PLOS ONE
RESEARCH ARTICLE
Perceived antimicrobial dispensing practices
in medicine outlets in Ghana: A maximum
difference experiment design
Eric Nyarko ☯ID* , Francisca Mawulawoe Akoto☯, Kwabena Doku-Amponsah☯
Department of Statistics and Actuarial Science, School of Physical and Mathematical Sciences, University of
Ghana, Legon, Accra, Ghana
☯ These authors contributed equally to this work.
* ericnyarko@ug.edu.gh
a1111111111 Abstract
a1111111111
a1111111111
a1111111111
a1111111111
Introduction
Antimicrobials are consumed among patients globally, but in developing and middle-income
countries, these drugs can be obtained without a prescription from pharmacies and licensed
drug stores due to inadequate regulation in the pharmaceutical sector. This study aimed to
OPEN ACCESS
assess antimicrobial dispensing practices in medicine sales outlets (i.e., pharmacies and
Citation: Nyarko E, Akoto FM, Doku-Amponsah K
licensed drug stores) to provide quantitative evidence for policy discussions to enhance
(2023) Perceived antimicrobial dispensing
practices in medicine outlets in Ghana: A maximum patient safety and care quality in Ghana’s pharmaceutical industry.
difference experiment design. PLoS ONE 18(7):
e0288519. https://doi.org/10.1371/journal.
pone.0288519 Method
Editor: Amal Al-Bakri, The University of Jordan The data for this study were obtained from a cross-sectional survey conducted in the
School of Pharmacy, JORDAN
Greater Accra region between July and August 2022. The survey was conducted through
Received: October 11, 2022 interviewer-administered questionnaires, and 200 staff members from medicine sales out-
Accepted: June 29, 2023 lets were randomly selected using a two-stage cluster and random sampling technique. The
Published: July 13, 2023 maximum difference experiment model, rooted in random utility theory, was used to analyze
© their antimicrobial dispensing practices.Copyright: 2023 Nyarko et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which
permits unrestricted use, distribution, and Result
reproduction in any medium, provided the original
author and source are credited. We found that medicine sales outlet staffs were highly concerned about following the drug
act and not dispensing antimicrobials without a prescription, and usually refer a patient to
Data Availability Statement: All relevant data are
within the paper and its Supporting Information get a prescription from a doctor if the patient has complications (like high fever, generalized
files. malaise, fatigue as symptoms, sinusitis). Stronger concerns were also observed for medi-
Funding: The author(s) received no specific cine outlet staff not dispensing antimicrobials without a prescription if the patient is pediatric
funding for this work. or geriatric with a severe infection. They also evaluated patients and dispensed antimicrobi-
Competing interests: The authors have declared als based on symptoms, not their age or gender. However, they tended not to dispense anti-
that no competing interests exist. biotics if the patient had a mild fever and requested it without a prescription.
PLOS ONE | https://doi.org/10.1371/journal.pone.0288519 July 13, 2023 1 / 15
PLOS ONE Perceived antimicrobial dispensing practices in medicine outlets
Conclusion
Our results provide insight into the need for a national surveillance system for monitoring
antimicrobial prescribing and dispensing practices at medicine sales outlets. Therefore, the
selection of antimicrobials for treating infectious diseases may be informed by evidence-
based antimicrobial prescribing and dispensing surveillance data and will help policymakers
to know the pattern of commonly consumed antimicrobials in the medicine sales outlets.
Introduction
It is a common practice for patients around the world to use antimicrobial agents, both with
and without a prescription [1,2]. Research shows that in low- and middle-income countries,
approximately two-thirds of the antibiotics purchased in pharmacies are consumed without a
prescription, and 20–50% of those are misused when treating bacterial infections [3,4]. These
countries typically have easy access to antimicrobials, which can be obtained without a pre-
scription from medicine sales outlets including pharmacies and licensed drug stores. Unfortu-
nately, these pharmaceutical sectors are poorly regulated [5,6], leading to inadequate
supervision, control, prescribing, and dispensing processes. This results in poor compliance
with antibiotic dispensing guidelines in developing and middle-income nations.
According to findings, antibiotics are among Africa’s most frequently prescribed medica-
tions. It was discovered that 90% of individuals with acute illnesses seek medical care outside
their homes, and 95% receive medication, with 36% being antibiotics. Most of these antibiotics
(75%) were Metronidazole, Cotrimoxazole, and Amoxicillin. Shockingly, one in four people
sought antibiotics from unofficial sources, and over 30% did so without a prescription [7]. In
Tanzania, many medications were supplied without a prescription and were either requested
by customers or suggested by pharmacists. Furthermore, dosage instructions were rarely given
by staff at medicine sales outlets, and when provided, the information was often inconsistent
with recommendations [8].
In Nairobi, a study found that 64% of retail pharmacies sold medications without a prescrip-
tion. Many of these stores also provided lower doses of antibiotics, particularly outside the city
center. The study also discovered that while 31% of guardians received antibiotics, 38% were
not prescribed by a doctor [9]. A study in Uganda’s Mukono district found that, 93.5% of drug
stores distributed antibiotics, including Amoxicillin and Trimethoprim-Sulfamethoxazole. The
study also revealed that overprescription was linked to the dispenser’s professional background.
