Accurate Assessment of Adherence (man vs. machine) [abstract]

A recent article in Chest1 demonstrates the value of electronic data collection in medication adherence. In this case it was inhalation therapy in cystic fibrosis patients. This type of article is important as we consider the future of electronic health records and where the data for such records should come. Remember, people are notoriously unreliable historians while computers don’t lie.


Background: People with cystic fibrosis have a high treatment burden. While uncertainty remains about individual patient level of adherence to medication, treatment regimens are difficult to tailor, and interventions are difficult to evaluate. Self- and clinician-reported measures are routinely used despite criticism that they overestimate adherence. This study assessed agreement between rates of adherence to prescribed nebulizer treatments when measured by self-report, clinician report, and electronic monitoring suitable for long-term use.

Methods: Seventy-eight adults with cystic fibrosis were questioned about their adherence to prescribed nebulizer treatments over the previous 3 months. Self-report was compared with clinician report and stored adherence data downloaded from the I-Neb [2] nebulizer system. Adherence measures were expressed as a percentage of the prescribed regimen, bias was estimated by the paired difference in mean (95% CI) patient and clinician reported and actual adherence. Agreement between adherence measures was calculated using intraclass correlation coefficients (95% CI), and disagreements for individuals were displayed using Bland-Altman plots.

Results: Patient-identified prescriptions matched the medical record prescription. Median self-reported adherence was 80% (interquartile range, 60%-95%), whereas median adherence measured by nebulizer download was 36% (interquartile range, 5%-84.5%). Nine participants overmedicated and underreported adherence. Median clinician report ranged from 50% to 60%, depending on profession. Extensive discrepancies between self-report and clinician report compared with nebulizer download were identified for individuals.

Conclusions: Self- and clinician-reporting of adherence does not provide accurate measurement of adherence when compared with electronic monitoring. Using inaccurate measures has implications for treatment burden, clinician prescribing practices, cost, and accuracy of trial data.

  1. CHEST August 2011 vol. 140 no. 2 425-432
  2. Detailed information on the I-neb AAD System.

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.