research paper

R E S E A R C H A R T I C L E

Impact of pharmacist-ledmedication reconciliation onadmissionusing electronic medical records on accuracy of discharge prescriptions Dona S. Lawrence1, Noman Masood1, Dawn Astles2, Claire E. Fitzgerald3, Ata Ul. Bari1

1 Pharmacy Department, Northern Beaches Health Service, Manly Hospital, Manly, Australia 2 Pharmacy Department, Northern Sydney Local Health District (NSLHD), Sydney, Australia 3 Pharmacy Department, Northern Beaches Health Service, Mona Vale, Australia

Abstract

Background: Unintentional medication errors are common at hospital discharge and have the potential to cause significant patient harm. Current electronic medical records systems offer the facility to change the process of medication reconciliation by pharmacists. Aim: To test the impact of pharmacist-led medication reconciliation at admission recorded on the electronic medication form, on the time- liness and accuracy of discharge prescriptions. Method:A prospective pre-andpost-interventionalobservationalstudywascarriedout fromJuneto October2013attwodistricthospitals. Pharmacists recorded admission medication using National Medication Management Plan (phase 1) and the electronic medication form in patients’ electronic medical records (phase 2). Data collected included time taken for the medical officer to complete the medication form in electronic medical records, the number of times the medical officer was contacted by the pharmacist completing the discharge rec- onciliation and unintentional discharge medication discrepancy types. Results: In total 118 patients were included: 66 patients in phase 1 and 52 in phase 2. Data were analysed using chi-squared test, Fisher’s exact test and Mann–Whitney test. There was a significant (33–13%, p < 0.0001) reduction in the proportion of medication orders with a discrepancy. This was because of the significant (25.5–1.9%, p < 0.0001) reduction in discrepancies relating to patients’ usual medication. Time taken for the medical officer to complete the medication form in electronic medical records decreased from 37 s/item (interquartile range, 29–48; n = 51) to 21 s/item (interquartile range, 11–35; n = 35) (p < 0.001). The number of telephone calls to medical officers decreased from 95 to 73%. Conclusion: This integrated approach to medication reconciliation has highlighted patient safety benefits, and reduced medical and pharmacy workload.

Keywords: medication, pharmacists, discharge.

INTRODUCTION

Medication reconciliation is a formal process of obtaining, verifying and documenting an accurate list of a patient’s current medicines at the time of admission and comparing this list with the admission, transfer and discharge orders, to identify and resolve the discrepancies.1

Prescribing at discharge remains a process with poten- tially significant risk for patient harm, and complete med- ication reconciliation lends itself to improving patient safety.1–4 A systematic review stated that patients with at least one medication discrepancy at hospital discharge var-

ied considerably from 25 to 80% and the proportion of med- ication orders with a discrepancy ranged from 8.4 to 16.3%. Of the 20 studies which reported the most frequent type of unintentional medication discrepancy, all but 3 reported that medication omission was the most frequent type iden- tified.5 Omission of medicine from the discharge summary list was associated with an increased risk (by a factor of 2.3) of hospital readmission or adverse medicine event.2

Previouslocalauditsundertakenattwodistricthospitals in the Sydney metropolitan area have shown that the risk of medication errors owing to incomplete medication recon- ciliation remains high at discharge with 68%6 and 73%7

of electronic discharges having at least one unintentional error. Errors of omission were the most frequent type, accounting for 45% in one study and the second most com- mon, accounting for 22% in the second study. Interestingly, handwritten medication history on admission (MHOA) forms completed by pharmacists were available to medical

Address for correspondence: Dona S. Lawrence, Pharmacy Depart- ment, Manly Hospital, 150 Darley Road, Manly, New South Wales 2095, Australia E-mail: [email protected]

Official Journal of the Society of Hospital Pharmacists of Australia

© 2015 Society of Hospital Pharmacists of Australia. Journal of Pharmacy Practice and Research (2015) 45, 166–173 doi: 10.1002/jppr.1091

officers for 65% of patients but it was observed that these forms had no impact on prescribing errors at discharge.6

