Medication adherence monitoring: implications for patients and providers

Introduction

A healthcare provider prescribes a medication and hands you the prescription. Where some patients head straight for the pharmacy after the visit, others do not. Some may put it off for another day or two, while others may never fill the prescription. In the meantime, the paper script could be misplaced or simply forgotten. In other cases, patients may ask their pharmacist to put the prescription on hold until they are ready to take it. Despite all these varied scenarios, in all cases, the prescriber has no confirmation of whether the patient has filled the prescription, when the medication was started, and whether or not the prescribed therapy is effective.

Studies have shown that up to one-third of prescriptions are never picked up by patients. Primary non-adherence, defined as the rate at which patients fail to fill and pick up new prescriptions, ¹ can result in adverse outcomes for the patient and unnecessary emergency department visits and hospital admissions. Surescripts (a company operating the largest electronic prescribing network in the United States) reports that 28% of all prescriptions either never make it to the pharmacy or are not purchased when patients realize the prohibitive cost of the medication, which can lead to adverse patient outcomes, thereby leading to increased health system costs. ² Fischer et al. showed that several factors were associated with non-adherence to e-prescriptions, including prescribing non-formulary medications and residence in a low-income area. ³ In a 2007 study, cost-related non-adherence was found to impact one in 10 Canadians. ⁴ Other factors that may predict non-adherence include forgetfulness, illiteracy, inability to understand the purpose of treatment, not perceiving the treatment as necessary, a lack of trust in the treatment, or a lack of knowledge about the effects of treatment. As well, psychiatric problems (including depression), cognitive limitations, missing visits, and a poor relationship with the healthcare provider have also been found to contribute to non-adherence. ⁵ In addressing the problem, it is also important to distinguish between intentional and unintentional non-adherence. Intentional non-adherence is a process in which the patient actively decides not to use treatment or follow treatment recommendations, whereas unintentional non-adherence may be the result of forgetfulness, not knowing exactly how to use medications, or other unplanned behaviour. ⁶

Non-adherence has many downstream effects on both individual patients and the health system as a whole. Patients risk their overall health by not filling their prescriptions and not adhering to their treatment. This can result in a worsening of their condition and, consequently, in higher utilization of the health system through more frequent emergency department visits and hospitalizations. According to a World Health Organization report, medication non-adherence accounts for 5% of Canadian hospital admissions, 5% of physician visits, and resulting in an additional $4 billion in healthcare costs annually. ⁷ In the United States, improving patient primary medication adherence can potentially lead to savings of between $140 and $240 billion over 10 years, measured in healthcare costs savings and improved health outcomes. ⁸ Realization of these benefits is dependent on pharmacies and in some cases, prescribers, following up with patients who have not picked up their medications. For patients who experience challenges in taking their medications as prescribed, adherence monitoring is essential to ensure that their condition does not deteriorate or that they do not experience adverse drug reactions. For example, patients on medication that is effective or safe within a narrow therapeutic range (eg, anticoagulants, insulin) would benefit significantly if monitored. Close monitoring by a healthcare professional allows for earlier intervention and avoids adverse outcomes that otherwise may send patients to acute care settings. In a study of patients with chronic disease, research participants indicated a desire for greater participation in decision-making concerning their healthcare and increased education about their illness and medication. Additionally, participants wanted individualized healthcare that recognized their preferences, along with their personal and emotional well-being. ⁹

Despite the benefits of monitoring medication adherence, patient advocates have raised ethical concerns over the paternalistic aspect of sharing information about a patient’s medication-taking behaviour, specifically adherence to prescribed medications. These patients feel that they are capable of taking their medication as prescribed and therefore do not need someone monitoring them. Conversely, if a prescriber is not informed of medication non-adherence, he or she may increase the dose, or change the therapy due to the belief that the medication is not working (based on other monitoring and testing results). This situation does not benefit the patient, physician, or the healthcare system. The patient is not receiving the treatment they need, the physician is expending time and resources on treating a patient with inadequate information, and there are unnecessary costs to the healthcare system.

