The use of electronic monitoring (MEMS(R)) to evaluate antipsychotic compliance in outpatients with schizophrenia
Remington G, Kwon J, Collins A, Laporte D, Mann S, Christensen B;
Commented by , 28 Feb 2007
Aim of the study
Non-compliance with medication is a major concern in most medical specialties, but studies have shown that in schizophrenia it is particularly prevalent. One problem is the measurement of compliance. In principle, there are five methods, none of which is perfect:
-
self-rating questionnaries filled in by the patient
-
the physician’s judgement after a clinical interview
-
pill-count
-
measurement of plasma-levels and
-
Medication Event Monitoring System (MEMS) technology.
The latter method consists in special medication containers which incorporate an electronic device which records every opening of the container.
Method
In their study the authors compared four of the methods described above, self-report, clinician rating, pill-count and MEMS. 52 outpatients with schizophrenia or schizoaffective disorder were followed up for four weeks. One important design feature was that the patients were not specifically informed about the use of MEMS. Compliance was assessed using a threshold of 80% intake, but also as a continuous outcome.
Results
The clearly highest noncompliance rate was detected by MEMS (52%), followed by pill-count (25%), clinician rating (24%) and self-report (3%). Similar results were obtained when compliance was assessed as a continuous measure. When the ability of the psychiatrists to predict the compliance of their patients was assessed, there were many wrong guesses in terms of both overestimating and underestimating, underlining how difficult it is to really tell whether patients take their medication.
More severe illness, older age, longer duration of illness and complexe dosing schemes (i.e. more than one dose per day) were associated with non-compliance, while perceived social support and a good alliance with the treatment team predicted a good compliance.
Dr Leucht's comments
From a clinical perspective the main finding of the study could be the difficulty of the treating physician to tell about patients' compliance. There are other examples in the literature showing that both self-reports and clinician ratings on the average underestimate the compliance. But the three more objective methods to assess compliance listed above have limitations, as well:
-
pill-count is very unreliable, because the patient can simply throw the medication away instead of bringing it back to the hospital. This method is mainly used in clinical trials
-
plasma-levels are expensive and only allow for a quantification as to whether the patient has taken medication at all or not. Furthermore, some patients may decide to take medication just in the days before the blood-test
-
the MEMS technology is expensive and therefore not useful in clinical routine. Furthermore, only the opening of the container is recorded, not whether the patient actually took the medication.
In this study this problem was limited by not informing the patients about the use of the MEMS technology, but for ethical reasons this would be impossible in routine care.
In summary, we need to be aware about the difficulty to judge upon our patients' compliance. For research purposes the MEMS technology is the most appropriate one and will hopefully be used in future large studies on this major problem in the treatment of schizophrenia.