Partner notification for the control of sexually transmitted infections

Abstract

Assisting patients in disclosing their diagnosis to partners is the biggest priority

In 2005 about 340 million people globally acquired new infections of the four most common curable sexually transmitted infections (gonorrhoea, chlamydia, syphilis, and trichomoniasis) and 4.1 million acquired HIV.¹,² Partner notification is essential to prevent reinfection of index patients, decrease the pool of infectious people, and prevent the transmission of HIV.³,⁴

Provider referral, where health service personnel trace and notify partners, is practised in parts of the developed world. Patient referral, where index patients are encouraged to inform their partners of the need for treatment, is universal practice in the developing world, where provider referral is neither feasible nor affordable.

In this week's BMJ, Trelle and colleagues report a systematic review of strategies to improve patient referral,⁵ as observational studies and randomised controlled trials indicate that current patient referral practices fail to reach many partners of people with sexually transmitted diseases in both the developed and developing world.⁶,⁷,⁸ Fourteen randomised controlled trials, four of which were conducted in countries with low average incomes, were reviewed. The trials evaluated two novel patient referral strategies: patient delivered partner therapy, where the index patient is given drugs or a prescription for their partner(s); and home sampling, where index patients with chlamydia give partners kits for collecting urine specimens, which are posted to a laboratory for testing. Meta-analysis of five trials (four conducted in the United States and one in Uganda) showed that supplementing patient referral with patient delivered partner therapy slightly reduced persistent or recurrent infection with gonorrhoea and chlamydia in index patients (risk ratio 0.73; 95% confidence interval 0.57 to 0.93) and increased the proportion of partners treated. Two Danish studies showed that home sampling increased the proportion of partners' specimens being tested.

Patient delivered partner therapy and home sampling are attractive strategies to increase partners' access to treatment or testing, because they are quick and simple for clinicians to implement. Increasingly, patient delivered partner therapy is being used in developed⁸ and developing countries.⁹ However, the current review shows that patient delivered partner therapy forms only one part of an effective patient referral strategy. The beneficial effects were modest, and they were susceptible to selection bias and measurement bias (in 23-70% of index patients a measurement of the primary outcome could not be obtained). The review also shows that patient delivered partner therapy can be substituted by patient delivered partner information (a booklet of tear out cards with treatment guidelines) with equal effect.

A home sampling strategy holds some promise in developed countries, but it needs further research because Trelle and colleagues' review could not determine whether increases in specimen testing translated into increases in the treatment of infected partners. In most developing countries, diagnostic testing of sexually transmitted infections is neither affordable nor feasible, and a syndrome based approach to their diagnosis and treatment has been adopted.⁷ This avoids the need for diagnostic testing for most curable sexually transmitted infections, and renders a home sampling approach of little value.

Neither of these two novel interventions tackles the fundamental barrier to patient referral strategies: the difficulty people have telling their partners that they have a sexually transmitted infection. In contrast, counselling and educational interventions can be tailored to deal with the barriers patients experience in relation to disclosure, and they can begin to tackle the gender inequities that influence whether and how partners communicate about sexually transmitted infections.

The review by Trelle and colleagues included two African randomised controlled trials evaluating one to one counselling and education for index patients; it found that more partners were notified or treated than with simple patient referral. Unfortunately, the trials did not measure infection rates in index patients. Novel strategies that aim to increase partner access to treatment might produce bigger effects if used in combination with counselling and education interventions for index patients. One of the two African trials used lay counsellors.

Current evidence leaves important questions unanswered. In developing countries where the syndromic approach is used, diagnostic specificity is lacking, especially in women with vaginal discharge. This leads to the unnecessary notification of partners and potential harms, including violence against women,⁷ about which little is known. Trelle and colleagues found no trials that investigated improving patient referral for HIV. Observational research in people with HIV suggests that continuous rather than one off counselling services are best for tackling the difficulties index patients have in disclosing to their partners.¹⁰

While patient delivered partner therapy and home sampling alone improve patient referral to some extent, strategies that promote and assist disclosure to partners are urgently needed as part of a comprehensive approach to patient referral.