See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/349603403

Impact and acceptance of pharmacist-led interventions during HIV care in a

third-level hospital in Spain using the Capacity-Motivation-Opportunity
pharmaceutical care model: The IR...

Article in European Journal of Hospital Pharmacy · February 2021

DOI: 10.1136/ejhpharm-2020-002330

CITATIONS READS

14 72

4 authors, including:

Maria de las Aguas Robustillo-Cortes Antonio Gutiérrez-Pizarraya

Hospital Universitario Nuestra Señora de Valme Junta De Andalucía
52 PUBLICATIONS 254 CITATIONS 58 PUBLICATIONS 1,435 CITATIONS

SEE PROFILE SEE PROFILE

Ramon morillo verdugo

Servicio Andaluz de Salud
177 PUBLICATIONS 1,000 CITATIONS

SEE PROFILE

All content following this page was uploaded by Antonio Gutiérrez-Pizarraya on 07 September 2021.

The user has requested enhancement of the downloaded file.

 Original research

Eur J Hosp Pharm: first published as 10.1136/ejhpharm-2020-002330 on 24 February 2021. Downloaded from http://ejhp.bmj.com/ on March 1, 2021 by guest. Protected by copyright.
Impact and acceptance of pharmacist-­led
interventions during HIV care in a third-­level hospital
in Spain using the Capacity-­Motivation-­Opportunity
pharmaceutical care model: the IRAFE study
M Gracia Cantillana-­Suárez,1 Maria de las Aguas Robustillo-­Cortés,2
Antonio Gutiérrez-­Pizarraya,3 Ramón Morillo-­Verdugo3
1
Pharmacy, Hospital ABSTRACT comorbidities that appear earlier than in the general
Universitario de Cáceres, Introduction In recent decades, HIV has become a population.2–4 Consequently, they usually take more
Cáceres, Spain
2
Pharmacy, Hospital Juan chronic disease with which the HIV specialist pharmacistnon-­antiretroviral drugs, and their drug therapy is
Ramón Jiménez, Huelva, plays a fundamental role. The traditional pharmaceuticalmore complex. In addition, clinical management is
Andalucía, Spain care model followed to date relied excessively on the complicated by the greater risk of drug interactions
3
Pharmacy, Hospital medication, obviating the uniqueness of each patient. and adverse events, adherence problems, falls, and
Universitario Virgen de Valme,
The purpose of this study was to determine the influencealso a greater risk of hospitalisation.5 6
Sevilla, Spain
and acceptance of a Capacity-­Motivation-­Opportunity The multidisciplinary approach to these patients
Correspondence to (CMO)-­based structured pharmaceutical care (PC) is ideal in such cases and the HIV specialist pharma-
Dr Antonio Gutiérrez-­ intervention in a multidisciplinary team for improving cist plays a fundamental role in the care of people
Pizarraya, Pharmacy, Hospital healthcare results. living with HIV.7 The traditional pharmaceutical
Universitario Virgen de Methods Prospective single-­centre study of a care (PC) model relied excessively on the medica-
Valme, 41014 Sevilla, Spain; ​
boticariors@​gmail.​com
structured health intervention with patients living withtion, obviating the uniqueness of each patient.8 For
HIV who attended hospital between January 2017 and that, this concept focused implicitly on the phar-
Received 28 April 2020 June 2018 for any cause. Pharmacotherapeutic follow-­up maceutical activity in the search for individual and
Revised 17 November 2020 was applied according to the CMO PC model based on transversal intervention.9 A redefined model of
Accepted 26 January 2021 three key elements, namely stratification, motivational care must be constructed in which the orientation
EAHP Statement 4: Clinical interview and new technologies. To assess differences into the individual and population needs, efficiency,
Pharmacy Services. the variables collected before and after the intervention,
technical quality, involvement and co-­responsibility,
Student’s t-­test or Wilcoxon test, and McNemar’s test accessibility, and professional integration are the
were used for quantitative and dichotomous variables, key elements.10 In fact, factors such as educational,
respectively. cognitive-­functional, demographic, or use of health
Results A total of 349 patients were included, 76.1% resources, among others, should be taken into
of which were men. The acceptance of pharmacist account when focusing on providing increased value
intervention by both doctors and patients was high to those patients in greater need. Therefore, there
[336 (97.7%) and 321 (93.3%)] and the adherence is a need to stratify our population in order to be
rate to antiretroviral therapy before intervention was able to organise and prioritise resources. Addition-
lower than that observed afterwards (85.6%±33.7% vs ally, a pharmacotherapy-­based relationship must be
96.4%±17.7%; p<0.001). No differences were found established with patients, in which the motivational
between median viral load pre- versus post-­interventioninterview should be used as a key work tool.
[1175 (62.75–26 050) copies/mL vs 274 (76.75– Lastly, we should move away from the idea of
5542) copies/mL], although the undetectability rate was PC being carried out in the presence of the patient,
recorded as higher after intervention compared with the as we can then carry out our activities not only in
previous period [294 (85.5%) vs 274 (79.7%); p<0.001]. the hospital scenario, and not in an episodic way
Conclusions Our results could help HIV pharmacy clinic but continuously and in accordance with patients'
specialists to recognise high-­risk patients and to develop
needs. In addition, real-­time decision-­making with
personalised follow-­up care, thereby ensuring good the support of technology will allow us to be much
adherence and response to treatments. more efficient than previously.
© European Association of
Hospital Pharmacists 2021. No Based on this, Morillo-­ Verdugo et al11 have
commercial re-­use. See rights defined a new PC model denoted by the acronym
and permissions. Published INTRODUCTION “CMO” (for Capacity-­Motivation-­Opportunity),
by BMJ. In recent decades, the life expectancy of people according to three key elements (namely stratifi-
To cite: Cantillana-­ living with HIV (PLWHIV) has increased consid- cation, motivational interview and new technolo-
Suárez MG, Robustillo-­ erably, and now the disease can be considered as gies) that has already been applied successfully in
Cortés MdlA, Gutiérrez-­ 12 13
a chronic one.1 This is a victory on a worldwide ambulatory HIV patients. The CMO model has
Pizarraya A, et al.
scale but is also a challenge, since it is necessary to been used to date to improve health outcomes in
Eur J Hosp Pharm Epub
ahead of print: [please maintain patients’ autonomy and independence as the cardiovascular field.
include Day Month Year]. they grow older. This study contributes a more in-­depth descrip-
doi:10.1136/ In this context of progressive ageing, HIV-­ tion of the different interventions carried out in
ejhpharm-2020-002330 infected people have a greater prevalence of the HIV field and other concomitant pathologies,
Cantillana-­Suárez MG, et al. Eur J Hosp Pharm 2021;0:1–7. doi:10.1136/ejhpharm-2020-002330    1
 Original research

