Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Since it was developed in the United States in the 1930s contact tracing, also known as partner notification, has come to be regarded as a key element in the control of sexually transmitted infections (STIs) and has been implemented in many parts of the world.1 2 The sexual partners of individuals with STI or HIV infection are notified, counselled on their exposure, and offered medical services.3 Sexual partners may be informed by the index patient (partner referral), a healthcare worker (provider referral), or a conditional approach, where initial partner referral is followed by provider referral after an agreed interval. Partner referral involves a health education interview with the index case possibly supplemented by contact slips; provider referral involves healthcare workers tele-phoning, writing to, or visiting contacts. Possible benefits include reduction in morbidity in the contacts through earlier treatment; reduction in the transmission of STIs by contacts through treatment and health education; and gaining insight into patterns of spread.
Sexual health services now face critical appraisal, both from purchasers of services and from non-specialists who are screening for STIs in areas such as primary care and family planning clinics. What evidence is there for the benefits of contact tracing? How should contact tracing evolve to meet the needs of these new patient groups, who may be asymptomatic, unsuspecting of infection, and unprepared to approach partners?
Is contact tracing an effective intervention? The evidence from published studies is generally supportive but mixed. It is possible to bring significant numbers of partners with previously undiagnosed STI, both bacterial (for example, gonorrhoea and syphilis4) and viral (for example, HIV5), into contact with health services through contact tracing. However other studies have found contact tracing to be only partially successful especially when index cases deny having sufficient information to contact partners. In a report on contact tracing for infectious syphilis in Florida, only 2236 (19.8%) out of 11 272 potentially exposed partners were located mainly due to insufficient information being provided by the index cases.6 Potterat7 argued that this disappointing outcome might have resulted from poorly trained and motivated staff working in a difficult environment where many patients were using “crack” cocaine with consequent hypersexuality.
Hard evidence on other outcome measures is scarce. It has not been proved that contact tracing results in a decrease in the prevalence of STIs at the community level although this is not surprising given the difficulty in establishing accurate community prevalences8 for STI let alone demonstrating clear changes in response to interventions. It has not been shown in a randomised controlled trial that contact tracing results in positive behavioural change although uncontrolled observations have suggested that contact tracing for HIV may lead to a reduction in the number of sexual partners of both HIV infected and uninfected contacts.5 Completing a particular contact tracing process does not equate to a successful outcome. Ramstedt et al found that although there was better adherence to protocols for contact tracing for chlamydial infection in women over time in Gothenburg, Sweden, the overall reinfection rate by their regular partners did not fall.9
There is a lack of good evidence on the cost effectiveness of contact tracing. In an analysis of a hypothetical cohort Howell et al showed that contact tracing for chlamydial pelvic infection would be cost effective provided at least 11% of the named female partners of male index cases, or 43% of the named male partners of female index cases, received treatment.10 These thresholds might not be met in all population groups; Oh et al could only verify that 27% of the male partners of adolescent females with chlamydial or gonococcal infection had received treatment following contact tracing.11 Howell et al only addressed the direct costs of pelvic infection and ignored the benefits of preventing further transmission, thus probably underestimating the true cost effectiveness of contact tracing. No study of the cost effectiveness of contact tracing for HIV has been published. Reports giving figures for its cost suggest it is a relatively expensive way of identifying new cases of HIV infection.12 For example, three papers from the United States published in the period 1991–3 reported costs per new case of HIV infection identified in the range $1625–3205 (at contemporary prices) through partner notification.13–15 This should not be taken as suggesting that it is not cost effective given the high, and rising, costs of caring for cases.
If we accept that contact tracing is useful in the control of STIs, how is it best done in the field? Again there is less evidence than we would like. In a systematic review of contact tracing published in 1994 Oxman et al identified only eight comparative studies of methods in contact tracing which they believed to be methodologically strong.16 They found little evidence to indicate which approach was superior for syphilis, gonorrhoea, or chlamydial infection although one good study showed provider referral to be more effective than patient referral in contact tracing for HIV.17 A subsequent study found no difference between conditional referral and provider referral strategies for syphilis.6 Oxman et al also remarked that none of the studies had attempted to measure any of the potential negative psychosocial effects of contact tracing such as anxiety, or the precipitation of domestic violence, or inquired of those traced as to the most acceptable method of approach. It has been shown that patient referral can work well for long term partners but casual sexual contacts are traced better by provider referral.18
If and how contact tracing is done outside STI clinics is poorly documented. In a Scottish primary care audit only 13% of patients diagnosed as having chlamydia attended specialist services for contact tracing.19
How should contact tracing develop in the future? Firstly, we must attempt to evaluate the processes used in a more systematic manner in order to convince critical non-specialists of the value of contact tracing. Given the almost universal agreement on, and implementation of, this intervention in STI services it is surprising how many basic questions remain unanswered. Secondly, contact tracing services must adapt to meet the challenges of community screening. Acceptability of methods becomes much more important when a large proportion of infections are unsuspected and threaten the stability of long term relationships. The expertise required for successful contact tracing needs to be made available in non-specialist areas—for example, through closer liaison with specialist sexual health services. This area should therefore provide fertile ground for clinical research studies in the future.