Enhancing Gonorrhea Surveillance to Guide Program and Policy The Pros and Cons of Using Incentives in Behavioral Surveillance

1
Enhancing Gonorrhea Surveillance to Guide Program
and Policy The Pros and Cons of Using
Incentives in Behavioral Surveillance

• May 9, 2006National STD Prevention Conference
• Kathryn E. Macomber,1 MPHOana Vasiliu,2 MS
• Jennifer Bissette2
• 1 Michigan Department of Community Health
• 2 Virginia Department of Health

2
Objectives

• To compare data from an state offering incentives
  (VA) to a state that did not use incentives (MI)
• Methods
• Logistics
• IRB
• Survey Response
• Missing Values
• Unknown Values

3
Methods-MI

• 6 local health department (LHD) STD clinics
• Offered to all clinic attendees but only data for
  GC
• Self-administered
• Initially 2 health departments in July of 2003
• Added 3 more health departments in 2004

4
Methods-VA

• Conducted at Richmond City STD Clinic
• Attempted to solicit all attendees
• 80 of eligible attendees
• Trained interviewer with established interview
  guide
• Excluded from survey participation
• those under 14
• correctional inmates
• treatment only patients

5
Logistics-MI

• Questions integrated into existing intake form
• Revised instrument and copies provided by state
• Mailing labels and envelopes provided by state
• Surveys from GC patients mailed by LHD once a
  month
• Original copy remained in patients chart

6
Logistics-VA

• Surveys were entered regardless of test result
• Surveys were kept at the state level
• Summary of findings presented to LHD
• Lab results obtained from LHD staff 2-4 weeks
  after survey completion
• Surveys were linked to lab results by unique
  identifier

7
Sample Form-MI
8
Sample Form-VA
9
Incentives

• MI-no incentives
• VA patients provided with 10 MacDonalds gift
  certificate (01/15/03-08/01/05)
• Survey continued without incentives through
  9/23/05
• Total cost of incentives 72,225

10
IRB

• MI IRB determined to be surveillance
• Exempt from review
• VA Reviewed by VDH IRB
• Consent forms required
• All survey modifications underwent review

11
Overall Response
MI VA- incentive VA-no incentive
Date Range 07/01/2003-12/31/2005 01/15/2003-08/1/2005 08/03/2005-9/23/2005
Response Rate 76.7 (54.0-100) 92.0 80.0
Surveys Received 1,609 7,140 310
Completed surveys 88.0 99.3 99.7
12
Completeness by Variables
Variable MI VA
How many male partners? (3 mo.) 98.6 99.6
Sex while drunk/high? (last yr) 97.9 99.6
Meet sex partners on Internet? (last yr) 97.8 99.6
Use Drugs? (last yr) 92.2 99.6
HIV Status? 88.5 95.8
Been in Jail? (last yr) 96.2 99.6
13
Percent Unknown by Variable
Variable MI VA
Are you Hispanic? 27.3 0.0
Been in Jail?(last yr) 3.6 0.3
Had gonorrhea?(last yr) 6.2 0.7
14
Limitations

• Different Methodology
• Self-administered (MI) vs. Trained Interviewer
  (VA)
• VAs non-incentive time period occurred after
  having provided incentives
• Response Rates not completely comparable
• MI GC surveys/GC tests
• VA All Participants/All Solicited STD Clients

15
Conclusions

• Use of incentives associated with an increased
  cost
• Providing incentives positively impacted response
  rates
• Urban et al (2005)
• Completion rates (overall and by variable) were
  similar to MIs without an incentive
• Differences in response rates and completion
  rates may be more due to methods of survey
  administration (interviewer)
• Greenberg et al (1999)