VALUE IN HEALTH REGIONAL ISSUES 17 (2018) 119–125

Available online at www.sciencedirect.com

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

HEALTH POLICY ANALYSIS

Potentially Inappropriate Medications in Elderly Ambulatory
Patients: A Comparative Study between a Primary Health Care
Center and a Community Pharmacy
Luciana C. Chiapella1, Jorgelina Montemarani Menna2, María Eugenia Mamprin, PhD1,*
1
Área Farmacología, Facultad de Ciencias Bioquímicas y Farmacéuticas, Universidad Nacional de Rosario, CONICET, Rosario, Santa
Fe, Argentina; 2Atención Primaria de la Salud, Nodo Rosario, Ministerio de Salud de la Provincia de Santa Fe, Rosario, Santa Fe, Argentina

AB STR A CT

Objective: This study aims to compare, both qualitatively and quan- dispensed to 51 patients (19.39%) were identified as PIMs. In the PHC,
titatively, the medication dispensed to elderly patients in a primary 72 prescriptions (8.13%) dispensed to 27 patients (28.42%) were
health care center (PHC) and a community pharmacy (CP) in Argentina identified as PIMs. In patients with major polymedication the possi-
and to identify the prescription of potentially inappropriate medica- bility of consuming these drugs was 2.55 times higher in the CP and
tions (PIMs). Methods: A cross-sectional observational study. Data 2.60 times higher in the PHC. The percentage of PIM prescriptions was
were acquired from 886 prescriptions in the PHC and 2368 in the CP significantly higher in the PHC, although the percentage of patients
between February and April 2015. Dispensed medications were coded receiving them did not differ significantly. Conclusions: The prevalence
according to the Anatomical, Therapeutic, and Chemical (ATC) clas- of PIMs found in this population is relevant enough to implement
sification system. The frequency of prescriptions for each of them was measures that address the problem in an integral way, to improve the
determined. The number and monthly average of drugs dispensed quality of prescriptions and the health outcomes of patients.
were calculated for each patient. The use of PIMs was identified using Keywords: aged, drug use, inappropriate medication, polypharmacy,
Beers Criteria. Results: In both institutions, the means of medications prescription of drugs.
dispensed per individual and month were similar: 3.69 ± 1.93 in the
PHC and 3.46 ± 2.18 in the CP. Most of the medications corresponded Copyright & 2018, International Society for Pharmacoeconomics and
to cardiovascular system agents. In the CP, 111 prescriptions (4.69%) Outcomes Research (ISPOR). Published by Elsevier Inc.

and physiological changes developing with age, which modify the

Introduction
pharmacodynamics and pharmacokinetics of drugs. This leads to an
The World Health Organization has recognized the need to establish increase in the consumption of medications, which in turn enhances
a national drug policy and the importance of an associated research the probability of the presence of unfavorable pharmacological
strategy that includes studies on the use of medicines [1]. This policy interactions and adverse reactions [2]. It has also been reported that
should aim at reaching a rational use of drugs, understood as the as many as 31% of hospital admissions of geriatric patients are
one with the greatest benefit to the patient, at the lowest cost directly related to problems with drug therapy. Such drug-related
possible for both the patient and the community. morbidity represents a substantial economic burden [3].
Medicines are therapeutic tools to cure acute diseases, stabi- Agents in which the risk of adverse events in older adults
lize chronic conditions, and sometimes save a patient’s life or exceeds the expectations of clinical benefits (although there are
improve its quality. Medications can, however, also cause more effective and safer alternatives) are usually denominated
adverse reactions, interact with other drugs, and sometimes in potentially inappropriate medications (PIMs) [4,5]. Several criteria
some patients be contraindicated. have been developed to identify PIM usage in older people, with
In particular, elderly people form a population that requires special Beers Criteria [4] being the most widely used. These criteria
attention as to the quality of drug prescription. This group is require low-level training and do not need a review of the clinical
characterized by pluripathology, functional and cognitive impairment, records of the patients [6].

