Sorry, you need to enable JavaScript to visit this website.
Skip to main content

Patient Safety and Adherence to Self-Administered Medications

Harriette Gillian Christine Van Spall, MD; Robby Nieuwlaat, PhD; and R. Brian Haynes, MD, PhD | July 1, 2011
View more articles from the same authors.

The Case

A 30-year-old man, with HIV and a recent diagnosis of central nervous system (CNS) toxoplasmosis, returned to the emergency department (ED) of a community hospital for the third time in two weeks.

Two months earlier, the patient was diagnosed at a university hospital with toxoplasmosis, based on characteristic brain lesions on MRI and presence of the parasite's DNA in his cerebrospinal fluid (via polymerase chain reaction). He was discharged on standard treatment for CNS toxoplasmosis (pyrimethamine, sulfadiazine, and folinic acid) and HIV disease (anti-retroviral medications). Two weeks before this ED visit, he was admitted to the community hospital with seizures. Imaging showed no change in the brain lesions. He was discharged on the anti-seizure medication phenytoin. The day before this ED visit, the patient had a second seizure that led to a motor vehicle accident. He was brought to the community hospital's ED, an MRI of his brain showed no change, and he was discharged home.

He returned the next day, this time complaining of new onset right-sided weakness. Compared to the MRI one day earlier, CT showed increased swelling of his brain and compression of his frontal lobe. Medical staff were now worried that this patient had treatment-resistant toxoplasmosis or another brain disease, and transferred him to the university hospital where the original diagnosis had been made.

During the workup at the university hospital, the patient said that his symptoms had improved after he was initially discharged two months earlier. However, one month earlier, he ran out of his medications for toxoplasmosis. He did not have any refills and did not know how to obtain refills. He continued the anti-retroviral HIV medications for which he had refills. Soon after stopping the toxoplasmosis medications he began to experience headaches, then seizures and other symptoms that led to his multiple presentations over the prior two weeks. The patient was restarted on appropriate therapy for CNS toxoplasmosis and his symptoms improved.

The Commentary

"Drugs don't work in people who don't take them." – C. Everett Koop

This case portrays low treatment adherence that is life-threatening. Adherence, the extent to which patients take their treatments as prescribed, is surprisingly low with prescriptions for medications, and even lower with prescribed lifestyle changes. Even when their medications are fully covered by insurance, more than 40% of patients are non-adherent.(1) Patients who are non-adherent usually take too few rather than too many doses, and are at increased risk of death, disability, and hospitalization.(2,3) Adherence to treatment may represent a surrogate marker for healthy behavior, which confers a survival advantage that is independent of the benefits of the treatment itself and is known as "the healthy adherer effect."(2) Of all medication-related hospital admissions in the United States, up to 69% may be due to poor medication adherence, with an estimated cost of $100 billion per year.(4)

Effectively managing low adherence in individual patients begins with detection. Practical methods for monitoring adherence include: (i) Watching for non-attendance: did this patient have a follow-up appointment? If so, did anyone take action when he did not attend? (ii) Watching for a less-than-expected response (clearly evident in this case). (iii) Measuring serum drug or metabolite levels (e.g., for this man's anticonvulsants). (iv) Asking the patient: "People often have difficulty taking all of their pills. During the past week, did you miss any of your medications?" An affirmative response to this question is associated with less than 60% adherence.(5) Other methods of monitoring adherence include proprietary medication electronic monitoring systems (which are expensive), prescription refill counts (which require pharmacy collaboration) and manual pill counts during home visits, comparing pill numbers with prescribed rates of intake.(5)

A myriad of factors can cause low adherence to self-administered prescriptions.(2-4) Demographic attributes such as age, gender, socioeconomic status, education, IQ, race, or religion do not affect adherence, and, therefore, should not influence who we pay attention to or how we manage adherence. And yet these stereotypes continue to arise when clinicians ask, "Why didn't the patient take their meds?" Studies show that clinicians are not able to reliably judge the adherence rates of their patients even if they feel that they know them well.(6) The key practical factors that affect adherence are inadequately communicated instructions, complex or lengthy (> 3 month) regimens, behaviorally demanding regimens (e.g., dieting, exercising), inadequate follow-up or supervision of care, and patients with limited capacity for self-care. All of these might affect our patient. The ostensible reason for loss of adherence to a 3-medication regimen for toxoplasmosis in this young man is that "…he ran out of his medications for toxoplasmosis. He did not have any refills and did not know how to obtain refills." That he is non-adherent can be taken at face value, given his clinical course. When asked about adherence, patient self-report of adherence is often incorrect, but self-report that acknowledges non-adherence is seldom incorrect (i.e., self-report of non-adherence has low sensitivity but high specificity as a diagnostic test).(5) Of interest, he indicates that he has continued to take his HIV meds, but this report may not be accurate (self-report is an insensitive measure!).

As for the patient's stated reasons for non-adherence, these seem plausible and could represent failures of the health care providers and system: no refills provided, failure to inform the patient that refills were needed and where they could be obtained, and failure to provide scheduled follow-up. Indeed, it is easy to believe that he has fallen into the gap between the university hospital and the community hospital, given his bouncing back and forth between them. These errors are all potentially solvable through provision of refills at the time of prescription, clear communication with the patient about the goals and duration of therapy, and arrangements for follow-up evaluation. Non-adherence due to complex regimens may be tempered by the use of weekly dosette-packaged medications as well as the use of a single pharmacy where medications could be tracked and reconciled. Transitions in care may be made safer by electronic data sharing between physicians, health care centers, and pharmacies, which may bridge the gap between the islands of information within which we deliver health care.

