research paper

Contents lists available at ScienceDirect

Research in Social and Administrative Pharmacy

journal homepage: www.elsevier.com/locate/rsap

Impact of medication reconciliation on health outcomes: An overview of systematic reviews A.B. Guisado-Gila,b,∗, M. Mejías-Truebaa, E.R. Alfaro-Laraa, M. Sánchez-Hidalgob, N. Ramírez-Duquec, M.D. Santos-Rubiod a Unidad de Gestión Clínica Farmacia. Hospital Universitario Virgen del Rocío, Sevilla, Spain b Departamento de Farmacología. Facultad de Farmacia, Universidad de Sevilla, Sevilla, Spain c Unidad de Gestión Clínica Medicina Interna. Hospital Universitario Virgen del Rocío, Sevilla, Spain d Unidad de Gestión Clínica Farmacia. Hospital Juan Ramón Jiménez, Huelva, Spain

A R T I C L E I N F O

Keywords: Systematic review Medication reconciliation Outcome assessment Evidence-based practice

A B S T R A C T

Background: Recent systematic reviews and meta-analyses suggest that medication reconciliation (MR) is ef- fective in decreasing the risk of medication discrepancies. Nevertheless, the association between MR and sub- sequent improved healthcare outcomes is not well established. Objectives: This systematic review of reviews set out to identify published systematic reviews on the impact of MR programs on health outcomes and to describe key components of the intervention, the health outcomes assessed and any associations between MR and health outcomes. Methods: PubMed, EMBASE, Cochrane Library, Cumulative Index to Nursing and Allied Health Literature (CINAHL) and SCOPUS were searched from inception to May 2019. Systematic reviews of all study designs, populations, intervention providers and settings that measured patient-related outcomes or healthcare utiliza- tion were considered. Methodological quality was assessed using A Measurement Tool to Assess Systematic Reviews 2 (AMSTAR 2). Two investigators performed study selection, quality assessment and data collection in- dependently. Results: Five systematic reviews met the inclusion criteria: 2 were rated as low quality and 3 as critically low quality. Reviews included primary studies in different settings (hospitals, the community and residential aged care facilities) that reported the impact of MR on mortality, length of stay, Emergency Department (ED) visits, readmissions, physician visits and healthcare utilization. Only one review reported results on mortality. However, healthcare utilization, which usually included ED visits and readmissions, was communicated in all reviews. Meta-analyses were conducted in all reviews except one. Medication reconciliation was not consistently found to be associated with improvements in health outcomes. Conclusions: Few systematic reviews support the value of MR in achieving good patient-related outcomes and healthcare utilization improvements. The quality of the systematic reviews was low and the primary studies included commonly involved additional activities related to MR. There was no clear evidence in favor of in- tervention in mortality, length of stay, ED visits, unplanned readmissions, physician visits and healthcare uti- lization.

Introduction

Care transitions are described as changes in care settings. Poor quality transitions may result in risks to patients’ safety, discontinuity in care plans and patient dissatisfaction with care.1 Therefore, transi- tional care interventions encourage positive healthcare goals,2 require coordination with healthcare professionals in both primary and sec- ondary care, and provide patients with accessible information on post-

transition.3 In this context, the process of medication reconciliation should account for any alteration made in the medication taken by patients, and should ensure that patients or their caregivers have been made aware of these alterations.4

According to the Institute for Healthcare Improvement, medication reconciliation (MR) refers to the process of identifying the most accu- rate list of all medications a patient is taking and using this list to provide correct medications for patients anywhere within the health

https://doi.org/10.1016/j.sapharm.2019.10.011 Received 13 July 2019; Received in revised form 1 October 2019; Accepted 15 October 2019

∗ Corresponding author. Unidad de Gestión Clínica Farmacia. Hospital Universitario Virgen del Rocío, Avenida Manuel Siurot, 41013, Sevilla, Spain. E-mail address: [email protected] (A.B. Guisado-Gil).

Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx

1551-7411/ © 2019 Elsevier Inc. All rights reserved.

Please cite this article as: A.B. Guisado-Gil, et al., Research in Social and Administrative Pharmacy, https://doi.org/10.1016/j.sapharm.2019.10.011

system.4 This review process identifies medication discrepancies. Un- intended medication discrepancies that represent reconciliation errors are responsible for more than half the medication errors occurring during transitions in care, and up to one-third could potentially cause harm.5

Previous primary research studies have evaluated the effect of MR on medication discrepancies, patient-related outcomes and healthcare utilization during care transitions. However, interpreting the evidence in relation to the impact of MR is a challenge due to the variation in study designs, interventions and settings. In recent years, a growing number of systematic reviews and meta-analyses relevant to the impact of MR on health outcomes have been published. Their results suggest that MR provided by pharmacists is effective in decreasing the risk of medication discrepancies.6–8 Nevertheless, the association between MR and the improvement in health outcomes, while plausible, is not well established. An overview of systematic reviews can play a role in summarizing existing research or highlighting the absence of evidence, improving access to specific information and supporting decision- making by clinicians, policy makers and developers of clinical guide- lines.9 The Cochrane Collaboration recommends an overview of sys- tematic reviews to summarize the evidence of existing systematic re- views that address different outcomes for a single intervention.10

A previous overview of systematic reviews has measured the impact of MR on health outcomes.11 This overview, conducted by Holte et al. and published in 2015, considered health-related outcomes (read- missions, adverse events and unwanted events), but also outcomes re- lated to the process of performing medication reconciliation (percen- tage performing medication reconciliation and medication discrepancies). It included seven moderate-quality systematic reviews which concluded that MR probably reduces the number of medication discrepancies, but that the impact on clinical outcomes is unclear. Neither publication specifically focuses on the impact of MR on patient- related outcomes and healthcare utilization. Moreover, this overview did not consider recent systematic reviews of MR.

