mHealth interventions improve medication adherence and self-efficacy in cancer patients, meta-analysis finds

This systematic review and meta-analysis examined the effectiveness of mobile health (mHealth) interventions on medication adherence and related outcomes in patients with cancer. The analysis included 17 randomized controlled trials with a total of 1309 participants. The specific setting for these trials was not reported. The population consisted exclusively of patients with cancer, though specific cancer types, stages, or treatment regimens were not detailed in the summary data provided. The average follow-up duration across studies was 3.0 months.

The intervention consisted of various mHealth technologies, including mobile applications, websites, and text messaging services, aimed at supporting patients. The comparator group was standard care or other control conditions, with specific protocols varying by included study. Dosing or specific intervention protocols were not detailed in the available summary, indicating variability in how mHealth tools were implemented across the different trials analyzed.

For the primary outcome of medication adherence, the meta-analysis found statistically significant improvements. Medication adherence rates showed an odds ratio of 3.47 (95% CI 1.92-6.26; P=.002), indicating patients receiving mHealth interventions were over three times more likely to be adherent. Medication adherence scores, measured as a continuous outcome, showed a standardized mean difference (SMD) of 1.01 (95% CI 0.51-1.52; P=.001). These represent large effect sizes, though absolute adherence rates were not reported.

Key secondary outcomes also showed significant benefits. Self-efficacy improved with an SMD of 0.90 (95% CI 0.29-1.51; P=.01). Service satisfaction was reported as significantly improved, though specific effect size data were not provided in the summary. Symptom burden was significantly reduced with an SMD of -0.38 (95% CI -0.61 to -0.14; P=.008), representing a small to moderate effect. In contrast, health literacy showed no significant effect, with an SMD of 0.51 (95% CI -1.50 to 2.52; P=.29).

Safety and tolerability findings were notably absent from the reported results. Adverse events, serious adverse events, discontinuation rates, and general tolerability assessments were all listed as 'not reported.' This represents a significant gap in the evidence, as digital interventions, while generally considered low-risk, can have unintended consequences including increased anxiety, technology-related stress, or privacy concerns that were not captured in this analysis.

These results contribute to a growing body of literature on digital health interventions in oncology. Prior studies have shown mixed results for mHealth tools, with some demonstrating benefits for symptom monitoring and quality of life, but less consistent evidence for hard endpoints like adherence. This meta-analysis suggests potentially stronger effects on adherence than some individual trials have shown, though the low certainty of evidence tempers this comparison. The finding of improved self-efficacy aligns with theoretical models suggesting mHealth tools empower patients in self-management.

The analysis has several important methodological limitations. The authors noted substantial heterogeneity among studies, indicating that the effects varied considerably across different interventions, populations, and settings. The overall risk of bias across included studies was rated as moderate. Most critically, the certainty of the evidence was assessed as low using GRADE criteria, meaning further research is very likely to change the confidence in the effect estimate. Subgroup analyses by intervention duration, format, or cancer type reportedly had wide confidence intervals or crossed zero for some comparisons, limiting specific conclusions about which approaches work best.

For clinical practice, these findings suggest mHealth interventions could be considered as adjunctive tools to support medication adherence and self-efficacy in cancer patients. However, given the low certainty of evidence, they should not replace established adherence support methods. Clinicians should be selective about recommending specific apps or platforms, considering factors like ease of use, evidence base for the particular tool, and patient digital literacy. The lack of effect on health literacy indicates these tools may not improve patients' fundamental understanding of their health information.

Several important questions remain unanswered. The optimal components, duration, and frequency of mHealth interventions for cancer patients are unclear. Whether benefits differ by cancer type, treatment regimen, or patient demographics requires further investigation. Long-term effects beyond 3 months are unknown. The cost-effectiveness of implementing these interventions at scale was not addressed. Most importantly, the absence of safety data represents a critical evidence gap that future studies must address to ensure these tools do not cause harm.