When it comes to treatment adherence, HIV is far from the easiest disease to manage. The sheer number of medications, combined with the precision with which they must be taken, can stymie patients and physicians alike. While this fact has been extensively documented in developing countries, where adherence to medication is listed high on the difficulties in treating HIV, the same problem exists in the U.S. among even the most carefully monitored patients.
“HIV is not unlike many other chronic conditions. Adherence to medications is very important,” explained Keith McInnes, ScD, a health services researcher at the University of Boston, while describing his research at this year’s VA Health Services R&D conference. “But for HIV, adherence may be the linchpin of treatment. Adherence is so important because, at the individual patient level, keeping the virus under control requires a very high level of adherence to their meds. And on the public health level, patients with poor adherence may create drug-resistant strains of HIV.”
Nevertheless, physicians have difficulty determining which patients are most at risk for being non-adherent, identifying only a quarter of those patients not taking their medications. Even when physicians know which of their patients are not taking their medication, recent studies have shown they are not equipped with the skills to counsel them adequately, McInnes said.
McInnes and his research partners have been attempting to combat non-adherence by developing a computerized assessment tool for patients—a self-administered exam that can help identify which patients are not taking their medications and why.
That assessment tool comes in the form of a tablet computer loaded with a computer-assisted self-interview (CASI). The CASI is given to patients when they first step into the waiting room, and they answer the questions while waiting for the doctor to see them.
The ongoing research project began by looking at 72 HIV-positive veterans spread out among 11 providers. The CASI led each patient through a series of branching questions:
- Which medications they are on;
- How often they are supposed to take them;
- How often they actually do take them; and
- If they are not 100% adherent, what keeps them from being so.
For example, early during the exam, the patient is shown a page with pictures of every type of pill or injection prescribed for HIV. Patients are asked to touch the ones they are taking. Then, they are asked a series of questions about each medication, such as what dosage they take and what times of day they are supposed to take it.
The computer is superior to paper and pencil, McInnes said, because the program is equipped with branching logic functions that lead the patient to the next appropriate question. It also is preferable to in-person questioning, because it gives a sense of anonymity and privacy.
When the patient is finished with the questionnaire, the results are printed and provided to the clinician before the patient is seen. The goal is not merely to see who is non-adherent but to identify gaps in knowledge about medication and socio-economic factors that keep patients from taking their medications.
“If a patient in the questionnaire says, ‘Yes, I’m less than 100% adherent,’ they can choose from 20 common problems with adherence, such as sleeping through a dose, or their schedule is too hectic,” McInnes said. “Once problems are identified, adherence care management is recommended.”
Care reccomendations are printed out along with the patient’s answers in the CASI report given to the provider. The report also alerts the physician to any self-reports of past injury, concerns with alcohol or substance abuse, or whether a patient has self-reported that he or she is thinking of self-injury.
The study also has an adherence care manager (ACM) component. An experienced nurse or pharmacist is on hand to provide telephone counseling with patients, McInnes said. “We want to make this a clinical tool used routinely by patients and providers. But it’s just one piece of the whole picture. We have to make sure there is support for patients and providers after we identify non-adherent patients.”
Identifying Barriers to Adherence
Testing of the CASI prototype showed early success in identifying non-adherent patients. All patients in the initial study successfully completed the CASI, with 77% reporting it was very easy to use and 21% indicating somewhat easy.
Using the accepted HIV medication cutoff (95% of doses), 37% of veterans reported non-adherence over the previous three days, and 53% reported 30-day non-adherence. Most notably, in 22% of cases, providers incorrectly judged patients as adherent, when, in fact, patients reported non-adherence.
Patients were also less likely to overestimate their own adherence when using the CASI. About 40% reported themselves suboptimal in terms of adherence, and 25% of patients had at least one of three different errors in medication knowledge: incorrect dose or frequency, reporting taking a medication not prescribed, or failing to list a prescribed medication.
The top four reasons for missing doses were that the patient was away from their medication (43%), forgot (38%), slept through a dose (27%), and schedule was too busy (22%). Other common reasons included stress, lack of pills, being high, being depressed, fear, and did not want to be seen taking HIV medication.
“I think, so far, we’re quite pleased to see the tool is working,” McInnes said. “We are finding that patients do report and indicate mistakes with medication. And patients are willing to self-report that they’re not perfectly adherent. They’re also reporting a large number of barriers to adherence, and we’ve being able to link a large number of those patients to phone counseling.
Since that initial project, the HIV CASI system was further rolled out in two other VA HIV clinics and one urban safety-net hospital.