Table 3.

Key conclusions and limitations of the reviewed interventions and recommendations for future research and practice.

	Key Conclusions and Limitations	Research and Practice Recommendations
Participant Characteristics	Demographic Characteristics
	EMI can be used to treat various psychological and physical symptoms and health behaviors EMI successfully implemented with men and women from teens to 60s	Older adults and individuals unfamiliar with mobile technology may require additional training to use electronic devices
	Technology Literacy
	EMI can be implemented with people who have varying levels of familiarity with technology	Most participants need extensive training to use palmtop computer-based EMI; mobile phone EMI studies require less training for most people (especially when using their own mobile phone)
	Sampling Bias
	People uncomfortable using mobile electronic devices may self-select out of EMI studies, an issue not empirically addressed in any studies	Research identifying person-level characteristics (e.g., age, education, familiarity with technology) that influence willingness to participate in EMI studies is needed, so appropriate dissemination methods can be developed
Research Design Issues	Study Design
	A variety of designs were used including case reports, uncontrolled trials, and RCTs	More RCTs are needed and comparison conditions should be selected to test EMI against existing treatments Elicit participant feedback during intervention development to improve treatment design and delivery
	Assessment Times
	Most studies use relatively short treatment follow-up periods (less than 6 months) Results from EMA data are not often reported	Longer-term follow-up evaluations are needed to establish the efficacy and utility of EMI EMA data collected during the intervention can address more nuanced research questions regarding EMI implementation and efficacy
	Additional Design Considerations
	Previous research shows limited evidence for measurement reactivity to EMA Demand characteristics are not discussed; results of one study suggest demand may account for treatment outcome	Limit reactivity by designing protocols using EMI with similar frequency and duration to the exposure participants might encounter in everyday life Reduce demand characteristics by using objective outcome measures (if possible) and matching demand across conditions
EMI Features	Delivery Methods
	Automated EMI delivery systems require little ongoing involvement by researchers or clinicians Ideal intervention duration and EMI frequency depend on study sample and treatment objectives	Need feedback from participants regarding the balance between study duration and EMI frequency (preferably during the design phase of the intervention so the EMI protocol can be adjusted as needed)
	Intervention Components
	EMI elicit symptom and behavior changes when implemented with other interventions (e.g., CBT, psychoeducation, support group) Few studies test the efficacy of EMI provided without any additional treatment or intervention	Investigations of the efficacy of EMI administered with minimal additional treatment support are necessary to identify for whom such EMI may be most (or least) beneficial
	EMI Content
	EMI included a variety of treatment approaches (psychoeducation, motivational techniques, behavioral interventions, cognitive restructuring)	Relative contribution of EMI content (i.e., psychoeducation, motivation, behavioral, cognitive) remains to be determined
	EMI Tailoring
	Content and timing of EMI can be individually tailored based on pre-intervention evaluations or momentary assessment data (i.e., EMA) Non-tailored (standard) EMI may not be well received by participants No clear pattern emerged regarding the relationship between EMI tailoring and treatment efficacy	Consider tailoring EMI content and timing based on data from momentary assessments (e.g., EMA, ambulatory physiological measures, environmental sensors) Need studies experimentally manipulating the extent to which EMI content and timing are tailored to determine how tailoring influences treatment acceptability and efficacy
Intervention Efficacy	Feasibility and Acceptability
	Adequate compliance with EMI procedures was seen EMI treatments were perceived as credible and acceptable by participants	Need research examining the relationship between EMI compliance and treatment efficacy
	Symptom and Health Behavior Outcomes
	Several studies demonstrate EMI can be effectively implemented as part of smoking cessation, weight loss, and anxiety interventions There is more limited (but promising) evidence for the use of EMI in diabetes management Other health behaviors including healthy eating, physical activity, and alcohol use have been effectively targeted using EMI Study design limitations prevent conclusions regarding treatment efficacy EMI for eating disorder symptoms	Research should move beyond only developing more cost effective treatments; use EMI to develop interventions with superior efficacy to existing therapies Conclusions regarding the efficacy of weight loss EMI are limited to overweight women and additional research with men is necessary RCTs or studies using superior within subject designs are needed to test the efficacy of EMI (particularly for eating disorder symptoms and diabetes management) Future research using mobile technology to deliver EMI as “booster” treatments may provide a cost effective method for improving the long-term efficacy of psychosocial interventions
Data Analysis Issues	Few studies estimated the cost-effectiveness of EMI Less than half the studies reported the clinical significance of outcomes Some research domains (e.g., smoking cessation) conduct intent-to-treat analyses, but most others do not	More cost-effectiveness analyses are need to help researchers, clinicians, patients, and third-party payers make more well-informed treatment decisions Need to conduct clinical significance and intent-to-treat analyses so researchers and clinicians can better anticipate treatment outcomes in clinical settings
Clinical Implementation	Mobile technology could be used to deliver EMI to people without access to behavioral healthcare (e.g., in developing countries or rural areas, home-bound) Before EMI are disseminated for clinical use, treatment programs must be developed and marketed and clinicians' cost concerns addressed Unique ethical concerns (practitioner competence, patient confidentiality) arise with mobile technology use	Research is need to determine if EMI are a cost-effective alternative to developing infrastructure for traditional medical and psychological interventions in developing countries and rural settings Manualized EMI treatment programs must be developed before EMI can be widely used in clinical settings

Note. CBT=cognitive behavior therapy, EMA=Ecological Momentary Assessment, EMI=Ecological Momentary Intervention, RCT=randomized control trial.