The authors have indicated that they have no conflicts of interest with regard to the content of this article.
* Address correspondence to: María Eugenia Mamprin, Área Farmacología, Facultad de Ciencias Bioquímicas y Farmacéuticas,
Universidad Nacional de Rosario, CONICET, Suipacha 531, Rosario S2002LRK, Santa Fe, Argentina.
E-mail: mmamprin@fbioyf.unr.edu.ar
2212-1099$36.00 – see front matter Copyright & 2018, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
https://doi.org/10.1016/j.vhri.2017.12.009
120 VALUE IN HEALTH REGIONAL ISSUES 17 (2018) 119–125

It has been reported that polypharmacy increases the risk of polypharmacy, drug interactions, and adverse drug reactions,
consuming PIMs and that some drug interactions may intensify thereby improving the risk–benefit ratio of medication regimens
the risk of problems related to the health of the users [7]. In in at-risk people. They include 53 medications and medication
Argentina, there has been a fast growth in the elderly population classes divided into three categories: PIMs and classes to avoid in
(according to the 2010 Census, they represented approximately all older adults, PIMs and classes to avoid in older adults with
12.8% of the total population, whereas, according to recent certain diseases and syndromes that the drugs listed may
estimates [8], they represent approximately 17.4%) and only a exacerbate, and medications to be used with caution in older
few studies, in institutionalized elderly people, have focused on adults. In this study, we considered drugs not recommended for
the evaluation of the prescription of PIMs in this population [9]. use in the elderly, with high and moderate quality of evidence
Based on the importance of identifying PIMs and associated and independence of diagnosis.
factors, as well as on the lack of published articles in Argentina
and South America in general, in the present study, we aimed to
compare, both qualitatively and quantitatively, the medications Statistical Analysis
dispensed to elderly patients in a primary health care center Descriptive statistics were used to summarize the characteristics
(PHC) and in a community pharmacy (CP) in the city of Rosario, of patients and the use of medications.
Santa Fe, Argentina. For this purpose, in the first section of this Quantitative variables were reported as means ± SDs and
article, we describe the characteristics of patients. Second, we median (min–max), and categorical variables as proportions (%).
determine the frequency of prescription of dispensed drugs and For quantitative variables, Student’s t test for two independent
describe the medication consumption. Finally, we identify PIMs, samples was used to compare means, and the Mann–Whitney
estimate their prevalence, and evaluate their possible relation U test was performed to contrast medians between patients for
with polypharmacy, sex, and age of patients. both dispensation sites. Pearson’s chi-square test was used to
analyze the possible relation between qualitative variables. If the
association was statistically significant (P o 0.05), logistic regres-
sions were used to determine the odds ratio (OR) [14] and their
Methods
confidence interval (95% CI).
To contrast the characteristics of PIM users and nonusers
Design and Data Collection within each dispensation sites, we used the same methods
A cross-sectional observational study was conducted between Feb- described in the foregoing. Multivariate logistic regressions
ruary and April 2015. Data were acquired from 2368 prescriptions in were used to determine patient factor associated with the use
the CP from 263 patients affiliated with the National Institute of of at least one PIM, considering the variables that were found
Social Services for Retirees and Pensioners (INSSJP)—the Argentine to be significant (P o 0.1) in the bivariate analysis. We also
public agency specializing in the care of the elderly—and 886 calculated the OR and its 95% CI to describe the significant
prescriptions corresponding to 95 patients treated in a PHC that associations between the use of PIM and the characteristics of
did not have trade union–run medical insurance. All patients were 60 the patients. Statistical analysis was performed using SAS
years old and older (the cut-off age for elderly people in developing University Edition.
countries), and data recorded from each prescription were year of
birth and sex of the patient, date of dispensation, and medications
dispensed. The prescriptions of psychotropic drugs were excluded
Results
because they are not dispensed at PHCs.
The medications prescribed at the PHC are included in the
Provincial Therapeutic Formulary of Santa Fe, Argentina [10], Patients’ Characteristics
which consists of a list of essential medicines based on recom- We analyzed 886 prescriptions from the PHC and 2368 from the CP.
mendations of the World Health Organization and are available The mean age of the 95 patients of the first group was 67.47 years old
for free. In the case of the dispensations made at the CP to (SD ¼ 6.85 years old), with values between 60 and 94 years old, and
members of the National Institute of Social Services for Retirees 50% of the patients were aged 65 or younger. In the CP, the mean age
and Pensioners, doctors may prescribe any medication within the of the 263 patients was 75.84 years old (SD ¼ 7.90 years old), with
National Vademecum [11], an official source in permanent values between 60 and 94 years old, and 50% of the patients were
update, in which all the medicines currently marketed in Argen- aged 75 or younger. The differences between the mean ages as well
tina are published. These patients must receive their medicines as the median were statistically significant (P o 0.001 in both cases).
for free or with a 50% discount at the pharmacy. Of the patients who attended the PHC, 56.84% were women, whereas
65.40% of those who attended the CP were women. The data did not
provide sufficient evidence to reject the hypothesis that the propor-
Classification of Drugs and Identification of PIMs tion of women attending both institutions is the same (P ¼ 0.145).
The dispensed medications were coded according to the Ana- At the PHC, one to nine drugs per patient were dispensed
tomical, Therapeutic, and Chemical classification (the ATC code), monthly, with an average of 3.69 medications per individual and
version 2016 [12]. For each patient, the number of drugs dis- month (SD ¼ 1.93) and a median of three drugs. Considering the
pensed monthly and the monthly average were determined. A levels defined by Bjerrum et al. [13], 58.95% of the patients presented
mean value of two to four drugs was considered minor poly- minor polymedication (two to four drugs), whereas 29.47% presented
medication, whereas a mean value of five or more drugs was major polymedication (five or more drugs). At the CP, 1 to 14
considered major polymedication [13]. medications were dispensed per month and patient, with an average
To detect the use of PIMs, we used the Beers Criteria corre- of 3.46 medications per individual and month (SD ¼ 2.18) and a
sponding to the review published in 2012 [4]. The Beers Criteria median of three drugs. Considering the levels defined by Bjerrum
for Potentially Inappropriate Medication Use in Older Adults, et al. [13], 53.61% of the individuals presented minor polymedication
commonly called the Beers List, are guidelines for health care and 27.00% presented major polymedication.
professionals to help improve the safety of prescribing medica- No significant differences were found in the average number
tions for older adults. They emphasize deprescribing medications of medications dispensed monthly per patient or in the median
that are unnecessary, which helps to reduce the problems of (P ¼ 0.338 and P ¼ 0.210 respectively). Also, no significant
 VALUE IN HEALTH REGIONAL ISSUES 17 (2018) 119–125 121

