Use of findings from a pilot trial and qualitative studies to refine MiQuit, a text message self-help intervention for pregnant smokers
Felix Naughton (UKCTCS)
14.15, Tuesday 9 July. The Clubhouse, level 4
Background
Low access rates to traditional forms of smoking cessation support and high interest in self-help among pregnant smokers has resulted in a renewed interest in self-help interventions. This presentation will describe how findings from a pilot trial and two qualitative intervention development studies were used to refine an individually tailored and automated text message based smoking cessation support system for pregnant smokers (MiQuit).
Materials and methods
In a pilot trial (n=207), the following data were collected and used to refine a prototype intervention: participants’ responses to text questions, their perceptions of texts, whether they set a quit date and their use of “on-demand”, instant support texts. Relevant data from the qualitative studies included participants’ experiences of quitting during pregnancy and preferences regarding the content and delivery of texts.
Results
Five main refinements are described. Firstly, as there were satisfactory response rates to question texts, tailoring information in the revised intervention is now collected by automated text questions rather than by questionnaire. Secondly, in pilot data, there was a high correlation between perceptions of receiving too many texts and these being annoying (r=0.88, p<0.001), so a facility for users to increase or decrease the text frequency at any time was introduced. Thirdly, the finding that MiQuit increased women’s propensity to set quit dates supported the addition of a facility which encouraged users to text in a quit date to get an additional set of support texts tailored around that date. Fourthly, the finding that women perceived risk information texts as the most helpful type received but that such information sent unsolicited could lead to defensive thinking that justified continued smoking, prompted us to provide an additional set of risk information texts which users request on-demand only. Finally, a low use of on-demand support texts in the pilot yet a perceived need for practical activities to help manage cravings and replace the reward provided by smoking, identified by qualitative analysis, led us to develop a QUIZ facility. This allows users to request and answer multiple-choice quiz messages on-demand as a form of delay or distraction from smoking.
Conclusions
Using multiple data sources to inform intervention refinement can help accommodate user preferences, experiences and actual intervention use and avoid over-reliance on user preferences alone shaping refinement. The likely interest in and impact of the refined MiQuit system will be assessed in two future NIHR-funded evaluation studies.
Low access rates to traditional forms of smoking cessation support and high interest in self-help among pregnant smokers has resulted in a renewed interest in self-help interventions. This presentation will describe how findings from a pilot trial and two qualitative intervention development studies were used to refine an individually tailored and automated text message based smoking cessation support system for pregnant smokers (MiQuit).
Materials and methods
In a pilot trial (n=207), the following data were collected and used to refine a prototype intervention: participants’ responses to text questions, their perceptions of texts, whether they set a quit date and their use of “on-demand”, instant support texts. Relevant data from the qualitative studies included participants’ experiences of quitting during pregnancy and preferences regarding the content and delivery of texts.
Results
Five main refinements are described. Firstly, as there were satisfactory response rates to question texts, tailoring information in the revised intervention is now collected by automated text questions rather than by questionnaire. Secondly, in pilot data, there was a high correlation between perceptions of receiving too many texts and these being annoying (r=0.88, p<0.001), so a facility for users to increase or decrease the text frequency at any time was introduced. Thirdly, the finding that MiQuit increased women’s propensity to set quit dates supported the addition of a facility which encouraged users to text in a quit date to get an additional set of support texts tailored around that date. Fourthly, the finding that women perceived risk information texts as the most helpful type received but that such information sent unsolicited could lead to defensive thinking that justified continued smoking, prompted us to provide an additional set of risk information texts which users request on-demand only. Finally, a low use of on-demand support texts in the pilot yet a perceived need for practical activities to help manage cravings and replace the reward provided by smoking, identified by qualitative analysis, led us to develop a QUIZ facility. This allows users to request and answer multiple-choice quiz messages on-demand as a form of delay or distraction from smoking.
Conclusions
Using multiple data sources to inform intervention refinement can help accommodate user preferences, experiences and actual intervention use and avoid over-reliance on user preferences alone shaping refinement. The likely interest in and impact of the refined MiQuit system will be assessed in two future NIHR-funded evaluation studies.