a) Interactions with field workers
Training and support supervision for KGBC field workers was planned carefully from the outset, given their key role in informed consent. These activities were jointly conducted by the KGBC research team and members of the Community Liaison group. In addition to training on the study itself, field workers attended a one week participatory workshop addressing communication skills and participants' rights in research, building on studies conducted previously in this community on local perceptions of research [28
]. During training it became clear that, despite having worked for KWTP for some time, several field workers had a low understanding of the nature of health research, often conflating treatment and research activities at the centre, and demonstrating the same therapeutic 'misconceptions' of research that were commonly described by community members. Of particular importance, the wider context of national and international research review processes and the existence of national and internationally agreed research ethics principles were generally unknown. Instead, field workers discussed their work of recruitment in the context of a very challenging aspect of a job for which they were employed by KWTP. The scientific basis for KGBC was also difficult to comprehend, exacerbated by the conceptual gap between the research procedures experienced (a single blood sampling) and the putative social value (future malaria drugs and vaccines). Strategies were developed to address these issues during the training workshop, including participatory methods such as small group discussions and role plays. Following the initial training, one of the authors (AM) met with the fieldworkers on a weekly basis to discuss their experiences of recruitment, assess the community's response to the study and facilitate problem solving.
From the weekly meetings, a particularly important issue that emerged over time concerned the high degree of stress experienced by field workers in relation to recruitment, given the difficulties in communicating about the study. They had participated in a lengthy participatory training with experienced facilitators to acquire understanding of research and KGBC themselves, but now felt limited in their own skills and the time they should take in explaining these to potential participants. In part this occurred as a response to recognising community needs, that is, for busy mothers not to spend time on activities unnecessary to their families' wellbeing. In part field workers' sense of a pressure of time arose from their perceptions that rapid recruitment of large numbers of participants would reflect positively on their performance. In many instances, they met worried or even hostile responses from community members who were, as will be described later, suspicious of the unfamiliar heel prick sampling and the involvement of children who were healthy. Frustration was experienced by some field workers in relation to participants seen as 'difficult' if unable to understand information or reluctant to engage on the topic at all. Field workers found the information sheet used as part of the informed consent process limited in supporting communication about the research, due to its length and their perceptions of the messages being not particularly important.
Given these difficulties, field workers tended to use their own explanations in place of the information sheet. In doing this they often put more emphasis on SCD screening than genomics research, an approach they believed would help participants to understand the 'value of the research'. These issues were discussed during weekly feedback meetings, including the importance of the informed consent process and principles of voluntarism and informed choice. Counteracting beliefs that higher recruitment rates reflected more effective field worker performance appeared important in shifting attitudes and reducing a perceived pressure of time on home visits. Discussing recruitment challenges and positive experiences supported peer learning and underlined the inherent, as opposed to personal, nature of these difficulties. One such difficulty was balancing every person's right to refuse to participate with checking that refusals were not being made on the basis of simple misunderstandings that could be easily addressed.
A follow up training workshop was planned four months into recruitment to address these issues more comprehensively. This was an important opportunity to re-develop the study information and consent form, based on the experiences of the first months of communicating about this study. Using group work, the field workers developed a revised version in Kiswahili, drawing on their own communication skills, local knowledge and a structured tool to guide content. Their inclusion in this process reportedly increased their understanding of the study, and is likely to have strengthened their skills to explain it and increased their motivation to provide information in this agreed, standardised way; all of which potentially contribute to strengthening the informed consent process. The revised information sheet and consent form were subsequently submitted to and approved by the institutional and national ethics review committees.
b) Consultations with chiefs
Before implementation, and following local and national science and ethics approval, members of the Community Liaison group explained and discussed the research with study area chiefs. This entailed providing information on the study, assessing chiefs' perceptions and attitudes towards it and asking for their recommendations, during a one day workshop conducted as part of a wider community engagement agenda. The findings are summarised below.
• Chiefs expressed general support for the research as a novel approach to vaccine and drug development in the future.
• Their main reservation was the acceptability of the heel prick; they expressed surprise at this method of taking blood and considered that people might not accept the procedure.