Several Australian studies have investigated the impact of electronicdischargereferral (eDR)systems onthequality of medication prescribing at discharge. Garret and McCor- mack8 found that the number of discharges with one or moreprescribing errors decreasedsignificantly after imple- mentation of the eDR system (57.6 vs 34.8%, p < 0.001). Ng et al.9 reported the percentage of electronic discharge med- ication profiles with changes made after reconciliation ran- ged from 19 to 40%. Callen et al.10 found that 13.3% of electronic summaries contained medication errors with medication omission the most common type.

In the last few years, there has been considerable uptake of electronic solutions to the admission medication recon- ciliation process, largely owing to increasing infiltration of vendors using sophisticated medication history tools.11

The currently available electronic medical records (eMR) system implemented across Northern Sydney Local Health District offers the facility of maintaining medication forms by a pharmacist, in addition to medical staff. This medica- tion form is automatically imported by the medical officer into the electronic discharge summary. The local Sustaina- ble Access Committee at the hospital identified the discharge medication management process as a high prior- ity for clinical redesign. With the aim that, in addition to improving patient safety, a redesign could potentially improve patient flow and may support the organisation in meeting its National Emergency Access Target (NEAT).

To the best of our knowledge, no published studies com- pare the effect of pharmacist-documented best possible medication history (BPMH) recorded in the electronic med- ication form at admission, which is then used as the basis of the electronic discharge medication profile, on uninten- tional medication discrepancies at discharge and time for the medical officer to complete the medication form on the eDR. Our study aims to address this gap in knowledge by testing the impact of pharmacist-led medication recon- ciliation at admission recorded on the electronic medica- tion form, on the timeliness and accuracy of discharge prescriptions.

METHODS

This is a prospective pre- and post-intervention observa- tional study carried out over a 4-month period from June until October 2013 at two district hospitals. A 150-bed pub- lic hospital, comprising mostly general medical, general surgical and orthopaedic patients, serviced by 5.5 full-time equivalent pharmacists. Second, a 200-bed public hospital, comprising mostly general medical, general surgical and aged-care specialities, serviced by 5.6 full-time equivalent

pharmacists. Consecutive adult patients were included if they were reviewed by a pharmacist (working hours, 08:30–17:00 hours, Monday–Friday), on one or more med- ications, the BPMH was completed and recorded in the emergency department (ED) by the ED pharmacist, the patient was admitted to hospital, and the patient’s eDR was screened and reconciled by the ward pharmacist. Rec- onciliation by the clinical pharmacist was carried out as per the Society of Hospital Pharmacists of Australia Standards of Practice for the Provision of Medication Reconciliation.12

Phase 1 (Normal Process)

For the first 8 weeks (24 June 2013–20 August 2013), the ED pharmacist used the handwritten National Medication Management Plan when recording BPMH and reconciling with the patient’s National Inpatient Medication Chart (NIMC) see Figure 1. Data were collected from patients presenting to the EDs at the two hospitals, who were then admitted to hospital and discharged from a medical ward only. The ED pharmacist recorded the time taken to com- plete the medication history. Once completed, the pharma- cist selected ‘other’ in the NIMC for additional charts option, and annotated ‘Med History form complete’. The ED pharmacist recorded if there was a medication form in the patient’s eMR (Cerner Firstnet/Powerchart; Cerner, Kansas City, MO, USA) from a previous admission, or if the ED medical officer recorded a BPMH on the medication form in the patient’s eMR at admission. At discharge, data were collected on the following basis: time taken for the medical officer to complete the medication form on the electronic discharge, time taken for the clinical pharmacist to complete discharge reconciliation, number of times the medical officer was contacted by the clinical pharma- cist, thenumber of medication orderswith anunintentional discrepancy, and the number and type of discrepancies (each medication order may have more than one type of discrepancy).

