Recommending a Strategy

Document Sample
Recommending a Strategy Powered By Docstoc
					Population-based Interventions to Improve Sexual Health: Development and Evaluation
Colleen A. Redding, Ph.D.
Cancer Prevention Research Center University of Rhode Island

How many people get new STI’s in the U.S. every year?
Disease All STI’s HPV (Genital warts) Trichomoniasis Chlamydia HSV (Genital herpes) Gonorrhea Syphilis HIV New cases/year 18.9 million 5.5 million 5 million 3 million 1 million 650,000 70,000 >40,000

Public Health Cost?
 $9.3–15.5
8

Billion per year in direct medical costs only
STIs (HIV, HPV, HSV2, HepB, Chlamydia, is Y2000 $
Gonorrhea, Syphillis, Trichomoniasis)
 estimate

 $6.5

Billion among 15-24 yr. olds

(Chesson et al., 2004)

The Transtheoretical Model
Intentional Behavior Change • Stages of Change • Decisional Balance • Situational Efficacy / Temptations • Processes of Change

Different variables important for each stage transition

What are Expert Systems?

 A computer software program that codifies the reasoning of human experts into decision rules or algorithms
 Integrates assessment and feedback consistently using decision rules

Different Levels of Targeting/Tailoring
Group Level - Same intervention for all

Stage Level - Targeted interventions
Precontemplation Contemplation Preparation Action Maintenance

Intermediate Level - Tailored interventions

Individual Level-Expert system interventions

Stage-targeted vs. TTM-Tailored
univariate group feedback clinical decisions multivariate (10-15) individual feedback empirical decisions normative comparison ipsative comparison thousands of types interactive algorithms vary by Stg

5 different types

Benefits of Expert System Interventions
Provide highly individualized feedback  Appropriate for those at all stages of change, (not only prepared to change)  Potentially cost-effective  Integrate multiple risk behaviors  Multimedia components  Confidentiality  Force explicit (testable) decision rules  High Fidelity


Efficacy of TTM-Tailored Interventions for Single Health Behaviors
       

Smoking Cessation Healthy Diet Physical Activity Sun Protection Medication Adherence Stress/Depression Management Mammography screening School Bullying

Efficacy of TTM-Tailored Tx with Multiple Behaviors
Smoking, Diet, Sun Protection  Smoking, Diet, Sun, Mammography  Smoking, Diet, Blood Glucose Monitoring  Smoking, Diet, Physical Activity, Stress  Diet, Physical Activity


Steps in the Intervention Development Process


Focus Groups


Learn language and how participants think about the area.



Pilot Sample
Validate measurement structure of constructs  Normative database  Assess variables that differentiate stages


Develop Prototype – test - retest  Pilot test intervention  Efficacy/effectiveness trial (s)


CA Redding1, JO Prochaska1, JS Rossi1, K Armstrong2, D Coviello2, UE Pallonen1, K Evers1, WF Velicer1, & L Ruggiero1
1 - Cancer Prevention Research Center, University of RI 2 - Family Planning Council, Philadelphia, PA

Human Papillomavirus - HPV


The most prevalent STI in the U.S.  Prevalence highest among 18-24 year old women (14% - 50%) (men not studied well)  Some HPV subtypes cause genital cancers  > 99% of cervical cancers have HPV DNA detected within the tumor  HPV associated with penile, anal, and oral cancers  New HPV vaccine protects against 4 types

Step By Step: Steppin’ for Healthier Teens
4 urban family planning clinics – Philadelphia metropolitan area  About 75% participation rate among eligible adolescents  833 female nonpregnant 14-17 y.o.  Teens - informed assent/consent parental consent not needed  Randomized clinical trial


Sample Diversity (N=831)
Race/Ethnicity
Black / African-American

%
81.0

White / European-American
Hispanic / Latina Native American Other / Multiracial

7.3
7.8 1.4 1.8

Sexual Risks
%
Age of sexual debut 13-14 y.o. Hx. Chlamydia Hx. Gonorrhea Hx. HPV, Herpes, or Syphilis Hx. Pregnancy (at least one) 62.7 20.5 10.3 9.4 36.0