• Many chiefs expressed ideas indicating a good understanding of the heritability of physical disorders and illnesses, volunteering albinism as an example. These were described as being passed down through family lines, or clans (Kiswahili "ukoo"), through substances in the blood, blood cells or (for a few participants) in spermatozoa.
• Messages that centered on the concept of inherited resistance to disease, rather than on explanations about the way that transmission occurs through genes (or anything else), were thought to be most effective in communicating about the study. A range of approaches using farming analogies were presented and discussed, given the prominence of subsistence farming as a livelihood in this community. Chiefs identified a well known genetic predisposition and resistance to illness of certain breeds of goats and maize, reflected in their commercial value, as supportive of explanations for KGBC.
• SCD was not widely recognized either as an illness or a syndrome apart from by a few individuals with direct experience. The chiefs supported the inclusion of SCD screening as part of the research and suggested that carriers should also be identified and informed. Several expressed a view that SCD screening should be available to the public and encouraged for couples prior to marriage. Facilitators raised a number of ethical issues for genetic screening programmes and disclosure of carrier status, including the inability of children to give consent, the potential for stigmatization of carriers and misaligned paternity. The general consensus therefore supported the planned process to disclose SCD results to families affected by the disorder, and not to carriers.
• The export of samples to the UK for genotyping as part of this study and the archiving of blood samples for potential future research (contingent on local and national review and approval) were explained. No questions or discussion were generated by chiefs around these aspects of the research during this meeting or in subsequent informal conversations.
c) Community meetings
Within the first six months of the start of this study, a series of 40 public meetings were organized by chiefs for community sensitization. Around 8000 people attended in total, with between 50 and 300 people per meeting. Messages on health research and KGBC were presented, alongside information on two other studies, and community facilitators addressed any questions raised. The same topics were later addressed in a series of 70 small group meetings with village elders, religious leaders and selected individuals seen as key opinion leaders. Reports of some religious leaders' negative attitudes to KEMRI's work led to their specific inclusion in this engagement activity. The main findings of these early community meetings are summarised below.
• Although attitudes were largely positive towards KEMRI, many questions reflected therapeutic 'misconceptions' of health research and consequent concerns about procedures, as has been reported for Kilifi elsewhere [11
• The most frequently asked questions and strongest reaction, of surprise and concern, were about the heel prick procedure, as predicted by the chiefs. The heel prick was described as unfamiliar and therefore as raising fear in the community. One village elder wondered if his grandchild's death some time after a heel prick was associated with the procedure. Others raised concerns that it would 'make children unable to walk'. In one small group discussion, some mothers were said to 'hide their children when they see KEMRI field workers approaching their homes'. Importantly, the lack of familiarity with heel pricking as a procedure was also encountered amongst health providers, particularly nurses in peripheral health facilities, whose negative attitudes to the procedure at the outset of the study may have strengthened community concerns. Facilitators' explanations centred on safety and minimising discomfort, including the ways in which participants' rights and safety are protected in research.
• Clarification was often sought on the reasons that healthy rather than ill children would be recruited into the study. Facilitators' responses drew on descriptions of the differences between health research and medical treatment to explain that KGBC would involve healthy children because it did not have a primary individual therapeutic goal.
• There were many questions asked about the nature and signs of SCD, requiring facilitators to describe and explain this disorder. In small group meetings where more interaction was feasible, a component of the research that generated strongly positive comments was the return to each home of individual participants' results from the SCD screening test.
• Occasional questions were asked about samples being sent overseas. Responses to these based on the lack of availability of appropriate technology in Kenya were readily accepted.
• Although the archiving of samples for potential future research was described and protections outlined, no questions were raised about this procedure.
Many people arrived and left over the course of the large public meetings, and while some chose to ask questions, the understanding and reactions of the majority could not be assessed. Small group meetings were highly interactive and lively, generating many questions and sometimes extending discussions far beyond the areas anticipated. Overall, facilitators noted a positive change in participants' attitudes over the course of small group discussions. For example, a group that heatedly raised negative issues about KEMRI at the outset went on to include the welfare of KEMRI staff in closing prayers at the end of the meeting.