Eur J Hosp Pharm: first published as 10.1136/ejhpharm-2020-002330 on 24 February 2021. Downloaded from http://ejhp.bmj.com/ on March 1, 2021 by guest. Protected by copyright.
METHODS
A prospective single-­ centre study of a structured pharmacist-­ led
health intervention among PLWHIV patients who attended hospital
between January 2017 and June 2018 for any cause was conducted.
Patients received the pharmacotherapeutic follow-­ up routinely
applied to ambulatory care patients according to the CMO PC
model. Patients were excluded if they were participating in a clinical
trial or did not give their written informed consent.
A flowchart illustrating the PC intervention is shown in figure 1
and the schedule of visits and procedures is outlined in figure 2.

Definitions
CMO PC model
This was a pharmacotherapeutic follow-­up of all medication taken
by the patient in order to detect and work towards the achievement
of pharmacotherapeutic objectives related to their prescribed drugs
as well as making recommendations, for example, for improving
diet, exercise and smoking cessation. Patients were given information
leaflets on non-­adherence and healthy habits (including information
regarding smoking cessation) and an individual motivational inter-
view to enhance this particular aspect. Finally, patients were peri-
odically contacted by sending text messages with content related to
healthy living habits and health promotion. Patients who failed to
attend two prearranged pharmacotherapeutic follow-­up visits were
withdrawn from the study and considered as dropouts and were not
replaced by new participants.

Adherence
Adherence to antiretroviral therapy (ART) and concomitant medi-
Figure 1 Flowchart illustrating the pharmaceutical care intervention. cation were measured with the Simplified Medication Adherence
CRVF, cardiovascular risk factors. Questionnaire (SMAQ) and the Morisky–Green questionnaire,
respectively. In addition, pharmacy dispensing records were also
consulted. In both cases, patients were considered adherent if
they obtained a positive score using the appropriate measurement
and also demonstrates the acceptance of these interventions by instrument.
both the patients and the rest of the multidisciplinary team, in a The SMAQ is a questionnaire based on the Morisky–Green–
routine clinical practice setting. Levine questionnaire and developed in our setting that consists of
The purpose of this study was to determine the influence six items that evaluate forgetfulness, routine, adverse events and
and acceptance of a CMO-­ based structured pharmaceutical missing doses.14 The Morisky–Green–Levine questionnaire consists
care intervention within a multidisciplinary team for improving of four items that evaluate forgetfulness, routine, adverse events
healthcare outcomes in PLWHIV patients. and, in contrast to the SMAQ, evaluates the impact of feeling better

Figure 2 Chronology of study follow-­up. PO, pharmacotherapeutical objectives.