Phase 2 (Intervention Process)

In the second 8 weeks (27 August 2013–24 October 2013), the ED pharmacist recorded the BPMH on the electronic medication form in eMR and reconciled with the patient’s NIMC (see Figure 1). Data were collected from patients presenting to the ED at one of the hospitals who were then admitted to the hospital and discharged from either the medical, surgical, emergency medical unit or orthopaedic wards. The ED pharmacist followed a locally approved procedure when entering the BPMH into the patient’s eMR and saved it to a pharmacy specific consult note. Once completed, the pharmacist selected ‘other’ in the NIMC for additional charts, and annotated ‘eMR Med History

Pharmacist-led medication reconciliation 167

© 2015 Society of Hospital Pharmacists of Australia. Journal of Pharmacy Practice and Research (2015) 45, 166–173

complete’. For this project, the ward pharmacists did not update the electronic medication chart throughout the patients hospital stay. All data collected were as per phase 1.

Clinical pharmacists completed an audit tool and classi- fied the discrepancies for the discharges they reviewed during their daily activities. A review of the published literature found considerable variation in the type of discrepancies or transcription errors identified on

discharge;8,10,13,14 therefore, our audit tool was developed based on the previous local audit tool and amended follow- ing consultation with clinical pharmacists and the medica- tion safety pharmacist. A description and examples of the discrepancies are shown in Table 1. For the purpose of the present investigation a discrepancy, or transcription error, was defined as any unintended inconsistency between the medication the patient should be discharged

Normal Process (Phase 1) Intervention Process (Phase 2)

On admission, pharmacist records a handwritten BPMH on the NMMP.

Pharmacist uses NMMP to reconcile medication on NIMC.

On discharge, MO create s a Medication Form on eMR. Medication from the most recent past admission will auto-populate the Medication Form (often very out of date). Date last updated is displayed in Medications Last Updated field at the bottom of the profile. MO refers to NMMP and NIMC before transferring relevant medication in to the Medication Form in eMR for current admission. Adds/amends all medications on admission (status = Medication continued – dose Unchanged).

On admission, pharmacist records BPMH on Medication Form in eMR and saves to a Pharmacy Consult note.

Pharmacist uses printed version of Medication Form in the Pharmacy Consult note to reconcile against NIMC.

On discharge, MO create s a Medication Form on eMR. Medication list from the pharmacists BPMH will auto-populate the Medication Form. MO refers to NIMC before transferring relevant medication in to the Medication Form in eMR for current admission.

MO continues to review medication for current admission. Adds any medications commenced this admission (status = New Medication). Lists any medications discontinued (status = Medication ceased). Lists any changes to dose (status = Dose increased or Dose reduced).

Once all medication r eviewed and updated, the Medication Form is imported in to the electronic discharge referral. MO saves as “interim”. Printed, signed by MO and handed to pharmacist for discharge reconciliation. Electronic medication profile can be updated again prior to discharge if necessary. Final version is imported into the electronic discharge summary and a final reconciliation and sign -off is completed by pharmacist before dispensing.

BPMH, best possible medication history; eMR , electronic Medical Records ; MO, medical officer; NIMC, National inpatient medication chart; NMMP, National medication management plan

Figure 1 Medication reconciliation process.

168 D. S. Lawrence et al.

Journal of Pharmacy Practice and Research (2015) 45, 166–173 © 2015 Society of Hospital Pharmacists of Australia.

with and the medication form in the eDR, identified by the clinical pharmacist, discussed and verified with the pre- scriberandreconciled bythechangebeing re-importedinto the final discharge summary byeither the medical officer or the pharmacist. The pharmacists timed themselves when recording the BPMH and reconciling the discharge referral. Time taken for the medical officer to complete the medica- tion form on the electronic discharge was noted by the investigational pharmacist by looking at the timeline on eMR.