Urban Female Teens (N=828) Stages of Condom Adoption
35 30 25 20 15 10 5 0 PC C PR A M

13.6% N=113

31.0% N=257

15.0% N=124

17.3% N=143

23.1% N=191

Pros and Cons of Condom Use
Weight of the positive and negative aspects of behavior change

PROS BENEFITS

CONS COSTS

REASONS To use condoms NOT to use condoms

Functional Relationship Stages & Pros + Cons
56 54 52 50 48 46 44 PC C PR A M Pros Cons

Baseline Sample - Pros & Cons of Condom Use (T-scores) by Stage
55

Pros
50

Cons
45

40 PC
n=113

C
n=257

P
n=124

A
n=143

M
n=191

N=828

Baseline Sample - Confidence in Condom Use (T-scores) by Stage
60

55

50

45

40 PC
n=113

C
n=257

P
n=124

A
n=143

M
n=191

N=828

TTM-Tailored Expert Systems for Condom Use & Smoking
For use in Family Planning Clinics  Mouse input (no keyboard!)  On-screen and printed feedback  Printed feedback for both participant and her clinic counselor  Smoking system appropriate for both smokers (cessation) and nonsmokers (prevention)


TTM Tailored Intervention Package


Interactive assessment and expert system feedback (onscreen & printed)
 

Condom Use Promotion Smoking Cessation OR Prevention Stages of change Pros & Cons Confidence or Temptation Processes of Change



Tailored feedback based on:
   



Stage-Matched Counseling

Standard Care Intervention Package
Identical computer-delivered assessment and generic feedback to use condoms, condom tips, and advise to either quit smoking or avoid starting to smoke.  Standard family planning counseling on birth control and condom use.


Stage-matched Counseling
Can be used with teens at all stages of change, not only those ready for action  Comparable to Motivational Interviewing  Counselors match Process exercises to stage using Manual  Counselor received printed output from computer with client’s stage of change and processes to work on


Processes of Change
people change • cognitive, emotional, behavioral, interpersonal strategies/techniques used to change behavior • different processes mediate transitions between stages • process-to-outcome research • foundation of intervention design
• HOW

Processes of Change
Experiential Processes Behavioral Processes

Thinking, Feeling or Experiencing Consciousness Raising Dramatic Relief Environmental Reevaluation Self Reevaluation Social Liberation

Doing
Counterconditioning Helping Relationships Reinforcement Management Self Liberation Stimulus Control

Newer Interpersonal Processes



Condom Communication - talking about condom use Condom Assertiveness - insisting on condom use Eroticizing Condoms - finding ways of making using condoms more enjoyable
Partner Support - getting partner’s support for condom use Interpersonal Systems Control - avoidance of challenging people and/or social/sexual situations


 

Experiential Processes of Change For Condom Use By Stage
60 55 CR DR ER SO SR

50

45

40

35 P C D A M

N=113

N=257

N=124

N=143

N=191

N=828

Interpersonal Processes of Change By Stage
60

55 AS CO EC PS

50

45

40

35 P C D A M

N=113

N=257

N=124

N=143

N=191

N=828

Retention Rates
Assessment/Intervention
Baseline

N
833

%
100

Time 2
Time 3

470
437

56.4
52.5

Time 4
12 months 18 months

442
530 500

53.1
63.6 60.0

% A/M - Condom Use in Baseline nonusers by Group by Time
50 45 40 35 30 25 20 15 10 5 0 Baseline 6 months 12 months 18 months

TTM Std. Care

% A/M – ITT Condom Use by Group by Time (N=494)
30 25 20 15 10 5 0 Baseline 6 months 12 months 18 months TTM Std. Care

% A/M - Baseline condom users by group by time
120 100 80 60 40 20 0 Baseline 6 months 12 months 18 months TTM Std. Care