Table 1 – Characteristics of the elderly population than for men (OR ¼ 1.81, 95% CI 1.01:3.25). Similarly, at the PHC,
studied. the chance of using nonsteroidal anti-inflammatory drugs
(NSAIDs, group M01A) was eight times greater for women than
Characteristics Primary health Community for men (OR ¼ 8.00, 95% CI 1.01:65.79), whereas no significant
care center pharmacy effect of sex on this group was found at the CP. Regarding lipid-
(N ¼ 95) (N ¼ 263) modifying agents (group C10A), the odds of consumption at the
CP was 1.88 higher for men than for women (OR ¼ 1.88, 95%CI
Gender: Female 54 (56.84%) 172 (65.40%)
1.07:3.29). No relationship was found between the use of NSAIDs
Age: ≥75 years 12 (12.63%) 133 (50.57%)*
and antacids (group A02B) (P ¼ 0.972 for the CP and P ¼ 0.060 for
Mean (SD) 67.47 (6.85) 75.84 (7.90)*
the PHC).
Median (min–max) 65 (60–94) 75 (60–94)*
The 10 most dispensed drugs are presented in Figure 1A,B.
Monthly mean
Enalapril was the most dispensed drug at both dispensation sites,
number of drugs:
followed by losartan at the CP and hydrochlorothiazide at the
2–4 56 (58.95%) 141 (53.61%)
PHC. These results are in agreement with those shown in Table 2.
≥5 28 (29.47%) 71 (27.00%)
At the PHC, the 10 most dispensed medications represented 81%
Mean by individual/ 3.69 (1.93) 3.46 (2.18)
of the dispensations, whereas at the CP, these concentrated only
month (SD)
32% of the dispensations. This fact may be related to the greater
Median (min–max) 3.00 (1–9) 3 (1–14)
availability of drugs that can be prescribed by physicians whose
* P o 0.001. patients obtain their medication at the CP. At the PHC, the most
prescribed drugs from group A02B (agents for peptic ulcer) were
ranitidine and omeprazole, whereas at the CP, the most pre-
difference was detected in the percentage of patients in each
scribed agents for peptic ulcer were pantoprazole, omeprazole,
level of polymedication (P ¼ 0.226). These data are shown in
ranitidine, lansoprazole, esomeprazole, rabeprazole, and dexlan-
detail in Table 1.
soprazole. Similar results were obtained for group C10A (lipid-
modifying agents). Simvastatin and fenofibrate were exclusively
Medication Consumption prescribed at the PHC, whereas atorvastatin, rosuvastatin, sim-
vastatin, gemfibrozil, fenofibrate, ezetimibe, and cholestyramine
Table 2 shows the 10 groups of drugs with the highest relative
were exclusively prescribed at the CP (data not shown). It is
frequency of prescription, considering the third level of the ATC
important to note that the drug of group A10B (oral blood glucose
code. As expected, at both institutions, most of the groups
lowering drugs) most dispensed at the CP was metformin,
corresponded to drugs that act on the cardiovascular system
whereas those most dispensed at the PHC were metformin and
(group C), followed by those that act on the digestive system
glibenclamide.
(group A). Nine out of the 10 most prescribed pharmacological
groups were the same at both the CP and the PHC. Group C03A
(low-ceiling diuretics, thiazides) was not present at the CP and Use of PIMs
was in the third position at the PHC. Similarly, group C09C Table 3 reports the PIMs recorded at the CP and the PHC. By
(angiotensin II antagonists, plain) was not present at the PHC following Beers Criteria, 111 (4.69%) of the prescriptions made at
and was in the fourth position at the CP. Also, at the PHC, the 10 the CP, which included 21 different drugs dispensed to 51 patients
groups concentrated 86% of the prescriptions, whereas at the CP (19.39%), were identified as PIMs for the elderly. Considering the
these represented 57% of the prescriptions. The possible relation- established degree of polypharmacy, the possibility of consuming
ship between the chance to use one drug of the nine coincident these drugs was 2.52 times higher in patients with major poly-
pharmacological groups and the sex of the patients was eval- medication than in those with minor polymedication (OR ¼ 2.52,
uated. At the PHC, the odds ratio of using drugs for peptic ulcer 95% CI 1.33:4.78, P ¼ 0.005). At the PHC, 72 out of the 886
and gastroesophageal reflux disease (group A02B) was 3.59 times prescriptions dispensed, which included three different drugs and
higher for women than for men (OR ¼ 3.59, 95% CI 1.21:10.74), corresponded to 27 patients (28.42%), were identified as PIMs (8.13%).
whereas at the CP, this odds ratio was 1.81 higher for women The possibility of receiving PIMs was 2.60 times larger in patients