Data Analysis

Categoricalvariableswerecomparedusingchi-squaretests or Fisher’s exact test when appropriate. Time taken to rec- ord BPMH is presented as median (interquartile range) and differences betweentimesinphase1andphase2werecom- pared using Mann–Whitney tests. Analysis was performed using Microsoft Office Excel 2007 and GraphPad Software QuickCalcs 2014. Phase 2 data were analysed for all wards and for medical wards only owing to the disparity in meth- odology used between phase 1 which collected data on medical ward discharges only and phase 2 which collected data on medical, surgical, emergency medical unit and orthopaedic ward discharges. Unintentional discrepancies at discharge were analysed as a whole and also further stra- tified into patient’s usual medication discrepancies (i.e. from BPMH at admission) and any new discrepancies

(i.e. medication changes made in the hospital; new medica- tion missing; frequency/dose/route of administration missing or incorrect for a new medication; or new changes to admission medication).

RESULTS

Data were collected from 66 patients in phase 1 and 52 patients in phase 2.

Admission Medication Reconciliation

The median number of items per BPMH was 11 (8.14) in phase 1 and 12 (8.14) in phase 2 (p = 0.94). The ED medical officers did not complete any medication forms in patient’s eMR at admission in phases 1 or 2. The percentage of med- ication forms in patient’s eMR from previous admissions was similar between phase 1 (64%) and phase 2 (65%) (p = 0.84). The median time for ED pharmacists to take and record a BPMH was 54 s/medication order (40,73) in phase 1 and 50 s/medication order (39,70) in phase 2 (p = 0.42).

Discharge Medication Reconciliation

AsseeninTable 2,thenumberofdischargeswithatleastone discrepancy was reduced by approximately one-third (97–65%, p < 0.0001). When comparing phase 1 and phase

Table 1 Discrepancy types and examples

Type of discrepancy Examplea

Omission (unintentional deletion of a drug)

Clopidogrel commenced in hospital following two transient ischaemic attacks. Not on discharge prescription.

Commission (unintentional addition of a drug)

Amoxycillin/clavulanic acid 875 mg/125 mg for 3 days and doxycycline for 3 days on 2013 discharge prescription. Carried over from the 2011 discharge summary in error.

Incorrect or missing dose Cholecalciferol dose increased in hospital as vitamin D level was 46 nmol/L (50–130 nmol/L), but discharge prescription was for pre-admission dose.

Incorrect or missing frequency Clindamycin 300 mg twice daily on discharge prescription. Increased to 8 hourly after pharmacist intervention.

Stopped medication not listed under the ceased section

Buprenorphine patch ceased in hospital and patient commenced on oxycodone SR. Buprenorphine patch still listed as current medication on discharge prescription.

Length of therapy missing or inappropriate

Length of antibiotic course on discharge prescription missing on numerous occasions.

Route of administration incorrect or missing

Tiotropium 18 micrograms listed as oral on discharge prescription instead of inhalation puffer.

Ceased (medication incorrectly listed as ceased)

Fentanyl12 micrograms/h patchandtramadolSR100 mglisted asceasedmedication onthedischarge prescription. The patient was not on either medication at hospital admission. The entry was carried over from an old discharge summary.

Status incorrect Levodopa/carbidopa 100 mg/25 mg listed as ‘medication ceased’ on the discharge prescription when it should have been ‘dose increased’.

Formulation/dosage form Diltiazem 180 mg daily on discharge prescription. Slow-release preparation was not specified.

a Examples were obtained from this study.

Pharmacist-led medication reconciliation 169

© 2015 Society of Hospital Pharmacists of Australia. Journal of Pharmacy Practice and Research (2015) 45, 166–173

2 discharges, there was a significant reduction in the propor- tion of medication orders with a discrepancy (33–13%, p < 0.0001) and a significant reduction in total discrepancies per medication order (36–14%, p < 0.0001). The latter was calculated because medication orders may have more than one discrepancy type. When comparing phase 1 and phase 2 discharges for medical wards only, there was a significant reduction in the proportion of medication orders with a dis- crepancy (33–21%, p = 0.005) and a significant reduction in totaldiscrepanciespermedicationorder(36–23%,p < 0.001).

