Qualitative Assessment of the Application of a Discrete Choice Experiment With Community Health Workers in Uganda: Aligning Incentives With Preferences

Figures & Tables

TABLE 1.

Discrete Choice Experiment Attributes and Levels, Uganda, 2011

Attributes	Definition	Level
Supervision	Frequency and location of supervisory meetings	Monthly CHW meetings at health center Same as (1) + quarterly visit by health center staff in the community
Training	Frequency and duration of initial and refresher training	5-day initial training and 3-day supervised practicum at health center Same as (1) + 3-day refresher training once a year
Transportation refund	Transportation refund received for each supervisory meeting attended	5,000 UGX per meeting 10,000 UGX per meeting
Start-up package	Items received upon joining the CHW program (one-time)	CHW kit with gumboots, raincoat, job aids, and stationery CHW kit + T-shirt + badge CHW kit + T-shirt + badge + bicycle
Communication	One-time provision of a mobile phone to communicate with program staff	No mobile phone Mobile phone without airtime

TABLE 2.

Illustrative Pair of Job Profiles Presented to CHWs During the Discrete Choice Experiment

Attributes	Job A	Job B
Supervision	Monthly CHW meetings at health center	Monthly CHW meetings at health center + quarterly visit by health center staff in the community
Training	5-day initial training and 3-day supervised practicum at health center + 3-day refresher training once a year	5-day initial training and 3-day supervised practicum at health center
Transport refund	5,000 UGX per meeting	10,000 UGX per meeting
Start-up package	CHW kit with gumboots, raincoat, job aids, and stationery + T-shirt + badge + bicycle	CHW kit with gumboots, raincoat, job aids, and stationery + T-shirt + badge
Communication	Mobile phone without airtime	No mobile phone

TABLE 3.

Number and Characteristics of CHWs Participating in a Discrete Choice Experiment via In-Depth Interview, by Type of Family Planning Program

	Public (n = 13)	NGO (n = 16)	Former NGO (n = 14)	Total (N = 43)
Age, mean, years	43	43	41	43
Number of living children, mean	6	5	6	6
Marital status, %
Single	0	0	7	2
Married or cohabitating	100	69	79	81
Divorced, widowed, or separated	0	31	14	16
Gender, %
Male	31	38	43	37
Female	69	62	57	63
Educational level, %
Primary	23	50	21	33
Secondary or higher	77	50	79	67
Number of years of service, mean	7	5	10	7
Contraceptive methods provided, %
Condoms only	0	6	0	2
Condoms and pills	0	25	0	9
Condoms, pills, and injectables	100	69	100	88^a

Abbreviation: CHW, community health worker.
↵^a One CHW indicated providing pills and injectables but not condoms. The sum of CHWs reporting which contraceptive methods they provided does not total to 100% due to rounding errors.

TABLE 4.

Percentage Distribution of Survey and IDI Participant Responses to the Discrete Choice Experiment, by Block of Jobs Presented

Pairs of Jobs/Attribute-Level Combinations	Block 1		Block 2		Block 3		Block 4
Pairs of Jobs/Attribute-Level Combinations	Survey (n = 47)	IDI (n = 12)	Survey (n = 41)	IDI (n = 10)	Survey (n = 50)	IDI (n = 10)	Survey (n = 45)	IDI (n = 11)
Pair 1
Job A	9	0	27	30	12	10	11	0
Job B	91	100	73	70	88	90	89	100
Pair 2
Job A	81	67	15	20	81	70	27	9
Job B	19	33	85	80	19	30	73	91
Pair 3
Job A	35	42	14	0	73	70	62	73
Job B	65	58	86	100	27	30	38	18
No response								9^a

Abbreviation: IDI, in-depth interview.
Percentages of respondents selecting each job in the choices presented to them are reported for each sample (survey or IDI). Weighted percentages are reported for survey participants.
↵^a One IDI participant did not select an option for the third pair in this set.