3 Impact of Pharmacist Led Interventions Adherence HTA
3 Impact of Pharmacist Led Interventions Adherence HTA
3 Impact of Pharmacist Led Interventions Adherence HTA
1,2 Background: The aim of this study was to provide a scoping review of the impact of
Mohamed Hassan Elnaem
1 pharmacist-led interventions on medication adherence and clinical outcomes in patients with
Nor Fatin Farahin Rosley
For personal use only.
submit your manuscript | www.dovepress.com Journal of Multidisciplinary Healthcare 2020:13 635–645 635
DovePress © 2020 Elnaem et al. This work is published and licensed by Dove Medical Press Limited. The full terms of this license are available at https://www.dovepress.com/terms.
http://doi.org/10.2147/JMDH.S257273
php and incorporate the Creative Commons Attribution – Non Commercial (unported, v3.0) License (http://creativecommons.org/licenses/by-nc/3.0/). By accessing the
work you hereby accept the Terms. Non-commercial uses of the work are permitted without any further permission from Dove Medical Press Limited, provided the work is properly attributed. For
permission for commercial use of this work, please see paragraphs 4.2 and 5 of our Terms (https://www.dovepress.com/terms.php).
Examples of modifiable risk factors include tobacco smok- Identifying the Research Questions
ing, raise blood lipid levels, hypertension, physical inactivity, This review focussed on studying the pharmacists-led
unhealthy diet, and obesity. A family history of CVD, age, interventions in promoting medication adherence among
gender, and socioeconomic status is considered as non-mod- patients with CVD, particularly hypertension and hyperli-
ifiable risk factors.1 Although evidence recommends guide- pidemia. The following research questions guided the pro-
lines-based pharmacotherapies for reducing the risk of CVD, cedure of including studies, extract and summarize the
Journal of Multidisciplinary Healthcare downloaded from https://www.dovepress.com/ by 94.158.22.224 on 10-Aug-2020
adjunctive lifestyle modification is equally important in redu- data type, assessment, and impact of the designed
cing the risk of repeated cardiovascular events.2
interventions:
Typical examples of important cardiovascular medica-
tions include statins, antihypertensives, and antiplatelet
1. What are the characteristics of pharmacists-led
drugs.3,4 Adherence to medication regimen; the extent to
interventions used in the included studies?
which people take medications as prescribed by their
2. What methods were used to measure the impact of
healthcare providers; is essential to attaining successful
interventions on medication adherence and clinical
clinical outcomes.5 Adherence is used interchangeably
outcomes among patients with hypertension and
with compliance; however, the motivational level seems
hyperlipidemia?
to be varied between them, where adherence denotes better
3. What was the overall effect of these interventions
patients’ engagement with the instructions of healthcare
on medication adherence and clinical outcomes?
providers.6 Unfortunately, nonadherence to cardiovascular
For personal use only.
Table 1 Inclusion Criteria extraction of the first five studies. A meeting was then
Category Inclusion Criteria conducted to discuss the extent of consistency of data
extraction against the study-specific extraction form.
Language of English
publication
Year of 2009–2019 Collating, Summarizing, and Reporting
publication
the Results
Journal of Multidisciplinary Healthcare downloaded from https://www.dovepress.com/ by 94.158.22.224 on 10-Aug-2020
Data charting was done in an iterative process using a cant improvement in medicine adherence (P< 0.05) and
study-specific extraction form that was aligned with the studies with non-significant or no improvement in adher-
research question. The included variables were study ence (P> 0.05). The included variables included study
author, publication year, study design, study population, design, study population, type of intervention, outcome
type of intervention, adherence measurement, and key measures, method of adherence measurement, and key
findings. Two independent researchers completed the data results.
Figure 1 PRISMA flowchart for selecting the studies according to the systematic scoping review methodology.
