Definition And Measurement Techniques

It would be ideal to have a specific and standard definition of posttransplant noncompliance, yet no such definition exists. Authors may dichotomously group recipients as compliant or noncompliant using a variety of definitions. Others use a percentage of prescribed doses correctly taken, but as above these numbers are often crude estimates. More importantly, it is unknown if there is a "compliance threshold" associated with deteriorating graft function or loss. Is missing 20%, 10%, or even 5% of doses a problem? Is missing a certain percent of doses more of a problem early versus later posttransplant? Is missing successive doses ("drug holiday") more of a problem than intermittently missing single doses? No studies have clearly addressed any of these questions. Similarly, there is no consensus on which end points are most informative. Some studies have used acute rejection episodes as an endpoint, some have looked at the incidence of chronic rejection or graft loss, while still others have measured glomerular filtration rate.

Also a variety of techniques have been used to assess compliance, each has some disadvantages. Pill counts performed at clinic visits assume that the patient has taken the missing pills. But there are numerous reports of patients discarding (or hoarding) medication. Blood levels, especially when a drug has a short half-life (e.g., cyclosporine), can be manipulated unpredictably. Noncompliance has been noted in recipients with both low and high cyclosporine levels (either missing doses or taking an extra dose just before the clinic visit). Chart reviews and patient surveys routinely underestimate the incidence of noncompliance. A chart review can only document noncompliance if it has been previously recorded. Patient interviews or questionnaires are only as accurate as the information being provided. Patients simply can not remember and report forgotten medication doses! Furthermore, personal embarrassment and the desire to "please the physician" lead to substantial under-reporting of consciously missed medications. Precisely for these reasons the most credible patient reports are those documenting noncompliance, clearly an under-reporting bias. Even anonymous questionnaires are of limited use since if anonymous, the degree of compliance cannot be related to an outcome. In addition, while studies using these techniques may be important in elucidating risk factors, they provide data too late to be of any value in an intervention trial.

0 0

Post a comment