Table 2 – Frequency of the dispensing of the 10 most prescribed pharmacological groups.

Primary health care center Community pharmacy

Code Group N (%) Code Group N (%)

C09A ACE inhibitors, plain 179 (20.20) C07A Beta-blocking agents 196 (8.28)
A10B Oral blood glucose lowering drugs 126 (14.22) A02B Drugs for peptic ulcer 180 (7.60)
C03A Low-ceiling diuretics, thiazides 95 (10.72) C10A Lipid-modifying agents. Plain 178 (7.52)
C10A Lipid-modifying agents, plain 82 (9.26) C09C Angiotensin II antagonists. plain 149 (6.29)
C07A Beta-blocking agents 67 (7.56) A10B Oral blood glucose lowering drugs 140 (5.91)
B01A Antithrombotic agents 63 (7.11) B01A Antithrombotic agents 138 (5.83)
A02B Drugs for peptic ulcer 59 (6.66) M01A Anti-inflammatory and antirheumatic 118 (4.93)
products, nonsteroids
C08C Selective calcium channel blockers 38 (4.29) C09A ACE inhibitors, plain 106 (4.48)
H03A Thyroid preparations 29 (3.27) C08C Selective calcium channel blockers 89 (3.76)
M01A Anti-inflammatory and antirheumatic 27 (3.05) H03A Thyroid preparations 89 (3.76)
products, nonsteroids
ACE, angiotensin-converting enzyme inhibitors.
122 VALUE IN HEALTH REGIONAL ISSUES 17 (2018) 119–125

Fig. 1 – Frequency of dispensing of the 10 most prescribed drugs. (A) PHC; (B) CP.

with major polymedication than in the rest of the elderly (OR ¼ 2.60, Table 4 shows a comparison of the individual characteristics of
95% CI 1.01:6.68, P ¼ 0.044). The percentage of PIMs prescriptions PIM users versus nonusers for both dispensation sites. In both cases,
was significantly higher at the PHC (P ¼ 0.001), although the the mean and the median of the individual monthly average of
percentage of patients receiving them did not differ significantly medications, and the percentage of patients who showed major
(P ¼ 0.068). polymedication, were significantly higher for PIM users. The average
At the PHC, the most frequently prescribed PIM was gliben- age and the percentage of patients older than 75 years of age who
clamide, followed by ibuprofen and digoxin (Table 3). At the CP, were PIM users were higher at the CP than at the PHC.
the variety of pharmacological groups that can cause PIMs was To evaluate the impact of patient characteristics on the
greater, and the most frequently prescribed drugs were diclofe- probability of using PIM, within each dispensation site, logistic
nac, meloxicam, digoxin, amiodarone, and glibenclamide. regressions were used considering the variables that were found

Table 3 – Prescribed drugs included in Beers Criteria 2012.