Figure 2 shows the distribution of unintentional discre- panciesondischargeprescriptions.Thesewerefurtherstra- tified into patients’ usual medication discrepancies, i.e. from BPMH at admission (Figure 3), and new discrepan- cies, i.e. new changes made in hospital (Figure 4).

When comparing phases 1 and 2, there was a significant reduction in patients’ usual medication discrepancies on discharge prescriptions (25.5–1.9%, p < 0.0001). In phase 1, the most frequent discrepancy types on discharge for patients’ usual medication were frequency (52/216), status (48/216) and omission (47/216) compared with errors of omission (5/12) in phase 2.

There was no statistically significant difference in new discrepancies on discharge prescriptions between phases 1 and 2 (11–12%, p = 0.66). In phase 1, the most frequent new discrepancy type was newly stopped medication not listed as ceased (28/93) compared with omission (16/75), frequency (14/75) and dose (13/75) in phase 2. Omission was the most common unintentional discharge medication discrepancy overall (80/396 discrepancies).

Time on Discharge

The median time for medical officers to complete the medication form inthe patient’s eMRfor the discharge refer- ral was almost halved between phases 1 and 2 (37 s/

medication order (interquartile range (IQR), 29–48) in phase1(n = 51),21 s/medicationorder(IQR,11–35)inphase 2 (n = 35) (p < 0.001) and 20 s/medication order (IQR,16–23) in phase 2 medical ward only (n = 11) (p < 0.001)). The num- ber of times the medical officer was contacted to sort out dis- crepancies in the discharge prescription was reduced by almost a quarter (95% in phase 1 to 73% in phase 2, p < 0.001). The median time for the clinical pharmacist to recon- cile the discharge referral was 103 s/medication order (IQR, 54–150) in phase 1 (n = 63) and 81 s/medication order (IQR, 55–136) in phase 2 (n = 46) (p = 0.36).

DISCUSSION

There are many international studies13–17 and several Aus- tralian studies8–10 describing eDR or medication manage- ment systems and the effect on discharge medication

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%

100%

P e

rc e

n ta

g e

Discrepancy Type

Phase 2 (n=87) Phase 1 (n=309)

Figure 2 Distribution oftotal discrepancies ondischargeprescriptions.

Table 2 Discrepancies at discharge

Phase 1 (n = 66

discharges)

Phase 2 (n = 52

discharges) p Value (phase 1

vs phase 2) Phase 2 medical ward only

(n = 17 discharges) p Value (phase 1 vs phase 2

medical wards only)

Discharges with at least one discrepancy

64 (97%) 34 (65%) <0.0001 14 (82%) 0.056

Number of medication orders for reconciliation

847 641 211

Mean (±SD) medication orders/ discharge

12.8 (±5.1) 12.3 (±4.6) 12.4 (±4.4)

Medication orders with discrepancy (%/medication order)

278 (33%) 83 (13%) <0.0001 44 (21%) 0.005

Total discrepancies (%/medication order)

309 (36%) 87 (14%) <0.0001 48 (23%) <0.001

170 D. S. Lawrence et al.

Journal of Pharmacy Practice and Research (2015) 45, 166–173 © 2015 Society of Hospital Pharmacists of Australia.

errors. To the best of our knowledge, this is the first pro- spective study to describe the impact of pharmacist record- ing BPMH on the electronic medication form at the time of admission, on the rate of unintentional discharge medi- cation discrepancies and time for doctor to complete the medication form at discharge. Previous studies pub- lished involve doctors manually typing the medications into the medication form at the time of discharge. In this study, the ED pharmacist has already entered the BPMH at the time of admission on the medication form which is automatically imported into the discharge referral, and the doctor on discharge only needs to annotate any changes or new medication, avoiding duplication of work.