Yet, the patient's statements could also be a deflection from other, more relevant reasons or even be self-serving. We don't know what he was told or what follow-up arrangements were made. Was he not told to refrain from driving a car after the onset of his seizures? Is he competent to make decisions for self-care? If not, who are his caregivers and were they informed of the exact regimen?

If he is non-adherent for reasons other than professional or system failure, these could include a very complex regimen (could it be simplified?), the patient's belief that the regimen had already worked (was he given written instructions on how long the medication needs to be taken? how to acquire refills?), and concern regarding actual or potential adverse effects from the medication (does he have any such concerns that might lead to premature discontinuation?). The only way to know is to ask. Again, these are potentially remediable and should be addressed with each patient.

While it is best to get to the root causes of low adherence for individuals, this may not be feasible as there are no validated approaches to identifying the reason for non-adherence. Moreover, trials of adherence interventions for chronic disease medical regimens show that there is no magic bullet for improving adherence, but combining two or more of the following may help: convenient care, information, reminders, self-monitoring, reinforcement, counseling, family therapy, psychological therapy, crisis intervention, manual telephone follow-up, and supportive care.(7) Alas, even the most effective interventions do not lead to large improvements in adherence and treatment outcomes. Thus, this is a vital missing link in translation of evidence-based treatments into effective practice, a topic in search of better solutions. One can imagine a rosier future, in which integrated information systems (hospital-clinic-lab-pharmacy-patient) provide timely feedback to patients and providers alike on expected and achieved processes of care, adherence to appointments and treatments, and patient-important outcomes. With modern information technology, this state is feasible; the limits lie in the up-front cost and in "human factors"—how willing and able are we and our patients to collaborate in achieving optimal treatment goals?

Take-Home Points

  • Low adherence is ubiquitous and should be considered in all patients.
  • Clinicians are not able to reliably judge the adherence rates of their patients even if they know them well.
  • Low adherence can be detected by watching for nonattendance, watching for less than expected response, asking the patient, and monitoring serum drug levels or prescription refills.
  • When asked "did you miss any of your medications during the past week," patient self-report of adherence is often incorrect, but self-report that acknowledges non-adherence is seldom incorrect.
  • Adherence can be improved by combinations of convenient care, clear communication, reminders, patient self-monitoring with reinforcement from clinicians, counseling, family therapy, psychological therapy, crisis intervention, telephone follow-up, and supportive care.

Harriette Gillian Christine Van Spall, MD FRCPC (Cardiology)

Division of Cardiology, Department of Medicine

Hamilton Health Sciences

McMaster University

Hamilton, Ontario

Robby Nieuwlaat, PhD

Assistant Professor of Clinical Epidemiology & Biostatistics

DMcMaster University Faculty of Health Sciences

Hamilton, Ontario

R. Brian Haynes, MD, PhD, FRCPC

Professor of Clinical Epidemiology and Medicine

Chief, Health Information Research Unit

McMaster University Faculty of Health Sciences

Hamilton, Ontario


1. Doshi JA, Zhu J, Lee BY, Kimmel SE, Volpp KG. Impact of a prescription copayment increase in lipid-lowering medication adherence in veterans. Circulation. 2009;119:390-397. [go to PubMed]

2. Simpson SH, Eurich DT, Majumdar SR, et al. A meta-analysis of the association between adherence to drug therapy and mortality. BMJ. 2006;333:15. [go to PubMed]

3. Rasmussen JN, Chong A, Alter DA. Relationship between adherence to evidence-based pharmacotherapy and long-term mortality after acute myocardial infarction. JAMA. 2007;297:177-186. [go to PubMed]

4. Osterberg L, Blaschke T. Adherence to medication. N Engl J Med. 2005;353:487-497. [go to PubMed]

5. Haynes RB, Taylor DW, Sackett DL, Gibson ES, Bernholz C, Mukherjee J. Can simple clinical measurements detect patient noncompliance? Hypertension. 1980;2:757-764. [go to PubMed]

6. Stephenson BJ, Rowe BH, Haynes RB, Macharia WM, Leon G. The rational clinical examination. Is this patient taking the treatment as prescribed? JAMA. 1993;269:2779-2781. [go to PubMed]

7. Haynes RB, Ackloo E, Sahota N, McDonald HP, Yao X. Interventions for enhancing medication adherence. Cochrane Database Syst Rev. 2008;2:CD000011. [go to PubMed]


This project was funded under contract number 75Q80119C00004 from the Agency for Healthcare Research and Quality (AHRQ), U.S. Department of Health and Human Services. The authors are solely responsible for this report’s contents, findings, and conclusions, which do not necessarily represent the views of AHRQ. Readers should not interpret any statement in this report as an official position of AHRQ or of the U.S. Department of Health and Human Services. None of the authors has any affiliation or financial involvement that conflicts with the material presented in this report. View AHRQ Disclaimers
Related Resources From the Same Author(s)
Related Resources