them abreast with necessary



Vol.3, No.8, 2013

16

information. Man hours had no statistical significance with delivery of level one health services with majority of the respondents working eight days per month on average.

Feedback information had a significant statistical association with delivery of health services at level one. This information assisted the respondents to re-plan and address gaps within their mandate. It can be hypothesized that planning improved the outcome of their work hence enhanced the confidence of the CHWs. The Channels of communication applied by CHWs in relaying health messages had no significant value in this study. The use of multiple means to delivery information was the most efficient while home visit came second. This may be because the CHWs use home visit to deliver personal and private services while public messages are relayed through different channels. The varied means ensured more coverage of services.

5.4 Knowledge, practice and attitude and the effect on delivery services

Knowledge of disease conditions by the respondent was significant in this study, with 90% of the CHWs being aware of common signs and symptoms of basic conditions such as diarrhoea. This finding is supported by the study done by Frank (2009) who reported that the tendency of patients to doubt or question advice offered by medical practitioners may also contribute to delivery of health services. Thus a knowledgeable CHW would deliver services more. Supervision, rewards and proximity to health facility were significant in the delivery of level one health services. None of the unsupervised CHWs met the threshold of delivery of level one health services. Support supervision increased the morale and confidence of the CHWs to deliver services. This was reiterated in a study on Community based distributors of contraceptives in Ethiopia (WHO, 2009) which showed that service delivery was enhanced by supervision. The best supervisory service was offered by Community Health Committees, then Community Health Extension Worker and the least was by an NGO representative. This is probably because Community Health Committees are local people, always available, friendly to the CHWs and may not require any resources to interact with CHWs.

This study reports that the community subscribes to the use of traditional medicine and other non conventional medicine to manage common illness. These practices and subscriptions to alternative medicine may be a hindrance to the services offered by CHWs. The community members resort or consult community health workers when they do not respond to their first line treatment which is alternative medicine. The belief in witchcraft and religious fanaticism is highly subscribed to by the Aembu community. The people attribute the attacks by common health related conditions (miscarriage, cough, malaria and diarrhoea) to witch craft and bad omen therefore consult witch-doctors and religious people for solutions. This finding concurs with that of Turin (2010) which reported that the use of traditional medicines and traditional doctors is not included in healthcare delivery data in Kenya yet there are strong cultural associations with traditional forms of healthcare.

The study found out that the community highly upholds the practice of FGM and any female who has not undergone the cut is socially not acceptable. Since majority of the CHWs are women this practice may be restricting non-circumcised female CHWs from rendering services. The study found that male chauvinism is high within the Aembu community. This norm restricts the interaction of people of opposite sex therefore female CHWs may be restricted and limited to female colleagues in the delivery of level one health services. This is supported by the study done by Gabrysh and Campbell (2009) which reported that in communities where women are not expected to mix freely with men, delivery of health services by opposite sex may be impeded.

5.5 Community factors and the delivery of services

Appreciation by community had statistical significance with delivery of level one health services. This agrees with another study done in Bangladesh where CHWs felt that they are needed by the community (Rahman et al, 2010). However a significant number of respondents (46.6%) did not meet the threshold of delivery of level one health services despite receiving community appreciation. This can be attributed to the fact the delivery of a service relies on some other factors not only community appreciation. The other factors (community support, communication and security) in this objective had no statistical significance. It can be hypothesized that the low service delivery is due to the community practices and beliefs.

Age was also an issue with young people being under estimated in the society while the elderly are respected and accepted in the society. This may explain why CHWs of over 46 years of age performed well in the delivery of level one health services. This study contrasts with one done in Busia by Ndedda (2010) which showed that those CHWs aged 30 to 40years were doing better.

Insecurity and lack of motivation was the key challenge hindering the delivery of level one service among the respondents. A study in Nigeria by Khan (2006) on the reasons for high CHWs turnover cited low salaries, lack of support for personal development and poor supervision as the key reasons for dropout by CHWs. The issue of



Vol.3, No.8, 2013

17

motivation may be the reason why CHWs scored poorly in targets requiring personal input and score highly in targets with public input such as the number of barazas and health education forums. Motivation may also be the reason for most of the CHWs working eight days per month. Insecurity, terrain, cultural values and family issues were the other challenges mentioned.

