Smoking lapses early on during a quit attempt are highly predictive of failing to quit. A large proportion of these lapses are driven by cravings brought about by situational and environmental cues. Use of cognitive-behavioral lapse prevention strategies to combat cue-induced cravings is associated with a reduced risk of lapse, but evidence is lacking in how these strategies can be effectively promoted. Unlike most traditional methods of delivering behavioral support, mobile phones can in principle deliver automated support, including lapse prevention strategy recommendations, Just-In-Time (JIT) for when a smoker is most vulnerable, and prevent early lapse. JIT support can be activated by smokers themselves (user-triggered), by prespecified rules (server-triggered) or through sensors that dynamically monitor a smoker's context and trigger support when a high risk environment is sensed (context-triggered), also known as a Just-In-Time Adaptive Intervention (JITAI). However, research suggests that user-triggered JIT cessation support is seldom used and existing server-triggered JIT support is likely to lack sufficient accuracy to effectively target high-risk situations in real time. Evaluations of mobile phone cessation interventions that include user and/or server-triggered JIT support have yet to adequately assess whether this improves management of high risk situations. While context-triggered systems have the greatest potential to deliver JIT support, there are, as yet, no impact evaluations of such systems. Although it may soon be feasible to learn about and monitor a smoker's context unobtrusively using their smartphone without burdensome data entry, there are several potential advantages to involving the smoker in data collection.
Implications: This commentary describes the current knowledge on the potential for mobile phones to deliver automated support to help smokers manage or cope with high risk environments or situations for smoking, known as JIT support. The article categorizes JIT support into three main types: user-triggered, server-triggered, and context-triggered. For each type of JIT support, a description of the evidence and their potential to effectively target specific high risk environments or situations is described. The concept of unobtrusive sensing without user data entry to inform the delivery of JIT support is finally discussed in relation to potential advantages and disadvantages for behavior change.