Impact of pharmacy-led medication reconciliation on admission to internal medicine service: experience in two tertiary care teaching hospitals

Medication errors remain a significant patient safety concern and a financial burden in hospitalized patients in the United States and Europe [1‐3].At transitions of care, more than 50% of medication errors occur, and up to 67% of medication histories contain one or more errors [4, 5]. With the rising number of patients receiving multiple medications and the complex medication therapy management, medication reconciliation becomes warranted. In many organizations, the implementation of medication reconciliation at all transitions of care (admission, transfer, and discharge) helped prevent Adverse Drug Events (ADEs) [6] and reduce medication errors [7]. In a recent randomized controlled study conducted in Oman, the implementation of medication reconciliation on admission and discharge reduced rates of preventable adverse drug events from 16 to 9.1%, p = 0.009 [8]. In another recent prospective interventional study conducted to determine the effect of medication reconciliation in two ICU settings in the Netherlands, the proportion of patients with ≥1 medication transfer error at ICU admission was reduced from 45.1 to 14.6% (OR_adj 0.18 [95% CI 0.11–0.30]) and after ICU discharge from 73.9 to 41.2% (OR_adj 0.24 [95% CI 0.15–0.37]) [9].

Effective January 2018, the Joint Commission named maintaining and communicating accurate patient medication information as a National Patient Safety Goal for hospitals [10].

The process of medication reconciliation performed by healthcare professionals (HCPs) has been previously defined [11, 12]. During this process, HCPs obtain the best possible medication history (BPMH), from patients and their families in order to identify and resolve any unintended discrepancies [10‐12]. To date, there is no consensus with regards to who should have ownership of the process and who should be assigned to complete the medication list [13‐15]. Other challenges include the absence of a standardized process, time limitation, and patients’ or physicians’ lack of familiarity with some medications [15]. The Institute for Healthcare Improvement identifies reconciliation as the shared responsibility of nurses, pharmacists, and physicians, where each has a defined role [15].

The American Society of Health-System Pharmacists advocates for the pharmacist’s role in medication reconciliation [14]. In recent studies, the pharmacist’s involvement in medication reconciliation at admission and discharge was shown to decrease the number of medication errors, improve patient safety, and reduce costs associated with health resource utilization [16‐24]. In a meta-analysis that compared the effectiveness of pharmacist-led medication reconciliation interventions to usual care, the pooled analysis from 17 studies involving 21,342 adult patients showed a substantial reduction of 67, 28 and 19% in adverse drug event-related hospital revisits (RR 0.33; 95% CI 0.20 to 0.53), emergency department (ED) visits (RR 0.72; 95% CI 0.57 to 0.92) and hospital readmissions (RR 0.81; 95% CI 0.70 to 0.95) in the intervention group as compared to the usual care group, respectively [24]. Furthermore, studies proved that pharmacy students and technicians are accurate and efficient when included in the medication reconciliation process; they decrease costs and provide support to other HCPs [25].

In Lebanon, a well-defined practice for medication reconciliation does not exist. The implementation of medication reconciliation in Lebanese hospitals does not keep pace with international standards. Nurses or physicians are currently involved in taking the medication history, while the role of the pharmacist is underestimated. A recent survey evaluating general hospital pharmacy practice in Lebanon revealed that around 41% of hospitals perform medication reconciliation upon admission, transfer of care, and discharge. Teaching hospitals were more likely to perform medication reconciliation than non-teaching hospitals; and pharmacy services in teaching hospitals seemed to be more advanced cooperating with affiliated medical schools [26]. A previous study, not directly focusing on medication reconciliation, showed that student pharmacists and faculty identified more than 1000 medication-related problems, 4.3% of which were related to medication reconciliation [27]. A study assessing hospital pharmacy medication reconciliation practice in Jordan (a neighboring country) showed relatively low awareness of the concept and policy of medication reconciliation process among Jordanian pharmacists [28]. To our knowledge, there are no published studies addressing medication reconciliation in Lebanon.

The primary objective of this study was to assess the clinical impact of pharmacy-led medication reconciliation performed on adults patients admitted to the Internal Medicine (IM) services with ≥1 chronic medication, on day one of hospital admission, measured by the incidence of unintended medication discrepancies identified. The secondary objectives were to assess the time needed for medication history, and the information sources used to complete the BPMH. Other secondary objectives were to classify the unintended discrepancies by medication class and route of medication administration, by potential ADE severity, and by proximal cause leading to the discrepancy. The study also assessed potential determinants of unintended discrepancies.

This is a pilot prospective study conducted in two tertiary care teaching hospitals in Beirut, Lebanon (Saint George Hospital-University Medical Center and Makassed General Hospital) from March through June 2018.

The study included 204 patients. The patients had equal gender distribution and a mean age of 70.5 years. The majority of the study population (82.4%) took ≥5 home medications. Approximately 16% of patients reported having a history of allergy to ≥1 medication, and around 8% reported experiencing a previous adverse drug reaction. Table 1 details the sociodemographic and baseline characteristics of the patient population.

