Appendix A: recommendations for policy and practice and supporting evidence statements

This appendix sets out the recommendations and the associated evidence statements taken from three reviews of effectiveness (see appendix D for the key to study types and quality assessments). It also sets out a brief summary of findings from the economic appraisal and the fieldwork.

The three reviews of effectiveness are:

  • Review 1: 'Contraceptive advice and provision for the prevention of under 18 conceptions and STIs: a rapid review'.

  • Review 2: 'Rapid review of the evidence for the effectiveness of screening for genital chlamydia infection in sexually active young women and men'.

  • Review 3: 'Rapid review of the evidence for the effectiveness of partner notification for sexually transmitted infections including HIV'.

Recommendations are followed by the evidence statement(s) that underpin them. For example: (evidence statement 1.1) indicates that the linked statement is numbered 1 in the review 'Contraceptive advice and provision for the prevention of under 18 conceptions and STIs: a rapid review'; (evidence statement 2.1) indicates that it is numbered 1 in the 'Rapid review of the evidence for the effectiveness of screening for genital chlamydia infection in sexually active young women and men'

The reviews are available on the NICE website. Where a recommendation is not directly taken from the evidence statements, but is inferred from the evidence, this is indicated by IDE (inference derived from the evidence).

Recommendation 1

Who is the target population?

Key groups at risk of STIs including:

  • men who have sex with men

  • people who have come from or who have visited areas of high HIV prevalence.

Behaviours that increase the risk of STIs include:

  • misuse of alcohol and/or substances

  • early onset of sexual activity

  • unprotected sex and frequent change of and/or multiple sexual partners.

Who should take action?

Health professionals working in:

  • general practice

  • genito-urinary medicine (GUM)

  • community health services (including community contraceptive services)

  • voluntary and community organisations

  • school clinics.

What action should they take?

  • Identify individuals at high risk of STIs using their sexual history. Opportunities for risk assessment may arise during consultations on contraception, pregnancy or abortion, and when carrying out a cervical smear test, offering an STI test or providing travel immunisation. Risk assessment could also be carried out during routine care or when a new patient registers.

  • Have one to one structured discussions with individuals at high risk of STIs (if trained in sexual health), or arrange for these discussions to take place with a trained practitioner.

(Evidence statement 1.1, 1.2, 1.3, 1.4, 2.20, 2.21, 2.26, 2.29, IDE)

Recommendation 2

Who is the target population?

Key groups at risk of STIs including:

  • men who have sex with men

  • people who have come from or who have visited areas of high HIV prevalence.

Behaviours that increase the risk of STIs include:

  • misuse of alcohol and/or substances

  • early onset of sexual activity

  • unprotected sex and frequent change of and/or multiple sexual partners.

Who should take action?

Health professionals trained in sexual health who work in:

  • general practice

  • GUM

  • community health services (including community contraceptive services)

  • voluntary and community organisations

  • school clinics.

What action should they take?

  • Have one to one structured discussions with individuals at high risk of STIs. The discussions should be structured on the basis of behaviour change theories. They should address factors that can help reduce risk-taking and improve self-efficacy and motivation. Ideally, each session should last at least 15–20 minutes. The number of sessions will depend on individual need.

For details of a range of behaviour change theories see 'Predicting health behaviour' (Conner and Norman 2005).

(Evidence statement 1.1, 1.2, 1.3, 1.4, IDE)

Recommendation 3

Who is the target population?

Patients with an STI

Who should take action?

  • Health professionals working in general practice, GUM and community health services (including community contraceptive services), voluntary and community organisations and school clinics. (However, they may need to refer the patient to a specialist.)

  • Specialists with responsibility for helping to contact, test and treat partners of patients with an STI (partner notification). They may be sexual health advisers, general practitioners (GPs) or practice nurses providing enhanced sexual health services, chlamydia screening coordinators or GUM clinicians.

What action should they take?

  • Help patients with an STI to get their partners tested and treated (partner notification), when necessary. This support should be tailored to meet the patient's individual needs.

  • If necessary, refer patients to a specialist with responsibility for partner notification. (Partner notification may be undertaken by the health professional or by the patient.)

  • Provide the patient and their partners with infection-specific information, including advice about possible re-infection. For chlamydia infection, also consider providing a home sampling kit. .