PIM Community pharmacy Primary health care center

Prescriptions (%) Patients (%) Prescriptions (%) Patients (%)

Anticholinergics
Chlorpheniramine 2 (0.08) 2 (0.76)
Cyproheptadine 1 (0.04) 1 (0.38)
Diphenhydramine 1 (0.04) 1 (0.38)
Oxybutynin 3 (0.13) 1 (0.38)
Cardiovascular drugs
Amiodarone 8 (0.34) 4 (1.52)
Flecainide 1 (0.04) 1 (0.38)
Digoxin 40.125 mg/d 9 (0.38) 3 (1.14) 5 (0.56) 2 (2.11)
Nifedipine 7 (0.30) 3 (1.14)
Spironolactone 425 mg/d 1 (0.04) 1 (0.38)
NSAIDs
Diclofenac 32 (1.35) 15 (5.70)
Ibuprofen 2 (0.08) 2 (0.76) 27 (3.05) 10 (10.53)
Meloxicam 25 (1.06) 15 (5.70)
Piroxicam 1 (0.04) 1 (0.38)
Naproxen 4 (1.17) 4 (1.52)
Ketorolac 1 (0.04) 1 (0.38)
Dexketoprofen 1 (0.04) 1 (0.38)
Others
Nitrofurantoin 1 (0.04) 1 (0.38)
Glibenclamide 7 (0.30) 4 (1.52) 40 (4.51) 16 (16.84)
Carisoprodol 1 (0.04) 1 (0.38)
Chlorzoxazone 1 (0.04) 1 (0.38)
Metoclopramide 2 (0.08) 2 (0.76)
Total 111 (4.69) 51 (19.39) 71 (8.13) 27 (28.42)
PIM, potentially inappropriate medication.
 VALUE IN HEALTH REGIONAL ISSUES 17 (2018) 119–125 123

Table 4 – Characteristics of users of PIMs versus nonusers.

Characteristic Primary health care center Community pharmacy

PIMs users PIMs nonusers PIMs users PIMs nonusers

N 27 (28.41%) 68 (71.59%) 51 (19.39%) 212 (80.61%)

Age (years)
60–74 25 (92.59%) 58 (85.29%) 24 (47.06%) 106 (50.00%)
≥75 2 (7.41%)‡ 10 (14.71%) 27 (52.94%) 106 (50.00%)
Mean (SD) 67.15 (6.17)‡ 67.60 (7.14) 75.55 (7.36) 75.91 (8.05)
Median (min–max) 64.00 (60–84)‡ 65.50 (60–94) 75.00 (64-93) 74.50 (60–94)
% Female 18 (66.67%) 36 (52.94%) 39 (76.47%)† 133 (62.74%)
Monthly mean number of drugs
2–4 15 (55.56%) 41 (60.29%) 26 (50.98%) 115 (54.25%)
≥5 12 (44.44%) 16 (23.53%) 22 (43.14%)* 49 (23.11%)
Mean (SD) 4.63 (2.00)* 3.32 (1.78) 4.59 (2.49)* 3.22 (2.01)
Median (min–max) 4 (2–8)* 3 (1–9) 4 (1–14)* 3 (1–11)
PIM, potentially inappropriate medication.
PIM users versus nonusers: *P o 0.05; †P o 0.10.
Primary health care center PIM users versus community pharmacy PIM users: ‡P o 0.001.