The time needed to obtain medication history ranged from 3 to 35 min. The student pharmacist used a combination of different information sources to complete the BPMH for each patient: patient interview (70.1%), caregiver/family member interview (57.4%), examination of home medication bottles or boxes (55.9%), review of previous medical record (25.5%), and communication with the physician (3.9%).

Following the critical analysis of patient cases, 94 patient cases (46.1%) had no discrepancies with a complete one-to-one match, and 7.8% included intended discrepancies explained by the patient’s clinical condition. There was 195 unintended medication discrepancies identified in 94 patients by the pharmacy team, with a maximum number of 7 discrepancies per patient case (Table 2).

Most of the unintended discrepancies (67.7%) involved a prescription medication, and 32.3% involved an over-the-counter medication. The most common discrepancies consisted of medication omission (71.8%), wrong dose (12.8%), wrong medication (7.2%), and wrong frequency (5.1%).

The proximal causes leading to the discrepancies included the patient’s forgetfulness or lack of knowledge (57.9%), the clinician’s lack of knowledge or familiarity with the involved medication (10.3%), a confusion regarding the dosage form of the involved medication (0.5%), and name similarity of the involved medication with another medication (0.5%). In 30.8% of the cases, the hospital pharmacist did not identify a reason behind the discrepancy.

When classifying the unintended discrepancies by route of administration, 92.3% were found to involve an oral medication. Only 3.6% of the unintended discrepancies involved an inhaled medication and 1.5% involved a subcutaneous medication.

The most common agents involved in unintended discrepancies consisted of dietary supplements (27.7%). Other medication classes involved were anti-hyperlipidemic agents (7.2%), medications for reflux disease (7.2%), and medications for asthma/COPD (7.2%). Further details are included in Table 3.

Moreover, around 8% of the unintended discrepancies involved a high-alert medication, including oral hypoglycemic agents, insulin, anti-arrhythmic agents, and anticoagulants. When assessing potential severity of all medication-related discrepancies by the investigators, 122 discrepancies out of 195 were judged as clinically insignificant, 71 (36.4%) were judged as clinically significant, and only 2 (1%) were judged to be serious. No life-threatening interventions were identified.

Based on the unintended discrepancies found, the pharmacy team recommended a total of 195 medication-related interventions. Amongst these interventions, 64.6% were accepted, 25.6% were rejected, and around 10% pending review. The most common types of medication-related interventions included: the addition of a medication (71.8%), dose adjustment (12.8%), discontinuation or switching of a medication (7.2%), and frequency adjustment (5.1%). Other interventions were giving medication from home supply (1.5%), and adjustment of medication route (1%).

In the multivariable analysis, the number of home medications was the only variable significantly associated with the occurrence of unintended discrepancies (ORa 1.11 (1.03–1.19); p = 0.007) (Table 4).

In this pilot study, the clinical impact of pharmacy-led medication reconciliation performed upon admission to the IM services is examined at two hospitals in Lebanon. During the study period, 204 patients were included, and 195 unintended discrepancies were identified in 94 patients.

The most intricate part of medication reconciliation is obtaining the BPMH. In this study, the student pharmacist used different information sources to complete the BPMH for each patient and mostly relied on patient interviews, caregiver/family member interviews, and examination of home medication boxes. Consulting at least one information source other than the patient is crucial to determine the accuracy of a patient’s medication history list [11]. Other sources that could have been used include outpatient pharmacies, and prescriber offices [16, 33]. In Lebanon, these sources may not provide excellent information to improve BPMH due to the practice of medicine in the country where there is a tendency for patients to seek care directly from specialists (often multiple) without the oversight of a primary-care provider [34]. Furthermore, many patients might not have loyalty to a single community pharmacy and to date, a prescription drug monitoring program database does not exist in Lebanon [35]. The number of information sources remained in the final model of the multivariable analysis for unintended discrepancies (ORa 1.37, p = 0.08). The lack of statistical significance may be attributed to the small sample size.

The majority of patients included in this study were polymedicated. In the multivariable analysis, the number of home medications was significantly associated with the identification of unintended discrepancies (ORa 1.11; p = 0.007). Based on previous studies, the number of pre-admission medications correlates with admission order errors and unintentional discrepancies [33].

Most of the unintended discrepancies in this study consisted of drug omissions (72%) as similarly reported in the literature [16, 18, 36‐39]. In such cases, the patient is failing to receive a pre-admission home medication that was deemed necessary by a HCP. The clinical significance of this omission depends on the drug omitted. The clinical significance of the unintended discrepancies was evaluated on a case-by-case basis. On another note, discrepancies in medication history, such as the omission of medications or incomplete information about adherence or non-adherence, may delay or prevent clinicians from timely identification of drug-induced diagnoses that may have led to the current hospital admission [16].

