A novel tool to monitor adherence to oral oncolytics: a pilot study

Abstract Background: Non-adherence is an important issue in cancer care as more oral cytotoxic and targeted agents become available. Although oral therapies may be more convenient for patients, measuring and optimizing adherence is challenging. The Nomi system records real-time medication taking behavior from a “smart” prescription bottle and displays the data on a web-based interface. Nomi can also communicate with patients via text message to intervene in cases of non-adherence. Aims: The objective is to report the results of a 28-patient pilot study aiming to assess Nomi’s ability to assist patients taking capecitabine, an oral chemotherapy agent with a complex, cyclical regimen. Methods: Eligible patients were prescribed capecitabine for breast, colorectal, pancreatic, or biliary cancer. The study had a pre-intervention stage, during which patients were monitored, and an intervention stage, in which the text messaging feature was enabled. Adherence was defined as the number of correct doses (both timing and quantity) over the total number of prescribed doses. Conversions were events in which patients took a dose after receiving a text intervention (from Nomi). Adherence throughout the study was calculated from the data that the bottles collected – we calculated adherence scores for each patient, during each cycle and study period (pre vs. post-intervention), defined as the number of correct doses (both timing and quantity) over the total number of prescribed doses. We defined three categories of patients by percent change in adherence: category 1 (>8%), category 2 (−8% to 8%), and category 3 (< −8%). Results: We collected data from 28 patients (24 pre/post and 4 pre-only). On average, patients were 84% adherent (N = 28; SD = 11%). During pre-intervention, patients had a self-adherence of 89% (SD = 12%), and afterwards, they had an average adherence of 90% (SD = 6%). Most of the patients in category 1 demonstrated a substantial conversion rate (> 35%). Patients in category 1 tended to live in regions with lower average household income (Mean = $58,937) than those in category 2 (Mean = $77,482) and category 3 (Mean = $90,972). Of survey respondents, 56% indicated that they would want to continue using Nomi, while 67% indicated that they would recommend it to others Conclusions: This innovative technology is able to monitor, measure and intervene for patients taking capecitabine in real-time. Adherence overall was high, and some patients appeared to benefit more from text message interventions. Future work should focus on patients deemed high risk for non-adherence.