The objectives of this overview were to identify published sys- tematic reviews on the impact of MR programs on health outcomes and to describe key components of the intervention, the health outcomes assessed and any associations between MR and health outcomes.

Methods

Eligibility criteria

This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.12 The completed PRISMA checklist is included as Supplementary File 1. Inclusion criteria according to PICOS (Popula- tion, Intervention, Comparison, Outcome and Study design) for the systematic review were:

Population: adults and/or pediatric patients experiencing a

transition of care. Intervention: medication reconciliation. The intervention involved a healthcare professional and was performed before, during or after a care transition. Comparison: a control group that received usual care. Outcomes (at least one of the following): patient-related outcomes (mortality) and healthcare utilization (length of stay, unplanned readmissions, Emergency Department visits and/or primary or sec- ondary care visits). Study design: systematic reviews.

We excluded reviews investigating additional interventions not fo- cused on MR. Those exclusively reporting other outcomes (medication discrepancies, adverse drug events [ADEs] with potential to cause in- jury, preventable ADEs, medication adherence and unanticipated in- creased workload) were also excluded.

Information sources

An electronic literature search was performed using 5 Healthcare Databases: PubMed, EMBASE, Cochrane Library, Cumulative Index to Nursing and Allied Health Literature (CINAHL) and SCOPUS, with no language or publication date restrictions up to 30 September 2018. Search terms included a mixture of MeSH terms and free text (key- words, synonyms and word variations) combined with Boolean opera- tors. Filters were used to limit the results of the search to systematic reviews (see Table 1 for the complete search strategy). The last full search was run on 1 May 2019 in order to identify new results.

Study selection

Two independent reviewers (ABGG and MMT) screened the titles and abstracts of all eligible reviews for possible inclusion, with any disagreements settled by consensus or with a third reviewer (ERAL). Where there was uncertainty, full-length publications were evaluated before a final decision on inclusion was made.

Quality assessment

The quality-assessment tool known as A Measurement Tool to Assess Systematic Reviews 2 (AMSTAR 2)13 was used to assess the quality and risk of bias in the studies included. It consists of 16 items whose re- sponse options were “yes”, “no” or “partial yes”. Items 4, 7, 11, 13 and 15 are considered critical domains. The overall quality can be rated as high (no or one non-critical weakness), moderate (more than one non- critical weakness), low (one critical flaw with or without non-critical weakness), and critically low (more than one critical flaw with or without non-critical weakness). Two independent reviewers (ABGG and MMT) conducted the quality assessment, and any disagreements on quality ratings between reviewers were discussed and a consensus reached.

Table 1 Complete search strategy for different databases.

Healthcare Databases Search strategy

PubMed (“medication discrepancies” [All Fields] OR “reconciliation discrepancies” [All Fields] OR (“medication reconciliation” [MeSH Terms] OR “medication reconciliation” [All Fields])) AND ((“impact” [All Fields] OR “health outcomes” [All Fields] OR “Health Impact Assessment” [Mesh] AND systematic [sb]))

EMBASE (“health outcomes”/exp OR “health outcomes”) AND (“medication therapy management”/exp OR “medication therapy management”) AND [systematic review]/lim

Cochrane Library “effect”:ti,ab, kw and “reconciliation”:ti,ab,kw Limits: Cochrane Reviews

CINAHL AB (medication reconciliation) AND (AB health outcomes OR AB impact or effect or influence or AB outcome or result or consequence) Type of publication: systematic review

SCOPUS (impact OR effect OR health AND outcome) AND (medication AND reconciliation OR reconciliation AND discrepancies) AND (LIMIT-TO (DOCTYPE,“re”)) AND (LIMIT-TO (SUBJAREA,“HEAL”))

A.B. Guisado-Gil, et al. Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx

2

Data collection

Reviewer ABGG independently extracted data and MMT examined all extraction sheets to ensure their accuracy. We also directly com- municated with the authors to obtain details not included in the pub- lished reports. If there were any missing data from a review, we ex- plicitly stated this. For each systematic review, the following variables were registered:

- Author and year of publication. - Aim of systematic review. - Number of primary studies. - Design of primary studies: randomized controlled trial (RCT), pro-

spective controlled trial, before-and-after, cohort study, observa- tional post-intervention, retrospective observational and cross-sec- tional studies.

- Number of participants. - Type of participants: adults and/or pediatric patients. - Settings and transitions of care involved. - Intervention providers: pharmacists, nurses and/or physicians. - Key components of intervention. - Health outcomes: mortality, length of stay, unplanned readmissions,

Emergency Department visits and primary and secondary care visits. - Other outcomes collected in systematic reviews: medication dis-

crepancies, ADEs with potential to cause injury, preventable ADEs, medication adherence and unanticipated increased workload.

Both narrative findings and meta-analyses of primary study data included in the systematic reviews were synthesized. The measures of association between MR and health outcomes included the risk ratio (RR) and difference in means (MD), with consistency (I2) reported by individual reviews and meta-analyses.

Results

The electronic search revealed 86 citations; 8 were removed using Mendeley via duplicate checking. Additionally, 56 reviews were ex- cluded following title and abstract filtering because they did not meet the eligibility criteria. This left 22 reviews that were potentially re- levant and retrieved in full-text: 17 were excluded before data extrac- tion and 5 met the inclusion criteria (Fig. 1). A list of the 17 publica- tions excluded after full-text evaluation and the reasons for exclusion are provided in Supplementary File 2.

