Mobilizing community action to improve maternal health in a rural district in Tanzania: lessons learned from two years of community group activities

Figures & data

Table 1. Process and outcome evaluation indicators and data collection methods.

Evaluation	Indicator	Data collection method	Data Source
Socio-demographic details of the community groups	Facilitators characteristics	Document review	Application and interview notes of 15 facilitators
	Community groups socio-demographics of participants	Document review	Attendance lists of all community groups
	Community groups establishment and continuation	Document review	Overview of active groups including number of meetings held.
Community group outcomes	Problem and strategies identification	Document review	Facilitator reports of 1922 group meetings
	Decision making process on problems and strategies	FGDs	Community groups (N = 9)
		Interviews	Facilitators (N = 18), group members (N = 21) and leaders (N = 3)
Logistics and supervision of the overall project	Local leadership involvement	Document review	Reports of 65 quarterly meetings
	Community meetings process	Document review	Facilitator reports of 1922 group meetings
		Observations	Research team reports of 51 community group meeting observations
	Facilitators development	Document review	Training reports of 4 one week trainings and 1 day workshop
		Document review	Reports of 16 supervision visits
	Project coordination	Document review	Reports of 66 project team meetings
		Interviews	Government officials (N = 3), project staff (N = 6), community members (N = 6)
	Evaluation of implementation costs	Document review and analysis	Quarterly and yearly financial reports
Experience and perceptions on the overall project	Leadership experience	Interviews	Local leaders (N = 11)
	Facilitator experience	Questionnaire Open ended (anonymous)	Facilitators (N = 15)
	Community group members experience	FGDs	Community Groups (N = 6)

Table 2. Characteristics of community groups at baseline and end-line.

	Baseline (onset of community groups,	End-line (evaluation of community groups,
	2014)	2016)
No of groups
Active groups	93	77
Additional groups	9	N/A
Terminated groups	N/A	25
Number of Registered members in community groups	2071	1365
Average no of members per group	23.8	18.2

N/A: not applicable

Table 3. Socio-demographics of community group participants at baseline and end-line.

Variable	Baseline (onset of community groups) 2014	End-line evaluation of community groups) 2016
Total number of participants	2071	1365
Gender
Female	1313 (63.4%)	929 (68.1%)
Male	714 (34.5%)	435 (31.9%)
Unknown	44 (2.1%)	1 (0.1%)
Marital status
Single	223 (10.8%)	110 (8.1%)
Married	1633 (78.9%)	1094 (80.1%)
Widow	119 (5.7%)	111 (8.1)
Divorced	8 (0.4%)	10 (0.7%)
Unknown	88 (4.2%)	40 (2.9%)
Level of education
None	330 (15.9%)	160 (11.7%)
Primary (not completed)	140 (6.8%)	65 (4.8%)
Primary (complete)	1374 (66.3%)	1038 (76.0%)
Secondary school	114 (5.5%)	51 (3.7%)
Higher education	3 (0.1%)	1 (0.1%)
Unknown	110 (5.3%)	50 (3.7)
Age
Mean (SD)	38.4(14.24)	39.75(12.64)
<20 years	61 (2.9%)	26 (1.9%)
20–34	846 (40.8%)	475 (34.8%)
35–49	700 (33.8%)	532 (39.0%)
50–65	332 (16%)	234 (17.1%)
>65 years	118 (5.7%)	61 (4.5%)
Unknown	14 (0.7%)	37 (2.7%)
No of children
Mean (SD)	4.86(3.09)	4.95(2.92)
None	140 (6.8%)	72 (5.3%)
1 or 2	383 (18.5%)	220 (16.1%)
3 or 4	443 (21.4%)	333 (24.4%)
5 or 6	488 (23.6%)	327 (24.0%)
7 or 8	332 (16.0%)	220 (16.1%)
9 or more	249 (12.0%)	160 (11.7%)
Unknown	36 (1.7)	33 (2.4%)

Table 4. Prioritized problems, solutions and actions taken by the community groups*.