Table 2 Medicine Adherence Measurement Methods and Their Characteristics Reported in the Included Studies
Adherence Self-Reported Refill Record-Based Prescription Manual Pill Count N=1
Measurement Adherence Adherence N=8 Abandonment N=1
Method (MMAS and Its
Related Scales)
N=7
● Common ● Common in measuring ● Known as primary medica- ● More objective than self-reporting
Journal of Multidisciplinary Healthcare downloaded from https://www.dovepress.com/ by 94.158.22.224 on 10-Aug-2020
Main in
Characteristics measuring adher- adherence of hyperli- tion nonadherence medication adherence
ence of antihy- pidemia medications ● Abandons to prescriptions ● Regarded as the gold standard for
pertensive drugs ● Common types:- portrayed that patients do validating other methods
● Simple and easy Medication possession not adhere to their ● It can be done without informing
to be used ratio (MPR) medications patients to avoid the disadvantage
● Self-reporting -The proportion of ● Reasons for prescriptions of pill dumping, where patients
style days covered (PDC) abandon: low perceived manipulate the number of pills.
● Subjected to ● Quantitative measures benefit due to asympto-
overestimation ● Less subjected to over- matic nature of the
of adherence estimation of adher- condition
level ence level.
● High chance of ● Subjected to a possibi-
response bias lity that medications
For personal use only.
example, studies with counseling sessions as the main meanwhile, the remaining three studies were carried
interventions that have used telephone reminders in the out among patients receiving medications for
follow-up period were classified under the “counseling” hyperlipidemia.14,19,29 Almost half of the interventions asso-
class of interventions.21 Overall, conventional face-to-face ciated with improvement in adherence involved direct phar-
counseling was the most frequently used intervention in macist-patient counseling sessions. A multi-faceted
the included studies (n=8), where patients were provided pharmacist intervention was employed in three studies,
Journal of Multidisciplinary Healthcare downloaded from https://www.dovepress.com/ by 94.158.22.224 on 10-Aug-2020
counseling on disease management, adverse effects, stra- while three studies relied on involving remote reminders or
tegies to improve medicine adherence, and advice on life- telephone-designed interventions. The non-significant
style modifications. All studies conducted follow-up improvement in medicine adherence was reported in two
counseling sessions; however, the structure, frequency, telephone-based interventions, three face-to-face counseling
and duration of the follow-up sessions were variable sessions, and one multi-faceted intervention. In addition, of
across the studies (see Tables 3 and 4). Telephone-based the twelve studies that assessed the impact of the interven-
or remotely conducted interventions were used in five tions on the clinical outcome measures, only four studies had
studies. As reported in the face-to-face counseling, there been associated with significant impact for the pharmacists’
was considerable variability in the delivery of these inter- interventions on the overall assessment of clinical outcome
ventions. For example, one study implemented behavioral measures. Of these, three studies were involving antihyper-
interviewing as a part of a multicomponent remote tensive medications and only one study involved lipid-low-
intervention.22 Another example was the use of successive ering therapy.
For personal use only.
Ho et al, RCT involved 241 Multi-faceted intervention Comparator: usual care. The intervention increased
201423 (USA) patients admitted with comprising medication Follow-up: 12 months. adherence to statin (93.2 vs.
Journal of Multidisciplinary Healthcare downloaded from https://www.dovepress.com/ by 94.158.22.224 on 10-Aug-2020
ACS and then discharged. reconciliation, patient education, Outcomes: Primary (refill 71.3%) (p=<0.001) and ACEI/
collaborative care, and voice adherence >0.8, % of ARB regimens (93.1% vs. 81.7%),
messaging. adherent patients), (p=0.03) significantly.
Secondary (% of patients There was no significant changes
achieved BP and LDL-C in percentage of target BP or
targets). LDL-C levels attainments.
Lyons et al RCT involved 677 T2DM Two telephone-based Comparator: standard care. The intervention group has less
201627 (UK) patients prescribed with Intervention, with medicine Follow-up: 6 months. percentage of nonadherence
LLT. chart reminder. Outcomes: Primary (self- compared to control group
reported nonadherence), (10.6% vs. 19.6%, p=0.010).
Secondary (LDL-C levels). There was no associated
significant difference in the
clinical outcomes.
For personal use only.
Taitel et al Retrospective cohort Two Face-to-face counselling Comparator: control group. The statin adherence has
201216 (USA) study included 2056 session including a motivational Follow-up: 12 months. improved in the intervention
patients who were newly interview Outcomes: medication group compared to control
initiated on statin adherence (MPR ≥80%). group is (61.8% vs. 56.9%,
medications. p<0.01).
There was no assessment for
clinical outcomes.
