Keeping community health workers in Uganda motivated: key challenges, facilitators, and preferred program inputs

Aurélie Brunie; Patricia Wamala-Mucheri; Conrad Otterness; Angela Akol; Mario Chen; Leonard Bufumbo; Mark Weaver

doi:10.9745/GHSP-D-13-00140

Figures & Tables

Tables

View popup

TABLE 1. Program Inputs and Levels Used in the Discrete Choice Experiment (DCE)

Program Inputs	Level 1	Level 2	Level 3
Training	5-day initial training and 3-day supervised practicum at health center	Same as Level 1 + 3-day refresher training once a year	N/A
Supervision	Monthly CHW meetings at health center	Same as Level 1 + quarterly visit by health center staff in the community	N/A
Incentives	CHW kit with gumboots, raincoat, job aids, and stationery	CHW kit + T-shirt + badge	CHW kit + T-shirt + badge + bicycle
Transportation refund	5,000 UGX for each meeting	10,000 UGX for each meeting	N/A
Communication	No mobile phone	Mobile phone without airtime	N/A

Abbreviations: CHW, community health worker; N/A, not applicable; UGX, Ugandan Shilling.

View popup

TABLE 2. Program Characteristics and Number of Interviews Conducted, by Program Type

	Active CHWs			Former CHWs
Program Type	No. of Surveys^a	No. of IDIs	Total No. Interviewed/ Total No. in Program	No. of IDIs
Public sector (2 districts)	35	13	48/48	3
Former NGO (3 districts)	82	14	96/112	0
NGO (2 districts)	66	16	82/203	2
Total	183	43	226/363	5

Abbreviations: CHWs, community health workers; IDIs, in-depth interviews; NGO, nongovernmental organization.
↵a Surveys included a Discrete Choice Experiment (DCE).

View popup

TABLE 3. Item Means, Grouped by Factor, by CHW Level of Activity

	Total (N = 183)	Highly Active (n = 88)	Less Active (n = 69)
Motivational Outcomes
Overall, I am very satisfied with being an FP CHW.	3.78	3.73	3.85
I would recommend to my children that they become CHWs.	3.50	3.57	3.48
I am glad to be working as an FP CHW rather than in some other volunteer position in the community.	3.20	3.13	3.40
I feel very little commitment to the FP CHW program.^a	1.35	1.39	1.33
I enjoy working in my community to make it better, even if it is without pay.	3.77	3.81	3.74
I do not think that it makes good sense to spend any time working in my community without payment.^a	1.55	1.46	1.57
I have no intention to keep doing my FP CHW work without pay.^a	1.42	1.36	1.40
Motivational Determinants
Barriers
My FP CHW work takes so much time that I am worried about how to support myself and my family.^a	2.13	2.03	2.07
My family complains about the demands of my FP CHW work.^a	1.53	1.65	1.39
Serving as an FP CHW will not help my chances of getting a better job in the future.^a	1.45	1.51	1.47
Talking about FP with my clients is very embarrassing.^a	1.27	1.24	1.33
It is difficult to help clients find the method that is right for them.^a	1.54	1.37	1.59
Contraceptive use often makes people sick.^a	2.16	2.21	2.05
Facilitators
Overall, my supervisors provide helpful feedback and support.	3.77	3.72	3.74
I can probably perform most of my activities without additional training.	2.51	2.50	2.58
The FP CHW program provides all the equipment and material I need to do my job well.	3.19	3.12	3.35
It is easy to find women who are interested in receiving an FP method in this community.	3.67	3.64	3.75

Abbreviations: CHW, community health worker; FP, family planning.
Level of activity of CHWs was based on data from service statistics, which were available for 157 CHWs.
Items were scored from 1 = “Disagree a lot” to 4 = “Agree a lot.” Nonresponses varied across items. Weighted means are reported.
↵a Item was reverse-scored before factor analysis.

