4.3 Method of study
4.3.2 Quantitative: Cluster randomised controlled trial
A cluster randomised controlled trial was used to collect quantitative data to evaluate participant outcomes in the current study. Cluster randomisation is the process in which larger communities are randomly assigned to intervention or control groups to measure the effects of an intervention. ‘Cluster randomisation is appropriate where an intervention is meaningful only at the level of the larger entity or when it is impractical to assign individuals randomly to different treatments’ (Michalopoulos, 2005, p. 23). It also controls for the contamination effect, where outcomes for people exposed to an intervention have the potential to affect outcomes for other people not exposed to it. Cluster randomisation is also effective when interventions are designed to affect the behaviour of groups of interrelated people rather than individuals. In addition, cluster randomisation maintains the integrity of a study because researchers can physically separate intervention and control groups (Bloom, 2005). For example, Reavley, McCann, Cvetkovski, and Jorm (2014b) conducted a cluster randomised controlled trial at a university with nine campuses, in Melbourne, Australia. The purpose of the study was to assess if a multifaceted intervention could improve students’ mental health literacy, facilitate their help-seeking, and reduce their psychological distress and alcohol misuse. The campuses were paired, with one of each pair assigned randomly to the intervention or control group cluster. The intervention was messages delivered to students about mental disorders and alcohol misuse and appropriate responses towards them. Messages were disseminated in different ways; Facebook pages of students’ groups, booklets and
fact sheets about mental disorders, emails to student email addresses, awareness programmes on mental health, posters on the back of campus toilet doors and mental health first aid (MHFA) training (www.mhfa.com.au) delivered by officers of the student counselling service. The intervention was conducted over two academic years and data were collected at three time-points (baseline, end of academic Year 1, end of academic Year 2). The findings showed that the intervention had some benefits in improving students’ overall mental health literacy (Reavley et al., 2014b).
In the present study, the districts were treated as clusters and randomised into intervention or control clusters. Conclusions were drawn based on how the training programme affected participants’ mental health literacy levels (Campbell & Walters, 2014). The researcher opted to use a cluster randomised controlled trial because it was unsuitable to randomly assign persons within each cluster (district) to intervention or control groups. This situation would have tainted the effectiveness of the intervention because participants who had been trained would be capable of influencing the results of participants in the control group because they would be discussing the intervention among themselves (Michalopoulos, 2005). Hence, it was suitable to randomise the programme by cluster rather than by individuals. Table 4.1 provides an outline of the cluster randomised controlled design of the mental health literacy programme. It summarises the population, unit and method of randomisation in which the clusters were paired, based on particular variables, treatment for cluster groups and data collection.
Table 4.1 Outline of cluster randomised controlled design of mental health literacy programme. Adapted from (Campbell & Walters, 2014).
Design item Description
Population Assembly members within the Brong Ahafo region in Ghana
Intervention cluster Mental health literacy programme (three hours) delivered to the intervention cluster by the researcher (PhD candidate)
Comparator (control cluster) Given a basic brochure about mental health Unit of randomisation District Assemblies within the region as clusters
Method of randomisation Simple randomisation performed by the Principal Supervisor of the research, assigning districts to intervention or control cluster groups. The researcher was blinded to this part of the exercise. Data collection Outcome instruments were administered at baseline (Week 0)
4.3.2.1 Treatment fidelity
Ensuring treatment fidelity of an intervention is a key procedural consideration in any trial examining interventions for behaviour change (Resnick et al., 2005), such as the present cluster randomised controlled trial. Century, Rudnick, and Freeman (2010) defined fidelity as the extent to which the significant parts of an intended intervention are present when that intervention is delivered.Treatment fidelity is the continual evaluation, monitoring and improvement of the reliability and internal validity of a study, and comprises two general components (Borrelli, 2011): (i) treatment integrity, the level to which a treatment is enacted as planned; and (ii) treatment differentiation, the extent to which the arms of two or more studies differ along important features.
