Codevelopment of a Text Messaging Intervention to Support Adherence to Adjuvant Endocrine Therapy in Women With Breast Cancer: Mixed Methods Approach

Background: Adjuvant endocrine therapy (AET) reduces breast cancer recurrence and mortality in women with early-stage breast cancer. Unintentional nonadherence to AET is common (eg, forgetting to take medication). Forming habits surrounding medication taking could reduce reliance on memory and improve AET adherence. SMS text messaging interventions may offer a low-cost approach for promoting medication-taking habits. To optimize the likely effectiveness of such SMS text messages, the content should be developed using a transparent approach to ensure fidelity to relevant psychological theory and with user input to increase acceptability.

Objective: This study aimed to develop a pool of brief SMS text messages promoting habit formation to support AET adherence, which are acceptable to women with breast cancer and show fidelity to theory-based behavior change techniques (BCTs).

Methods: According to published literature, we selected 6 BCTs derived from the habit formation model: action planning, habit formation, restructuring the physical environment, adding objects to the environment, prompts/cues, and self-monitoring of behavior. In study 1, behavior change experts (n=10) created messages, each based on 1 of the 6 BCTs, in a web-based workshop and rated the fidelity of the messages to the intended BCT. In study 2, women with experience of taking AET discussed the acceptability of the messages in a focus group (n=5), and the messages were refined following this. In study 3, women with breast cancer rated the acceptability of each message in a web-based survey (n=60). In study 4, additional behavior change experts rated the fidelity of the remaining messages to the intended BCT in a web-based survey (n=12). Finally, a consultant pharmacist reviewed a selection of messages to ensure that they did not contradict general medical advice.

Results: In study 1, 189 messages were created targeting the 6 BCTs. In total, 92 messages were removed because they were repetitious, unsuitable, or >160 characters, and 3 were removed because of low fidelity (scoring <5.5/10 on a fidelity rating scale). Following study 2, we removed 13 messages considered unacceptable to our target population. In study 3, all remaining messages scored above the midpoint on an acceptability scale (1-5); therefore, no messages were removed (mean 3.9/5, SD 0.9). Following study 4, we removed 13 messages owing to low fidelity (scoring <5.5/10 on a fidelity rating scale). All the remaining messages showed fidelity to the intended BCTs (mean 7.9/10, SD 1.3). Following the pharmacist review, 2 messages were removed, and 3 were amended.

Conclusions: We developed a pool of 66 brief SMS text messages targeting habit formation BCTs to support AET adherence. These showed acceptability to women with breast cancer and fidelity to the intended BCTs. The delivery of the messages will be further evaluated to assess their effect on medication adherence.