1. Home
  2. Infectious disease topics
  3. HIV infection and AIDS
  4. Surveillance and disease data
  5. Behavioural surveillance toolkit
  6. Methods
  7. Specific populations
  8. STI clinic clients

STI clinic clientsArchived

Behavioural surveillance in STI clinics is venue-based sampling. The examples of designs for data collection in the table below include various methods that can be used in different settings and sampling populations. Each of the data collection methods described can, in principle, be applied to any of the settings or study populations, unless otherwise specified. The examples cited are not comprehensive.

Selected features of available methods for collecting behavioural data among STI clinic attenders

 

Design Advantages Disadvantages Comments

 Data collection methods
Continuous data collection    
  • Can be automated, routine data collection might improve completion rates
  • Allows collection of serial annual trends
  • Requires sophisticated IT infrastructure
  • Questions need to be highly structured
  • Might increase consultation time

 

Periodic surveys, e.g. during one week per year, three months per year (a), or alternate years  

 Limits workload for clinic staff Time period selected might not be representative of all periods

 

Unlinked anonymous data collection (b,c,d) 
  • Should reduce bias associated with refusal to participate
  • Can be targeted to specific populations
  • Requires infrastructure to link result with questionnaire and then remove identifiers
  • Requires universal testing for specific infections  
 Only for use with populations having specific tests, e.g. HIV tests done on all patients having syphilis serology

 

Settings for data collection 
 All clinics (e)

National data

 Logistically intensive and expensive for linking behavioural and test data

 
Sentinel clinics/sites (a,b,c,d,f)  More efficient as number of clinics is limited
  • Provides trend data only; cannot estimate national burden of STI unless clinics are a random selection
  • Changes in composition of participating clinics can make interpretation of trends difficult

 
Populations
All STI clinic attenders (b,c,f)  Allows comparison between people with STI and people without  Requires sophisticated IT infrastructure to link behavioural data with STI/HIV test data   
STI clinic attenders having specific tests, e.g. HIV tests (f)  
  • Defined population; suitable for unlinked anonymous surveillance
  • Allows collection of trend data
  • No information about behaviour of those not tested
  • Affected by changes in policies for HIV testing, e.g. shift from opt-in to opt-out
  
STI clinic attenders with specific diagnosed infections, e.g. syphilis, gonorrhoea (a), LGV, or a range of infections (g)  
  • Allows collection of trend data for behaviours associated with specific infections
  • Makes use of enhanced surveillance systems that are needed to assist control efforts
  • No information for comparison with those uninfected
  • If system relies on request for data collection after diagnosis, then often incomplete
  

 

 

Examples(a) UK: Gonococcal Resistance to Antimicrobials Surveillance Programme. Since 2003 in sentinel clinics.

(b) UK: Unlinked Anonymous Survey of Genitourinary Medicine Clinic Attendees. Since 1995 in 16 sentinel clinics.

(c) Netherlands: Unlinked anonymous HIV testing. Since 1991 in Amsterdam and Rotterdam.

(d) Spain: Unlinked anonymous survey on HIV prevalence. 1998-2002, 7 sentinel STI clinics around the country.

(e) Netherlands: Behavioural surveillance in STI clinic attenders having HIV tests. National since 2004.

(f) Spain: EPI-VIH survey. Since 1991, now includes 19 clinics

(g) Germany: Sentinel surveillance system for STI. Since 2002, sentinel sites. STD-Sentinel des RKI: Ausgewählte Ergebnisse unter dem Aspekt der Migration und Prostitution. Epid Bull 2007; 4:23-27;

Bremer V, Marcus U, Hofmann A, Hamouda O. Building a sentinel surveillance system for sexually transmitted infections in Germany, 2003. Sex Transm Infect. 2005 Apr;81(2):173-9.