Intended for healthcare professionals


Partner notification for the control of sexually transmitted infections

BMJ 2003; 327 doi: (Published 18 September 2003) Cite this as: BMJ 2003;327:633
  1. Sarah Hawkes, lecturer (sarah.hawkes{at},
  2. David Mabey, professor,
  3. Phillippe Mayaud, senior lecturer
  1. London School of Hygiene and Tropical Medicine, London WC1E 7HT

    Effectiveness in resource poor countries is unproved

    Partner notification (also termed partner management or contact tracing in some settings) is a well established public health activity in programmes to control sexually transmitted infection.1 The approach is based on the premise that the sexual partners of people with sexually transmitted infections are likely to be infected but may be asymptomatic and may not otherwise seek care. Partners can be reached through several different strategies including those led by infected “index” patients (patient led), by health providers (provider led), or by a combination of approaches (conditional referral–index patients are encouraged to ensure that partners attend by an agreed date, after which the provider will notify the partner). The public health objectives of each of these strategies are the same: to increase the coverage of care of sexually transmitted infections–by identifying and when necessary treating those people known to be at high risk for sexually transmitted infections 2; to interrupt the cycle transmission of infection; and to reduce the incidence of new infections and the overall burden of disease. This is the theory. In practice, do partner notification strategies work?

    The evidence for the effectiveness of strategies for partner notification has recently been reviewed.3 This systematic Cochrane review is the first to include studies of partner notification in low income countries–an economic definition, but one that includes many countries where sexually transmitted infections contribute substantially to the overall burden of adult disease.4 The authors of the review found 11 randomised controlled trials, which compared two or more strategies for partner notification, and only two of these studies had been undertaken in poor countries, both of them in southern Africa. This imbalance raises questions about transferring study findings (and by implication conclusions and recommendations for policies and programmes) to settings far different from those where the original research was undertaken–a concern raised by the authors of the review.

    The systematic review concluded that provider led referral (or a choice between patient led and provider led referral) is more likely to result in partners presenting for medical care than patient led referral, and that conditional referral for patients with gonorrhoea is more effective than patient led referral. In addition a slight increase in partners treated was seen in one study, which used nurse led health education and lay counselling as the intervention.

    What are the resource and programmatic implications of these findings for public sector health systems in settings that are poor in resources–those, for example, struggling to provide adequate care and attention in an average of well under five minutes per patient? Will providers in such settings really have time to undertake the steps needed to ensure that partner referral programmes are successful? Given the myriad competing concerns on health budgets, are partner notification interventions for sexually transmitted infections likely to be cost effective? Moreover, what are the implications of implementing either provider led or patient led referral programmes in settings of widely different cultural, social, and economic environments from the ones where most studies are undertaken–settings where, for example, there will be highly differential rates of power in sexual relationships5 or a risk of gender based violence?6 Should strategies found to be (just) effective in one type of setting be so easily recommended in an entirely different milieu?

    Concerns over the feasibility and effectiveness of partner notification for sexually transmitted infections in resource poor settings are compounded by the relative lack of specificity of many diagnoses of sexually transmitted infections in these settings. In the absence of highly sensitive and specific diagnostic tools at low cost, providers rely on approaches that may result in relatively high levels of overdiagnosis of sexually transmitted infections, especially in women.7 Although these approaches may sometimes be justified in public health terms, should they be the basis for recommending management of partners if we are not sure that the individual known as the index patient is truly infected?

    Partner notification has come a long way since its inception in the 19th century but has much further to go in terms of knowing what is effective in resource poor settings. While many studies concentrate on the issue of effectiveness before considering allocation of resources, it is time to build on the findings of this review and carry out methodologically sound trials to determine what is appropriate and acceptable to individuals in a variety of resource poor communities. This should be the first step in deciding whether partner notification is justified for programmes to control sexually transmitted infections globally.


    • Competing interests None declared.