View popup

TABLE 4. Selected Characteristics of Survey Respondents, by Level of Activity

Characteristics	Total	Highly Active	Less Active	P Value
Characteristics	(N = 183)	(n = 88)	(n = 69)	P Value
Age, y, mean (SE)	41.3 (0.8)	41.1 (1.1)	40.9 (1.2)	.89
No. of living children, mean (SE)	5.5 (0.2)	5.2 (0.2)	6.0 (0.5)	.23
Education, %
Primary	27.4	26.1	37.4	.17
Secondary or higher	72.6	73.9	62.6
Sex, %
Female	49.8	59.1	48.8	.47
Male	50.2	40.9	51.2
No. of years as an FP CHW, mean (SE)	5.6 (0.4)	5.5 (0.4)	5.3 (0.4)	.83
Provides other services besides FP, %
No	7.3	4.8	9.4	.31
Yes	92.7	95.2	90.6
Prior volunteer experience, %
No	10.9	12.2	5.3	.05
Yes	89.1	87.8	94.7
Received refresher training in past year, %
No	26.1	27.2	32.9	.50
Yes	73.9	72.8	67.1
Ever received supervision from HC staff, %
Never	34.0	39.4	36.7	.82
Ever	66.0	60.9	63.3
Received supervisory visits in community in the past year, %
No	41.3	36.0	50.8	.23
Yes	58.7	64.0	49.2
Problems with supplies, %
No	36.1	26.7	41.2	.01
Yes	63.9	73.3	58.8
Received recognition/support from community in the past year, %
No	10.9	10.0	12.0	.75
Yes	89.1	90.0	88.0
Received incentive from NGO or district in the past year, %
No	46.2	45.2	47.4	.81
Yes	53.8	54.8	52.3
Collaboration with other CHWs, %
No	37.1	46.8	24.7	.03
Yes	62.9	53.2	75.3
Time to HC, h, mean (SE)	1.2 (0.1)	1.3 (0.1)	1.3 (0.1)	.88
Motivation, mean (SE)	-0.08 (0.1)	-0.05 (0.1)	-0.04 (0.1)	.95
Barriers, mean (SE)	-0.05 (0.1)	-0.04 (0.1)	-0.04 (0.1)	.96
Facilitators, mean (SE)	-0.02 (0.1)	-0.05 (0.1)	0.07 (0.1)	.41

Abbreviations: CHW, community health worker; FP, family planning; HC, health center; NGO, nongovernmental organization; SE, standard error.
Level of activity of CHWs was based on data from service statistics, which were available for 157 CHWs.
Nonresponses varied across items. Weighted percentages and means are reported.

View popup

TABLE 5. Factors Associated With CHW Level of Activity in Logistic Regression Analysis (N = 156)

Characteristics^a	Odds Ratio (95% CI)
Characteristics^a	Odds Ratio (95% CI)	Demographic
Male^b	0.63 (0.23–1.76)
Secondary or higher education^b	2.30 (0.75–7.01)
Age, y	1.04 (0.99–1.08)
Situational
Prior volunteer experience^b	0.67 (0.26–1.72)
Travel time to health center, h	0.98 (0.68–1.43)
Work
Problems with supplies^b	2.22 (1.32–3.75)
Collaboration with other CHWs^b	0.33 (0.13–0.86)
Motivational outcomes and determinants
Motivation	1.25 (0.70–2.23)
Barriers	1.07 (0.67–1.72)
Facilitators	0.67 (0.40–1.14)

Abbreviations: CHW, community health worker; CI, confidence interval.
↵a Control variables were education, sex, age, time to health center, motivation, barriers, and facilitators.
↵b Indicator variable. For male/sex, the reference is female. For education, the reference is primary education. Other variables are yes/no binary variables, with “no” as the reference level.

View popup

TABLE 6. Mixed Logit Model Results for Program Inputs Influencing CHW Preferences in the Discrete Choice Experiment (N = 182)

Program Input	Model Coefficients
Program Input	Mean Estimate (SE)	Standard Deviation Estimate (SE)
T-shirt, badge, and bicycle	3.90^a (1.41)	1.24 (1.45)
Mobile phone, no airtime	0.99^b (0.41)	…
10,000 UGX transport refund	0.77^b (0.35)	1.06 (0.77)
Yearly refresher training	0.73^b (0.34)	1.22 (0.63)
T-shirt and badge	1.97 (1.02)	1.51 (0.86)
Quarterly supervisory visits in community	0.70 (0.38)	1.15 (0.70)