Within this context, the implementation of the change in medication reconciliation process resulted in a significant 32% absolute reduction in the number of discharges with at least one unintentional discrepancy and a significant 20% absolute reduction in the proportion of medication orders atdischarge with anunintentional discrepancy. This indicates that pharmacists entering BPMH electronically improves discharge medication accuracy and potential safety. This reduction in unintentional discrepancies at dis- charge was due to a significant reduction in patients’ usual medication discrepancies (i.e. from BPMH at admission), with no difference in new discrepancies (i.e. medication changes made in hospital).

The proportion of discharges with at least one discrep- ancy (97% in phase 1 and 65% in phase 2) is higher than that of previous studies describing a frequency of 25–80%.5

Marked variation in the frequency reported in published studies is likely due to methodological differences, types of discrepancies collected, and inconsistencies in the inter- pretation of ‘error’ which is a major limitation in previous trials as well as in this study. The higher frequency can be explained as we collected discrepancies on more para- meters than previous local audits6,7 and published studies,8,10,13,14 including status incorrect (see Figure 1), formulation/dosage form, and medication incorrectly listed as ceased. In keeping with previous studies,5 error of omission was the most commonly reported discrepancy in this study. Other limitations include the lack of an inde- pendent/multidisciplinary panel to review reports and risk of subjectivity in the application of error classification, low numbers, therefore finding could be due to chance, and data could only be collected during pharmacy working hours (08:30–17:00 hours, Monday–Friday).

Another major limitation of the study was the disparity in methodology followed between phases 1 and 2. Phase 1 collected data across two hospitals and unintentional med- ication discrepancies on medical ward discharges thereby involving two-ward pharmacists only, reducing the time required for training and the chance of inconsistencies in interpretation of discrepancies. Phase 2 collected data at one hospital site which was unavoidable owing to phar- macy staffing issues in the ED and therefore had to include discharges from medical, surgical, emergency medical unit and orthopaedic wards to complete the project in a reason- able time frame. A previous local audit7 found that the emergency medical unit and orthopaedic ward had the highest frequency of medicationdiscrepancies ondischarge; therefore, looking at medical ward only in phase 1 should not bias the results. Despite this, we analysed the results of the medical ward discharges in phase 2 separately which still showed a significant reduction in discrepancies. The small number of patients in phase 2 medical ward only (n = 17) may limit the generalisability of our results.

82%

84%

86%

88%

90%

92%

94%

96%

98%

100%

P e

rc e

n ta

g e

Discrepancy Type

Phase 2 (n=12) Phase 1 (n=216)

Figure 3 Distribution of patient’s usual medication discrepancies on discharge prescriptions.

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%

100%

P e

rc e

n ta

g e

Discrepancy Type

Phase 2 (n=75) Phase 1 (n=93)

Figure 4 Distribution of new discrepancies on discharge prescriptions.

Pharmacist-led medication reconciliation 171

© 2015 Society of Hospital Pharmacists of Australia. Journal of Pharmacy Practice and Research (2015) 45, 166–173

The potential clinical significance of the discharge med- ication discrepancies was not evaluated and is a limitation of this study. A systematic review on the impact of medica- tion reconciliation and review on clinical outcomes found evidence to demonstrate that medication reconciliation hasthepotentialtoidentifymanymedicationdiscrepancies and reduce potential harm, but the impact on actual clinical outcomes, such as reductions in hospital readmissions, is less clear.5 Of all the medication discrepancies that were assessed in the systematic review for their reported poten- tial impact on clinical outcomes, 18.4 –80.5% had the poten- tial to cause no or minimal harm, 17.9–78.1% had the potential to cause moderate harm, and 0–24.3% had the potential to cause severe or life-threatening harm.5 One study found a significant reduction in the 30-day rate of hospital readmission (17.8–12.3%, p = 0.042) in high-risk patients, after implementation of a team-based medication reconciliationattransitionsofcare,whichcouldyielddirect cost savings.18

