TABLE 2.

Comparison of approaches to measuring adherence for clinicians

Measurement method	Strengths	Weaknesses	Recommendation
Self-report w/standardized instruments	Inexpensive, easy to administer, can determine reasons for nonadherence and lead to intervention planning, represents patients’ experience/views; can assess multiple elements of the posttransplant medication regimen, including medication-taking, clinic appointment attendance, completion of blood work, following diet and exercise requirements, use of substances (eg, alcohol, tobacco), and any other monitoring requirements (eg, blood pressure, temperature, lung function)	May be less sensitive, specific, or precise than desired, subject to desirability response bias and recall bias; in the case of medication adherence, does not assess key components of medication-taking execution, for example taking, timing, or drug holidays	We recommend use of a self (patient)-reported standardized instrument that indicates reasons for nonadherence (eg, intentional or unintentional). Efforts should be made to reduce bias in reporting, such as explaining to the patient that it may be difficult to stay totally adherent to all prescribed regimen components
Provider report	Inexpensive, easy to administer, can determine reasons for nonadherence and lead to intervention planning	May be less sensitive, specific, or precise than desired, subject to bias due provider misperceptions; often underestimates nonadherence; in the case of medication adherence, does not assess key components of medication-taking execution, for example taking, timing, or drug holidays	We do not recommend the sole use of provider reports
Medication refill reportsa	Not subject to reporting bias, potentially accurate, potentially inexpensive	Indirect measure—pills obtained by patients may not have been taken, not taken on time, and/or not taken for multiple days (ie, drug holiday), clinics may not have access to all pharmacies used by their patients to obtain refill records, may not reflect dose changes	We recommend use of refill reports if a clinic has resources in place to access these metrics
Medication blood levelsa	Direct measure of whether medication has been taken	Potentially difficult or expensive to implement, relatively intrusive to patient; only reflects adherence within a few days prior to level (ie, skipped for multiple days drug holiday); potentially less accurate because many factors other than adherence impact blood level results	We recommend use of blood levels if a clinic has resources in place to collect blood samples and data
Electronic medication event monitoring devicesa	Can provide precise determination of concordance between pills used and prescribed regimen (eg, taking, timing, drug holidays), continuously records pill usage over desired time period and can quantify length of gaps between pills used, not subject to reporting bias	Expensive to obtain, potentially difficult and costly to implement by providers and use by patients, indirect measure—pills removed from device by patient may not have been taken; patients often only willing to use for short periods of time and do not see these devices as suitable replacements for weekly pill boxes for all medications	Recommended if a clinic has resources in place to purchase and implement these devices
Electronic medical record review of adherence behavior (eg, clinic appointments, laboratories completed)	Can provide data on appointment and laboratory testing adherence; easily accessible by transplant staff; adherence assessment can be performed quickly	Requires access to electronic medical record	We recommend that this method should be implemented for laboratory tests and clinic visits
Remote spirometry monitoring (in lung transplantation)	Can provide precise determination of spirometry use; not subject to reporting bias	Potentially difficult to implement by providers and for use by patients, indirect measure	We recommend this method as a complement to patient report if a clinic has resources to commit to its use
Other biological assays (eg, smoking, alcohol use indicators)	Direct measure of whether prohibited substances have been used	Potentially difficult or expensive to implement, relatively intrusive to patient	We recommend this method as a complement to patient report if a clinic has resources to commit to its use

^aOnly applicable to medication adherence.