By Linda Dodds, Director, Medicines Use and Safety Division, East and South East England

Introduction

By Linda Dodds, Director, Medicines Use and Safety Division, East and South East England Specialist Pharmacy Services

Incomplete or inaccurate communication at the time of admission has been shown to lead to prescribing errors in up to 67% of all admissions. Studies which have included multidisciplinary review of unintentional discrepancies between the pre-admission and admission medication lists have concluded that between 30-50% of intercepted errors have the potential to translate into adverse events which impact on patient outcomes, leading to increased monitoring and length of inpatient stay or worse. Uncorrected errors on admission to hospital not only impact on the inpatient stay, but may be continued after discharge and adversely affect long term care, resulting in further patient harm, increased contact with healthcare providers and even death. For these reasons, accurate medicines reconciliation at admission to hospital has been endorsed as a key patient safety initiative in many health economies, with England and Wales producing a joint NICE/NPSA Safety solution in 2007.

As well as improving patient safety, accurate medicines reconciliation also contributes to the QIPP programme, by reducing length of stay, and the risk of readmission

The challenges of providing medicines reconciliation to all patients shortly after their admission are however well recognised, and most organisations are still working to optimise delivery of a pharmacy led service within the limited staff resource available. Recent published literature on medicines reconciliation can not only help share learning on how best to ensure accurate medicines reconciliation on admission to an inpatient setting, but also act as a lever to support renewed efforts to deliver this important safety initiative.

Synopsis

Medicines reconciliation has been defined by the Institute for Healthcare Improvement as “the process of identifying the most accurate list of a patient’s current medicines – including the name, dosage, frequency and route – and comparing them to the current list in use, recognizing and discrepancies, and documenting any changes, thus resulting in a complete list of medications, accurately communicated”.

In the United States, the Joint Commission on Accreditation of Healthcare Organisations (JCAHO) made medicines reconciliation on admission to hospital (or other transfers of care) a National Patient Safety Goal (No 8) in 2005. The JCAHO Sentinel Event Alert No.35 (25 Jan 2006) refers to the Massachusetts Coalition for the Prevention of Medical Errors advising that reconciliation should take place within specified time frames – “within 24 hours of admission; shorter time frames for high-risk drugs, potentially serious dosage variances, and/or upcoming administration times”. The NPSG states that health care organisations should consider “creating a process for reconciling medications at all interfaces of care (admission, transfer, discharge) and determining reasonable time frames for reconciling medications. Patients, and responsible physicians, nurses and pharmacists should be involved in the medication reconciliation process.”

In the United Kingdom, guidance on medicines reconciliation has been given by the National Institute for Clinical and Healthcare Excellence, in conjunction with other national bodies (see Key Publications listed below), and most National Health Service organisations have produced their own policies for implementing this guidance locally.

Medicines reconciliation remains a very active field for both academic and practice-based research worldwide, and this annotated bibliography lists (in order of the first author’s name) papers published during the last two years.

Medication discrepancies warranting reconciliation have been found by studies across many countries, in patient admission and discharge, and in a range of situations, including emergency units (Caglar et al., De Winter et al. (i) and (ii), Hummel et al., Kimland et al., Knight et al., Mazer et al., Mills et al., Soler-Giner et al., Thomas), critical/intensive care (Corbett and Shulman, Hatch et al.), mental health/psychiatry (Morrison, Nelson et al., Paton et al.), kidney dialysis (St Peter), orthopaedics (S Williams et al.), paediatrics (Rappaport et al., Stone et al., Terry et al., White et al.) and the elderly (Gizzi et al., Perennes et al., Steurbaut et al., Stitt et al., Villanyi et al.).

High rates of discrepancies have been found at both admission (Gleason et al., Lindved Karkov et al., Tschantz Unroe et al.) and discharge (Climente-Marti et al., Herrero-Herrero and Garcia-Aparicio), while a study in Norway found incomplete information about medication and medication changes in both admission and discharge letters (Frydenberg and Brekke).

All sources of information about medicines have been found to be more or less inaccurate (Balon et al., Elrod et al.), including electronic medical records (Platte et al.), and a study in Sweden found major discrepancies between information on medicines in different electronic systems (Ekedahl et al.).