(Evidence statement 3.1, 3.2, 3.8, 3.16, IDE)

Recommendation 4

Who is the target population?

Population served by a PCT

Who should take action?

PCT commissioners

What action should they take?

  • Ensure that sexual health services, including contraceptive and abortion services, are in place to meet local needs. All services should include arrangements for the notification, testing, treatment and follow-up of partners of people who have an STI (partner notification).

  • Define the role and responsibility of each service in relation to partner notification (including referral pathways).

  • Ensure staff are trained.

  • Ensure there is an audit and monitoring framework in place.

(Evidence statement 3.1, 3.2, 3.8, 3.16, IDE)

Recommendation 5

Who is the target population?

Vulnerable young people aged under 18. This may include young people:

  • from disadvantaged backgrounds

  • who are in – or leaving – care

  • who have low educational attainment.

For a more detailed definition of vulnerable young people see Department for Education and Skills (2006) 'Teenage pregnancy: accelerating the strategy to 2010'.

Who should take action?

  • GPs, nurses and other clinicians working in healthcare settings such as primary care, community contraceptive services, antenatal and postnatal care, abortion and GUM services, drug/alcohol misuse and youth clinics, and pharmacies.

  • GPs, nurses and other clinicians working in non-healthcare settings such as schools and other education and outreach centres.

What action should they take?

  • Where appropriate, provide one to one sexual health advice on:

    • how to prevent and/or get tested for STIs and how to prevent unwanted pregnancies

    • all methods of reversible contraception, including long-acting reversible contraception (LARC) (in line with NICE clinical guideline 30)

    • how to get and use emergency contraception

    • other reproductive issues and concerns.

  • Provide supporting information on the above in an appropriate format.

(Evidence statement 1.3, 1.4, 1.18, 1.19, IDE)

Recommendation 6

Who is the target population?

Vulnerable young women aged under 18 who are pregnant or who are already mothers. This may include young women:

  • from disadvantaged backgrounds

  • who are in – or leaving – care

  • who have low educational attainment.

For a more detailed definition of vulnerable young people see Department for Education and Skills (2006) 'Teenage pregnancy: accelerating the strategy to 2010'.

Who should take action?

Midwives and health visitors who provide antenatal, postnatal and child development services

What action should they take?

  • Regularly visit vulnerable women aged under 18 who are pregnant or who are already mothers.

  • Discuss with them and their partner (where appropriate) how to prevent or get tested for STIs and how to prevent unwanted pregnancies. The discussion should cover:

    • all methods of reversible contraception, including LARC (in line with NICE clinical guideline 30), and how to get and use emergency contraception

    • health promotion advice, in line with NICE guidance on postnatal care (NICE clinical guideline 37)

    • opportunities for returning to education, training and employment in the future.

  • Provide supporting information in an appropriate format.

  • Where appropriate, refer the young woman to the relevant agencies, including services concerned with reintegration into education and work.

(Evidence statement 1.17, IDE)

Evidence statements

Evidence statement 1.1

In summary, the evidence on the effectiveness of one to one interventions for the prevention of STIs is mixed but, on balance, marginally supports the interventions. There is evidence from Project RESPECT a large (++) US study (Kamb 1998) that both a two session and a four session one to one counselling intervention can reduce STIs in the long and very long term in heterosexuals, and from one (+) study that STIs in men can be reduced in the long term after one 90 minute session (Kalichman). However, the effect appears to decrease over time, with one study finding a reduction in effect after 6 months (Kamb 1998).

RESPECT intervention model

This comprised brief or enhanced counselling sessions. The brief intervention consisted of two, 20 minute, client-focused interactive sessions with a counsellor. It involved negotiating an acceptable and achievable risk-reduction plan that focused on condom use. The enhanced counselling consisted of four interactive sessions with a counsellor, based on the theory of reasoned action. The sessions took place over a 2 week period. The first lasted 20 minutes, the remainder were 60 minutes long. They involved negotiating a long-term plan for behaviour change. The aim was to ensure condoms were consistently used. Both types of counselling helped change the attitudes and self-efficacy (determining intention) of women who attended. Only the more intensive counselling was effective for men. The models of behaviour change underpinning RESPECT were the social cognitive theory and the theory of reasoned action.