	Problem (Phase 1)	No of groups (%)	Solution (Phase 2)	No of groups (%)	Actions (Phase 3)	No of groups (%)	Evaluation (Phase 4)	No of groups (%)
1	Health services too far	19 (25)	Providing health education	29 (38)	Providing health education to individuals/families, at health facility or during meetings	63 (82)	Behavioural change women/husbands (e.g. women are more prepared, more ANC visits)	47 (61)
2	Lack of service in health facility (e.g. equipment, medication)	15 (19)	Construct health facility	19 (25)	Setting up a community health fund	33 (43)	Increase of knowledge (of women, men, community)	44 (57)
3	Lack of health workers at the health facility	14 (18)	Increase health seeking behaviour for ANC	19 (25)	Collaborating with the government (inform leaders, have meetings, write letters, ask for help)	21 (27)	Behavioural change health workers (e.g. less corruption, better behaviour, transfer of HCW)	18 (23)
4	Medical problems related to pregnancy (e.g. bleeding, convulsions)	11 (14)	Promote facility birth	17 (22)	Preparing and taking action for construction work (e.g. collecting stones, water, cement)	18 (23)	Health outcomes improved (e.g. less maternal deaths, less still births)	17 (22)
5	Lack of health worker housing	10 (13)	Promote health insurance	9 (12)	Raising awareness on disrespectful behaviour of health providers	11 (14)	Good response from or collaboration with the community	17 (22)
6	Disrespectful behaviour of health providers	10 (13)	Construct health care worker housing	9 (12)	Encouraging community member to join the health insurance	9 (12)	Services improved (e.g. more equipment, supplies, shorter waiting time)	14 (18)
7	Lack of health education (e.g. poor knowledge or awareness)	8 (10)	Renovate existing health facility	4 (5)	Generating income for the community (chicken program/honey business)	7 (9)

* Some groups had indicated more than 1 prioritized problem or solution so the total percentage can be more than 100%. Not all groups prioritized problems aligned with prioritised solutions. Rows are ordered based on percentages of groups and therefore do not always relate to one topic. No = Number. HCW = Health care worker. ANC = Antenatal care.

Table 5. Process evaluation of implementation process based on Draper et al. (2010) [31].

Indicators	Evaluation
Leadership	Project initiated by project staff members in collaboration with community leadership and in response to findings of the needs assessment. Leadership structure formed through a top down approach based on local protocol. At district level a steering committee was formed including representatives of the district council, ward leadership and project staff. At ward level the ward district council was responsible to oversee the project implementation. Village/hamlet leaders were asked to provide support. Facilitators played an important role in voicing the concerns of community members.
Planning and Management	Project staff informed the community how to participate and decided the projects focus (on maternal health), goals and activities (community groups). Community invited to participate within a predetermined structure, but free in how and what they wanted to do. Activities reflect community priorities and involved local people including existing groups. Through training of facilitators some transfer of skills occurred.
Women’s involvement	The active participation of both men and women in positions of decision-making and responsibility was a main objective. Maternal health projects are often mainly focused on women involvement, active involvement of men was a conscious decision. At the same time, the project ensured facilitators and leaders (where possible) were female. Also the project staff had both male and female members.
Resource mobilisation	The majority of funding for the intervention was from outside the community. The initial intention was to get financial buy-in from the community leadership, but this never materialized. Communities offered non-financial resources including time, space and local expertise. Professionals allocated the use of resources, although this was done in consultation with the steering committee. Facilitators played an active role in development of their materials and how the coordination of the project should be structured.
Monitoring and evaluation	Project staff design evaluation approach and performed analyses, but facilitators and their groups were involved in data collection. A broad definition of ‘success’ was used, based on the groups own perspectives of their achievements. Responses to monitoring findings are jointly decided and community feedback is both sought and given.

Figure 1. Community group intervention participation scores according to dimensions in Draper et al.’s process evaluation framework [31].

Draper AK, Hewitt G, Rifkin S. Chasing the dragon: developing indicators for the assessment of community participation in health programmes. Soc Sci Med. 2010;71:1102–1109.

 PubMed Web of Science ®Google Scholar