Scenario 1: The case of Mr. A

Mr. A began frequenting the pharmacy about a year ago. He had a few prescriptions from his family doctor. He asked the pharmacist to put his cholesterol medication, atorvastatin 10 mg, on hold as he “had enough at home.” Three months later, after again seeing his family doctor, Mr. A visited the pharmacy. While talking to the pharmacist, again Mr. A advised he did not need his atorvastatin. The pharmacist inquired if he realized his doctor had increased his dose to 20 mg. Mr. A was very nonchalant and advised “oh, yeah we talked about it, I’m just going to use up what I have at home.” Another 3 months later, Mr. A brought in another prescription. Along with his atorvastatin 20 mg, the doctor had prescribed another new medication to further help with his cholesterol. By now, the pharmacist had gotten to know Mr. A pretty well and approached him, non-judgementally, and inquired as to why he was not taking his cholesterol medication. Mr. A, appreciating the pharmacist’s approach, confessed that he had read about the side effects of the medication and was apprehensive to take his medication. He did not want to affect his relationship with his doctor, as he had great respect for him, so he never told him about the doubts he had about taking the medication. The pharmacist thanked Mr. A for his candour and provided education on the probability of him experiencing side effects and the importance of him taking the medication since he has a heart condition. The pharmacist asked Mr. A’s permission to inform his doctor of what had occurred and asked for the original prescription (atorvastatin 10 mg) to be started. Mr. A was happy with the outcome and knew that the pharmacist would help maintain his relationship with his doctor.

If Mr. A’s doctor had known that the medications he had prescribed were not filled, he could have begun a dialogue with Mr. A regarding his non-adherence. Furthermore, if prescription data were available to the pharmacist, and not only what prescriptions Mr. A had filled (ie, dispense data), the pharmacist would also have been able to provide a more timely intervention.

Current methods of identifying medication adherence

In 2016, a Canadian survey of community pharmacists found that five in 10 prescriptions they receive are handwritten (either brought in by the patient or faxed) and four in 10 are electronically generated and were either brought in by the patient or in some cases electronically faxed (auto-faxed) from the prescriber’s Electronic Medical Record (EMR). ¹⁰ The study reveals that a majority of prescribers rely on patients to bring their prescriptions to the pharmacy, therefore, resulting in a greater likelihood of losing prescriptions and increasing the potential for fraud and misuse. Additionally, with a lack of comprehensive medication utilization information at the point of care, clinicians are not enabled to fully assess a patient’s adherence and or otherwise address-related concerns in collaboration with the patient.

Currently, most provinces have Drug Information Systems (DIS), which store data related to prescriptions dispensed by pharmacies. However, these databases, do not have information related to what was prescribed. These data reside in hospital and prescriber systems and often cannot be accessed by other healthcare providers involved with the patient’s care. Clinicians have information about which medications are dispensed but no information for the roughly 30% of prescriptions that are generated but never filled.

Pharmacy Management Systems (PMS) can be used to track whether or not patients pick up their medications from the pharmacy. This allows pharmacists to follow-up with patients who have repeat prescriptions or for whom new prescriptions were transmitted ahead of time by prescribers. However, there is no method currently in place for this information to flow back to prescriber system, so that they could reach out to their patients, if they noticed that prescribed medications were not dispensed. A 2010 study published in the International Journal of Medical Informatics reported that 77% of physicians would call patients if the software alerted them when patients failed to pick up a prescription that would lead to serious medical consequences if not taken on time. ¹¹

Scenario 2: the case of Mrs. B

Mrs. B has been a long-time patient at the pharmacy. She is now 71-years old, in good health overall, and lives on her own. The pharmacist sees Mrs. B at the end of her shift, and Mrs. B begins to tell her about her grandson’s upcoming piano recital. The pharmacist notices Mrs. B is in a bit of pain; “It’s just my arthritis acting up” Mrs. B remarks. The pharmacist congratulates Mrs. B on her grandson’s achievement, and the pharmacy assistant steps in to take Mrs. B’s prescription, as the pharmacists are changing shifts. One month later, the pharmacist sees Mrs. B at the pharmacy again, in a great deal of pain. She hands the pharmacist a prescription for an opioid pain reliever. The pharmacist says she is sorry to see her in so much pain. Mrs. B says “I should have paid for the prescription I brought in a month ago, maybe I wouldn’t be in this state now.” The pharmacist reviews her file and sees there was a prescription put on hold for Tylenol Arthritis, which is not covered by her seniors plan. Mrs. B remarks, “at least this one’s covered.” The pharmacist asks Mrs. B if she told her nurse practitioner she did not take the Tylenol Arthritis. Mrs. B said “Oh, I never thought to tell her, I don’t like to waste her time.” The pharmacist discusses with Mrs. B that by her not taking medication for her pain, she had likely aggravated her condition, which is primarily arthritis, and a milder pain reliever along with perhaps some physiotherapy would be a better alternative for her. Although the new opioid pain medication is covered, there are more potential risks for her, especially as she lives on her own. Mrs. B gives permission for the pharmacist to speak with her nurse practitioner, who agrees with the recommendation.