Quality of the systematic reviews

Table 2 reports on the AMSTAR 2 response option for each domain. The overall quality of the included studies assessed with AMSTAR 2 was poor. Of the 5 reviews, none was rated as high or moderate quality, 2 were rated as low quality,15,17 and 3 were critically low quality.7,14,16

All reviews presented similarities with respect to responses in cri- tical and non-critical domains, except for Lehnbom et al.14 who did not carry out a meta-analysis. Regarding non-critical flaws, none showed that they had worked with a written protocol with independent ver- ification before the review was undertaken (item 2), or documented the funding sources for each study included in the review (item 10), except Redmond et al.17 As for critical flaws, none provided a complete list of potentially relevant studies with justification for the exclusion each one (item 7). Redmond et al.17 only included a selection of their excluded articles. In addition, Lehnbom et al.14 did not assess the risk of bias (RoB) in individual studies or include a discussion of its impact on the interpretation of the results (item 13) and, together with the other two critically low quality reviews,7,16 did not perform a sensitivity analysis to determine publication of bias (item 15).

Characteristics of included systematic reviews and meta-analyses

Full details of the included studies are shown in Table 3. All the review articles7,14–17 aimed to identify MR interventions and to test their association with clinical outcomes. One of them14 also evaluated separately the effectiveness of medication review. They included vari- able numbers of primary studies: Kwan et al.7 18 studies, Lehnbom et al.14 40 studies, McNab et al.15 14 studies, Cheema et al.16 18 studies and Redmond et al.17 25 studies. The 5 reviews cited a combined total of 87 original research articles: 40 RCTs, 4 prospective controlled trials, 7 before-and-after, 9 cohort studies, 22 observational post-intervention, 3 retrospective observational and 2 cross-sectional studies. In total, 65912 patients were studied, with people recruited from hospitals, from the community and from residential aged care facilities (RACF). Despite contacting the authors, McNab et al.15 were unable to provide data on the number of patients enrolled and the number of participants under 18, in two studies included in their review.

Pharmacists were primarily responsible for delivering the inter- vention, working closely with other healthcare professionals (physi- cians and nurses) in some cases.7,14,17 Interventions beyond MR in- cluded patient counselling,7,14,16,17 creation of post-discharge medication lists,7,14,16,17 post-discharge communication7,14,17 and medication review.17 Reviews reported the control group's intervention to consist of usual care in the context in which the study took place, except in two studies18,19 included in Kwan et al.'s publication,7 which compared two forms of MR.

The reviews included a combined total of 37 primary research studies that reported the impact of MR on different health outcomes: mortality,17 length of stay,14,17 Emergency Department visits,14,15,17

readmissions14,15,17 and physician visits,14,15 without restrictions on follow-up periods. All reviews7,14–17 also studied the impact of MR on healthcare utilization, a composite variable made up of two or more health measures. Meta-analysis was conducted except in one case14 if available data allowed pooling of results, and the variables measured were homogenous. Additional outcomes identified were: medication discrepancies,14–17 ADEs with potential to cause injury,7,16 preventable ADEs,16,17 medication adherence17 and primary care workload.15

Effectiveness of medication reconciliation on health outcomes (see Table 4)

Mortality One review17 reported mortality with no statistically significant

differences between MR and standard care (RR 0.75, 95% CI 0.27 to 2.08) based on the results of one RCT.

Length of stay Two reviews14,17 reported no statistically significant differences

between MR in hospitals and standard care regarding length of stay. Pooled results of 2 RCTs included in the meta-analysis performed by Redmond et al.17 were MD 0.48 (95% CI -1.04 to 1.99) with some evidence of heterogeneity between these studies (I2 = 52%; p = 0.15). Medication reconciliation in RACF demonstrated significantly shorter hospital stays (p = 0.026) according to one before-and-after study in- cluded by Lehnbom et al.14

Emergency Department visits In relation to the hospital setting, Lehnbom et al.14 communicated

no statistically significant differences regarding ED visits within 72 h, 14 days or 30 days after discharge based on the results of one before- and-after study, whereas Redmond et al.17 included the results of one small RCT with a moderate risk of bias that reported reduced ED visits within 30 days after discharge (RR 0.07, 95% CI 0.00 to 1.07). On the other hand, McNab et al.15 included 2 non-RCTs with no difference observed between groups, indicating no clear effect of MR on ED visit rates.

A.B. Guisado-Gil, et al. Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx

3

Unplanned readmissions Two reviews14,17 communicated no statistically significant differ-

ences regarding unplanned rehospitalization rates in the hospital set- ting. The pooled results of 5 RCTs included in the meta-analysis per- formed by Redmond et al.17 were RR 0.72 (95% CI 0.44 to 1.18) with a follow-up range of 5–30 days and some evidence of heterogeneity be- tween these studies (I2 = 45%; p = 0.12). In the community setting, Lehnbom et al.14 included one cohort study where the readmissions rate decreased at 7 (p = 0.01) and 14 days (p = 0.04) but not at 30 days (p = 0.29) after discharge. McNab et al.15 included 7 studies in the meta-analysis (4 RCTs and 3 cohort studies), and the pooled RR was 0.91 (95% CI 0.66 to 1.25) with high heterogeneity (I2 = 71%; p = 0.002). The follow-up period was 30 days, except in 2 RCTs in which it was 6 months.

Physician visits One prospective controlled study in the community setting included

by Lehnbom et al.14 found that control patients had a lower rate of discrepancy resolution and reported no significant trend towards more planned and unplanned physician visits compared with intervention patients. One RCT in McNab et al.'s review15 reported a significant increase in general practitioner visits of 43% (p = 0.002) in the MR group, while another RCT reported no significant difference at 6 months.

