Before therapy adherence could be measured, it first had to be clearly defined. Through extensive research and expert consultations, CROPLAND and Novellas Healthcare developed a tailored definition, drawing on the WHO’s categories affecting adherence (health system, condition, treatment regimen, socioeconomic environment and the patient), the AMA’s threshold of at least 80% of prescribed medication taken, and the APA’s factors such as education, medication cost, belief in treatment and fear of side effects. This resulted in a definition of therapy adherence as a patient’s commitment to follow the agreed treatment plan, including dosage and timing, while accounting for symptoms, side effects, cost and disease knowledge. CROPLAND analyzed Novellas Healthcare’s datasets, focusing on adherence programs and factors such as age, location, disease and dropout rates, while Novellas Healthcare developed Dynamic Forms to capture structured patient data securely. Add sensitive data and an initially vague description of therapy adherence to that mix, and it is clear that analysing this was no easy task. With a custom data pipeline, CROPLAND’s data engineers were able to transform unstructured patient surveys into workable data. This data was then used to test the correlation between various factors and therapy adherence. Interpretation of the results was supported by Novellas Healthcare experts, who were continuously involved in the project. This iterative approach created a synergy between both partners, leading to better results than either could have achieved separately.