Hepatitis C: The case against routine partner notification

Role of partner notification

Partner notification has been defined as ‘… the process of contacting the sexual partners (or drug injecting partners) of an individual (source or index client or patient) who has an STI including HIV, and advising the partners that they have been exposed to infection. By this means, people who are at high risk of STI/HIV, many of whom are unaware that they have been exposed, are contacted and encouraged to attend for counselling, testing and other prevention and treatment services. Where possible, confidentiality is maintained by not providing the name of the source client to partners ¹¹ (p. 32).

In England, partner notification is usually carried out by trained health advisers attached to GU medicine clinics. ¹² Partner notification can be performed in one of three ways (see Table 2).

OPEN IN VIEWER

Table 1

Scenario and rate of transmission

Scenario	Rate of transmission (%)
Regular sexual partner	0–2
Sexual partner of HIV-positive gay man	7
Mother to baby	5
Persons shared injecting equipment	47

OPEN IN VIEWER

Table 2

Methods of partner notification

Patient referral	Patients may inform partners themselves
Provider referral	Patients supply details for the health worker to notify the partner without naming the patient and infection
Contract referral	A combined approach whereby a time frame is agreed for patients to inform partners before the health worker notifies those who have not presented for care

In this context partner notification is a ‘formal’ process where the details of partners are recorded and their test results verified where possible. This is distinct from ‘informal’, where details of partners are not recorded and their test results not followed up. In this review article we focus on ‘formal’ partner notification.

Within GU medicine clinics patients diagnosed with hepatitis C should have a partner notification interview as recommended by the Manual for Sexual Health Advisers, but this is written in the context of it being an STI. ¹² Thus from the guidance it is unclear whether drug-using contacts will be sought; this ambiguity manifests in practice with some health advisers in GU medicine clinics tracing injecting acquaintances and others not. In a subsequent chapter of the manual on HIV partner notification it is recommended that drug-using contacts are traced. There is no recommendation for hepatitis C.

For persons screened and diagnosed with hepatitis C outside GU medicine clinics partner notification will not be undertaken. In hepatitis C clinics in England infected people are asked to notify their sexual partner but the outcome is not verified. The children of mothers with hepatitis C will be referred to hospital for testing for the virus.

Effectiveness of partner notification

There is little evidence available on the effectiveness of partner notification for hepatitis C. However, an evaluation of partner notification using patient referral (with monetary incentives) in injection drug users (IDUs) mainly with hepatitis C in Seattle (USA) found: of 447 contacts (from 23 index cases), 160 (36%) were sought and 17 (10%) contacts came forward; of these, 10 had hepatitis C. ¹³ However, of these 10 contacts not all were linked to an index case, and would be picked up through alternative, more cost-effective means. This suggests that less than 1% of contacts would be uniquely identified using this contact tracing route. This 1% might well contain people for whom, ‘Partner notification is also likely to be ineffective if health services are either inaccessible or unacceptable to clients’ ¹⁴ (p. 249). It is a common finding that many people with hepatitis C do not wish to engage with health services and thus remain untreated.^15–18

As such, the Seattle study suggests contact tracing is unlikely to be cost-effective. How does this compare with contact tracing for STIs, including HIV?

By contrast in the setting of HIV, four observational studies on partner notification included IDUs. ¹⁹ In summary, IDUs elicited higher numbers of contacts compared with other patients, with more contacts being screened. There was no indication that the proportion of contacts that were tested was lower in IDUs compared with other patients.

A systematic review confirmed that partner notification is a means of newly detecting STIs in the partners of index patients. ²⁰ The review concluded that provider referral or contract referral, when compared with patient referral increases the rate of partners presenting for medical evaluation, and that patient referral enhanced by verbal health education together with patient-centred counselling by lay workers, when compared with standard care, results in small increases in the rate of partners treated.