Healthcare utilization Healthcare utilization after hospital discharges usually included ED

visits and readmissions. In this respect, reported data from 4 RCTs in- cluded in Cheema et al.'s meta-analysis16 showed no significant

reduction in favor of the MR group, with a pooled RR of 0.78 (95% CI 0.61 to 1.00) and no heterogeneity between the studies (I2 = 0%; p = 0.54). Four RCTs included in Redmond et al.'s meta-analysis17 re- ported no certainty of the effect of the intervention (RR 0.78, 95% CI 0.50 to 1.22) and some evidence of heterogeneity (I2 = 48%; p = 0.12). On the contrary, the pooled results from 3 RCTs from Kwan et al.'s meta-analysis7 achieved a statistically significant reduction in health- care usage (RR 0.77, 95% CI 0.63 to 0.95), but heterogeneity was not assessed. Medication reconciliation by a clinical pharmacist at admis- sion, and medication review by a clinical pharmacist during hospitali- zation, also offered no improvements in terms of number of ED visits, hospital readmissions and mortality rates compared with standard care, according to one prospective controlled study included by Lehnbom et al.14 However, in RACF one RCT reported fewer ED visits and readmissions (p = 0.035) when compared to control patients. McNab et al.15 communicated the results of 2 articles (1 RCT and 1 before-and- after study) which measured healthcare utilization as readmissions and ED or general practitioner visits, with no statistically significant dif- ferences between groups.

Discussion

This is the first overview of systematic reviews that specifically fo- cuses on the impact of MR on patient-related outcomes and healthcare utilization. Two independent reviewers systematically reviewed the literature with no publication date or language restrictions, and eval- uated the quality of the systematic reviews using the review instrument AMSTAR 213 validated for randomized or non-randomized studies of healthcare interventions, and extracted data from the publications

Fig. 1. Preferred Reporting Items for Systematic Reviews and Meta-Analyses literature search and study selection flowchart.12

A.B. Guisado-Gil, et al. Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx

4

T ab le 2

Q ua

lit y

of sy

st em

at ic

re vi

ew s

ba se

d on

th e

16 -i te

m A

M ST

A R

2 C he

ck lis

t. 1 3

It em

K w

an et

al . 20

13 7

Le hn

bo m

et al

. 20

14 1 4

M cN

ab et

al .

20 18

1 5

C he

em a

et al

.2 01

81 6

R ed

m on

d et

al . 20

18 1 7

1. D

id th

e re

se ar

ch qu

es ti

on s

an d

in cl

us io

n cr

it er

ia fo

r th

e re

vi ew

in cl

ud e

th e

co m

po ne

nt s

of PI

C O

? Y es

Y es

Y es

Y es

Y es

2. D

id th

e re

po rt

of th

e re

vi ew

co nt

ai n

an ex

pl ic

it st

at em

en t

th at

th e

re vi

ew m

et ho

ds w

er e

es ta

bl is

he d

pr io

r to

th e

co nd

uc t

of th

e re

vi ew

, an

d, di

d th

e re

po rt

ju st

if y

an y

si gn

ifi ca

nt de

vi at

io ns

fr om

th e

pr ot

oc ol

? Pa

rt ia

l ye

s Pa

rt ia

l ye

s Pa

rt ia

l ye

s Pa

rt ia

l ye

s Y es

3. D

id th

e re

vi ew

au th

or s

ex pl

ai n

th ei

r se

le ct

io n

of th

e st

ud y

de si

gn s

fo r

in cl

us io

n in

th e

re vi

ew ?

Y es

Y es

Y es

Y es

Y es

4. D

id th

e re

vi ew

au th

or s

us e

a co

m pr

eh en

si ve

lit er

at ur

e se

ar ch

st ra

te gy

? Y es

Y es

Y es

Y es

Y es

5. D

id th

e re

vi ew

au th

or s

pe rf

or m

st ud

y se

le ct

io n

in du

pl ic

at e?

Y es

Y es

Y es

Y es

Y es

6. D

id th

e re

vi ew

au th

or s

pe rf

or m

da ta

ex tr

ac ti

on in

du pl

ic at

e? Y es

Y es

Y es

Y es

Y es

7. D

id th

e re

vi ew

au th

or s

pr ov

id e

a lis

t of

ex cl

ud ed

st ud

ie s

an d

ju st

if y

th e

ex cl

us io

ns ?

N o

N o

N o

N o

N o

8. D

id th

e re

vi ew

au th

or s

de sc

ri be

th e

in cl

ud ed

st ud

ie s

in ad

eq ua

te de

ta il?

Y es

Y es

Y es

Y es

Y es

9. D

id th

e re

vi ew

au th

or s

us e

a sa

ti sf

ac to

ry te

ch ni

qu e

fo r

as se

ss in

g th

e R

oB in

in di

vi du

al st

ud ie

s th

at w

er e

in cl

ud ed

in th

e re

vi ew

? Y es

N o

Y es

Y es

Y es

10 . D

id th

e re

vi ew

au th

or s

re po

rt th

e so

ur ce

s of

fu nd

in g

fo r

th e

st ud

ie s

in cl

ud ed

in th

e re

vi ew

? N

o N

o N

o N

o Y es

11 .I

f m

et a-

an al

ys is

w as

pe rf

or m

ed , di

d th

e re

vi ew

au th

or s

us e

ap pr

op ri

at e

m et

ho ds

fo r

th e

st at

is ti

ca l co

m bi

na ti

on of

re su

lt s?