Choudhry A Pragmatic cluster RCT Telephone-based behavioral Comparator: usual care. The intervention showed a
et al, 201822 involved 4078 patients’ interviewing, text messaging, and Follow-up: 12 months. significant improvement of 4.7%
(USA) non-adherent to their progress reports. Outcomes: Primary (95% CI, 3.0–6.4%) in medication
hypertension and (medication adherence, % adherence.
hyperlipidemia of covered days), Secondary There were no significant
medications. (LDL-C & SBP). changes in the overall
assessment of clinical outcomes.
Hedegaard RCT included 532 Tailored medication review, Comparator: control group. Nonadherence was higher in the
et al, 201521 patients prescribed with patient interview, followed by Follow-up: 12 months. control group (30.2% vs. 20.3%,
(Denmark) AHT and LLT. telephone reminders. Outcomes: Primary p=0.01) as compared to the
(medication adherence, intervention group.
MPR ≥80%), Secondary (BP There were no significant
& hospital admission). differences in the evaluated
clinical outcomes.
Ramanath RCT involved 52 patients Counseling sessions using Comparator: control group. The overall adherence increased
et al, 201226 on AHT. patient information leaflets and Follow-up: 1 month (twice). significantly in the intervention
(India) telephone reminders. Outcomes: Primary (self- group compared to the control
reported adherence, MMAS group.
& MARS), Secondary (BP There was no significant impact
control). on BP control.
(Continued)
Table 3 (Continued).
Author, Study Design and Pharmacist Intervention Comparator, Follow-Up Key Findings
Year, Population and Outcome Measures
Country
Morgadoet al RCT included 197 Counseling and educational Comparator: usual care. There was a statistically
201113 patients receiving AHT. sessions. Follow-up: 9 months. significant improvement in blood
Journal of Multidisciplinary Healthcare downloaded from https://www.dovepress.com/ by 94.158.22.224 on 10-Aug-2020
Benbrahim RCT included 176 Face-to-face (written and oral) Comparator: usual care. The adherence was increased
et al 201317 patients on AHT. tailored educational Follow-up: 6 months. significantly to 95.5% (baseline
(Spain) intervention. Outcomes: medication 86%) in the intervention group
adherence (pills count). compared to 86.5% (baseline
85.4%) in the control group
(p=0.011).
There was no assessment for
clinical outcomes.
For personal use only.
Fischer et al, RCT included 124 131 Live telephone calls with tailored Comparator: control group. The live pharmacy-based
201418 (USA) patients with newly educational messages Follow-up: 30 days following interventions decreased primary
prescribed cardiovascular index date. medication adherence by 4.8%
medications. Outcomes: Primary (P< 0.0001) compared to
medication adherence control group.
(prescription There was no assessment for
abandonment). clinical outcomes.
Stewart et al, Cluster RCT involved 395 Multi-faceted intervention Comparator: control group. No significant difference in % of
201424 patients who were taking consisted of motivational Follow-up: 6 months. adherent patients between
(Australia) at least one AHT. interviews, refill reminders, Outcomes: Primary (self- control (57.2% vs. 63.6%) and
training on BP monitoring and reported adherence), intervention (60% vs. 73.5%)
medication reviews. Secondary (BP control) groups at baseline and 6 months,
respectively.
Non-adherence decreased from
61.8% at baseline to 39.2% at 6
months in the intervention
group (P = 0.007).
There was a significant SBP
reduction in the intervention
group (p=0.01).
Svarstad et al, Cluster RCT included 567 Team Education and Adherence Comparator: control (only Participants in the intervention
201325 (USA) patients taking one or Monitoring program involved patient information). group had better adherence
more AHT. tailored counselling and Follow-up: 6 and 12 (60% vs 34%, p<0.001) and BP
education using take-home months. control (50% vs. 36%, p=0.01)
toolkit, leaflets and medication Outcomes: adherence (refill compared to the control group.
box. ≥80%) and BP control.