When assessing potential severity, most of the unintended discrepancies (63%) were judged as clinically insignificant, where the error was less likely to cause harm, such as the omission of a dietary supplement or the omission of an antacid [40]. Around 36% of unintended discrepancies were judged as clinically significant, where the error had the potential to cause harm and may have required additional monitoring, such as omission/wrong dose of a beta-blocker. Discrepancies involving high alert medications such as the omission of an oral hypoglycemic agent were also considered as significant [29]. Two serious discrepancies found in this study involved antiarrhythmic and antithrombotic therapy. The first case involved the omission of amiodarone. The potential severity was classified as serious since antiarrhythmic therapy is useful in reducing the frequency and shortening the duration of arrhythmias, and decreasing arrhythmia-related hospitalizations [41]. Although amiodarone has a long half-life, its omission may be serious as it puts the patient at an increased risk of developing another arrhythmia, or worsening of the original arrhythmia. This also requires additional monitoring including laboratory or diagnostic tests, which may increase length of hospital stay and associated costs. The second serious unintended discrepancy involved the detection of a wrong dose of apixaban, a new oral anticoagulant (NOAC). This severity classification was based on a review of previous studies which showed that off-label dosing of NOACs was associated with morbidity and mortality, and underdosing NOACs was associated with increased cardiovascular hospitalization, and higher stroke rates [42, 43].

Most of the medication-related recommendations relayed by the pharmacy team regarding the unintended discrepancies were accepted (64.6%), and resulted in an alteration in the patient’s pharmaceutical care plan. This finding further supported the role of pharmacists and student pharmacists in the medication reconciliation process. Most of the rejected recommendations were related to the omission of a dietary supplement, which were assessed by the physicians as unnecessary during acute illness and hospitalization.

Furthermore, the investigators estimated the proximal cause of each unintended discrepancy identified. Evaluation of proximal causes helps identify areas of weakness in the medication reconciliation process, where special attention is paid to prevent future discrepancies [19]. Patient forgetfulness or lack of knowledge was the most common cause (57.9%), followed by clinician lack of knowledge or familiarity with the medication (10.3%). The lack of expertise of non-pharmacists in obtaining a proper medication history, and patients who are poor historians are reported as common barriers for medication reconciliation [44, 45]. In this study, there was no identified reason for the medication discrepancy in 30% of the cases. The latter may be attributed to the time-consuming nature of the BMPH process, the lack of collaboration between the different stakeholders, and the limited access to technology and health information systems [44, 45]. In fact, electronic-based systems were shown to improve medication reconciliation processes by facilitating documentation and access to information [46]. Lebanon is an upper middle-income country [47], where the implementation of electronic-based health information systems is lagging behind, across the healthcare sector.

Moreover, in 2019, the Ministry of Public Health in Lebanon in collaboration with the Haute Authorité de Santé (HAS France) updated the Accreditation Standards for Hospitals in Lebanon and emphasized the need for medication reconciliation at admission, and discharge. The medication management chapter within the standards also reinforces that this reconciled list should be shared with the healthcare providers and the pharmacy [48]. Owing to their distinct knowledge and skills as medication experts, pharmacists are uniquely qualified to lead interprofessional efforts, and establish and maintain an effective medication reconciliation process in hospitals [49]. An interprofessional medication reconciliation process revolves around communication, teamwork, and involves the different healthcare professionals in the medication reconciliation process to ensure optimal patient outcomes [50]. For example, pharmacists taking BPMH should not work in silos and need to communicate with the treating physicians to clarify medication orders, and relay any interventions. Moreover, pharmacists are expected to reach out to the patients’ nurses to verify monitoring parameters.

To our knowledge, this is the first published data on medication reconciliation in Lebanon. The study further highlights the importance and the clinical significance of the pharmacist’s role in medication reconciliation and supports the involvement of student pharmacists in medication history taking. The integration of student pharmacists in the collection of patient medication history allows pharmacists some free time to focus on other in-depth clinical activities.

In context of the updated mandatory hospital accreditation standards, the findings of this pilot study were reported to the individual hospital sites, to further support the integration of the pharmacy team in the medication reconciliation process. In collaboration with the study sites, the authors are further planning to assess patient outcomes on transitions of care when comparing pharmacist-driven medication reconciliation versus standard of care medication reconciliation (completed by nurse, physician etc.) in each site. In today’s challenging and fast-paced healthcare environment, healthcare institutions are urged to optimize resources and eliminate duplication of efforts from the patient care process [51]. Although outside the scope of this study, a future research focusing on the cost effectiveness of pharmacist-led medication reconciliation, with or without student pharmacists, in comparison to other health care professionals, may also bring insight to hospital administrators on to how to allocate limited resources.

Unintended medication discrepancies were common on hospital admission to the internal medicine services. Pharmacy-led medication reconciliation upon admission, with student pharmacist involvement and physician communication helped reduce unintended discrepancies and improve medication safety and patient outcomes. Further studies are warranted to assess the impact of pharmacy-led medication reconciliation across transitions of care.