2 Cantillana-­Suárez MG, et al. Eur J Hosp Pharm 2021;0:1–7. doi:10.1136/ejhpharm-2020-002330
 Original research

Eur J Hosp Pharm: first published as 10.1136/ejhpharm-2020-002330 on 24 February 2021. Downloaded from http://ejhp.bmj.com/ on March 1, 2021 by guest. Protected by copyright.
Calderón-­Larrañaga et al18 who classified those patterns according
Table 1 Baseline features of patients included in the study
to the type of disease they were intended to treat (cardiovascular,
Total cohort
Baseline feature (n=349) CI (95% CI)
depression-­anxiety, acute respiratory infection, chronic pulmonary
disease, rhinitis-­asthma, pain and menopause). After categorising a
Demographics
drug according to the anatomical therapeutic chemical classification
Age (years) (mean±SD) 48.3±10.7 47.4 to 49.6
system (ATC) up to the first three levels, a patient was categorised
Gender (male) 265 (76.1)
CDC classification (AIDS)* 84 (24.3) 19.8 to 28.8
to a specific pattern when he/she was dispensed at least three drugs
Acquisition risk factor*
included in the pattern.
 Sexual habits 202 (58.4) 52.6 to 62.9
 ADVP‡ 140 (40.5) 35.1 to 45.3
Health outcomes
 Vertical transmission 4 (1.2) 0.45 to 2.91
The consequences of pharmaceutical intervention on health
Comorbidities
outcomes were established with the measure of compliance with
Comorbidities (mean±SD)† 1.53±1.41 1.38 to 1.67
certain objectives, such as those related to dyslipidaemia, hyperten-
 Cardiometabolic 121 (35.2) 29.8 to 39.8
 Geriatric depressive 41 (11.9) 8.8 to 15.5
sion, diabetes and hepatitis C treatment, and defined according to
 Thyroid mechanic 7 (2) 0.9 to 4.1
the corresponding scientific societies' criteria.19
 Various 73 (21.2) 16.9 to 25.4
All information was recorded from the patient’s clinical records
 None 102 (29.7) 24.7 to 34.2 except for the evaluation of therapy compliance that was determined
Multipathological 222 (64.5) 58.4 to 68.4 by patient interview.
Medical acceptance 336 (97.7) 93.7 to 97.8 We recorded demographic data (age, sex), HIV infection control
Patient acceptance 321 (93.3) 88.6 to 94.3 variables as viral load (copies/mL) and CD4 count at the time of
Category inclusion (cells/μL), as well as comorbidities and pharmacological
 Capability 145 (41.5) 36.5 to 46.8 therapy.
 Opportunity 107 (30.7) 26.1 to 35.7 In addition, the latter included ART and other medication
 Motivation 97 (27.8) 23.3 to 32.7 for complications and/or comorbidities. Clinical variables and
Unless otherwise stated, values are given as number (percentage). pharmacotherapeutics, such as type of ART therapy, concom-
*Missing values=3.
†Missing values=5. itant medications prescribed and adherence (pre- and post-­
‡ADPV denotes parenteral drug addiction. intervention), switching treatment and polypharmacy, were also
CDC, Centers for Disease Control and Prevention.
recorded.
The rest of the required information was obtained during the
and does not evaluate missing doses; we used the Spanish validated interview, held at the pharmacy unit during the periodic ART medi-
version.15 cation dispensing visit in accordance with the methodology stipu-
Adherence rate was quantified as the proportion of days covered lated in the study protocol.
(PDC) during the 6 months prior to the study according to filled The main objective of the study was to assess the percentage
e-­prescriptions . We estimated the total days of supplies from the of intervention acceptance both by the patient and by the rest of
first to the last refill during the 6-­month observation period divided the multidisciplinary team, after 3 months of follow-­up after study
by the total days of the treatment interval, defined as the time elapsed inclusion. As secondary objectives we considered the following: the
from the date of the first refilled prescription until the end of the percentage of patients who increased adherence to HIV and non-­
observation period. A PLWHIV was considered as adherent if the HIV treatments and who achieved optimal virological control.
PDC was >95% and not positive on the SMAQ (where positive Additionally, we evaluated the influence on health outcomes such
means that there was a positive response to any of the qualitative as dyslipidaemia, hypertension, diabetes and hepatitis C coinfection,
questions), no more than two doses were missed over the past week, will be determined for each patient. The study was approved by the
or they had fewer than 2 days of total non-­medication during the ethics committee “Comité Ético de Investigación del Sur de Sevilla”
past 3 months. (Sevilla, Spain) (Code FAR-­VIH-2017–01).
To evaluate adherence to concomitant medication, we only consid-
ered disease-­ modifying medications (eg, treatment for diabetes,
cardiovascular disease, etc.) but not symptomatic treatments (eg, Statistical analysis
analgesics, medications for gastro-­oesophageal reflux, etc.). Adher- The quantitative variables were expressed as means±standard devia-
ence to concomitant medication was defined as a PDC >90% and tions (SDs), or medians and interquartile ranges (IQRs) when appro-
also a Morisky–Green–Levine questionnaire score ≥4. priate, and qualitative variables as counts (percentages).
To assess the differences in the variables collected before and
Patient’s stratification after the intervention, a Student's t-­test or Wilcoxon test for related
PC variables like stratification of patients was performed according groups was carried out to compare quantitative variables. McNe-
to the risk-­stratified model for PC in HIV patients of the Spanish mar’s test was applied to analyse changes in the dichotomous
Society of Hospital Pharmacy.16 Interventions were classified variables.
according to the taxonomy for pharmaceutical interventions in For analysis purposes, the CD4 level was dichotomised with a cut-­
HIV+ patients based on the CMO model.17 off point set at 350 cells/mL or more, the undetectability of the viral
load as a serum level of fewer than 50 copies/mL and adherence was
Polypharmacy considered as achieved when it was at least 95%. Significant differ-
Polypharmacy was defined as the use of six or more different ences were quantified with 95% confidence intervals. The threshold
drugs, including antiretroviral medication; major polypharmacy for statistical significance was defined as p<0.05. Data analysis was
was restricted to the use of ≥11 different drugs. To describe the performed with IBM SPSS 25.0 statistical software (IBM Corp.,
patterns of polypharmacy we used the categorisation proposed by Armonk, NY, USA).
Cantillana-­Suárez MG, et al. Eur J Hosp Pharm 2021;0:1–7. doi:10.1136/ejhpharm-2020-002330 3
 Original research

