Example of information that can be collected to evaluate the impact of medication adherence technologies on treatment outcomes
| Outcome | Potential definitions |
| Medication adherence (ie, dosing implementation and persistence on therapy) | Proportion of all expected doses that were missed during the full treatment course* (continuous outcome) |
| Proportion of patient months with >X%† of expected doses missed (continuous outcome) | |
| Proportion of patients who completed therapy with <X% of expected doses missed over the full treatment course (binary outcome)† | |
| Treatment interruptions | Proportion of patients who completely interrupt tuberculosis (TB) therapy for a short time period (eg, >1 month) or who are formally lost to follow-up (eg, >2 months) (binary outcome) |
| Treatment success | Proportion of patients who achieved cure or treatment completion (binary outcome) |
| Proportion of patients who achieved cure or treatment completion without extension of treatment duration due to non-adherence (binary outcome) | |
| Mean or median number of medication refills per patient as a proxy of months of treatment completed (continuous outcome) | |
| Post-treatment tuberculosis recurrence-free survival | Proportion of patients who complete TB therapy and achieve 1-year recurrence-free survival (binary outcome)ठ|
*An ‘ideal’ length of therapy could be used for assessing the number of expected and missed doses—for example, 182 expected treatment days for patients on daily therapy for drug-susceptible TB (see online supplementary appendix 1).
†Threshold of the percentage of expected doses missed can vary depending on baseline rates of adherence (eg, greater than or less than 10%, 20%, etc).
‡That is, patients who achieve treatment completion and do not experience post-treatment TB recurrence or death.
§Follow-up times can vary, though we recommend a minimum of 6 months of post-treatment follow-up.