Evidence statement 1.2

In addition EXPLORE, a large (++) US study of ten sessions of one to one counselling for MSM [men who have sex with men], found a 15.7% reduction in HIV infection but this was not statistically significant (EXPLORE 2004). The other studies found no effect on STIs, but may have been underpowered for this outcome.

EXPLORE intervention model

The intervention consisted of 10 core counselling modules delivered at one to one counselling sessions, over a 4–6 month period. Typically, one module was delivered per session. After the initial 10 modules, maintenance sessions were delivered every 3 months. The intervention was designed to address the individual, interpersonal and other factors associated with risk taking by some men who have sex with men. These factors include: the greater pleasure derived from risky sexual behaviour; negative mood states; communication difficulties; social norms that encourage misperceptions of risk and risk taking; use of alcohol or recreational drugs; and life events and environments that are catalysts for risk taking. The intervention was carried out by counsellors who had completed the required 40 hours of training specified by the intervention protocol.

Evidence statement 1.3

Interventions with adolescents appeared to be particularly effective. A subgroup analysis of Project RESPECT (Bolu 2004) found a significant reduction in sexually transmitted infections with both the four and two session interventions versus a didactic control. Although this was the only study to show a statistically significant difference, the general trend in this group of studies was towards a reduction in STIs.

Evidence statement 1.4

Twenty five studies reported condom use, of which only eight showed a statistically significant increase in condom use in the intervention group compared to the control. However, overall there is weak evidence (that is, it is mixed or conflicting but on balance marginally supports) that one to one STI/HIV prevention interventions can increase short and long-term condom use compared to control. Project RESPECT, a large good quality (++) US study found an increase in condom use in both the four and two session counselling intervention groups compared to a didactic control (Kamb 1998). However, several studies found the effect of an intervention appears to decrease or disappear over time. Greater uniformity is needed in the way in which condom use is measured in studies.

(See details of the RESPECT intervention model above)

Evidence Statement 1.17

Six studies evaluated interventions to support pregnant women or mothers. Although only two of the studies focused solely on adolescents (O'Sullivan 1992, Quinlivan 2003) all included at least 40% of adolescents and focused on disadvantaged, low income women. There is good evidence that multi-session support and home visiting for disadvantaged low income pregnant women or mothers can prevent repeat pregnancies with two (+) (Olds 2002; Olds 2004) and one (-) (O'Sullivan 1992) studies showing a significant reduction in repeat pregnancies in the intervention group compared to control. In addition one (-) study (Olds 1997) found a reduction in repeat pregnancies in poor unmarried women, although not in the sample as a whole.

Evidence Statement 1.18

In relation to the prevention of pregnancy, two (-) studies evaluated contraception advice and support in a clinic-based setting (Shlay 2003; Winter 1991). Neither found a significant reduction in pregnancies but both showed a trend towards a reduction in the intervention group compared to control.

Evidence Statement 1.19

Seven studies reported contraception use. This was measured in various different ways, including oral contraception, emergency contraception (EC) and condom use. Four studies showed a statistically significant effect on contraception use. Two increased oral contraceptive use. These were a (++) RCT (Quinlivan 2003) and a (+) RCT (Danielson 1990) that found one to one interventions with teenagers can improve contraception use in the long term. Of the two (++) studies of advanced provision of EC, one (Harper 2005) found an increase in the use of EC at 6 month follow-up and the other (Gold 2004) found a short term increase in EC use but this was no longer significant at 6 months. This study (Gold 2004) also reported an increase in condom use but no significant difference in use of the oral contraceptive pill (Gold 2004). In the other studies the general trend was towards an increase in contraception use although one (-) study found the effect on contraception use was no longer significant at 12 months (Winter 1991). Therefore, there is some evidence that one to one interventions with under 18s can increase contraception use.

Evidence Statement 2.20

There is evidence from two (+) controlled trials (one randomised, one non-randomised) that offering chlamydia testing in general practice increases the number of young women and men screened compared with usual care. This evidence applies to women and men under 30 years attending general practices.

Evidence Statement 2.21

There is evidence from two (+) randomised controlled trials (one large, one small) suggesting that changing systems of health service delivery can increase the numbers of teenage women screened opportunistically, and the number of chlamydia cases detected. This evidence applies to sexually active young women under 20 years attending general paediatric or teen clinics.

Evidence Statement 2.26

Descriptive studies in general practice (two studies, one ++, one +) suggest that offering GPs incentives [to screen patients] might increase acceptance rates by patients. There were too few studies to be able to say anything about the effects of incentives on effective screening rates.