Quit Rates in Smokers by Group at 18 months (n=88, ns)
40 30 20 10 0 18 Months TTM SC

Smoking Uptake among Baseline Nonsmokers by Group
20 15 10 5 0 18 months TTM Std. Care

Step by Step Conclusions
Results support the efficacy of the TTM Tailored expert system intervention & stage matched counseling package to increase condom use and reduce condom relapse in this high risk sample  Despite lack of statistical significance, smoking cessation results at 18 months replicated prior results with adults and adolescents.  No support for effectiveness of the smoking prevention intervention.  Significant initial increases in condom use were sustained over 18 months, however, control group caught up.


Remaining Questions?
Would these results generalize to at risk adults?  Would condom use results hold up without the counseling component?


Tailored intervention to increase dualmethod use: an RCT to reduce unintended pregnancies and STIs

Peipert JF1, Redding CA3, Blume JD2, Allsworth JE1, Matteson KA2,4, Lozowski F2,4 , Mayer KH2, Morokoff PJ3, Rossi JS3
1- Washington University, School of Medicine, St. Louis, MO 2 - Brown University, Providence, RI 3 - University of Rhode Island, Kingston, RI 4 - Women and Infants Hospital, Providence, RI 5 - Rhode Island Hospital, Providence, RI

Project PROTECT Study Dual Method Use
Recruited N=542 at risk women (13-35)  59% of eligibles recruited  Tested for STIs before enrollment



If +, treatment & test of cure before enrollment

English speaking  Avoid pregnancy X 2 years  < 13 y.o. required parental consent  RCT


PROTECT Study Timepoints


Baseline – full exam
TTM group - 1 + 2 months sessions Standard Care – no additional sessions

6 & 18 months phone survey  12 & 24 months – full survey & exam


PROTECT Baseline Sample Characteristics (N=542)
        

Median Age = 22 years 90% Single 25% < H.S. Education (*unbalanced) 22% Black & 17% Hispanic 47% History STI (*unbalanced) 49% History unplanned pregnancy 34% No contraceptive use 33% Hormonal contraceptive use 48% smokers

PROTECT Study Outcomes
TTM N=272 n (%)
Reported Dual Method Use Reported Consistent Condom Use Any STI or unintended pregnancy

Control N=270 n (%)

Unadjusted HRR (95% CI)

Adjusted for propensity score

86 (32)

71 (26)

1.38 1.70 (1.00, 1.89) (1.09, 2.66) 1.14 1.26 (0.89, 1.47) (0.88, 1.79) 1.08 1.19 (0.81, 1.44) (0.79, 1.79)

124 (46)

124 (46)

95 (35)

93 (34)

PROTECT Conclusions
TTM Tailored Expert system increased reported dual method use (~ 70%)  Smaller effect on condom use (~ 30% increase) – not significant  No effect on incident STIs and unplanned pregnancies


RI Project RESPECT
CA Redding1, PJ Morokoff1, JS Rossi1, KS Meier1, BB Hoppner1, K Mayer2, B Koblin3, P BrownPeterside3
1 – Psychology Department & CPRC, University of RI 2 - Miriam Lifespan Hosp. & Brown Univ., Providence, RI 3 – New York Blood Center, Bronx, NY

RI Project RESPECT



9 local sites in urban areas


Drug Tx. Programs, STD Clinics

1 site in the Bronx, NY Blood Center  RCT  TTM-Tailored ES Feedback compared to Generic feedback alone  Intervention at Baseline, 2, 4 months  Follow-up at 6, 12, 18 months

Participation Criteria
18 - 44 years old & English speaking  Heterosexually active in past 3 months


unprotected vaginal or anal sex  At least one opposite sex partner


Not pregnant or trying to get pregnant  Self report - HIV Negative


Participation Criteria continued


One of the following in the past year:  3 sexual partners  diagnosed with an STI (other than HIV)  a sex partner with 3 sex partners  a sex partner who is a bisexual male  a sex partner who has injected drugs  exchanged sex for money or drugs

Baseline Sample (n=315)
Age mean = 32.2 years (s.d. = 8.1)  Gender



28% Male



Employment


65% Unemployed  21% Full-time work  11% Part-time work  3% Other  Education


42.5% < H.S.

