17
Initial Outcomes of a Brief Motivational Interviewing-Based Intervention During Provider- Initiated HIV Testing and Counseling in Rural Uganda Susan M. Kiene 1,2 Haruna Lule 3 , Moses H. Bateganya 4 , Harriet Nantaba 3 , Rhoda K. Wanyenze 2 1 University of Connecticut School of Medicine, USA, 2 Makerere University School of Public Health, Uganda, 3 Gombe Hospital, Uganda, 4 CDC, Atlanta, USA Funded by NIMH, K01MH083536 23 July, 2014

Initial Outcomes of a Brief Motivational Interviewing-Based Intervention During Provider-Initiated HIV Testing and Counseling in Rural Uganda Susan M

Embed Size (px)

Citation preview

Initial Outcomes of a Brief Motivational Interviewing-Based Intervention During

Provider-Initiated HIV Testing and Counseling in Rural Uganda

Susan M. Kiene1,2 Haruna Lule3, Moses H. Bateganya4, Harriet

Nantaba3, Rhoda K. Wanyenze2

1 University of Connecticut School of Medicine, USA, 2 Makerere University School of Public Health, Uganda, 3 Gombe Hospital, Uganda,

4 CDC, Atlanta, USA

Funded by NIMH, K01MH083536

23 July, 2014

2

Background

Majority of those who are HIV positive are unaware of their status (WHO, 2013)

HIV testing and behavior change (e.g., Kamb et al., 1998; Marks et al., 2005)

Provider-initiated HIV testing Counseling

3

Intervention

Information-Motivation-Behavior Skills model (Fisher & Fisher, 1992, 2000)

Motivational Interviewing (Miller & Rollnick, 1991) Client-centered Identify patient’s risk behaviors Choose a behavior to change Discuss barriers to change Develop a risk reduction plan

Brief (<10 min.)

4

Hypothesis

MI-based client-centered counseling during provider-initiated HIV testing will be more effective than the standard-of-care counseling at reducing sexual risk behavior and related outcomes.

5

Gombe

6

RecruitmentBaseline Interview

Control n = 160 Intervention n = 173

3 and 6-month follow-up

HIV testStandard-of-care counseling

HIV testClient-centered

motivational interviewing counseling

RecruitmentBaseline Interview

7

Measures Number of sex events in prior 3-months with 3

most recent partners Condom use Partner type and knowledge of HIV status

Tested w/in prior 12 months “Risky” sexual events

Participants

8

Control n=160 Intervention n=173

% Female 51.9% 49.9%

Average age 33.5 (SD 9.5) 33.6 (SD 10.4)

Married 87.5% 86.1%

Education• Primary or <• Secondary (O level)• Adv. secondary (A level) and >

59.5%28.1%12.4%

67.6%24.3%8.1%

Employment• Never• Once in a while• Part of the year• Throughout the year

12.5%30.6%15.0%41.9%

10.4%24.9%28.3%36.4%

First time HIV testers 16.9% 17.4%

Retained through 6mo FU 90.2% 91.4%

Groups differed on employment

Baseline Descriptive Statistics

9

Control n=160 Intervention n=173

HIV positive 10.6% 8.7 %

Knows all partner(s)’ HIV status

20.0% 22.5%

Sex events M 20.79, SD 14.59 M 21.87, SD 17.56

Risky sex events M 14.91, SD 14.33 M 15.99, SD 16.41

Percent of sex events that were risky

73.7% 72.8%

Groups were not significantly different at baseline.

Percentage knowing their partner(s)' HIV status

10

Controlling for sociodemographics and HIV results

Time main effect: 3mo: χ2 20.09, OR 3.19 CI (1.92-5.26), p<0.001, 6mo: χ2 43.13, OR 6.76 CI (3.82-11.95),

p<0.001Time x Gender: 3mo: χ2 9.45, OR 0.44, CI (0.26-0.74), p=0.002

6mo: χ2 9.75, OR 0.39, CI (0.22-0.71), p=0.002

GEE regression model, autoregressive correlation structure, binomial distribution, logit link

11Time x study condition: 3mo: χ2 0.20 OR 0.85 CI (0.41-1.74) p=0.65, 6mo: 6mo: χ2 5.87 OR 0.42 CI (0.21-0.85) p=0.015GEE regression model, autoregressive correlation structure, binomial distribution (events w/in trials), logit link

12Time x study condition: 3mo: χ2 0.34 OR 0.90 CI (0.64-1.27) p=0.56, 6mo: 6mo: χ2 9.16 OR 0.59 CI (0.41-0.83) p=0.002GEE regression model, autoregressive correlation structure, Poisson distribution, log link

How did risk decrease?

13

Controlling for knowledge of partner(s)’ status: Time x study condition: χ2 9.22, df =2, p=0.01Time x study condition x knows partner status x gender: χ2 7.24, df=2, p=0.027

Risky sex events with serodiscordant vs. unknown status partners

14

Discussion and Conclusion

Women’s difficulty in getting their partners tested Intervention effectiveness

For both HIV negative and HIV positive Limitations

Study design Follow-up Self-report

Behavior change is possible!

15

AcknowledgementsParticipants

CollaboratorsRhoda WanyenzeMoses BateganyaHaruna Lule

FundingNIMH, K01MH083536

Research Assistants and StudentsRuth SessangaHarriet NantabaHajara Kagulire Farouk KimbowaRuth NamulemeNalongo KijjeKia JayaratneJoe JasperseRebecca SternKaty Sileo

Contact: [email protected]

Extra Slides

16

Changes in condom use

17