Identifying Participants andCommunity Capacity:Means to Improved Health Practicesor an End-in-itselfCarol Underwood, PhDAp...
CC - Background• The intervention– Health Communication Partnership in Zambia(HCPZ)– 5-year, USAID-supported project, 2005...
CC - Background• What is community capacity?– The characteristics of communities that influence theirability to overcome b...
Study Approach• Goals– Characterize & develop CC domains and indicators– Validate domains & indicators– Test validated com...
Phase I Results• Community members identified 11 uniquedomains:– sense of community belonging; effectivecommunity organisa...
Phase II Results• Quasi-probability sample of 720 individuals• Study found:– Social cohesion (7 indicators); alpha=0.621– ...
Phase III
Phase III: Hypotheses• We hypothesized a multi-step pathwayleading from the intervention activities tohealth behaviors thr...
Phase III: Methods• Probability sample of 2,462 women (15-49) & 2,354 men (15-59) from 24intervention & 12 comparison dist...
Phase III Results• Individuals living in interventioncommunities, regardless of the level ofintensity of these activities,...
Odds Ratios and Confidence Intervals from amultivariate logistic regression model predictingreported community action to a...
Phase III Results• Individuals reporting higher levels ofcommunity capacity were also morelikely to report that their comm...
Community capacity &health behaviors• Compared to individuals who did not reportthat their community worked together in th...
Community capacity &health behaviors (cont’d)• Controlling for community action, exposureto the intervention was not direc...
Total, direct and indirect sizes estimatedfrom mediation analysisStep A – B Step B – CFP Use Bednet Use HIV TestTotal effe...
Limitations• Qualitative phase: not generalizable, based onpurposively selected sample.• Post-test only (baseline data at ...
Conclusions• Enhanced capacity was signficantlyassociated with having taken communityaction for health – a community-level...
Conclusions (continued)• This demonstrates that buildingcommunity capacity, in this instance,was both a means to an end – ...
Identifying Participants andand now over to you . . .
Upcoming SlideShare
Loading in …5
×

Community Capacity Means to Improved Health Practices or an End-in-Itself_Carol Underwood_4.25.13

431 views
325 views

Published on

Published in: Technology, Health & Medicine
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
431
On SlideShare
0
From Embeds
0
Number of Embeds
0
Actions
Shares
0
Downloads
6
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Community Capacity Means to Improved Health Practices or an End-in-Itself_Carol Underwood_4.25.13

