Using intervention mapping for the development of a targeted secure web-based outreach strategy named SafeFriend, for Chlamydia trachomatis testing in young people at risk

The Intervention Mapping process starts with a need assessment to analyse a specific health problem within a target population and associated important stakeholders. Furthermore, behavioural and environmental risk factors that contribute to the health problem are identified and the underlying determinants are explored. At the end the programme outcomes are stated. To plan and guide the need assessment the PRECEDE-model is used [26].

The Dutch national STI policy targets were explored in order to identify expectations from practical and political stakeholders. A literature search was conducted reporting on Ct prevalence in different population groups, including young people, and existing strategies for Ct control, why these strategies failed or succeeded, and which methods were cost-effective to reach high risk young people. Furthermore, semi-structured interviews were held among Ct positive young people (8 women and 13 men) who were between 17 and 24 years old and visited the sexual health care centre South Limburg for Ct treatment. They were from various ages, sexual orientation, educational levels and ethnicity. They were interviewed about partner notification (PN) to find out the determinants to motivate peers to test for Ct. Ethical approval of the Regional Medical Ethics committee in the Netherlands was not necessary, because participants in these interviews were not “subjected to procedures or required to follow certain rules of behaviour” [27]. Based on the needs assessment using the PRECEDE-model (Figure 1), the health problem, behavioural and environmental factors, the underlying determinants and the target group were defined, resulting in two desired behavioural outcomes that were selected to be targeted by the intervention.

In step 2, each of the behavioural outcomes is broken down into performance objectives, i.e. the actions that expectedly lead to the desired outcome. For each performance objective, the most important and changeable determinants are stated based on theoretical models for behavioural change. Combining the performance objectives with the determinants identified results in matrices that include the change objectives that guide the intervention development process.

The performance objectives and most important and changeable determinants were selected based on previous studies, the results of the semi-structured interviews and the Precaution Adoption Process Model (PAPM) [28]. The model comprises of a stage theory that ranges from unaware of a problem to action and maintenance of a specific behaviour (e.g., Ct testing and motivating network members to test). This model was chosen, because it emphasizes that hazard responses are influenced by behaviours and opinions of peers, which is in line with the underlying of social and sexual networks surrounding Ct positives. A person’s decision concerning the need to test is influenced by the opinion and actions of peers and not just by specific beliefs about the hazard that lead to rational and independent decisions. This social influence may occur at several stages of the model.

In step 3, theory-informed intervention methods are selected to create changes in the determinants of the behavioural outcomes. The relevance of methods for a particular project and the changeability that can be realized with a particular method guides this selection process. Practical applications that fit the target population and context are then translated from these theory-informed intervention methods.

Before a final list of methods was determined on the basis of existing evidence and theory, theoretical methods were explored and analysed by a small panel of experts consisting of researchers, prevention experts, informatics-specialists, nurses and medical doctors. The experts then transformed the methods into practical applications taking into account the parameters of use, the literature review and conducted semi-structured interviews.

In this step, the practical applications are combined and translated into programme components and specific materials. During this process several end users are included to select the best program design that fits the user’s preferences. The intervention is then pilot-tested among some members of the target population. Eventually, necessary adjustments are made and the definitive intervention developed.

The intervention was constructed, tested and refined with close involvement of high risk young people and sexual health care nurses. The selected applications were incorporated into two websites: (1) a smart and highly interactive website for young people to communicate and motivate those at high risk in their social and sexual networks and (2) another website for sexual health care professionals to start recruitment-chains and oversee the process of home-test requests. The final name and look-and-feel of the intervention were set after pre-testing it among young people. The first pre-test of the intervention will be conducted among young people (age ≤ 25 years) visiting the public sexual health care centre South Limburg for Ct treatment. Evaluation will then be done to identify facilitating factors and barriers to use and the application is refined further (specific guidance for graphic designers and programmers).

The last two steps of the IM protocol, step 5 and 6, are outside the scope of this paper and are described briefly here.

An implementation plan is made to deploy the intervention to a successful outcome. End users and implementers of the intervention are identified and implementation objectives are stated considering dissemination, adoption, implementation and maintenance of the intervention.

Health care professionals (e.g., sexual health care nurses, medical doctors) and policy makers were involved from the start of the project by regular team meetings. They are the key linkage group for the success of the development and implementation and have therefore been involved in the entire research process. The ultimate goal is national implementation in sexual health care centres and the future expansion to GPs will be explored.

In the final step of intervention mapping, an evaluation plan and associated criteria are set up. The level of evaluation is chosen and indicators and measures determine to evaluate the effects and process of the intervention. All program objectives are defined in a measurable form.

To assess whether the intervention leads to changes in determinants and thereby in a change in behaviour of young people (i.e., Ct testing and motivate high risk young people to test for Ct) a plan for the quantitative and qualitative evaluation was made. Factors that may hinder or facilitate the use of the intervention will be reviewed in the course of this project and strategies will be explored to positively affect intervention use.