More Sample Description


Diversity (matches rates of HIV in RI)
38% White  33% African American  23% Hispanic




Relationship Status
86% Unmarried/Separated/Divorced/Widowed  14% Married or Living With Partner




85% Sexually Active in past 2 months

Behavioral Risks
%
Hx. Of STI Used injection drugs Not use condom @ last sex Age of sexual debut # sex partners in past 30 days 46.4 18 72 15.25 3.2

Baseline Stages of Condom Adoption (N=315) At-Risk Sample of M + F
45 40 35 30 25 20 15 10 5 0 PC C PR A M

40%

41%

14%

5%

-

Expert System Enhancements
New background pictures + recorded new adult male and female audio.  Gender-matched systems  Added new sections for main and other partner readiness to use condoms.


Respect Retention Rates
Assessment/Intervention
Baseline 2 months 4 months 6 months

N

%

12 months
18 months

527 409 338 359 324 278

100 77.6 64.1 68.1 62.2 52.8

Outcomes
- DVs:
 

# times unprotected sex in past 30 days (n = 267) % of times safe (includes those not sexually active in past 30 days)
(n=296) (N=292)

% A/M consistent condom use - Any Stage Progress (N=305)
-

Baseline to 6 Months X Group # times unprotected sex (N=267)
15 13 11 9 7 5 3 1 Baseline 6-Month Treatment Control

6 Mos. - % Time Being Safe
N=296
100% 80% 60% 40% 20% 0% Baseline 6-Month treatment control

% A/M at 6 months (n=292)
30% 25% 20% 15% 10% 5% 0% 0.246 0.14

Treatment

Control

 

Includes only Pre-Action Stages at Baseline (PC, C, PR) 10% more progress to A/M in Tx than in Control

Intent to Treat (ITT) Analysis of % A/M at 6 months (n=448)
30% 25% 20% 15% 10% 5% 0%

0.162 0.091

Treatment

Control

  

All baseline pre-action S’s included. Assumes no progress among lost-to-follow-up participants. Reduces effect from 10% to 7%, still statistically significant.

Any Stage Progress at 6 months
(n=305)
60% 50% 40% 30% 20% 10% 0% 0.551 0.451

Treatment

Control



10% more progress in Treatment than in Control

ITT Analysis of Stage Progress
(N=448)
60% 50% 40% 30% 20% 10% 0%

0.357

0.287

Treatment

Control



~ 7% more progress in Tx than in Control

RI Project RESPECT Conclusions
   

 

We were able to recruit a high risk sample of men and women from different sites. We were able to get good proportions (78%) of the sample to come back for at least 2 sessions. Retention was a concern. Results support the 6 month efficacy of the TTM Tailored expert system to increase condom use in this high risk sample. Longer term outcomes look like Step X Step… Durability of these effects over time?

Differences Across Studies?
Condom use STI Study N d [95% CI] d [95% CI] Protect 346 0.140 [0.07–0.35] 0.154 [0.06–0.36] RI Respect 292 0.461 [0.23–0.69] Step By Step 622 0.477 [0.32–0.64]

Noar SM, Black HG, Pierce LB. (2009). Efficacy of computer technology-based HIV prevention interventions: a meta-analysis. AIDS, 23, 107–15.

What’s next?
Process to outcome research  Compare cross-sectional to longitudinal findings  Examine predictors of changes over time  Enhance intervention outcomes



(More sessions? New variables? New behaviors? More ? )

Enhance retention  Generalize to additional at risk samples + settings  Dissemination & Translation


Useful Intervention Refinement Process
Focus Groups/ Formative Work

New questions?
Pilot Sample - Measurement work

Efficacy/effectiveness trials

Pilot test intervention


				
DOCUMENT INFO