Eur J Hosp Pharm: first published as 10.1136/ejhpharm-2020-002330 on 24 February 2021. Downloaded from http://ejhp.bmj.com/ on March 1, 2021 by guest. Protected by copyright.
Figure 3 Types of pharmacist intervention. ART, antiretroviral therapy.

RESULTS
A total of 349 patients were included in the study, of which 76.1%
Table 2 Pharmacotherapy pre- and post-­intervention were men with a mean age of 48.3±10.7 years. Of all the patients
Total (n=349) included in the study, 69.3% had at least one comorbidity, the
cardiometabolic type being the most frequently observed with a
Variable Pre-­intervention Post-­intervention P value
prevalance of 35.2%. At baseline, the percentage of patients with
ART status* concomitant medication was 76.8% with an average of 2.8±2.9
 Naive 40 (11.6) – – drugs per patient. The most prescribed pharmacotherapeutic groups
 Rescue (≤2 ART agents) 119 (34.5) – – were hypoglycaemic agents (18, 5.2%), antihypertensives (82,
 Multi-­failure (>3 ART agents) 186 (53.9) – – 23.8%) and lipid-­lowering drugs (85, 24.6%).
ART type† Regarding the primary endpoint, the level of acceptance of phar-
 ITIAN+Inin 95 (27.5) 104 (30.2) 0.078 macist intervention by the medical professionals and the patient was
 ITIAN+ITINN 76 (22) 73 (21.2) 0.791 high [336 (97.7%) and 321 (93.3%), respectively]. The remaining
 ITIAN+PI 50 (14.5) 48 (14) 1 demographic and clinical characteristics of the patients are shown
 Others 125 (36.1) 119 (34.6) 0.146 in table 1.
Adherence to ART (%) (mean±SD) 85.6±33.7 96.4±17.7 <0.001 The most common pharmacist interventions were: concomitant
Adherence to ART (>95%)* 278 (80.6) 330 (95.9)‡ <0.001 medication review and validation (121, 34.7%), direct communica-
ART duration (unit) (mean±SD) 89.5±6.8 -- – tion (98, 28.1%), commitment (49, 14%) and adherence (28, 8%).
Concomitant medication*§ 265 (76.8) – – The rest of the interventions are shown in figure 3.
 Lipid-­lowering drugs 85 (24.6) – – At the time of the intervention and afterwards there were no differ-
 Hypoglycaemic agents 18 (5.2) – – ences between stratification levels, which were respectively: N3 [274
 Antihypertensives 82 (23.8) – – (78.5%) vs 272 (77.9%)], N2 [38 (10.9%) vs 35 (10.0%)] and N1 [34
 Others 227 (65.8) – – (9.7%) vs 37 (10.6%)]. No differences were found between median
Polymedicated* 111 (32.2) viral load pre- versus post-­intervention [1175 (62.75–26 050) copies/
Concomitant drugs (mean±SD) 2.85±2.93 mL vs 274 (76.75–5542) copies/mL] although the undetectability
Unless otherwise stated, values are given as number (percentage). rate was recorded as higher after the intervention compared with
*Missing values=4. the previous period [294 (85.5%) vs 274 (79.7%); p<0.001]. The
†Missing values=3.
‡Intragroup differences were assessed pre- versus post-­intervention by McNemar test or
measurement of CD4 levels showed a lower level before pharmacist
Wilcoxon test for dichotomous and quantititative variables, respectively. intervention (271, 78.8% vs 294, 85.5%; p<0.001).
§59 patients had more than one concomitant medication. In relation to ART (table 2), the initial status distribution was
ART, antiretroviral therapy; Inin, integrase inhibitors; ITIAN, nucleoside analogue reverse
transcriptase inhibitors; ITINN, non-­nucleoside reverse transcriptase inhibitors; PI, protease observed as follows: naive (40, 11.6%), rescue (119, 34.5%) and
inhibitors. multi-­failure (186, 53.9%).
4 Cantillana-­Suárez MG, et al. Eur J Hosp Pharm 2021;0:1–7. doi:10.1136/ejhpharm-2020-002330
 Original research