Y es

N o

m et

a- an

al ys

is co

nd uc

te d

Y es

Y es

Y es

12 .I

f m

et a-

an al

ys is

w as

pe rf

or m

ed ,d

id th

e re

vi ew

au th

or s

as se

ss th

e po

te nt

ia l im

pa ct

of R

oB in

in di

vi du

al st

ud ie

s on

th e

re su

lt s

of th

e m

et a-

an al

ys is

or ot

he r

ev id

en ce

sy nt

he si

s? Y es

N o

m et

a- an

al ys

is co

nd uc

te d

Y es

Y es

Y es

13 .D

id th

e re

vi ew

au th

or s

ac co

un tf

or R

oB in

in di

vi du

al st

ud ie

s w

he n

in te

rp re

ti ng

/d is

cu ss

in g

th e

re su

lt s

of th

e re

vi ew

? Y es

N o

Y es

Y es

Y es

14 .D

id th

e re

vi ew

au th

or s

pr ov

id e

a sa

ti sf

ac to

ry ex

pl an

at io

n fo

r, an

d di

sc us

si on

of ,a

ny he

te ro

ge ne

it y

ob se

rv ed

in th

e re

su lt s

of th

e re

vi ew

? Y es

Y es

Y es

Y es

Y es

15 .I

f th

ey pe

rf or

m ed

qu an

ti ta

ti ve

sy nt

he si

s, di

d th

e re

vi ew

au th

or s

ca rr

y ou

t an

ad eq

ua te

in ve

st ig

at io

n of

pu bl

ic at

io n

bi as

(s m

al l st

ud y

bi as

) an

d di

sc us

s it

s lik

el y

im pa

ct on

th e

re su

lt s

of th

e re

vi ew

? N

o N

o m

et a-

an al

ys is

co nd

uc te

d Y es

N o

Y es

16 .D

id th

e re

vi ew

au th

or s

re po

rt an

y po

te nt

ia l so

ur ce

s of

co nfl

ic t

of in

te re

st ,i

nc lu

di ng

an y

fu nd

in g

th ey

re ce

iv ed

fo r

co nd

uc ti

ng th

e re

vi ew

? Y es

Y es

Y es

Y es

Y es

Q ua

lit y

le ve

l C ri

ti ca

lly lo

w C ri

ti ca

lly lo

w Lo

w C ri

ti ca

lly lo

w Lo

w

a A M

ST A

R =

A ss

es sm

en t

of M

ul ti

pl e

Sy st

em at

ic R

ev ie

w s.

PI C O

= Po

pu la

ti on

, In

te rv

en ti

on , C om

pa ri

so n,

O ut

co m

e. R

oB =

ri sk

of bi

as .

A.B. Guisado-Gil, et al. Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx

5

T ab le 3

Su m

m ar

y of

ch ar

ac te

ri st

ic s

of se

le ct

ed sy

st em

at ic

re vi

ew s

an d

m et

a- an

al ys

es .

A ut

ho r,

ye ar

A im

N um

be r

an d

de si

gn of

st ud

ie s

N um

be r

an d

ty pe

of pa

rt ic

ip an

ts

Se tt

in g

(t ra

ns it

io n

po in

t) In

te rv

en ti

on pr

ov id

er In

te rv

en ti

on H

ea lt h

ou tc

om es

O th

er ou

tc om

es

K w

an et

al .

20 13

7 To

su m

m ar

iz e

ev id

en ce

on th

e eff

ec ti

ve ne

ss of

ho sp

it al

-b as

ed M

R in

te rv

en ti

on s

5 R

C Ts

4 be

fo re

-a nd

-a ft

er 9

po st

- in

te rv

en ti

on

39 17

0 ad

ul ts

29 5

pe di

at ri

c pa

ti en

ts

H os

pi ta

l (a

dm is

si on

, di

sc ha

rg e

ho m

e, in

- ho

sp it

al tr

an sf

er )

Ph ar

m ac

is ts

A ls

o ph

ys ic

ia ns

an d

nu rs

es

M R

Pa ti

en t

co un

se lli

ng C re

at io

n of

po st

- di

sc ha

rg e

m ed

ic at

io n

lis ts

Po st

-d is

ch ar

ge co

m m

un ic

at io

n

H ea

lt hc

ar e

ut ili

za ti

on (E

m er

ge nc

y D

ep ar

tm en

t vi

si ts

an d

re ad

m is

si on

w it

hi n

30 da

ys )

C lin

ic al

ly si

gn ifi

ca nt

un in

te nd

ed m

ed ic

at io

n di

sc re

pa nc

ie s

(e qu

al to

A D

Es w

it h

po te

nt ia

l to

ca us

e in

ju ry

)

Le hn

bo m

et al

. 20

14 1 4

To ev

al ua

te ho

w eff

ec ti

ve M

R an

d m

ed ic

at io

n re

vi ew

ar e

in id

en ti

fy in

g an

d re

ct if yi

ng ha

rm fu

l di

sc re

pa nc

ie s

an d

m ed

ic at

io n-

re la

te d

pr ob

le m

s, an

d to

as se

ss th

ei r

im pa

ct on

cl in

ic al

ou tc

om es

5 R

C Ts

4 pr

os pe

ct iv

e co

nt ro

lle d

2 be

fo re

-a nd

-a ft

er 3

co ho

rt 21

po st

- in

te rv

en ti

on 3

re tr

os pe

ct iv

e ob

se rv

at io

na l

2 cr

os s-

se ct

io na

l

19 15

8 ad

ul ts

H os

pi ta

l (a

dm is

si on

, di

sc ha

rg e

ho m

e, in

- ho

sp it

al tr

an sf

er )

