Abstract

Nonadherence to medicines continues to be a challenge in healthcare and is further complicated in patients who have multiple long-term conditions (MLTCs). This research aimed to design a new intervention delivering automated two-way text messaging from community pharmacy (TIMELY), to support medication-taking in patients with MLTCs. A co-design approach using the Behaviour Change Wheel (BCW), Human Centred Design, and realistic evaluation, was used to iteratively develop a new complex intervention and associated programme theory for how the intervention may work.
Intervention design started with a narrative synthesis systematic review of automated two-way digital communication with or without wider components such as face-to-face consultations with healthcare professionals. Data extraction included delivery characteristics and behavioural coding of content. These were analysed for impact on clinical, medication adherence, and acceptability outcomes. Results suggested that automated two-way digital communication could improve reflective motivation and promote habit formation for medication-taking. The role of supplementary healthcare professional support was unclear.
Using findings from the systematic review and the BCW, a new intervention concept was designed and communicated in a series of prototypes representing intervention components. Feedback on the prototypes was then gathered from patients and healthcare professionals in five focus groups (n=21) and using modified Nominal Group Technique (NGT) to support a co-design process. The intervention concept was found to be acceptable and useful changes were identified.
A text message library - termed Alice - was then constructed for eight long-term conditions, including: asthma, chronic obstructive pulmonary disease, chronic pain, depression, ischaemic heart disease, heart failure, hypertension, and type 2 diabetes. Delivery components were also created. These were used to run a small simulation study of the intervention as part of a co-design process with patients (n=8). Feedback was gathered using modified diary-interviews. Amendments to the tailoring and monitoring processes were identified and data indicated that Alice was able to improve motivation for medication-taking, and some contextual factors which may affect this were identified. Findings also provided insight on how to support patient engagement with Alice, including the importance of intervention delivery from a community pharmacy setting. To co-design pharmacy training for the intervention, prototypes of training components were created and tested in a simulated training event, integrated with a focus group and modified NGT exercise. Pharmacy staff (n=4) found the proposed training helpful and were able to perform most of the required tasks. However, some tasks took longer than anticipated, and findings suggest that training should be expanded to cover some areas in more depth. Tools to support communication between pharmacies and general practice were also co-designed by creating and testing these in a focus group with general practice staff (n=7) and incorporating modified NGT. The tools themselves were broadly acceptable but additional information needs of general practice about the intervention were identified.
The new TIMELY intervention seems to be acceptable to patients and healthcare professionals. TIMELY may particularly benefit patients with MLTCs, though further research is needed to understand who the intervention benefits most and under what circumstances. The co-design process also offers a novel approach for designing similar complex healthcare interventions in the future.