Eur J Hosp Pharm: first published as 10.1136/ejhpharm-2020-002330 on 24 February 2021. Downloaded from http://ejhp.bmj.com/ on March 1, 2021 by guest. Protected by copyright.
The percentage of patients with undetectable viral load and
Table 3 Pre- and post-­intervention bivariate analysis
with seric CD4 count >350 cells/mL increased significantly after
Total cohort (n=349) the pharmaceutical intervention. Although the viroimmunolog-
Parameter Pre-­intervention Post-­intervention P value ical control of the patients is key, in recent years it is working,
Dyslipidaemia target 50/85 (58.8) 65/85 (76.5) <0.001 according to the specific guidelines, in increasing the therapeutic
achievement* results beyond merely control of the infection.20 This is why it is
AHT target 40/82 (48.7) 52/82 (70.7) <0.001 important to have a comprehensive vision of pharmacotherapy
achievement*† and direct all efforts towards the control of the disease and its
Diabetes mellitus target 9/18 (50.0) 14/18 (77.7) <0.001 comorbidities.
achievement‡ Importantly, we also determined that polypharmacy in HIV
Stratification§ patients is significantly related and that adherence among these
 N3 274 (78.5) 272 (77.9) NS patients might be particularly different. Thus, PLWHIV patients
 N2 38 (10.9) 35 (10.0) NS were more adherent to their ART drugs but less to comedication.
 N1 34 (9.7) 37 (10.6) NS It is known that patients receiving several concomitant drugs
CD4 level (≥350 cells/ 271 (78.8)†† 294 (85.5) <0.001 tended to have less adherence to other prescribed treatments.21
mL) The intake of numerous medications for comorbidities is a
Viral load (copies/mL) 1175 (62.75–26 274 (76.75–5542) 0.211 frequent event in patients receiving HIV treatment. Since discon-
(median, IQR)¶ 050)‡‡ tinuation of comedication may lead to major health problems,
Undetectable viral load 274 (79.7)** 318 (92.4) <0.001 it is essential to determine whether longitudinal PC interven-
(<50 copies/mL)
tions in a multidisciplinary team might influence adherence to
Unless otherwise stated, values are given as number (percentage). concomitant medications. It is particularly relevant to highlight
*Pre NA 248 (71.9%), post NA 245 (71.2%).
†Pre NA 279 (80.9%), post NA 278 (80.8%).
the longitudinal, non-­ transverse and patient-­ centred aspect
‡Pre NA 312 (90.4%), post NA 311 (90.4%). of this work model. In this sense, the clinical pharmacist is
§Missing values=3. uniquely positioned to help patients manage their medications
¶Patients with undetectable viral load were excluded from the calculation. and provide adherence, motivational and new technologies skills
**Differences were found between pre- vs post-­intervention undetectability rate
support.
(McNemar test=30.81, p<0.001).
††Differences were found between pre- vs post-­intervention CD4 level (McNemar In our study, cardiometabolic disease was the most prevalent
test=11.25, p<0.001). comorbidity. Accordingly, we found that the most commonly
‡‡No differences were found between pre- vs post-­intervention viral load (Wilcoxon prescribed comedications were lipid-­lowering drugs (85, 24.6%)
test p=0.211). and hypoglycaemic agents (18, 5.2%). This is consistent with
AHT, arterial hypertension; ART, antiretroviral therapy; IQR, interquartile range; NA,
not applicable; NS, not significant; PDA, parenteral drug addiction.
previous studies in which an association was found between
HIV infection and the appearance of vascular, cognitive and
metabolic comorbidities with similar prevalence rates.22–24 In
a previous multicentric study, with a very similar preliminary
The usage of different types of combinations were similar for
methodology study, it was shown that this PC intervention might
both pre- and post-­intervention periods, respectively: nucleoside
lead to improved health outcomes in HIV+ patients at greater
analogue reverse transcriptase inhibitors plus integrase inhibitors
cardiovascular risk.12
(ITIAN+Inin), 95 (27.5%) versus 104 (30.2%); ITIAN plus non-­
In our study, since non-­adherence may be related to poly-
nucleoside reverse transcriptase inhibitors (ITINN), 76 (22%)
pharmacy, which may negatively impact therapeutic success, it is
versus 73 (21.2%); ITIAN+protease inhibitors (PI), 50 (14.5%)
important to closely monitor patients at high risk for poor medi-
versus 48 (14%); and others, 50 (14.5%) versus 48 (14%). The
cation adherence, and to choose appropriate interventions to
polymedication rate at the beginning was 32.2%.
improve compliance. In this sense, the pharmacy clinic staff are
In the whole group of patients, the adherence rate to ART
uniquely positioned to help patients manage their medications
before intervention was lower than that observed afterwards
and provide adherence support. Another important character-
(85.6%±33.7% vs 96.4%±17.7%; p<0.001). In addition, the
istic of this model is that by stratifying the population served and
percentage of patients with an adequate adherence to ART (PDC
having defined what type of interventions to carry out specifi-
>95%) increased significantly by 15.4% after pharmacist CMO-­
cally for each level of complexity, this new way of attending to
based intervention compared with the pre-­intervention situation
patients allows more time to be dedicated to those who need it
[330 (95.9%) vs 278 (80.6%); p<0.001].
and enables better planning of the care bundle.
The analysis of the relationship between patients’ objectives
Given the known association between polypharmacy and low
achievement rates prior to pharmacist intervention and after-
adherence rates to concomitant medications, the choice of ART
wards is shown in table 3. The percentage of achievement of
treatment and concomitant drugs is another variable that nowa-
the dyslipidaemia objective was higher after pharmacist inter-
days should be considered, especially in patients with multiple
vention than before (76.5% vs 58.8%: p<0.001) and the same
comorbidities. Physicians should prescribe the antiviral treat-
results were recorded for the arterial hypertension (AHT) target
ment that includes the least number of pills whenever possible
achievement rate (48.7% vs 70.7%; p<0.001) and diabetes
to increase the likelihood of patient adherence to the concom-
mellitus objective (50% vs 77.7%; p<0.001). No patients
itant treatment. Pharmacists should guide physicians’ efforts to
receiving hepatitis C virus (HCV) therapy were recorded.
optimise pharmacotherapy. Conversely, it is essential that the
patient be able to deal with the complexity of the prescribed
DISCUSSION pharmacotherapy since it is related to adherence and, therefore,
Our study found that the CMO PC model applied to PLWHIV the achievement of pharmacotherapeutic objectives.25
has a positive influence on healthcare results and a high level of Nimarko et al recently demonstrated that pharmacists can
acceptance in a multidisciplinary HIV care team, and particularly decrease the frequency of antiretroviral (ARV) errors and the
impacts on adherence to ART and to concomitant medications. need to incorporate such reviews in well-­established stewardship
Cantillana-­Suárez MG, et al. Eur J Hosp Pharm 2021;0:1–7. doi:10.1136/ejhpharm-2020-002330 5
 Original research