Abbreviations: ACS, acute coronary syndrome; RCT, randomized controlled trial; LLT, lipid-lowering therapy; SBP, systolic blood pressure; T2DM, type 2 diabetes mellitus;
AHT, antihypertensive; LDL-C, low-density lipoprotein cholesterol.
outcome measures.18,21-23 Furthermore, five out of eight significant results.13,17,24-26 Finally, only two out of five
studies that only targeted improvement in the adherence studies aimed to enhance the adherence to lipid-lowering
outcome measures for antihypertensive drugs reported therapy reported a positive impact of the implemented
Eussen et al, A multicentre, open- Five structured Comparator: usual care. The intervention showed a lower
201029 label RCT included 899 counselling sessions over Follow-up: 6 and 12 months. discontinuation rate of that was
Journal of Multidisciplinary Healthcare downloaded from https://www.dovepress.com/ by 94.158.22.224 on 10-Aug-2020
(Netherlands) patients on statin a year. Outcomes: Primary (adherence, significant only at 6 months but not
medications. 1-year discontinuation rate), significant at 1 year.
Secondary (6 months Median MPR was not significantly
discontinuation rate, ≥90% MPR different between groups (99.5% vs.
and LDL-C levels). 99.2%, p=0.14).
Adherent patients were more likely
to achieve target LDL-c levels at 6
months (74% vs. 50%, p=0.01).
Kooy et al RCT included 299 Electronic reminder Comparator: control group. Overall, refill adherence was not
201319 elderly patients (65 device (ERD) with or Follow-up: 360 days. significantly improved with
(Netherlands) years or above) who without counselling Outcomes: adherence (refill counselling with ERD (69.25,
had started statins at sessions. ≥80%). p=0.55), ERD only (72.4%, p=0.18)
least one year. compared to control group (64.8%).
For personal use only.
Ma et al, RCT involved 689 Five Telephone Comparator: usual care. The intervention did not show
201014 (USA) patients with counselling calls. Follow-up: 12 months. significant improvement in statin
underlying CHD who Outcomes: Primary (% patients adherence (0.88 vs. 0.90, p=0.51).
had an LLT achieved LDL-C levels), Secondary It had no significant impact on
prescription. (adherence, continuous multiple- clinical outcomes (65% vs. 60%,
interval (CMA) from pharmacy p=0.29).
records).
Gums et al, Cluster RCT included Physician-pharmacist Comparator: usual care. There was no significant difference in
201520 (USA) 593 patients who had collaboration Follow-up: 9 months. the measures of medication adherence
at least one AHT. management (PPCM) Outcomes: Primary (Adherence, between the groups. Patients in the
self-reported questionnaire), intervention group experienced higher
Secondary (medication changes). medication changes compared to
control group (4.9 vs. 1.1, p=0.003).
There was no assessment for clinical
outcomes.
Wong et al, RCT included 274 Counselling sessions with Comparator: usual care (brief Overall, both percentage of patients
201315 (Hong patients taking at least structured patient drug advice). with optimal adherence and BP control
Kong) one long-term AHT and education and provision Follow-up: 3 and 6 months. were improved throughout study
having suboptimal of pillboxes and Outcomes: Primary (BP control), period.
compliance medication knives. Secondary (adherence, self- However, there were no significant
reported) differences between the groups in both
outcome measures.
Van der Laan RCT included 170 Two face-to-face Comparator: usual care. There were no significant differences
et al, 201828 patients who were on consultation (3 months Follow-up: 9 months. between intervention and control
(Netherlands) AHT. apart). Outcomes: Primary (self- groups in both outcome measures.
reported adherence), Secondary
(BP control).
pharmacist interventions.16,27 In a study that reported the overestimation of the adherence level induced by the self-
overall good impact of the intervention on medication reporting. Consequently, the improvement in adherence to
adherence, it was underpinned that the adherence measures the self-reporting component might be less valid compared to
for antihypertensive were more positively affected by the other quantitative methods.
intervention compared to lipid-lowering therapy.18 Pharmacist-led interventions were associated with
Moreover, among studies that failed to show the overall improved patients’ adherence to their medications but were
Journal of Multidisciplinary Healthcare downloaded from https://www.dovepress.com/ by 94.158.22.224 on 10-Aug-2020
impact of pharmacist intervention on the adherence mea- less likely to be consistently associated with the attainment of
sures, it was highlighted that women, particularly, were clinical outcomes. The findings of this study seem to be
able to attain a significant impact on adherence to lipid- influenced by the variations in the type and design of the
lowering therapy for secondary prevention compared to intervention, control group, decision on primary and second-
the control group.19 The findings suggest the potential ary outcomes, and tools for assessing adherence. Most of the
importance of the customization of the implemented inter- studies measured clinical outcomes as secondary outcomes
ventions to consider the medications and the patients’ following the assessment of medication adherence as primary
demographics carefully. outcomes. Measurement of clinical outcomes was missing in
Twelve studies assessed the impact of pharmacists’ some studies that made it challenging to investigate the
interventions on clinical outcomes including hypertension association between adherence and clinical outcomes.17
and hyperlipidemia. Of these, four studies were designed The findings of this suggested that interventions with
primarily to assess the impact of pharmacists’ interven- frequent follow-ups,13 the use of a variety of verbal and
For personal use only.