Several behavioural barriers for testing among young people have been identified from the literature search (i.e., Table 1). For example the level of knowledge of STI; Ct being an asymptomatic disease, the connection between Ct and fertility complications and current testing options [35‐37]. Another barrier includes misperceived vulnerability; young people do not think they are at risk for Ct [35, 36, 38]. Embarrassment about the need for testing and fear of a positive test result also plays a role [36, 38, 39]. Shame and fear may be enforced when the provider is a family doctor and young people are charged for testing (note that Dutch sexual health care centres test free of charge) [37‐39]. Home-based test were found to be acceptable among young people and could be used as a facilitator factor in testing [9, 39‐41].

Barriers related to motivate peers to get tested for Ct include fear of stigmatization among peers, repercussion (e.g. aggression), blaming the other, availability of contact details and the level of emotional commitment [38, 42‐44].

The analyses of the 21 semi-structured interviews showed several barriers and facilitators related to motivating peers to get tested for Ct. Ct positive young people can feel embarrassed about their positive Ct test result when motivating peers to get tested. Some indicated that in case of a casual sex partner they would like to notify anonymously. Fear of aggression of sex partners can also discourage Ct positive young people to motivate peers. Interviewees expressed emotional commitment to their sex partners and felt responsible for informing them about the Ct risk. They were aware of the possible health risks for direct and indirect sex partners, whereas others were afraid of re-infection. As stated by one interviewee: “Partner notification is self-evident, when you do not warn your steady partner, you will get re-infected”. Social influence plays also an important role. Several interviewees were notified about their Ct risk via their sex partners and then visited a sexual health care centre. They also indicated that friends visited a sexual health care centre after talking to them about Ct testing and their own positive test result. As noted by one interviewee: “We [friends] talk about everything. We do not have any problems to talk about Ct positive test results”.

For each of the 2 behavioural outcomes 4 performance objectives were defined (i.e., Tables 2 and 3). Firstly, young people should be able to appraise their own and others Ct risk behaviour. Secondly, if necessary, young people should decide to get tested and motivate peers who are at risk for Ct to get tested. Thirdly, planning for getting tested or motivating peers is another objective which will lead to the fourth objective of actually getting tested and motivating other young people at risk for Ct. The following determinants were identified as most important and changeable: knowledge, self-efficacy, risk perception, social norms and attitudes. Finally, a matrix of change objectives was developed for each behavioural outcome.

The treatment and follow up for re-testing is already included in the current sexual health care practice and is therefore not described within the matrices.

As a result from the initial needs assessment and the matrices of change in step 2, a web-based RDS was used to select peers to motivate young people and encourage them to attend Ct testing, thereby motivating high-risk young people through their own social and sexual network. Web-based RDS was judged highly promising and feasible in high-risk young people, because it has been successfully used before in a group of young people [18]. The concept of using peer influence and chain-referral sampling has already worked well to target “those in need” in complex networks but never used in Ct control in young people [19, 45, 46]. In our strategy sampling begins with Ct positive young people attending the South Limburg sexual health care centres and proceeds in waves recruiting more targeted contacts from hitherto-hidden sexual and social networks. Participants select friends based on their knowledge of the extent to which these friends need or use appropriate care. The person who begins the chain-referral sampling within a given high risk network is already tested for Ct, which increases perceived social norms. To improve the chance of the targeted contact to decide for testing, messages and home-based test kits will be used. Other supporting methods geared to improve testing and motivating include tailoring, personalize risk, modelling, consciousness raising, elaboration, mobilizing social support and self re-evaluation (i.e., Tables 4 and 5).

As a result of the selected methods and practical applications the deliverable will be an outreach Ct screening strategy based on a secure web-based RDS engine, with an online interface specially written for young people to recruit and motivate peers. This website is named http://​www.​SafeFriend.​nl. The use of the web was judged promising during this IM process, because many young individuals have an email address and/or mobile phone to reach them. Using messages via email and/or mobile phone young people can motivate their sex partners and/or friends. The message will be brief and contains carefully worded text to alert the receiver of possible Ct risk, provide options for Ct testing and modelling (i.e., peers who test). To increase the likelihood of use the exact content of this message can be partly adapted (i.e., name or anonymous, level of language) by the user who sends it. After login and system user verification the person is offered tailored information on sexual health and likely Ct risk. They will be prompted with a brief questionnaire and the answers are automatically assessed to decide to offer him or her a home-based Ct test. Home-based testing has proven feasible and effective in other screening programs, however, such internet-delivered testing is yet to be included in regular care in public sexual health care centres or other care settings. The new recruit is then encouraged to motivate at that moment or at a later stage others who will then go through the same procedure. It should be noted that within the website in some cases a young individual will be asked to visit the sexual health care centre as they are likely to require more sexual health care than Ct testing alone. Necessary facilities like computers and home-based Ct test kits will be in place at the moment the intervention is implemented in practice. Overall, the use of the SafeFriend website by young people is self-initiated and not guided by the care professionals. The exception is the first recruitment step where the Ct positive young people at the sexual health care centre are motivated and supported by the nurses to use the intervention. For this, a second designed interface is accessed by the nurse during the consultation visit (the moment that in routine care PN is discussed). This interface enables the nurse to start the recruitment-chain and to oversee the process of further home-test requests. The young Ct positive client can motivate others during the consultation (like in provider referral) and/or at a later moment at home (like in patient-referral). Of course the clients may choose not to motivate others, without any further consequences for their own care. As the actions on motivating others are standardly registered (coded) the intervention yields systematic data on PN, data which are currently lacking.