to be significant (P o 0.1) in the bivariate analysis. In the case of although international guidelines for the treatment of primary
CP, the model included as covariates the sex of patients and the hypertension (in patients aged 55 years or older) consider that
number of drugs used. The results for both variables were low doses of thiazides are the first choice of therapy—alone or in
statistically significant. The possibility of receiving PIMs was combination with other drugs [25]—diuretics remained underu-
2.17 times higher for women than for men (OR ¼ 2.17, 95% CI tilized by patients whose prescriptions were dispensed at the CP.
1.04:4.55, P ¼ 0.03). An increase in one unit in the number of drugs The percentage of PIM users found in the present study was
used implies that the odds of using PIMs increase 1.32 times slightly lower than that found in other studies [20,21,23,26]. This
(OR ¼ 1.32, 95% CI 1.15:1.52, P o 0.001). For the PHC, the model could be because we excluded the prescriptions of psychotropic
included only the quantity of drugs used. An increase in one unit drugs—not dispensed at the PHC—and self-medication. However,
in this value, implies that the odds of using PIMs increase 1.43 both at the CP and at the PHC, the differences in the mean and
times (OR ¼ 1.43, 95% CI 1.12:1.83, P ¼ 0.003). median number of medications per month between PIM users
and nonusers were statistically significant. In addition, in the two
cases, the percentage of polymedicated patients was higher
among PIM users. This is in line with other studies suggesting
Discussion
that the risk of inappropriate is greater in patients with poly-
Polypharmacy is usually observed in the elderly population pharmacy [5,27,28]. In addition, at the CP, we found an associa-
[2,15,16]. However, both polypharmacy and its incidence can be tion between female sex and the use of PIM, in agreement with
defined in different ways [17–19]. In this study, the proportion of other studies [5,22,26]. However, we found no correlation with
patients with major polypharmacy (defined here as five or more age.
drugs) was 29.47% at the PHC and 27% at the CP. These values Although only three drugs from the Beers Criteria were
were similar, although the average age was higher for patients detected at the PHC—as compared to the 21 drugs detected at
from the CP. Some recent studies on the drugs used by older the CP—the percentage of PIM prescriptions at the PHC was much
people have shown that the percentage of patients with major higher than that at the CP. This is partly due to the fact that more
polypharmacy varies between 35% and 90% [20–23]. We believe than half of the PIM prescriptions dispensed at the PHC corre-
that this difference between our results and the ones observed in sponded to glibenclamide, one of the most frequently prescribed
the cited literature may be due to the fact that we did not oral antidiabetic medicines for patients with type 2 diabetes in
consider the dispensation of psychoactive drugs or the self- primary health care centers in Argentina and a drug included
medication of these patients. The mean and median of medi- with metformin in the Provincial Therapeutic Formulary. This
cation by month found in our study are similar to those observed formulary does not provide gliclazide as an alternative to the
in other studies [20,21,24]. same class (sulfonylurea), which is indicated for elderly patients
Nine out of the 10 pharmacological groups with the highest because of its lower adverse reaction profile [29,30]. At the CP, the
frequency of dispensation were similar at both places. The more used oral blood glucose lowering drugs were gliclazide,
pharmacological group with the main proportion of dispensation, saxagliptin, vildagliptin, and metformin (data not shown).
both at the PHC and the CP, was the one related to the This study has some limitations. First, although the results
cardiovascular system. This was an expected outcome because obtained can be generalized to ambulatory elderly patients in
elderly people have a high index of cardiovascular comorbidity Argentina, they cannot be generalized to hospitalized patients.
[15,20,23,24]. In some patients, we found an association of two or Second, it would be interesting to investigate whether the
more drugs for cardiovascular disease. It is important to note that prevalence of PIMs is similar in the same population if other
the most frequent association at the PHC was that of enalapril criteria, such as the STOPP/START or Priscus Criteria, is used
with hydrochloride, while that at the CP was that of losartan with [31,32]. Beers Criteria are the gold standard to study inappropriate
amlodipine (data not shown). Although we are not aware of the medication [6], but present a number of disadvantages such as
existence of other risk factors associated with the underlying the fact that each drug is evaluated independently of the
pathologies, because the design of the study did not include the therapeutic context, and the need for continuous updating of
consultation of medical records, these results show that, criteria according to the available evidence [33,34].
124 VALUE IN HEALTH REGIONAL ISSUES 17 (2018) 119–125