On the other hand, a study in France with cardiovascular medicines showed good agreement between doctors’ prescribing records and patients’ self-reports (Grimaldi-Bensouda et al.).

Reconciliation failures have been found to depend on a number of factors, including the classes of drugs involved (Grimes et al. (iii)). Another study found that non-prescribed medicines, and topical and inhaled preparations were most likely to be overlooked (Owusu-Ankomah et al.).

Interviews with hospital-based pharmacists and doctors in the USA identified a number of barriers to medicines reconciliation (Boockvar et al. (ii)).

One study showed that handwritten and electronically generated discharge summaries had similar levels of error, possibly because in this case computerization was still in a relatively early stage, with a continuing need for transcription (Callen et al.). Transcription should be avoided where possible (Grimes et al. (ii)). Problems have also been found when different electronic systems were used in different departments of a hospital (Corbett and Shulman). One paper suggests that errors should be improved by using compatible IT systems based on the ISO/CEN EN 13 606 standard (Farfan Sedano et al.)). Another study showed little difference between the use of electronic or paper-based prescribing, or hybrid systems (Lee et al.).

Medicines reconciliation is regarded as essential for patient safety (Greenwald et al.) as it can reduce medicines discrepancies and potential adverse events (Platte et al., Villanyi et al.), although there is little evidence of its effect on actual adverse events (Etchells). Inadequate reconciliation has been shown to be a cause of medication errors in HIV patients (Snyder et al.). Implementation of a formal reconciliation process on hospital admission has been shown to have a favourable effect on outcomes, including reducing drug-related readmissions (Hellstrom et al.), but in another study documentation of correct discharge procedures did not affect readmission rates (Hansen et al.). Better control of blood pressure has been demonstrated in hypertensives with fewer medication discrepancies (Persell et al.). Medicines reconciliation reduced adverse events due to erroneous medication changes on admission, but not those caused by all medication changes (Boockvar et al. (i)). Reconciliation may reduce the level of inappropriate prescribing after patients have been discharged (Hatch et al. (ii)).

A number of approaches have been proposed to improve medicines reconciliation. Various electronic reconciliation tools have been shown to be effective (Gimenez Manzorro et al., Schnipper et al.) and such tools have been reviewed (Bassi et al.). Use of an automated filtering process has also been suggested as a means of reducing discrepancies (Hasan et al.).

Shared electronic health records may reduce discrepancies, but medication lists have been found to have errors, and it has been suggested that both primary care and hospital doctors would need to have read/write access (Moore et al.). The views of primary care doctors in the USA on the best ways to provide medication lists in primary care record systems have been surveyed (Weeks et al.).

Quality improvement methods with a range of interventions were found to be effective in increasing the proportion of patients for whom reconciliation was performed within 24 hours of admission at a paediatric hospital in the USA (White et al.). A multispecialty children's integrated health care network in the USA made successful use of an electronic medical record (EMR)-based quality improvement intervention to improve reconciliation rates from zero to 71% over six years (Rappaport et al.).

Education of doctors was found to reduce the rate of discrepancies on admission to hospital (Chan et al.), and specific training of non-medical, non-pharmacy staff in medicines reconciliation was also found to be useful in the mental health field (Morrison).

Standardised documentation seems helpful, for example the use of standard forms (Bedard et al.) or a standardised list of questions (De Winter et al. (ii)). At a French hospital, revised paper documentation was introduced as an interim measure pending computerisation, in order to reduce medication discrepancies on admission (Prevost et al.). Standard processes for nurses to document medications have been tried in both home (DL Green et al.) and hospital (Tessier et al.) settings.

Numerous studies have shown benefits from involving pharmacists in medicines reconciliation (Corrigan et al., De Winter et al. (i), Galvin et al., Grimes et al. (i), Hatch et al. (i), Hellstrom et al., Kennedy and Gatewood, Nynke Eggink et al., Peyton et al., Richards et al., Steurbaut et al., CD Williams et al., S Williams et al.). This task calls upon the pharmacist’s clinical skills (Thomas). Specific examples in the UK include the Safer Patients Initiative (Ashley) and the use of the PREVENT tool by pharmacists to identify high-risk patients at admission (Barnett et al.).