C om

m un

it y

(d is

ch ar

ge fr

om ho

sp it

al to

ho m

e) R

A C F

(a dm

is si

on fr

om ho

sp it

al )

Ph ar

m ac

is ts

A ls

o ph

ys ic

ia ns

an d

nu rs

es

M R

Pa ti

en t

co un

se lli

ng C re

at io

n of

po st

- di

sc ha

rg e

m ed

ic at

io n

lis ts

Po st

-d is

ch ar

ge co

m m

un ic

at io

n

Em er

ge nc

y D

ep ar

tm en

t vi

si ts

R ea

dm is

si on

s Le

ng th

of st

ay Ph

ys ic

ia n

vi si

ts H

ea lt hc

ar e

ut ili

za ti

on (E

m er

ge nc

y D

ep ar

tm en

t vi

si ts

, re

ad m

is si

on s

an d

m or

ta lit

y)

M ed

ic at

io n

di sc

re pa

nc ie

s

M cN

ab et

al .

20 18

1 5

To de

te rm

in e

th e

eff ec

ti ve

ne ss

of M

R in

ov er

al l di

sc re

pa nc

y id

en ti

fi ca

ti on

an d

re so

lu ti

on , th

e cl

in ic

al re

le va

nc e

of re

so lv

ed di

sc re

pa nc

ie s

an d

he al

th ca

re ut

ili za

ti on

5 R

C Ts

3 be

fo re

-a nd

-a ft

er 6

co ho

rt

36 42

ad ul

ts an

d pe

di at

ri c

pa ti

en ts

C om

m un

it y

(d is

ch ar

ge fr

om ho

sp it

al to

ho m

e, re

si de

nt ia

l un

it or

nu rs

in g

ho m

e)

Ph ar

m ac

is ts

M R

Em er

ge nc

y D

ep ar

tm en

t vi

si ts

R ea

dm is

si on

s Pr

im ar

y an

d se

co nd

ar y

ca re

co ns

ul ta

ti on

s

M ed

ic at

io n

di sc

re pa

nc ie

s Pr

im ar

y ca

re w

or kl

oa d

C he

em a

et al

. 20

18 1 6

To up

da te

th e

ex is

ti ng

ev al

ua ti

on of

th e

im pa

ct of

ph ar

m ac

is t-

le d

m ed

ic at

io n

re co

nc ili

at io

n on

he al

th ca

re ou

tc om

es

18 R

C Ts

60 38

ad ul

ts H

os pi

ta l (a

dm is

si on

, di

sc ha

rg e

ho m

e, in

- ho

sp it

al tr

an sf

er )

Ph ar

m ac

is ts

M R

Pa ti

en t

co un

se lli

ng C re

at io

n of

po st

- di

sc ha

rg e

m ed

ic at

io n

lis ts

H ea

lt hc

ar e

ut ili

za ti

on (d

ru g-

re la

te d

Em er

ge nc

y D

ep ar

tm en

t vi

si ts

an d

re ad

m is

si on

s)

M ed

ic at

io n

di sc

re pa

nc ie

s A

D Es

w it

h po

te nt

ia l to

ca us

e in

ju ry

Pr ev

en ta

bl e

A D

Es

R ed

m on

d et

al .

20 18

1 7

To as

se ss

th e

eff ec

t of

M R

on m

ed ic

at io

n di

sc re

pa nc

ie s,

pa ti

en t-

re la

te d

ou tc

om es

an d

he al

th ca

re ut

ili za

ti on

in pe

op le

re ce

iv in

g th

is in

te rv

en ti

on du

ri ng

ca re

tr an

si ti

on s

25 R

C Ts

69 95

ad ul

ts H

os pi

ta l (a

dm is

si on

, di

sc ha

rg e

ho m

e, in

- ho

sp it

al tr

an sf

er )

C om

m un

it y

(d is

ch ar

ge fr

om ho

sp it

al to

ho m

e) R

A C F

(a dm

is si

on fr

om ho

sp it

al )

Ph ar

m ac

is ts

A ls

o ph

ys ic

ia ns

an d

nu rs

es

M R

Pa ti

en t

co un

se lli

ng M

ed ic

at io

n in

fo rm

at io

n Po

st -t

ra ns

it io

n co

m m

un ic

at io

n M

ed ic

at io

n re

vi ew

Le ng

th of

st ay

M or

ta lit

y Em

er ge

nc y

D ep

ar tm

en t

vi si

ts R

ea dm

is si

on s

H ea

lt hc

ar e

ut ili

za ti

on (E

m er

ge nc

y D

ep ar

tm en

t vi

si ts

an d

re ho

sp it

al iz

at io

n)

M ed

ic at

io n

di sc

re pa

nc ie

s A

D Es

w it

h po

te nt

ia l to

ca us

e in

ju ry

Pr ev

en ta

bl e

A D

Es M

ed ic

at io

n ad

he re

nc e

R C Ts

= R

an do

m iz

ed C on

tr ol

le d

Tr ia

ls . M

R =

M ed

ic at

io n

re co

nc ili

at io

n. A

D Es

= ad

ve rs

e dr

ug ev

en ts

. R

A C F

= R

es id

en ti

al ag

ed ca

re fa

ci lit

ie s.

A.B. Guisado-Gil, et al. Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx

6

included. Medication reconciliation is proposed to avoid possible medication errors, and consequently harm to patients, however, our findings show that, compared with usual care, the intervention does not achieve a clear improvement for patient-related outcomes and health- care utilization.