Evidence Statement 2.29

Data from one (+) randomised controlled trial, one (++) descriptive study, and three (+) descriptive studies (one + contradictory study) show that less than half of women and men under 25 years attending general practice get screened for chlamydia because not all those who are eligible for screening are offered a test.

Evidence Statement 3.1

There is evidence from four large randomised controlled trials (two +; two -) that patient delivered partner therapy, plus additional information for partners, reduces persistent or recurrent infections in women and men diagnosed with gonorrhoea or chlamydia by approximately 5% compared to patient referral (either minimal or supplemented by contact card).

Evidence Statement 3.2

There is evidence from one large randomised controlled trial (-) that patient referral, supplemented by additional information about infection for index patients and partner(s), reduces persistent or recurrent infections in men diagnosed with gonorrhoea or chlamydia by approximately 5% when compared to minimal patient referral.

Evidence Statement 3.8

There is weak evidence from two randomised controlled trials (both -) that giving index patients diagnosed with chlamydia sampling kits for their partner(s) can increase the number of partners who get tested, when compared to getting the partner(s) to visit their doctor for testing.

Evidence Statement 3.16

There is evidence from one randomised controlled trial (++) that patient referral for patients with chlamydia conducted in general practice is at least as effective, in terms of partners who get treated, when compared to referring patients to a specialist health service.

Cost-effectiveness evidence

Overall, one to one interventions were found to be cost effective. The results of the cost-effectiveness analysis are summarised below.

STI counselling interventions

Most of the brief STI counselling interventions appear cost effective when compared with 'usual treatment' (using £30,000 per QALY as the threshold).

The incremental analysis demonstrated that brief interventions involving information giving or developing motivation and behavioural skills (particularly among women) produce the greatest benefits for the least cost. More intensive behavioural skills counselling and enhanced counselling appear to be least cost effective. These analyses apply to the general population, including vulnerable young women.

In the absence of data, no costs were attributed to 'usual treatment'. As a result, when interventions are compared against usual treatment the cost difference may be overestimated and the incremental cost-effectiveness ratios may be artificially high.

The loss of quality of life (QALYs lost) is particularly important in the analysis. The cost per QALY may be high (if low values are assigned to the change in quality of life) but brief STI counselling falls below a £30,000 per QALY threshold (based on 0.1 of a QALY change).

Partner notification at GP clinics

Partner notification by a practice nurse in a general practice costs the same as in a GUM setting – but more patients can be treated in a GP setting.

Fieldwork Findings

Fieldwork aimed to test the relevance, usefulness and the feasibility of implementing the recommendations and the findings were considered by PHIAC in developing the final recommendations. The fieldwork was conducted with practitioners and commissioners involved in sexual health services. They included practitioners working across youth, community and clinical settings in the NHS, local authorities and the voluntary sector. The fieldwork report is online.

Fieldwork participants who work with young people were overwhelmingly positive about the recommendations and their potential to help reduce STIs and under 18 conceptions. Many participants stated that the recommendations were already part of current practice. (Those working in general practice were least likely to be involved in STI prevention.)

The recommendations were viewed as reinforcing aspects of the modernisation agenda for sexual health, particularly in relation to the:

  • identification of asymptomatic sexually transmitted infections

  • role of primary and community services in providing level 1 and 2 sexual health services, as defined by the 'National strategy for sexual health and HIV'

  • provision of choice of referral for STI testing and treatment

  • promotion of the use of long-acting reversible contraception methods and information about the availability and use of emergency contraception

  • increased integration of STI prevention and community contraceptive services.

While practitioners and commissioners did not view the recommendations as offering a new approach, these interventions have not been implemented universally. Wider and more systematic implementation would be achieved if there was/were:

  • a clearer definition of the nature of one to one interventions (offering details such as, 'what should be done, who should carry it out and where?')

  • information about the relative effectiveness of one to one interventions versus other options

  • information about how the recommendations might help meet the national GUM 48 hour access target

  • recognition of the need for open access to STI testing, and the general need for acceptable, accessible and confidential sexual health services

  • recognition of the need for local flexibility in service provision

  • incentives to encourage GPs to get involved in STI prevention

  • provision of and/or access to training opportunities to develop competencies for the delivery of sexual health interventions.