In both scenarios described above, there was a lack of timely information available for clinicians to make informed decisions for the patient. Data flow between prescriber and pharmacy systems could make this information more consistently available and improve overall communication between providers and patients.

The role of e-prescribing in improving adherence

The Canadian Medical Association and the Canadian Pharmacists Association jointly define e-prescribing as “the secure electronic creation and transmission of a prescription between an authorized prescriber and a patient’s pharmacy of choice, using clinical Point of Service (POS) solution, in a manner which integrates clinical workflow and software.” Such a system provides an opportunity to track initial prescriptions and identify non-adherence that may have previously been undetected. A 2011 study in the United States observed that non-adherence occurred less often when e-prescriptions were transmitted directly to the pharmacy rather than printed to give to patients. ³ A 2017 study of dermatology patients supported those findings by observing a 47% reduction in primary nonadherence if the prescription was in electronic format compared with a paper prescription.

In Canada, there is a coordinated effort between the provinces, Health Canada, and Canada Health Infoway to implement e-prescribing at a national level. The current information gap at the time of prescribing will be filled by augmenting data already collected by provincial DIS with prescribing data, to help clinicians detect non-adherence with prescribed medications. As more providers use e-prescribing to both generate and transmit prescriptions, the more valuable this information will become from a population health viewpoint. Additional functionality such as dispense notification from a pharmacy system to a provider’s EMR will further close that information gap. Incorporating provincial formulary information into EMRs will enable providers to prescribe medications that are covered by public insurance plans and potentially minimize cost as a barrier to adherence for patients.

With the rise of connected health information made possible by interoperable patient care systems such as EMRs, PMS, and electronic prescribing platforms such as Surescripts in the United States and PrescribeIT^TM in Canada, electronic notification of patient adherence to medication (or lack thereof) is technically feasible. However, are there ethical implications to consider when a prescriber is notified of whether or not a patient picked up their medication? Would such a system infringe upon the autonomy of individuals or violate their expectations of privacy?

Research has been conducted to gather the perceptions of patients on medication adherence with e-prescribing. Lapane et al. ¹² were the first to study the perceptions of geriatric patients. Researchers found that patients felt e-prescribing generated somewhat increased discussions about adherence with their provider compared to patients who did not receive e-prescriptions. ¹² The study concluded that regardless of whether the patients received an e-prescription, 75% of geriatric patients reported that they did not tell their physician if they did not want a prescription, and 85% reported that they never tell their provider if they do not intend to purchase the prescribed medication. ¹²

In evaluating and intervening to increase adherence to medical treatments, clinicians, and researchers must address ethical issues pertaining to best interest, autonomy, and privacy. ¹³ Further, if an easy and efficient monitoring system for medication adherence was available, this could provide benefit for patients, prescribers, and the healthcare system. However, this mechanism would need to accommodate the wishes and address privacy concerns of individuals. Once PrescribeIT^TM makes drug dispense status data available to physicians via their EMR, prescribers will be able to look up whether or not a prescription that they have generated has been dispensed and decide to follow up with the patient or inquire with other clinicians in the patient’s circle of care. From a privacy standpoint, this is not different from the current state, where prescribers could make that inquiry via phone, fax, or any other means of communication. A more active form of notification such as an alert or reminder to prescribers that their patient has not filled their prescription after a given amount of time would likely require more scrutiny since patients may not be aware that their behaviour is actively being monitored. An alternative solution that would mitigate some of these ethical concerns could be to remind patients directly that they have not picked up their medications. A study conducted in two primary care clinics in Omaha, Nebraska demonstrated that patient nonadherence rates decreased drastically after patients were sent an electronic reminder via an on-line patient portal seven days after a prescription was issued. ¹⁴ Regardless of the approach, clinicians should make patients aware that in an electronic environment, more complete patient data are more readily available to all providers within the patient’s circle of care.

Conclusion

Medication adherence in an era of EMRs requires complex analysis with diverse stakeholders to determine an ethical path forward. Although there are many factors that influence a patient’s decision to follow a prescribed medication regimen, evidence demonstrates that the negative impacts of medication non-adherence on patients, clinicians, and the healthcare system are significant. As a result, electronic tools that support clinicians to help improve patient adherence are often well received. Tools including e-prescribing, dispense notification to prescribers, electronic patient reminders, and related communication will allow patient data to flow across information systems, such as PMS and prescriber EMRs. Research further demonstrates that ensuring patients are informed, engaged, and empowered in decisions about their health is key to safeguarding the appropriate implementation of these tools. Relevant stakeholders, including patient advocates, clinicians, regulators, and privacy experts, need to further collaborate to address ethical concerns so the full potential of e-prescribing may be realized.