The reviews included were low15,17 and critically low7,14,16 in quality according to AMSTAR 213. The main results in critical and non- critical domains were consistent with Wolfe et al.20 Overall, the most frequent methodological shortcomings were: rarely providing evidence that the authors had worked with a written and registered protocol with independent verification; not providing a list of excluded studies and justification of exclusions; infrequently assessing the likelihood of publication bias, and neglecting to state the sources of funding of each primary study. A list of excluded studies and justification (item 7; cri- tical domain) was necessary to qualify publications as moderate quality. However, it is currently rare when publishing systematic reviews in journals to include this information beyond the study selection flow- chart.

Of the 5 systematic reviews included, all conducted further meta- analyses except Lehnbom et al.14 For the meta-analysis, authors used a random effects model. This model, as opposed to a fixed-effects model, is advised in the case of statistical heterogeneity in studies.10 Ad- ditionally, two reviews7,16 used a fixed model to validate their results.

The results of this review are in line with other authors21,22 who communicated that hospital-based care is the most commonly studied point of transition, follow by primary care and long-term care settings. Regarding intervention providers, the findings were consistent with Mueller et al.6 whose review found that the majority of, and most successful, interventions relied heavily on pharmacists. Growing evi- dence shows that medication lists obtained by pharmacists contain significantly fewer errors than those obtained through the usual means. In this sense, a new law in California, which came into effect on 1 January 2019, requires pharmacy staff at hospitals with more than 100 beds to obtain an accurate medication list for each newly admitted high-risk patient.23

Most studies compared the intervention group with usual care, but it is not always clear what usual care involved. For ethical reasons, most studies failed to evaluate MR versus “no medication reconciliation”, thus limiting their ability to detect a significant difference between groups. In addition, all reviews except McNab et al.15 included articles that bundled MR with other interventions aimed at improving transi- tions of care, but the specific effect of MR in these multifaceted inter- ventions is not found. In this overview, MR does not achieve a clear improvement in health outcomes. Only one review17 based on a single publication reported results on mortality. However, the impact of MR

on healthcare utilization was communicated in all reviews.7,14–17

Emergency Department visits and readmissions were other results with a follow-up period no longer than 30 days in most studies. In contrast, a trial of MR with no additional discharge interventions that used a longer follow-up (12 months) reported a significant reduction in ED visits and readmissions.24 This suggests the convenience of future re- search to consider a time point beyond the traditional 30-day mark.

The main strength of our study is that it was not limited to specific study designs, population, intervention providers, settings or health outcomes measured, and was intended to provide an outline of current evidence related to MR. The results summarize the effectiveness of the intervention on mortality, length of stay, ED visits, unplanned read- missions, physician visits and healthcare utilization. The main limita- tion is that the low number of systematic reviews included did not make it possible to exclude reviews with insufficient quality, and moderate and high quality systematic reviews in the literature that met the in- clusion criteria were not found. In addition, because of evidence of substantial heterogeneity in study designs, settings, intervention com- ponents and health outcomes, statistical pooling of all primary studies that measured health outcomes was not always possible. Even those in which a meta-analysis was carried out, the value of I2 showed a non- negligible level of heterogeneity.

While all reviews in our overview addressed the impact of MR on patient-related outcomes and healthcare utilization, a limitation com- monly admitted by the authors of the reviews included is that few primary research articles specifically focused on these results. These were often listed as secondary or composite outcomes, and the studies were not powered to detect a significant difference between groups. This means that it is difficult to draw definitive conclusions from meta- analysis, other than to say that there is currently no firm evidence that MR improves health outcomes. Further research is needed that includes more studies that are robust and of adequate sample size to test the impact of MR on health outcomes.

Conclusion

Few systematic reviews support the value of MR in achieving good patient-related outcomes and healthcare utilization improvements. The quality of the systematic reviews were low and the primary studies included, mostly RCTs, often involved additional activities related to MR. There was no clear evidence in favor of the intervention in mor- tality, length of stay, ED visits, unplanned readmissions, physician visits and healthcare utilization, the latter being the most frequently com- municated clinical outcome.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Declaration of competing interest

None.

Acknowledgements

We would like to thank Fundación Andaluza Beturia para la Investigación en Salud (FABIS) who provided critical review and support in writing this review.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https:// doi.org/10.1016/j.sapharm.2019.10.011.

Table 4 Association between medication reconciliation and health outcomes.

Health outcomes Setting Number of systematic reviews

Number of primary studies

Association

Mortality Hospital 117 1 ND Length of stay Hospital 214,17 3 ND

RACF 114 1 + Emergency

Department visits

Hospital 214,17 2 ND Community 115 2 ND

Unplanned readmissions

Hospital 214,17 6 ND Community 214, 15 7 ND

Physician visits Community 214, 15 3 ‒/ND Healthcare

utilization Hospital 47, 14,16,17 10 +/ND Community 115 2 ND RACF 114 1 +

RACF = Residential aged care facilities. ND=No statistically significant dif- ferences (p > 0.05). + = Statistically significant differences in favor of medication reconciliation (p < 0.05). ‒ = Statistically significant differences against medication reconciliation (p < 0.05).

A.B. Guisado-Gil, et al. Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx

7

References

1. Moore C, McGinn T, Halm E. Tying up loose ends discharging patients with un- resolved medical issues. Arch Intern Med. 2007;167:1305–1311. https://doi.org/10. 1001/archinte.167.12.1305.

2. Chenoweth L, Kable A, Pond D. Research in hospital discharge procedures addresses gaps in care continuity in the community, but leaves gaping holes for people with dementia: a review of the literature. Australas J Ageing. 2015;34:9–14. https://doi. org/10.1111/ajag.12205.

3. Jack BW, Chetty VK, Anthony D, et al. A reengineered hospital discharge program to decrease rehospitalization: a randomized trial. Ann Intern Med. 2009;150:178–187. https://doi.org/10.7326/0003-4819-150-3-200902030-00007.