On the other hand, the PHC studied uses a limited list of [5] Holmes HM, Luo R, Kuo YF, et al. Association of potentially
medicines—because patients do not have trade union–run med- inappropriate medication use with patient and prescriber
characteristics in Medicare Part D. Pharmacoepidemiol Drug Saf
ical insurance—and is thus not completely adapted to the specific
2013;22:728–34.
needs of the elderly. This represents a significant restriction on [6] Steinman MA, Beizer JL, DuBeau CE, et al. How to use the American
the availability of drugs that professionals can prescribe to select Geriatrics Society 2015 Beers Criteria: a guide for patients, clinicians,
a better therapeutic option. Based on these findings, it would be health systems, and payors. J Am Geriatr Soc 2015;63:e1–7.
[7] Chau SH, Jansen AP, van de Ven PM, et al. Clinical medication reviews
important to contemplate the incorporation of some drugs into
in elderly patients with polypharmacy: a cross-sectional study on
the Provincial Therapeutic Formulary. drug-related problems in the Netherlands. Int J Clin Pharm
Despite the aforementioned limitations, this study is valuable 2016;38:46–53.
to reconsider the prescription of certain drugs in elderly patients [8] IDELAS-UCES. La población argentina aumentó su edad promedio
respecto de la media del mundo. Boletín Mensual IDELAS. Available
and, in the future, we envision the development of a specific
from: http://la5pata.com/2017/02/09/idelas-uces-la-poblacion-
screening tool for the detection of PIMs in our country. argentina-aumento-su-edad-promedio/ [Accessed November 1, 2017].
Also, from an economic point of view we have perceived that [9] Marzi MM, Diruscio AV, Nunez HM, et al. Analysis of medication
increased information and education at the pharmacy (pharma- prescription in an Argentinian geriatric hospital. Rev Med Chil
2013;141:194–201.
ceutical care) could produce large cost savings because of the
[10] Comisión Provincial de Medicamentos. Ministerio de Salud. Gobierno
reduced incidence of drug-related problems. de la Provincia de Santa Fe. Formulario Terapéutico Provincial. Listado
de Medicamentos para uso obligatorio en Instituciones Oficiales
Provinciales. Ley Provincial Nº 9524/84 y Decreto Nº 4219/84, 2014.
[11] Anmat. Vademecum Nacional de Medicamentos. Available from: http://
Conclusion anmatvademecum.servicios.pami.org.ar/ [Accessed November 1, 2017].
[12] WHO Collaborating Centre for Drug Statistics Methodology. Guidelines
In this study, we described and compared the profile of dispens- for ATC Classification and DDD Assignment 2016. 19th ed. Oslo,
Norway: World Health Organization, 2015.
ing medications to older people at a PHC and a CP in Rosario,
[13] Bjerrum L, Sogaard J, Hallas J, et al. Polypharmacy: correlations with
Argentina. Cardiovascular and alimentary drugs were the princi- sex, age and drug regimen: a prescription database study. Eur J Clin
pal classes used by the elderly, who showed a high prevalence of Pharmacol 1998;54:197–202.
polypharmacy, although we had no information about their self- [14] McHugh ML. The odds ratio: calculation, usage, and interpretation.
medication or use of psychotropics. We also verified the pre- Biochem Med 2009;19:120–6.
[15] Steinman MA, Landefeld CS, Rosenthal GE, et al. Polypharmacy and
scriptions of PIMs and their relationship with a greater amount of prescribing quality in older people. J Am Geriatr Soc 2006;54:1516–23.
medicines used. [16] Fulton MM, Allen ER. Polypharmacy in the elderly: a literature review.
It is important to highlight that in CP we found a higher J Am Acad Nurse Pract 2005;17:123–32.
number of PIM users in patients older than 75 years with greater [17] Le Cossec C, Sermet C, Perronnin M. Measuring polypharmacy in the
elderly: impact of the method on prevalence and therapeutic classes.
polymedication. These conditions could increase the risk of Questions d’économie de la santé 2015:213.
hospitalization through drug adverse effects, drug interactions, [18] Bjerrum L, Rosholm JU, Hallas J, et al. Methods for estimating the
or intolerability and should be considered by the prescribing occurrence of polypharmacy by means of a prescription database. Eur J
Clin Pharmacol 1997;53:7–11.
physician.
[19] Monégat M, Sermet C, Perronnin M, et al. Polypharmacy: definitions,
In conclusion, the prevalence of PIMs found according to the measurement and stakes involved review of the literature and
Beers Criteria in this population is relevant enough to implement measurement tests. Questions d’économie de la santé 2014;214.