  1. 1. Identifying Participants andCommunity Capacity:Means to Improved Health Practicesor an End-in-itselfCarol Underwood, PhDApril 25, 2013
  2. 2. CC - Background• The intervention– Health Communication Partnership in Zambia(HCPZ)– 5-year, USAID-supported project, 2005-2010– Key aspect: to enable individuals &communities to take positive health actions &strengthen community-based systems andnetworks– Community capacity building a central featureof HCPZ
  3. 3. CC - Background• What is community capacity?– The characteristics of communities that influence theirability to overcome barriers and find or cultivateopportunities to address social, economic, politicalissues– A social protective factor or “condition that canmitigate social ills” IOM• What does the literature tell us?– Extensive CC literature; identifies a range of domains,including participation, leadership, social and inter-organizational networks, sense of community,resource mobilization, among others– Yet, research has rarely explored how communitiesdefine and understand the concept
  4. 4. Study Approach• Goals– Characterize & develop CC domains and indicators– Validate domains & indicators– Test validated community capacity indicators• Three phases:• Phase I:– Literature review to inform qualitative study– Qualitative study: 16 FGDs with minors, adult women, adult men,urban & rural• Phase II: Field test & validation of identified indicators• Phase III: Evaluation• Co-authors:– Marc Boulay, Gail Snetro-Plewman, MubianaMacwan’gi, Janani Vijayaraghavan, MebeloNamfukwe, David Marsh
  5. 5. Phase I Results• Community members identified 11 uniquedomains:– sense of community belonging; effectivecommunity organisation and institutions;enhanced community participation; communitycooperation; strengthened community support;improved use of individual skills, knowledge andabilities; community power; social cohesion;resource mobilisation; leadership; and ability toraise awareness• International team met to vet the domains,augmented domains with key areas from theliterature, reduced 11 domains to 6 for fieldtesting
  6. 6. Phase II Results• Quasi-probability sample of 720 individuals• Study found:– Social cohesion (7 indicators); alpha=0.621– Collective efficacy (4 indicators); alpha=0.792– Conflict management (4 indicators);alpha=0.621– Type of leadership (5 indicators); alpha=0.785– Effective leadership (6 indicators);alpha=0.853– Participation (4 indicators); alpha=0.739
  7. 7. Phase III
  8. 8. Phase III: Hypotheses• We hypothesized a multi-step pathwayleading from the intervention activities tohealth behaviors through their effect onCommunity Capacity:– H1: the interventions will be associated with/influence Community Capacity [Step A];– H2: Capacity will then prompt CommunityAction [Step B], and– H3: Community Action will affect healthbehaviors [Step C].
  9. 9. Phase III: Methods• Probability sample of 2,462 women (15-49) & 2,354 men (15-59) from 24intervention & 12 comparison districts• Principal components analysis withvarimax rotation identified a single factorthat explained 60% of the variance acrossthe CC indicators• Therefore, a single scale to measure CCwas retained
  10. 10. Phase III Results• Individuals living in interventioncommunities, regardless of the level ofintensity of these activities, reportedsignificantly higher scores on theCommunity Capacity scale compared toindividuals living in the controlcommunities. H1 supported.
  11. 11. Odds Ratios and Confidence Intervals from amultivariate logistic regression model predictingreported community action to address a healthproblem in the past yearSource: 2009 HCP/Zambia Endline SurveyDifference from Referent Group: *p<0.05; **p<0.01; ***p<0.001
  12. 12. Phase III Results• Individuals reporting higher levels ofcommunity capacity were also morelikely to report that their communityworked together in the past year toaddress a health problem in theircommunity. H2 supported.
  13. 13. Community capacity &health behaviors• Compared to individuals who did not reportthat their community worked together in thepast year, individuals who lived incommunities that worked together to addresshealth problems were:– twice as likely to be currently using a moderncontraceptive method– 1.8 times more likely to have received an HIV testand to know the results of that test, and– 1.5 times more likely to have had their youngestchild sleep under a bed net to prevent malaria.
  14. 14. Community capacity &health behaviors (cont’d)• Controlling for community action, exposureto the intervention was not directly relatedto these behaviors and communitycapacity was directly related to the use ofa bed net, but not other health practices.• Yet . . . there were also indirect effects.
  15. 15. Total, direct and indirect sizes estimatedfrom mediation analysisStep A – B Step B – CFP Use Bednet Use HIV TestTotal effectsizeCoefficient 0.045* 0.077* 0.145* 0.096*95% CI 0.007, 0.082 0.041, 0.114 0.071, 0.217 0.061, 0.133Direct effectsizeCoefficient 0.020 0.022 0.122* 0.051*95% CI -0.018, 0.054 -0.017, 0.060 0.046, 0.195 0.014, 0.087Indirect effectsizeCoefficient 0.025* 0.055* 0.022* 0.046*95% CI 0.017, 0.034 0.041, 0.068 0.006, 0.043 0.036, 0.057Proportion oftotal effectmediated56.2 71.5 15.3 47.6Step A – B: Intervention  Community Capacity  Community ActionStep B – C: Community Capacity  Community Action  Health Behavior*Effect size different from 0, based on bootstrap using 500 iterations
  16. 16. Limitations• Qualitative phase: not generalizable, based onpurposively selected sample.• Post-test only (baseline data at both control andintervention communities would have strengthenedthe evidence)• Challenges associated with measuring macro-levelconcepts, such as community capacity, withindividual-level reports.• We suggest this approach is a useful step in thedevelopment of a tool for evaluating the effectivenessof community-based projects for health and socialdevelopment.
  17. 17. Conclusions• Enhanced capacity was signficantlyassociated with having taken communityaction for health – a community-leveloutcome. Thus, community members cametogether in an attempt to collectively improvehealth outcomes.• Moreover, by fostering community capacityand stimulating community action, theintervention appears to have had signficantindirect effects on such health behaviors ascontraceptive use, receipt of HIV tests, andbed net use among young children.
  18. 18. Conclusions (continued)• This demonstrates that buildingcommunity capacity, in this instance,was both a means to an end – improvedhealth behaviors and reported collectiveaction for health – and an end-in-itself,both of which are vital to socialdevelopment.
  19. 19. Identifying Participants andand now over to you . . .

×