Eur J Hosp Pharm: first published as 10.1136/ejhpharm-2020-002330 on 24 February 2021. Downloaded from http://ejhp.bmj.com/ on March 1, 2021 by guest. Protected by copyright.
programmes.26 It also reveals the shift in the use of PI-­based regi- methodology that takes into account their needs and individual
mens to Inin-­based regimens. With this shift, there remains a risk characteristics and that is not focused on the prescribed medica-
of harmful ARV errors during hospitalisations, with the most tion, stressing the importance of an effective patient care model
common of these being drug interactions. Other authors have to closely monitor high-­risk patients.
also gone in the same direction.27 28 Even Robustillo et al showed As PLWHIV are becoming increasingly complex, and include
how hospital admission is a factor associated with the increase in both young people and older patients, it is increasingly neces-
pharmacotherapeutic complexity and, therefore, a higher possi- sary to individualise healthcare by offering a work system more
bility of errors.29 Our PC model includes not only work on the oriented to patients' individual needs. We have shown that the
outpatient, but also the admitted patient and their risk of read- CMO PC model is a methodology that improves adherence and
mission, as one of the assessments to be carried out within the the achievement of pharmacotherapeutic objectives and has
multidisciplinary team.30 high acceptability to both patients and the rest of the multidis-
The taxonomy of interventions published for CMO does not ciplinary team.11 One of the most important pending challenges
specifically include a line on medication errors that is already for the future is to adapt this methodology to the needs of the
included in others such as the review and validation of both ART multidimensional approach needed by PLWHIV, especially those
and concomitant medication. Additionally, the enhancement of older patients, who are ageing. For this reason, it will be neces-
direct communication with patients, through new technolo- sary to incorporate new concepts and strategies of joint work to
gies, included in the concept of ‘opportunity’ of the model can carry out interventions of the type of deprescription.31 32
prevent the appearance of errors, by the pharmacist intervening In conclusion, this knowledge will help HIV pharmacy clinic
with patients before the error occurs. specialists to recognise high-­risk patients and to develop person-
This study has several strengths including a large sample size alised follow-­ up care, thereby ensuring good adherence and
and the evaluation of multiple variables related to adherence, response to treatments, thus increasing the value of the contri-
concomitant medications and other not previously assessed bution of the pharmacist within the multidisciplinary teams that
factors (motivational interview, coordination, etc.). However, it care for the PLWHIV population.
also has some limitations. Adherence rates were obtained using
pharmacy dispensing records and the SMAQ and Morisky score Twitter Maria de las Aguas Robustillo-­Cortés @awina87
which despite being widely used in clinical practice, are known Contributors MGC-­S, MAR-­C and RAM-­V substantially contributed to study
to overestimate rates. Also the study was validated at a time when conception and design. MGC-­S and MAR-­C contributed to data acquisition. RAM-­V
most patients were taking treatments based on protease inhibi- contributed to data acquisition, analysis and interpretation. AG-­P contributed to
tors, the use of which has now been reduced. Another limita- study design and analysis as well as data interpretation. All the authors critically
tion of the study is its single-­centre nature. Once the number of revised the manuscript and approved the final manuscript.
hospitals working with this methodology has been expanded, it Funding The authors have not declared a specific grant for this research from any