tions on the clinical outcomes of blood pressure control or written materials,17 and interventions implemented with sev-
achievement of target LDL-C levels.13–15,25 However, only eral subsequent components were more likely to achieve
three studies reported significant improvement in clinical positive intervention outcomes.24 Future studies involving
outcomes following pharmacists’ interventions with two of pharmacists’ interventions should involve frequent contact
them being multi-faceted. It is also important to highlight with patients as better-structured multicomponent interven-
that studies that assessed both outcome measures for both tions are more likely to achieve optimal adherence levels.
antihypertensive and lipid-lowering therapy were less Furthermore, future studies should focus on delivering inter-
likely to have a significant impact on the clinical outcomes ventions to patients who are more likely to benefit from the
despite their significant impact on medicine adherence. intervention. Thus, it is essential to recruit patients who are
Majority of the included studies used pharmacy-refill experiencing adherence issues at baseline to make efficient
records either as medication possession ratio (MPR) or the use of the available human and financial resources.25 Also, it
proportion of days covered (PDC) to measure the adherence would be critical to investigate the individual patients’ level
level of patients. The MPR calculates medicine adherence as characteristics such as gender that have the potential to affect
the days’ sum supply of the medications during the observa- the overall outcomes of the implemented interventions.19
tion period divided by the total days during that period.16 The This study has some limitations. Quality assessment of
PDC, on the other hand, takes into account the covered days, the included studies was not conducted. The searches were
which is calculated as the proportion of days covered of the confined to three databases that may have limited the
360 days following the index date divided by the total days’ opportunity to identify further eligible studies published
supply by the number of the study period.19 Both methods in other relevant databases. Furthermore, since our work
are calculative measures not subjected to a bias imposed by was limited to the inclusion of published literature only,
self-reporting. However, the interpretation of adherence dif- publication bias is probable. However, it can be argued
fered within studies with some studies regarding adherence that a significant number of the included studies were not
as MPR ≥ 90%, while some regarded adherence as MPR ≥ necessarily reporting positive findings. Nevertheless, the
80%. The other most popular method of measuring adher- review employed a comprehensive, transparent, and rigor-
ence was the Morisky Medication Adherence Scale (MMAS) ous method to identify studies.
that has also been a basis for developing derived scales.13,20
This adherence measurement method has widely been used Conclusion
in measuring adherence to antihypertensive medications due Pharmacist-led interventions were associated with
to its simplicity, ease of administration, and low cost.15 improved patients’ adherence to their medications but
However, this method has limitations, including an were less likely to be consistently associated with the
attainment of clinical outcomes. Face-to-face counseling 16. Taitel M, Jiang J, Rudkin K, Ewing S, Duncan I. The impact of
pharmacist face-to-face counseling to improve medication adherence
provided by pharmacists was found to be the most widely
among patients initiating statin therapy. Patient Prefer Adherence.
used; meanwhile, the multi-faceted interventions were 2012;6:323–329. doi:10.2147/PPA.S29353
more likely to be effective in improving the overall out- 17. Fikri-Benbrahim N, Faus MJ, Martínez-Martínez F, Sabater-
Hernández D. Impact of a community pharmacists’ hypertension-
come measures. The rigorous design of targeted interven- care service on medication adherence. The AFenPA study. Res Soc
tions with more frequent follow-ups, careful consideration Adm Pharm. 2013;9(6):797–805. doi:10.1016/j.sapharm.2012.12.006
18. Fischer MA, Choudhry NK, Bykov K, et al. Pharmacy-based inter-
Journal of Multidisciplinary Healthcare downloaded from https://www.dovepress.com/ by 94.158.22.224 on 10-Aug-2020