measures that address the problem in an integral way, to [20] Lopes LM, Figueiredo TP, Costa SC, et al. Use of potentially
inappropriate medications by the elderly at home. Cien Saude Colet
improve the quality of prescription and the health outcomes of
2016;21:3429–38.
the patient. There is also an urgent need for the health care [21] Aubert CE, Streit S, Da Costa BR, et al. Polypharmacy and specific
authorities to work toward the development of a national screen- comorbidities in university primary care settings. Eur J Intern Med
ing tool based on the many existing criteria. 2016;35:35–42.
[22] Cojutti P, Arnoldo L, Cattani G, et al. Polytherapy and the risk of
potentially inappropriate prescriptions (PIPs) among elderly and very
elderly patients in three different settings (hospital, community, long-
term care facilities) of the Friuli Venezia Giulia region, Italy: are the
Acknowledgments very elderly at higher risk of PIPs? Pharmacoepidemiol Drug Saf
2016;25:1070–8.
This study was supported by Universidad Nacional de Rosario [23] Cano Perez MD, Castell Alcala MV, Queipo Matas R, et al. Use of primary
(N° 445/2015) and a grant from Secretaría de Estado de Ciencia, care services, care specialized and drug use by population 65 years and
Tecnología e Innovación -SECTEI- (N°2010-174-14), Santa Fe, more in the community of Madrid, Spain. Rev Esp Salud Publica
Argentina. The authors would like to thank the Association of 2016;90:e1–11.
[24] Tôrres Faggiani F, Schroeter G, Luz Pacheco S, et al. Profile of drug
Pharmacists of the 2nd Circumscription of the Province of Santa utilization in the elderly living in Porto Alegre, Brazil. Pharm Pract
Fe for the collaboration provided for the realization of this work. (Granada) 2007;5:179–84.
[25] James PA, Oparil S, Carter BL, et al. 2014 evidence-based guideline for
the management of high blood pressure in adults: report from the
R EF E R EN CE S panel members appointed to the Eighth Joint National Committee (JNC
8). JAMA 2014;311:507–20.
[26] Arellano C, Saldivia G, Córdova P, et al. Using two tools to identify
potentially inappropriate medications (PIM) in elderly patients in
[1] Wade O. Concept of drug utilization studies. In: Bergman U, Grimsson southern Chile. Arch Gerontol Geriatr 2016;67:139–44.
A, Wahba A, Westerholm B, eds. Studies in Drug Utilization. [27] Cooper JA, Moriarty F, Ryan C, et al. Potentially inappropriate
Copenhagen: WHO Regional Office for Europe, 1979. prescribing in two populations with differing socio-economic profiles: a
[2] Hajjar ER, Cafiero AC, Hanlon JT. Polypharmacy in elderly patients. Am J cross-sectional database study using the PROMPT criteria. Eur J Clin
Geriatr Pharmacother 2007;5:345–51. Pharmacol 2016;72:583–91.
[3] Cahir C, Fahey T, Teeling M, et al. Potentially inappropriate prescribing [28] Chang CB, Yang SY, Lai HY, et al. Application of three different sets of
and cost outcomes for older people: a national population study. Br J explicit criteria for assessing inappropriate prescribing in older
Clin Pharmacol 2010;69:543–52. patients: a nationwide prevalence study of ambulatory care visits in
[4] Campanelli CM. American Geriatrics Society updated Beers Criteria for Taiwan. BMJ Open 2015;5:e008214.
potentially inappropriate medication use in older adults. J Am Geriatr [29] CADTH Rapid Response Reports. Glyburide, gliclazide or glimepiride
Soc 2012;60:616–31. for elderly patients with type 2 diabetes: a review of the clinical
 VALUE IN HEALTH REGIONAL ISSUES 17 (2018) 119–125 125

effectiveness and safety— an update. Ottawa, ON: Canadian Agency for [32] Holt S, Schmiedl S, Thurmann PA. Potentially inappropriate
Drugs and Technologies in Health, 2015. medications in the elderly: the PRISCUS list. Dtsch Arztebl Int
[30] Chan SP, Colagiuri S. Systematic review and meta-analysis of the 2010;107:543–51.
efficacy and hypoglycemic safety of gliclazide versus other [33] Fialová D, Onder G. Medication errors in elderly people: contributing
insulinotropic agents. Diabetes Res Clin Pract 2015;110:75–81. factors and future perspectives. Br J Clin Pharmacol 2009;67:641–5.
[31] O’Mahony D, O’Sullivan D, Byrne S, et al. STOPP/START criteria for [34] De Smet PA, Denneboom W, Kramers C, et al. A composite screening
potentially inappropriate prescribing in older people: version 2. Age tool for medication reviews of outpatients: general issues with specific
Ageing 2015;44:213–8. examples. Drugs Aging 2007;24:733–60.