Addressing treatment fidelity can enhance commitment to the delivery of an intervention as planned and also ensure that participants experience the complete treatment, thus strengthening statistical power (Resnick et al. (2005). Treatment fidelity can assist in explaining non-significant outcomes and provide a justification for further studies. It may prompt a researcher to review the implementation of an intervention and, if done, it assures confidence in the outcomes (Bellg et al., 2004). The process of ensuring treatment fidelity in a study necessitates attention to five areas. These are study design, aimed at ensuring that a study can sufficiently test its hypotheses in relation to the fundamental theory and clinical procedures; training providers, evaluating and improving the training of treatment providers so they are capable of delivering the intervention to research participants; delivery of treatment, monitoring whether delivery of the intervention is being done as intended; receipt of treatment, ensuring that the treatment has been received and understood by participants, and whether they can demonstrate similar cognitive strategies and behavioural skills in a given situation; and enactment of treatment skills, confirming that participants can perform treatment-related behavioural skills and cognitive strategies of the intervention in a real-life situation (Bellg et al., 2004; Resnick et al., 2005).
In this study, ensuring treatment fidelity in the intervention was guided by the framework provided by Resnick et al. (2005). The first area of focus, study design, was considered by designing the intervention as a cluster randomised controlled trial, supported by a
theoretical framework and well framed hypotheses35. The next area of focus for treatment fidelity was training providers. In this case, the student researcher was the provider of the training programme. He underwent training to ensure that he understood the concepts and was able to impart the knowledge. First, he was well educated in the subject matter of the two main mental disorders (depression and schizophrenia) of the training programme by his supervisors. In addition, he also undertook two formal training sessions: introduction to mental health issues and mental disorders and Youth Mental Health First Aid course. These preparations aided in making the trainer an effective facilitator during the programme, for example in presentations (relating concepts to real life situations), posing probing questions to encourage and direct discussions and answering questions from participants. Adhering to the delivery of treatment was also of concern to ensure treatment fidelity in this study. Attention was paid to the content of and length of time in presenting the intervention. The content of the intervention36 was prepared in a PowerPoint format and it was reviewed by the two supervisors of the project to confirm it met the objectives of the study. The length of time for delivery of the intervention was set at three hours. To ensure that this time was adhered to, a programme guide was prepared and distributed to participants indicating time allocated for each item on the programme (Appendix 4). In addition, a visible large wall clock was placed in the training centre to assist with time management. Furthermore, a participant also volunteered as time keeper to alert the trainer if deviations from the time limits were detected. These strategies were effective because they ensured that the training programme was delivered as intended. All items or concepts in the programme were presented in a reasonably appropriate time but with a few deviations in content and time. These deviations occurred particularly during the delivery of the training programme with the first intervention group. However, these deviations were duly noted and stricter time monitoring and management of unrelated issues during discussions were employed to correct such deviations in subsequent delivery to the other two groups. The next area for treatment fidelity is ensuring receipt of treatment as planned. This required an assessment that participants had received and understood the treatment, knew what to do and could apply this knowledge in real life contexts. Hence, the programme incorporated two 30-minute exercises that provided
35 See Chapter 4 section 1 and.2.
participants with the opportunity to practice the knowledge and skills acquired during the programme. In group format, participants were tasked with using the knowledge they acquired to address mental health problems and present their solutions to the whole group. Each group used the knowledge acquired to present their solutions. Common strategies used by groups to address the mental health problems were identified and emphasised to reinforce the knowledge and attitudes acquired. In addition, at the end of the training programme a survey was conducted with all participants based on the treatment implementation model of Lichstein et al. (1994) (See chapter 6 on results of this survey). Finally, to make an assessment of the enactment of the treatment skills, 12 weeks after the programme a sample of the participants were interviewed about the usefulness of the programme in their everyday lives. Participants stated they had used knowledge and attitude acquired from the programme in their lives (See results on qualitative process evaluation).