6 Conclusions and recommendations

6.1 Conclusion

Community health workers offer vital health services at the community level in Embu district. Generally, the level of service delivery is above average in all the key target areas of community service except in the area of visiting households. The main socio—demographic factor influencing the delivery of health services at level one is the availability of income and source of income. Age and gender influence service delivery whereby older CHWs are more active in service delivery than the younger ones. Male CHWs are more active than female CHWs. Leading programmatic factors influencing service delivery at level one are availability of supplies, period of refresher training course, use of feedback information and knowledge of the signs and management of common diseases. Community factors hindering health service delivery at level one are cultural values, customs, beliefs, traditional practices, insecurity and lack of motivation. Factors that enhance delivery of services at level one are appreciation by the community, supervision and motivation.

6.2 Recommendation

6.2.1 Policy i. There is need for the Ministry of Health to develop clear contextualized guidelines on the qualifications, recruitment, mandate, scope of work, motivation and supervision of CHWs who deliver health services at level one.

6.2.2 Practice

i. There is need to provide all trained CHWs with a sustainable kit with essential supplies.

ii. There is need to scale up the recruitment and training of all Community Health Committees since they play a pivotal role in the supervision of CHWs.

References

1.Abdallah S., Singer P., Taylor A., Ross E.G. Singh J., 2003, LaveryGrand challenges in Global Health: The Ethical, Social and Cultural program.

2. Addai I., 2000, Determinants of use of maternal-child health services in rural Ghana.

3. Duong D.V., Binns C.W., Lee A.H., 2004, Utilization off delivery services at the primary health care level in rural Vietnam. Soc Sci Med 2004, 59(12): 2585-2595.

4. Furuta M. and Salway S, 2006, Women’s position within the household as a determinant of maternal health care use in Nepal.

5. Gabrysch S. and Campbell O., 2009, “Still too far to walk”: Literature review of the determinants of delivery service use. BMC Pregnancy and childbirth 9(1):34

6. Goddard M. and Smith, 1998, Equity of access to health care services: Theory and evidence from the UK. Social Science and Medicine, Volume 53, Issue 9, Pg 1149-1162.

7. Haines A. and Largade M., 2007, Conditional cash transfers for improving uptake of health interventions in low and middle income countries: A systematic review. JAMA 298(16):1900-1910. J Biosoc Sci 2000, 32(1): 1-15

8. Ouma P., Hamel A., Sikuku E., Odhiambo F., Munguti K., 2005, Antenatal and delivery care in rural western Kenya: The effect of training health care workers to provide focused antenatal care.

9. Rahman M., Ali N., Jennnings L., Seraji H., Mannan I., Shah R., Al Mahmud A., Bari S., Hossain DDas M., Baqui A., Arifeen S., Winch P., 2010, Factors affecting recruitment and retention of community health workers in a newborn care intervention in Bangladesh. http://www.human-resources-health.com/content/8/1/12

10. Turin D., 2010, Health care utilization: Analyzing the Kenyan health system

This academic article was published by The International Institute for Science,

Technology and Education (IISTE). The IISTE is a pioneer in the Open Access

Publishing service based in the U.S. and Europe. The aim of the institute is

Accelerating Global Knowledge Sharing.

More information about the publisher can be found in the IISTE’s homepage:

http://www.iiste.org

CALL FOR JOURNAL PAPERS

The IISTE is currently hosting more than 30 peer-reviewed academic journals and

collaborating with academic institutions around the world. There’s no deadline for

submission. Prospective authors of IISTE journals can find the submission

instruction on the following page: http://www.iiste.org/journals/ The IISTE

editorial team promises to the review and publish all the qualified submissions in a

fast manner. All the journals articles are available online to the readers all over the

world without financial, legal, or technical barriers other than those inseparable from

gaining access to the internet itself. Printed version of the journals is also available

upon request of readers and authors.

MORE RESOURCES

Book publication information: http://www.iiste.org/book/

Recent conferences: http://www.iiste.org/conference/

IISTE Knowledge Sharing Partners

EBSCO, Index Copernicus, Ulrich's Periodicals Directory, JournalTOCS, PKP Open

Archives Harvester, Bielefeld Academic Search Engine, Elektronische

Zeitschriftenbibliothek EZB, Open J-Gate, OCLC WorldCat, Universe Digtial

Library , NewJour, Google Scholar

http://www.iiste.org/

http://www.iiste.org/journals/

http://www.iiste.org/book/

http://www.iiste.org/conference/

Determinants of delivery of health services by community health workers a case of embu district

Health & Medicine

delivery of services

health facility

community health committees

community health workers

public health officers

service delivery p

determinants of delivery

school of health sciences