The change in the medication reconciliation process resulted in a reduction in medical staff workload, shown by a 22% absolute reduction in the number of telephone calls to medical officers and a decrease from 37 s/medica- tion order (IQR, 29–48; n = 51) to 21 s/medication order (IQR, 11–35; n = 35) in the time taken for the medical officer tocompletetheelectronicmedicationform.Thelatterresult has limitations as medical officers were not aware that the timeline in the patient’s eMR would be used to determine how long they took to complete the medication form and they may have been interrupted. Also, for almost a quarter of the discharges in phase 1 (15/66) and a third of dis- charges in phase 2 (17/52), we could not determine the time taken by the medical officers to complete the medication form on the patient’s eMR as it had been opened on multi- ple occasions and/or over multiple days. Despite this, a studyin2011showed similarresults:pharmacytechnicians entering electronic medication histories on admission reduced the average time spent by doctors completing the discharge paperwork from an average of 30 to 23 min per discharge (p = 0.007).19

The observeddifference intime takenbytheED pharma- cists to take and record a BPMH when completing a hand- written National Medication Management Plan and electronic recording on a medication form in eMR was found to benon-significant. Pharmacists recordingmedica- tionhistoriesineMRandreconciling theprintedcopy ofthe electronic BPMH against the NIMC on admission is a rela- tively new and slightly different process. Although some initial delays were expected, they were not observed in this study. It is expected as the data repository expands over time and voice activation devices being currently trialled in our ED have increased the uptake, the electronic process will, in the long run, be significantly less time-consuming.

The change in process potentially results in more efficient use of pharmacist time, especially for patients seen by ED pharmacists who are discharged or transferred to another hospital, as the electronically recorded BPMH can be used by another hospital or used locally if the patient is readmitted at a later date. It also resulted in a clear and concise electronic medication reconciliation form that is printed and available at the bedside with the NIMC, as well as electronically avoiding poor hand- writing issues.

The difference in time taken for the clinical pharmacist to reconcile the discharge medication was non-significant despite the lower rate of medication discrepancies to resolve and fewer telephone calls to the medical officers in phase 2.

Although this integrated approach to medication recon- ciliation has highlighted clear patient safety benefits while reducing clinician (medical and pharmacy) work- load there is still a long way to go to reduce unintentional medication discrepancies at discharge from hospital to zero. Unfortunately, prescribing on discharge remains a process with ongoing potential significant risk for patient harm and pharmacist discharge reconciliation, and collab- oration with prescribers remains an important safety measure.

The modified process tested in this study has the poten- tial to be the basic framework for the introduction of a med- ication reconciliation component in any future integrated electronic medication management solution. This study highlights the importance of enhanced multidisciplinary focus on medication reconciliation, ongoing education of the medical officers and better utilisation of eMR to reduce errors further. As medication reconciliation is a mandatory criteria for EQuIP20 National Accreditation Survey 4, National Safety and Quality in Healthcare Standards (crite- rion 4.6), it is important for organisations to adopt system- based approaches that allow them to benchmark and mon- itor the progress of this important safety indicator. Elec- tronic recording of this information can provide an opportunity to have access to real-time data related to med- ication reconciliation as well as to reduce the potential bur- den of labour-intensive auditing.

ACKNOWLEDGEMENTS

The author thanks contributing pharmacists at the Northern Beaches Health Service for support with data col- lection. In addition, the author thanks Jillian Patterson (Biostatistician, Kolling Institute, Royal North Shore Hospital, St Leonards) for guidance and assistance with data analysis.

172 D. S. Lawrence et al.

Journal of Pharmacy Practice and Research (2015) 45, 166–173 © 2015 Society of Hospital Pharmacists of Australia.

Competing interests

None declared.

REFERENCES

1 Duguid M. The importance of medication reconciliation for patients and practitioners. Aust Prescr 2012; 35: 15–19.

2 Australian Pharmaceutical Advisory Council (APAC). Guiding principles to achieve continuity in medication management. APAC Secretariat: Canberra; 2005.