will be interesting, as a future line of research, to develop multi- funding agency in the public, commercial or not-­for-­profit sectors.
centric research that allows us to contrast the data obtained and, Competing interests None declared.
in addition, to further profile the interventions and the type of Patient consent for publication Not required.
patient that will potentially be a candidate for closer monitoring Provenance and peer review Not commissioned; externally peer reviewed.
by the multidisciplinary team. In addition, we recognise that the
Data availability statement Data are available upon reasonable request.
information regarding the change of concomitant medication
has not been recorded. However, when dealing with patients
with a high average age, the change seems unlikely given that REFERENCES
most are prescribed for chronic diseases. 1 Antiretroviral Therapy Cohort Collaboration. Survival of HIV-­positive patients starting
Despite these limitations, our study has successfully identified antiretroviral therapy between 1996 and 2013: a collaborative analysis of cohort
the HIV pharmacy clinic specialist interventions to be carried studies. Lancet HIV 2017;4:e349–56.
2 Guaraldi G, Orlando G, Zona S, et al. Premature age-­related comorbidities among
out more frequently and intensively in current patients, with a HIV-­infected persons compared with the general population. Clin Infect Dis
2011;53:1120–6.
3 Schouten J, Wit FW, Stolte IG, et al. Cross-­sectional comparison of the prevalence
What this paper adds of age-­associated comorbidities and their risk factors between HIV-­infected and
uninfected individuals: the AGEhIV cohort study. Clin Infect Dis 2014;59:1787–97.
4 Smit M, Brinkman K, Geerlings S, et al. Future challenges for clinical care of an ageing
What is already known on this subject
population infected with HIV: a modelling study. Lancet Infect Dis
►► The multidisciplinary approach to HIV patients is undoubtedly 2015;15:810–8.
the best approach for improving healthcare results. The 5 McNicholl IR, Gandhi M, Hare CB, et al. A pharmacist-­led program to evaluate and
traditional pharmaceutical care model of care followed to reduce polypharmacy and potentially inappropriate prescribing in older HIV-­positive
date relied excessively on the medication, obviating the patients. Pharmacotherapy 2017;37:1498–506.
6 Palepu A, Sun H, Kuyper L, et al. Predictors of early hospital readmission in HIV-­
uniqueness of each patient. infected patients with pneumonia. J Gen Intern Med 2003;18:242–7.
7 Schafer JJ, Gill TK, Sherman EM, et al. ASHP guidelines on pharmacist involvement in
What this study adds
HIV care. Am J Health Syst Pharm 2016;73:468–94.
►► This is the first study specifically to investigate the use of 8 Hepler CD, Strand LM. Opportunities and responsibilities in pharmaceutical care. Am J
Capacity-­Motivation-­Opportunity (CMO) methodology to Hosp Pharm 1990;47:533–43.
determinate its positive influence on healthcare results in HIV 9 Moltó-Puigmartí C, Vonk R, van Ommeren G, et al. A logic model for pharmaceutical
patients. care. J Health Serv Res Policy 2018;23:148–57.
10 Blackburn DF, Yakiwchuk EM, Jorgenson DJ, et al. Proposing a redefinition of
►► The results of the study have clearly illustrated that
pharmaceutical care. Ann Pharmacother 2012;46:447–9.
CMO methodology has a high level of acceptance in a 11 Morillo-­Verdugo R, Calleja-­Hernández Miguel Ángel, de Las Aguas Robustillo-­Cortés
multidisciplinary HIV care team, particularly impacting on M. A new pharmaceutical care concept: more capable, motivated, and timely. Hosp
adherence to antiretroviral therapy and to concomitant Pharm 2019;54:348–50.
medications. 12 Morillo-­Verdugo R, Robustillo-­Cortés MdeLA, Martín-­Conde MT, et al. Effect of a
structured pharmaceutical care intervention versus usual care on cardiovascular risk