4. Institute for Healthcare Improvement. Medication Reconciliation Review. 2007; 2007 http://www.ihi.org/resources/Pages/Tools/MedicationReconciliationReview.aspx 28.06.19.

5. Cornish PL, Knowles SR, Marchesano R, et al. Unintended medication discrepancies at the time of hospital admission. Arch Intern Med. 2005;165:424–429. https://doi. org/10.1001/archinte.165.4.424.

6. Mueller SK, Sponsler KC, Kripalani S, Schnipper JL. Hospital-based medication re- conciliation practices: a systematic review. Arch Intern Med. 2012;172:1057–1069. https://doi.org/10.1001/archinternmed.2012.2246.

7. Kwan JL, Lo L, Sampson M, Shojania KG. Medication reconciliation during transitions of care as a patient safety strategy: a systematic review. Ann Intern Med. 2013;158:397–403. https://doi.org/10.7326/0003-4819-158-5-201303051-00006.

8. Mekonnen AB, McLachlan AJ, Brien JE. Effectiveness of pharmacist-led medication reconciliation programmes on clinical outcomes at hospital transitions: a systematic review and meta-analysis. BMJ Open. 2016;6. https://doi.org/10.1136/bmjopen- 2015-010003 e010003.

9. Smith V, Devane D, Begley CM, Clarke M. Methodology in conducting a systematic review of systematic reviews of healthcare interventions. BMC Med Res Methodol. 2011;11(1):15–21. https://doi.org/10.1186/1471-2288-11-15.

10. Higgins JPT, Thomas J, Chandler J, eds. Cochrane Handbook for Systematic Reviews of Interventions Version 6.0 (updated July 2019). Cochrane; 2019http://www.training. cochrane.org/handbook, Accessed date: 29 September 2019.

11. Holte HH, Hafstad E, Vist GE. Overview of Reviews on Effect of Medical Reconciliation Oslo: National Knowledge Center for Health Service; 2015 Report from Knowledge Center Nº 7-2015.

12. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6. https://doi.org/10. 1371/journal.pmed.1000097 e1000097.

13. Shea BJ, Reeves BC, Wells G, et al. Amstar 2: a critical appraisal tool for systematic reviews that include randomised or non-randomised studies of healthcare

interventions, or both. BMJ. 2017;358:j4008. https://doi.org/10.1136/bmj.j4008. 14. Lehnbom EC, Stewart MJ, Manias E, Westbrook JI. Impact of medication re-

conciliation and review on clinical outcomes. Ann Pharmacother. 2014;48:1298–1312. https://doi.org/10.1177/1060028014543485.

15. McNab D, Bowie P, Ross A, et al. Systematic review and meta-analysis of the effec- tiveness of pharmacist-led medication reconciliation in the community after hospital discharge. BMJ Qual Saf. 2018;27:308–320. https://doi.org/10.1136/bmjqs-2017- 007087.

16. Cheema E, Alhomoud FK, Kinsara ASAL-D, et al. The impact of pharmacists-led medicines reconciliation on healthcare outcomes in secondary care: a systematic review and meta-analysis of randomized controlled trials. PLoS One. 2018;13. https://doi.org/10.1371/journal.pone.0193510 e0193510.

17. Redmond P, Grimes TC, McDonnell R, Boland F, Hughes C, Fahey T. Impact of medication reconciliation for improving transitions of care. Cochrane Database Syst Rev. 2018;23:CD010791. https://doi.org/10.1002/14651858.CD010791.pub2.

18. Koehler BE, Richter KM, Youngblood L, et al. Reduction of 30-day postdischarge hospital readmission or emergency department (ED) visit rates in high-risk elderly medical patients through delivery of a targeted care bundle. J Hosp Med. 2009;4:211–218. https://doi.org/10.1002/jhm.427.

19. Kripalani S, Roumie CL, Dalal AK, et al. PILL-CVD (Pharmacist Intervention for Low Literacy in Cardiovascular Disease) Study Group. Effect of a pharmacist intervention on clinically important medication errors after hospital discharge: a randomized trial. Ann Intern Med. 2012;157:1–10. https://doi.org/10.7326/0003-4819-157-1- 201207030-00003.

20. Wolfe D, Yazdi F, Kanji S, et al. Incidence, causes, and consequences of preventable adverse drug reactions occurring in inpatients: a systematic review of systematic reviews. PLoS One. 2018;13. https://doi.org/10.1371/journal.pone.0205426 e0205426.

21. Bayoumi I, Howard M, Holbrook AM, Schabort I. Interventions to improve medica- tion reconciliation in primary care. Ann Pharmacother. 2009;43:1667–1675. https:// doi.org/10.1345/aph.1M059.

22. Chhabra PT, Rattinger GB, Dutcher SK, Hare ME, Parsons KL, Zuckerman IH. Medication reconciliation during the transition to and from long-term care settings: a systematic review. Res Soc Adm Pharm. 2012;8:60–75. https://doi.org/10.1016/j. sapharm.2010.12.002.

23. Thompson CA. California pharmacists to ensure accuracy of high-risk patients' ad- mission medication lists. Am J Health Syst Pharm. 2018;75:1961–1963. https://doi. org/10.2146/news180076.

24. Gillespie U, Alassaad A, Henrohn D, et al. A comprehensive pharmacist intervention to reduce morbidity in patients 80 years or older: a randomized controlled trial. Arch Intern Med. 2009;169:894–900. https://doi.org/10.1001/archinternmed.2009.71.

A.B. Guisado-Gil, et al. Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx

8