The role of the patient is also important. Cancer patients have been shown to have a high level of knowledge of medications (Hwang et al.) and a study in Ireland suggested that patients and their carers were the best source of information on medicines at admission (Fitzsimons et al.), although a study in asthma patients showed only moderate agreement between patient self-reports and pharmacy claims data (Lim et al.). A study in France of concordance between patients’ and prescribers’ reports on the use of NSAIDs and coxibs indicated better concordance for specific or serious conditions (Fourrier-Reglat et al.). The health literacy of patients was found to be important in obtaining accurate information on admission (Mervanova et al.). A study of tetracycline exposure during pregnancy investigated the reliability of patients’ recall of medication use in interviews (Yau et al.).

In the context of who should take overall responsibility for maintaining an accurate medication list (Gandhi and Lee), it has been proposed that patients be accountable for this (Balon et al.) and that they should have a paper version of their medication list until it was available electronically (Frydenberg and Brekke). However, it appears that patient-held medication lists should be interpreted with caution (CF Green et al.) and another study found that while patients said that they found medication lists useful, they did not in fact improve their knowledge or adherence (Luca et al.). Information on drug allergies recorded on patient-held smart cards in Taiwan has been found to be incomplete (Hsu et al.).

Inclusion of data from a range of sources can be helpful, for example GP practice records (Nicholls and Wilcock). Use of information from community pharmacies (the MedCheck process) was found to improve pre-operative medication records (Leung et al.) and other studies have shown improved accuracy as a result of incorporating community pharmacy dispensing records (Glintborg et al.) or contacting community pharmacists (Perennes et al.).

Opportunities for medicines reconciliation in community pharmacy have been discussed (Johnson et al.) and a it has been suggested that home medicines review (as conducted in Australia) provides a further opportunity for reconciliation after discharge from acute care (Ellitt et al.). A study in Norway found a decrease in medication discrepancies in general practitioners’ electronic records after the implementation of multidose dispensing by pharmacies, although it is thought that this may be due to changes in workflow and greater focus on the medication process, rather than multidose dispensing per se (Wekre et al.)

Some Key Publications Before 2010

Technical patient safety solutions for medicines reconciliation on admission of adults to hospital

National Institute for Health and Clinical Excellence Patient Safety Guidance PSG001, Dec 2007

A systematic review of the effectiveness and cost-effectiveness of interventions aimed at preventing medication error (medicines reconciliation) at hospital admission

F Campbell, J Karnon, C Czoski-Murray, R Jones, University of Sheffield, School of Health and Related Research (ScHARR)

National Institute for Health and Clinical Excellence,Sep 2007, 75pp.

Medicines reconciliation: a guide to implementation

National Prescribing Centre, Mar 2008

Moving patients, moving medicines, moving safely: Guidance on discharge and transfer planning

Royal Pharmaceutical Society of Great Britain (RPSGB), Pharmaceutical Services Negotiating Committee, Primary Care Pharmacists’ Association, Guild of Hospital Pharmacists, Feb 2006

2010 and 2011 Papers on Medicines Reconciliation

How can effective medicines reconciliation be achieved?

M Ashley

Pharmacy Management, Jan 2010, vol. 26, no. 1, p. 3-7

This paper outlines the background and policy regarding medication histories, identifies the need for improvement in taking medication histories, describes a medicines reconciliation project at Musgrove Park Hospital, Taunton, that utilises the 'Safer Patients Initiative, phase 2' (SPI2) approach, and confirms outcomes of medicines reconciliation (MR) rates being consistently around 92%. (6 refs.)

Comparison of hospital admission medication lists with primary care physician and outpatient pharmacy lists

J Balon, SA Thomas

Journal of Nursing Scholarship Sep 2011;43(3):292-300

Purpose: Medication reconciliation is a process to reduce errors and harm associated with loss of medication information as the patient enters and moves through the healthcare system. This study examines medication list accuracy upon hospital admission.

Design: This prospective study enrolled 75 English-speaking medical and surgical patients (18 years of age or older) who were taking prescription medications. The study took place at a rural, tertiary teaching hospital in the northeastern United States. Data collection occurred from Nov 2006 to Mar 2009.