There appears to be limited evidence of studies on partner notification measuring cost-effectiveness. However, in the National Chlamydia Screening Programme in England, a modelling study suggested that increasing the effectiveness of partner notification is likely to cost less than increasing screening among men. ²¹

The majority of hepatitis C infections are diagnosed outside of GU medicine clinics. ⁸ As discussions about partners should be initiated early after diagnosis, and it is beneficial if the patient does not have to be referred to another clinic such as GU medicine, this raises the concern that untrained staff would have to notify partners. The cost of training this group would be high and given the low probability of finding many of the undiagnosed population; it is unlikely to be cost-effective.

But even if proven to be cost-effective, should we undertake partner notification in hepatitis C? Would it be acceptable?

Acceptability of routine partner notification

Even if partner notification might reach some previously unidentified persons (particularly those not attending drug services), would there be unwanted consequences? From evidence above, in the setting of HIV, it appears IDUs may find partner notification acceptable and cooperate with the process. ¹⁹ However, the evidence appears mixed, with more recent qualitative studies exploring the attitudes of HIV-infected drug users towards partner notification finding that a greater proportion reported negative reactions to their partners being notified as compared with positive attitudes, ²² and at least 50% would refuse to participate in partner notification of sex partners if provider referral would be employed instead of patient referral. ²³ Further, drug users were even less willing to participate in partner notification for needle-sharing partners than sexual partners. The participants reported that partner notification for needle-sharing partners is not practical and may result in ‘gossip’ in the community. Thus, persons with hepatitis C may find discussion about notifying their contacts intrusive, and they may be reluctant to share information about those with whom they have injected. If found or proven to be unacceptable to drug users, introduction of partner notification may result in individuals at high risk of infection putting off being tested which would be detrimental to the wellbeing of the individual and to public health.

Would partner notification for hepatitis C be practical? It is unlikely that people who put themselves at risk many years ago (not currently injecting) will be found. There is no reason to believe that they would necessarily be linked to people who have a diagnosis and, therefore, to believe they would be reached by contact tracing. Those who are current drug users and may be contacts of those with hepatitis C may nonetheless be hard to find due to not having permanent accommodation, being homeless or sent to prison.^24,25 Further, groups of drug users might inject in places away from their accommodation and not know where each other lives, thus making it difficult to locate and trace them. Therefore, it could be argued that partner notification is unlikely to find high-risk and hard-to-reach persons. In addition, outreach workers employed to work with high-risk and hard-to-reach persons might be better able to engage the group of those with undiagnosed hepatitis C. People with a previous negative test may be alerted through partner notification to recent risk behaviour with someone known to have the virus. However, the introduction of annual (or sooner where a risk is identified) screening within drug services of IDUs with a previous negative test would obviate the need for partner notification. This approach of repeat screening has been agreed locally within drug treatment services (a minimum of 2-yearly) (Sheffield Drug and Alcohol Action Team. Minutes of Blood Borne Virus Group Meeting 24 November 2011. Sheffield: Sheffield Health and Social Care NHS Foundation Trust, 2011, unpublished).

Another difficulty relates to the time-span of those defined as contacts. Some persons may have been infected with hepatitis C for many years and have not injected recently. The median time from first injection to diagnosis is 11 1/2 years (Sheffield Hepatitis C Database. Number of years from first injecting risk to diagnosis. September 2011, unpublished). This leads to a concern about the appropriate cut-off time. Is it reasonable and practicable to ask about and attempt to notify contacts from the distant past? One could argue that uncertainties about the look back period should not influence a decision on whether partner notification should be offered to all contacts. However, it will impact on time, and on the cost and efficacy of this activity (discussed previously). Requiring details of partners’ drug use and other activity over several years raises concerns about the consent of the original infected person – knowing that your life is to be investigated in such depth could put off those who would otherwise seek a test.

Way forward

This discussion of the feasibility of partner notification in hepatitis C should not detract from the current priority of getting current and former IDUs tested. This is in itself a challenge not least because many individuals have poor venous access and some health workers may lack the knowledge to discuss hepatitis C and gain informed consent of the individual.

Some areas for future research to better define the role of partner notification in hepatitis C are: •

the acceptability for IDUs of tracing drug using acquaintances;

In the meantime, without empirical evidence I would suggest the following conditions might be favourable for undertaking partner notification in hepatitis C: •

to focus on those persons at high risk, i.e. IDUs;