6 Cantillana-­Suárez MG, et al. Eur J Hosp Pharm 2021;0:1–7. doi:10.1136/ejhpharm-2020-002330

 Original research

Eur J Hosp Pharm: first published as 10.1136/ejhpharm-2020-002330 on 24 February 2021. Downloaded from http://ejhp.bmj.com/ on March 1, 2021 by guest. Protected by copyright.
in HIV patients on antiretroviral therapy: INFAMERICA study. Ann Pharmacother 22 López-­Centeno B, Badenes-­Olmedo C. Polypharmacy and drug-­drug interactions in
2018;52:1098–108. HIV-­infected subjects in the region of Madrid, Spain: a population-­based study. Clin
13 Morillo-­Verdugo R, Robustillo-­Cortés MA, Manzano García M, et al. Influence of Infect Dis 2020;71:353–62.
pharmacist intervention, based on CMO model, to improve activation in HIV patients. 23 Ellenbogen MI, Wang P, Overton HN, et al. Frequency and predictors of polypharmacy
Rev Esp Quimioter 2019;32:40–9. in US Medicare patients: a cross-­sectional analysis at the patient and physician levels.
14 Knobel H, Alonso J, Casado JL, et al. Validation of a simplified medication adherence Drugs Aging 2020;37:57–65.
24 Siefried KJ, Mao L, Cysique LA, et al. Concomitant medication polypharmacy,
questionnaire in a large cohort of HIV-­infected patients: the GEEMA study. AIDS
interactions and imperfect adherence are common in Australian adults on suppressive
2002;16:605–13.
antiretroviral therapy. AIDS 2018;32:35–48.
15 Morisky DE, Green LW, Levine DM. Concurrent and predictive validity of a self-­ 25 Manzano-­García M, Pérez-­Guerrero C, Álvarez de Sotomayor Paz M, et al.
reported measure of medication adherence. Med Care 1986;24:67–74. Identification of the medication regimen complexity index as an associated factor of
16 Morillo-­Verdugo R, Martínez-­Sesmero JM, Lázaro-­López A, et al. Development of a nonadherence to antiretroviral treatment in HIV positive patients. Ann Pharmacother
risk stratification model for pharmaceutical care in HIV patients. Farm Hosp 2018;52:862–7.
2017;41:346–56. 26 Nimarko K, Bandali A, Bias TE, et al. Impact of an antimicrobial stewardship team on
17 Morillo Verdugo R, Villarreal Arevalo AL, Alvarez De Sotomayor M, et al. Development reducing antiretroviral medication errors. Ann Pharmacother 2020;54:767–74.
of a taxonomy for pharmaceutical interventions in HIV+ patients based on the CMO 27 DePuy AM, Samuel R, Mohrien KM, et al. Impact of an antiretroviral stewardship team
model. Farm Hosp 2016;40:544–68. on the care of patients with human immunodeficiency virus infection admitted to an
18 Calderón-­Larrañaga A, Gimeno-­Feliu LA, González-­Rubio F, et al. Polypharmacy academic medical center. Open Forum Infect Dis 2019;6:ofz290.
patterns: unravelling systematic associations between prescribed medications. PLoS 28 Li EH, Foisy MM. Antiretroviral and medication errors in hospitalized HIV-­positive
One 2013;8:e84967. patients. Ann Pharmacother 2014;48:998–1010.
19 European AIDS Clinical Society (EACS). European guidelines for treatment of 29 Robustillo Cortés MdeLA, Morillo Verdugo R, Barreiro Fernández EM, et al. Influence
of hospital admission in the pharmacotherapy complexity of HIV+ patients. Farm Hosp
HIV-­positive adults in Europe, edition 10.0. Available: http://www.​eacsociety.​org/​
2017;41:518–26.
guidelines/e​ acs-​guidelines/e​ acs-​guidelines.​html [Accessed 13 Mar 2020].
30 Montes-­Escalante I, Monje-­Agudo P, Calvo-­Cidoncha E, et al. Design and validation
20 Del Amo J, Campbell C, Navarro G, et al. [HIV in Spain 2017: policies for a new of a predictive model for 1-­year hospital admission in HIV patients on antiretroviral
management of chronicity beyond virological control]. Rev Esp Salud Publica 2018;92. treatment. Eur J Hosp Pharm 2016;23:224–7.
[Epub ahead of print: 06 Sep 2018]. 31 Blanco J-­R, Morillo R, Abril V, et al. Deprescribing of non-­antiretroviral therapy in HIV-­
21 Cantudo-­Cuenca MR, Jiménez-­Galán R, Almeida-­Gonzalez CV, et al. Concurrent use infected patients. Eur J Clin Pharmacol 2020;76:305–18.
of comedications reduces adherence to antiretroviral therapy among HIV-­infected 32 Negredo E, Back D, Blanco J-­R, et al. Aging in HIV-­infected subjects: a new scenario
patients. J Manag Care Spec Pharm 2014;20:844–50. and a new view. Biomed Res Int 2017;2017:1–9.

Cantillana-­Suárez MG, et al. Eur J Hosp Pharm 2021;0:1–7. doi:10.1136/ejhpharm-2020-002330 7

View publication stats