3 Australian Commission on Safety and Quality in Health Care. Medication reconciliation. Available from <www.safetyandquality. gov.au/our-work/medication-safety/medication-reconciliation>. Accessed 19 October 2014.

4 Thompson-Moore N, Liebel MG. Health care system vulnerabilities: understanding the root causes of patient harm. Am J Health Syst Pharm 2012; 69: 431–6.

5 Lehnbom E, Stewart MJ, Manias E, Westbrook JI. Impact of medication reconciliation and review on clinical outcomes. Ann Pharmacother 2014; 48: 1298–312.

6 Fitzgerald C. Local audit. Discharge medication reconciliation: is the view any better at the far end? Mona Vale: Northern Beaches Health Service, Mona Vale Hospital; 2011.

7 Sumpter M. Local audit. Northern Beaches Health Service, Manly Hospital; 2012.

8 Garrett T, McCormack C. Does an electronic discharge referral system improve the quality of medication prescribing? J Pharm Pract Res 2014; 44: 29–34.

9 Ng C, Welch SA, Luddington J, Bui D, Glasson E, Richardson KL. Medication reconciliation challenges at discharge from hospital using anelectronic medicationmanagementsystemandelectronic discharge summaries. J Pharm Pract Res 2013; 43: 25–8.

10 Callen J, McIntosh J, Li J. Accuracy of medication documentation in hospital discharge summaries: a retrospective analysis of medication transcription errors in manual and electronic discharge summaries. Int J Med Inform 2010; 79: 58–64.

11 Forum AP. Report of the 2013 AMCP partnership forum on electronic solutions to medication reconciliation and improving transitions of care. J Manage Care Pharm 2014; 20: 937–47.

12 Burridge N. SHPA standards of practice for the provision of mediation reconciliation. J Pharm Pract Res 2007; 37: 231–3.

13 Wong JD, Barcar M, Wong CG, Alibhai SM, Huh JH, Cesta A, et al. Medication reconciliation at hospital discharge: evaluating discrepancies. Ann Pharmacother 2008; 42: 1373–9.

14 Cornu P, Steurbaut S, Leysen T, Baere ED, Ligneel C, Mets T, et al. Discrepancies in medication information for the primary care physician and the geriatric patient at discharge. Ann Pharmacother 2012; 46: 983–90.

15 Stewart AL, Lynch KJ. Identifying discrepancies in electronic medical records through pharmacist medication reconciliation 2012. J Am Pharm Assoc 2012; 52: 59–66.

16 Cornu P, Steurbaut S, Leysen T, De Baere E, Ligneel C, Mets T, et al. Effect of medication reconciliation at hospital admission on medication discrepancies during hospitalization and at discharge for geriatric patients. Ann Pharmacother 2012; 46: 484–94.

17 Grimes TC. Medication details documented on hospital discharge: cross-sectional observational study of factors associated with mediation non-reconciliation. Br J Clin Pharmacol 2011; 71: 449–57.

18 Anderegg SV, Wilkinson ST, Couldry RJ, Grauer DW, Howser E. Effects of a hospitalwide pharmacy practice model change on readmission and return to emergency department rates. Am J Health Syst Pharm 2014; 71: 1469–79.

19 Sherer AP, Cooper JB. Electronic medication reconciliation: a new prescription for the medications matching headache. Heart Lung 2011; 40: 372–3. 7th Annual Conference of the American Association of Heart Failure Nurses, AAHFN 2011, July-August 2011, Seattle, WA.

20 The Australian Council on Healthcare Standards. Available from <www.achs.org.au/programs-services/equipnational/>. Accessed 15 January 2015.

Received: 21 May 2014 Revised version received: 31 January 2015 Accepted: 19 February 2015

Pharmacist-led medication reconciliation 173

© 2015 Society of Hospital Pharmacists of Australia. Journal of Pharmacy Practice and Research (2015) 45, 166–173

Copyright of Journal of Pharmacy Practice & Research is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.