Methods: Nursing admission team medication lists were reconciled with primary care physician (PCP) and outpatient pharmacy (OP) lists. Outcome measures were accuracy of medication history generated by admission nurses (ANs) compared with PCP and OP lists, and identification of factors influencing probability of accurate medication list generation by ANs. The Generalized Estimating Equations modelling approach was used to compare AN, OP and PCP medication list accuracy. Additionally, sex and age were analysed as covariates and included in the model.

Findings:45 males and 30 females (N = 75) with a mean age of 60 years (SD 15) participated. 57subjects (76%) used over-the-counter or herbal medications, but the AN recorded only 31 (41%) cases. Patients received outpatient care from 1 to 12 providers. 40patients (67%) obtained medications from one pharmacy, 22 (29%) from two and 3 (4%) from three pharmacies. OP medication lists were completely accurate more often than PCP but not AN lists (19/75 (25%) OP vs 6/75 (8%) PCP vs 14/75 (19%) AN; 95% CI of the difference, 0.07 to0.50). No difference between AN and PCP list accuracy was found. Completely accurate AN lists were more than twice as likely with male and younger patients (95% CI of the difference, 1.07 to6.22 and 0.94 to0.99, respectively).

Conclusions: Like other studies, this study showed admission medication reconciliation lists are often inaccurate. Our results suggest that verification of admission medication lists with outpatient provider lists may improve accuracy. Patients, with guidance from outpatient care providers, should assume accountability for maintaining accurate medication lists. A secure, universal, interactive electronic medical record may be a future solution for organising and sharing medication data between providers.

Clinical Relevance: Medication reconciliation upon inpatient admission remains a high-volume and high-acuity problem. We found that not only hospital medication lists, but source lists, including those maintained by the patient, the PCP and the OP, are vastly inaccurate.

Medicines-related admissions: you can identify patients to stop that happening

N Barnett, D Athwal, K Rosenbloom

Pharmaceutical Journal 16 Apr 2011;286(7649):471-472

The Harrow Integrated Medicines Management Service was established in 2008 as a collaboration between Harrow Primary Care Trust and North West London Hospitals Trust to reduce preventable medicines-related problems and readmissions. As part of this service, the PREVENT tool was developed to identify patients at risk. Pharmacists use the tool when visiting patients on admission to a ward, as part of the consultation around medicines reconciliation. The PREVENT tool has helped to identify 147 patients considered to be at high risk of a medicines-related readmission.

Use of information technology in medication reconciliation: a scoping review

J Bassi, F Lau, S Bardal

Annals of PharmacotherapyMay 2010;44(5):885-897

Objective: To identify studies involving information technology (IT) in medication reconciliation (MedRec) and determine how IT is used to facilitate the MedRec process.

Data Sources: The search strategy included a database search of MEDLINE and Cumulative Index of Nursing and Allied Health Literature (CINAHL), hand-searching of collected material, and references from articles retrieved. The database search was limited to English-language papers. MEDLINE includes publications dating back to 1950 and CINAHL includes those dating back to 1982. The search included articles in both databases up to Mar 2009. Boolean queries were constructed using combinations of search terms for medication reconciliation, IT and electronic records.

Study Selection and Data Extraction: Three inclusion criteria were used. The study had to (1) involve the MedRec process, (2) be a primary study and (3) involve the use of IT. Selection was performed by 2 reviewers through consensus. Data related to study characteristics, focus and IT use were extracted.

Data Synthesis: The studies included described a range of IT used throughout the MedRec process, from basic e-mail and databases to specialised MedRec tools. A generic MedRec workflow was created and types of IT found in the studies were mapped to the workflow activities as well as to a set of functionalities based on the Institute of Medicine's Key Capabilities of an Electronic Health Record System. In the studies reviewed, IT was mainly used to obtain medication information. Although there were only a few MedRec tools in the studies, those that did exist supported the central activities for MedRec: comparison of medications and clarification of discrepancies.

Conclusions: MedRec is an important process to ensure patient medication safety. Evidence was found that IT can and has been used to facilitate some MedRec activities and new applications are being developed to support the entire MedRec process.