Terrence Steyer, M.D.
Chair, Department of Family Medicine
Vik et al. (2004) reviewed the state of measuring medication adherence through a literature search of MEDLIN, PubMed, and International Pharmaceutical Abstracts using the following keywords: elderly, adherence/nonadherence, compliance/noncompliance, medication/drug, methodology/measurement, and hospitalization. The authors state that although several methods are available for the assessment of adherence, accurate measurement continues to be difficult. At present there is no generally accepted "Gold Standard" for measuring adherence. Krousel-Wood et al. (2004) also provide a summary of methods to assess medication adherence.
Biological assays measure the concentration of a drug, its metabolites, or tracer compounds in the blood or urine of a patient. These measures are intrusive and often costly to administer. Patients who know that they will be tested may consciously take medication that they had been skipping so the tests will not detect individuals who have been nonadherent. Drug or food interactions, physiological differences, dosing schedules, and the half-life of the drugs may influence the results. Biological tracers that have known half lives and do not interfere with the medication may be used, but there are ethical concerns. All of these methods have high costs for the assays that limit the feasibility of these techniques. (Vik et al. 2004)
Counting the number of pills remaining in a patient's supply and calculating the number of pills that the patient has taken since filling the prescription is the easiest method for calculating patient medication adherence. Some data indicate that this technique may underestimate adherence in older populations (Grymonpre et al. 1998). Patterns of non-adherence are often difficult to discern with a simple count of pills on a certain date weeks to months after the prescription was filled. Because pill counts are often based upon the date a prescription is filled, patients who get prescriptions refilled prior to their first one running out and then combining pills into a single (and possibly non-original) bottle presents complications. Loss of data is common among many studies. (Vik et al. 2004)
Hess et al. (2005) used the weight of a topical medication remaining in a tube as a measure of adherence. When compared with patient log books of daily medication use, weight estimates of adherence were considerably lower than patient log estimates. They recommend that clinical trials involving topical applications incorporate medication weights as the primary measure of adherence. In a comparison of methods to measure adherence, Carroll et al. (2004) found that estimates calculated from medication logs and medication weights were consistently higher than those of electronic monitors.
The Medication Event Monitoring System (MEMS) manufactured by Aardex Corporation allows the assessment of the number of pills missed during a period as well as adherence to a dosing schedule. The system electronically monitors when the pill bottle is opened, and the researcher can periodically download the information to a computer. The availability and cost of this system could limit the feasibility of its use. On the Aardex web site there is a bibliography of publications that use the MEMS.
This method can be used primarily for medications that are taken for chronic illnesses (such as hypertension). Concerns regarding the completeness and reliability of these records have been expressed (Vik et al. 2004). These records provide only an indirect measure of drugs consumed. Patterns of over and under consumption for periods less than that between refills cannot be assessed.
Studies have consistently shown that third-party assessments of medication adherence by healthcare providers tend to overestimate patients' adherence (Vik et al. 2004). Interviewing patients to assess their knowledge of the medications they have been prescribed and the dosing schedule provide little information as to whether the patient is adherent with the actual dosing schedule. Subjective assessments by interviewers can bias adherence estimates. This method is rarely used in medical research to assess adherence (Vik et al. 2004).
Direct questioning of patients to assess adherence can be an effective method. Patients who admit to nonadherence are generally accurate in their assessment. However, patients who claim adherence may be underreporting their nonadherence to avoid caregiver disapproval (Vik et al. 2004). Other methods may need to be employed to detect these patients.
Morisky et al. (1986) developed a 4-item scaled questionnaire to assess adherence with antihypertensive treatment. Thier scale demonstrated acceptable psychometric properties.
Li et al. (2005) developed four instruments to measure antihypertensive medication adherence in a population of Chinese immigrants in the US. Their measures are culturally sensitive and demonstrate good reliability.
The Hill-Bone Compliance to High Blood Pressure Therapy Scale includes 14 items, 8 of which are directed at assessing medication taking behavior in hypertensive patients (Hill et al. 2000). Not only is this method relatively simple and economically feasible to use, but it has the added advantage of soliciting information regarding situational factors that interfere with medication adherence (e.g. forgetfulness, remembering to bring medications along when out of town) (Krousel-Wood et al. 2004).
The Compliance-Questionnaire-Rheumatology (CQR) is a 19-item questionnaire that has been favorably compared with electronic medication event monitoring (de Klerk et al. 2003). This instrument has good validity and reliability.
The purpose of the MASE scale is so clinicians and researchers can identify situations in which patients have low self-efficacy in adhering to prescribed medications. Link
Carroll CL, Feldman SR, Camacho FT, Manuel JC, Balkrishnan R (2004) Adherence to topical therapy decreases during the course of an 8-week psoriasis clinical trial: commonly used methods of measuring adherence to topical therapy overestimate actual use. Journal of the American Academy of Dermatology. 51(2):212-6.
de Klerk E, van der Heijde D, Landewe R, van der Tempel H, van der Linden S (2003) The compliance-questionnaire-rheumatology compared with electronic medication event monitoring: a validation study. Journal of Rheumatology. 30(11):2469-75.
Grymonpre RE, CD Didur, PR Montgomery, DS Sitar. (1998) Pill count, self-report, and pharmacy claims data to measure medication adherence in the elderly. Ann. Pharmacother. 32: 749-754.
Hess LM, K Saboda, DC Malone, S Salasche, J Warneke, DS Alberts (2005) Adherence Assessment Using Medication Weight in a Phase IIb Clinical Trial of Difluoromethylornithine for the Chemoprevention of Skin Cancer. Cancer Epidemiol Biomarkers Prev 14: 2579-2583.
Kim MT, Hill MN, Bone LR (2000) Development and testing of the Hill-Bone Compliance to High Blood Pressure Therapy Scale. Prog Cardiovasc Nurs 15:90–96.
Krousel-Wood, M, S Thomas, P Muntner, D Morisky (2004) Medication adherence: a key factor in achieving blood pressure control and good clinical outcomes in hypertensive patients. Current Opinion in Cardiology. 19(4):357-62.
Li W-W, AL Stewart, NA Stotts (2005) Cultural factors and medication compliance in Chinese immigrants who are taking antihypertensive medications: instrument development. J Nursing Meas. 13: 231-52.
Morisky DE, LW Green, DM Levine (1986) Concurrent and predictive validity of a self-reported measure of medication adherence. Med. Care 24: 67-74.
Vik, SA, CJ Maxwell, DB Hogan (2004) Measurement, correlates, and health outcomes of medication adherence among seniors. Annals of Pharmacotherapy 38, 303-312.
f you know of a measurement tool that should be included in this list please contact the Measurement Tools Site administrator: Mark Geesey