Optimising informed consent for participants in a randomised controlled trial in rural Uganda: a comparative prospective cohort mixed-methods study

摘要

Poor participant understanding of research information can be a problem in community interventional studies with rural African women, whose levels of illiteracy are high. This study aimed to improve the informed consent process for women living in rural eastern Uganda. We assessed the impact of alternative consent models on participants' understanding of clinical trial information and their contribution to the informed consent process in rural Uganda. The study applied a parallel mixed-methods design for a prospective comparative cohort, nested within a pilot study on the community distribution of an alcohol-based hand rub to prevent neonatal sepsis (BabyGel pilot trial). Women of at least 34?weeks' pregnancy, suitable for inclusion in the BabyGel pilot trial, were recruited into this study from their homes in 13 villages in Mbale District. As part of the informed consent process, information about the trial was presented using one of three consent methods: standard researcher-read information, a slide show using illustrated text on a flip chart or a video showing the patient information being read as if by a newsreader in either English or the local language. In addition, all women received the patient information sheet in their preferred language. Each information-giving method was used in recruitment for 1 week. Two days after recruitment, women's understanding of the clinical trial was evaluated using the modified Quality of Informed Consent (QuIC) tool. They were also shown the other two methods and their preference assessed using a 5-point Likert scale. Semi-structured interviews were administered to each participant. The interviews were audio-recorded, transcribed and translated verbatim, and thematically analysed. A total of 30 pregnant women in their homes participated in this study. Their recall of the trial information within the planned 48?h was assessed for the majority (90%, 27/30). For all three consent models, women demonstrated a high understanding of the study. There was no statistically significant difference between the slide-show message (mean 4.7; standard deviation, SD 0.47; range 4-5), video message (mean 4.9; SD 0.33; range 4-5) and standard method (mean 4.5; SD 0.53; range 4-5; all one-way ANOVA, p?=?0.190). The slide-show message resulted in the most objective understanding of question items with the highest average QuIC score of 100 points. For women who had been recruited using any of the three models, the slide show was the most popular method, with a mean score for all items of not less than 4.2 (mean 4.8; SD 0.6; range 4-5). Most women (63%, 19/30) preferred the slide-show message, compared with 17% (5/30) and 20% (6/30) for the standard and video messages, respectively. The reasons given included the benefits of having pictures to aid understanding and the logical progression of the information. Our results from this small study suggest that slide-show messages may be an effective and popular alternative way of presenting trial information to women in rural Uganda, many of whom have little or no literacy. ISRCTN, ISRCTN67852437 . Registered on 18 March 2018.