Another study revealed that, despite oral rehydration salts and zinc pills being recommended as
the first line of treatment for children with diarrhea, 29.4% of pharmacy staff members still
chose antibiotics as their first option. Only 8.2% of providers had received antibiotic training,
while 10.6% had received training in managing pneumonia cases [10].
A recent study in Northwest Cameroon analyzed the prescribing trends and factors influ-
encing antibiotic prescription in primary healthcare facilities. Results showed that 36.71% of
patients received antibiotics, with an average of 1.14 prescribed antibiotics per patient. Amoxi-
cillin was the most commonly prescribed antibiotic, accounting for 29.9% of prescriptions,
and respiratory tract infections were the primary reason for 21.27% of prescriptions. Nearly all
prescriptions (99.87%) were from the Essential Drug List and prescribed by generic name
(98.36%). The study found that laboratory results, patient involvement, and performance-
based financing strongly influenced the prescribing rates of antibiotics [11]. Additionally, a
separate study discovered that before law enforcement, 70.7% of dispensers sold antibiotics
PLOS ONE | https://doi.org/10.1371/journal.pone.0288519 July 13, 2023 2 / 15
PLOS ONE Perceived antimicrobial dispensing practices in medicine outlets
without a prescription. In comparison, 96.6% and 87.7% of dispensers sold antibiotics without
a prescription for pharyngitis and urinary tract infections, respectively [12].
In Ghana, the Health Professions Regulatory Act, 2013 (Act 857) only allows medical pro-
fessionals like medical doctors, physician assistants, midwives, and nurses who have received
prescribing training to write registered antibiotic prescriptions [13]. The Pharmacy Act of
1994 (Act 489) also includes guidelines for the dispensing and selling of drugs, including anti-
biotics. Pharmacies may dispense certain antibiotics like Amoxicillin, Flucloxacillin, Norfloxa-
cin + Tinidazole, Ciprofloxacin, Doxycycline, Tetracycline, Erythromycin, and Ampicillin
based on the recommendation of a practicing pharmacist. However, licensed chemical sellers
are limited to selling only Cotrimoxazole for treating specific infections like infective diarrhea,
urinary tract infections, and upper respiratory tract infections. Moreover, if a restricted drug is
prescribed, the supplier must enter the date and information on the prescription and retain it
for two years in a manner as to be readily available for inspection [13,14].
Although the Pharmacy Act exists in Ghana, antimicrobial agents are easily accessible in medi-
cine sales outlets, as in many other developing and middle-income countries. These outlets are
often the first point of contact for ill patients [1]. Here the phrase "outlets" refers to both pharma-
cies and licensed drug stores. Medicine sales outlets are registered with the Pharmacy Council of
Ghana [15], to offer pharmaceutical services. However, little is known about the prescribing and
dispensing of antimicrobials in these outlets, where most sick individuals seek treatment. Misuse
of antimicrobial drugs can lead to the emergence, dissemination, and persistence of resistant
microorganisms, which poses a global hazard [7]. Therefore, governments are concerned about
the excessive use of antimicrobials and the increasing prevalence of bad prescribing and dispens-
ing practices due to the implications of antimicrobial resistance habits [3,7].
Few recent studies have investigated the practices of prescribing and dispensing antimicro-
bial drugs in medicine outlets in Ghana [16–21]. However, they do not model or provide a
quantification of these practices. Our study aims to fill this knowledge gap by conducting an
in-depth analysis of the antimicrobial dispensing practices in medicine sales outlets (so-called
pharmacies and licensed drug stores) in Accra, Ghana. We utilized a best-worst scaling (BWS)
experiment (also known as maximum difference scaling) to quantify medicine sales outlet
staff’s antimicrobial dispensing practices. This research is crucial as it can help improve patient
safety and care quality in Ghana’s pharmaceutical industry by identifying areas that require
interventions to enhance antimicrobial sales in medicine outlets. The results of this study
could serve as a valuable guide for discussions and decision-making in the pharmaceutical sec-
tor in Ghana and other developing countries.
Methodology
Study area
The study was conducted in the Greater Accra region, bordered by the Eastern region to the
north, the Volta Region to the east, the Gulf of Guinea to the south, and the Central region to
the west. As the capital city of Ghana, the Greater Accra region is the country’s smallest region,
covering an area of 3,245 km2, which is 1.4% of Ghana’s total geographical area. It has 16
administrative districts and is the second most populous area after the Ashanti Region, with a
population of 5,455,692 people in 2021, which is 17.7% of Ghana’s total population. A high
proportion, 87.4% of its population, resides in urban areas [22].
Sample size and data collection
To conduct this study, we surveyed staff from 200 medicine sales outlets in Greater Accra
between July and August 2022. We used a two-stage cluster and random sampling technique
PLOS ONE | https://doi.org/10.1371/journal.pone.0288519 July 13, 2023 3 / 15
PLOS ONE Perceived antimicrobial dispensing practices in medicine outlets
to obtain the data collected through interviewer-questionnaire administration. The minimum
permissible sample size for simple random sampling depends on the required degree of accu-
racy of computed probabilities. Let g denote the true choice proportion of the study popula-
tion, a be the allowable deviation as a percentage between ĝ and g, q = 1−g, α be the level of
confidence in the estimations such that Prðjĝ  gj � agÞ � a for any given sample n [23]
under the standard normal distribution is given by
q h � a�i2
n � F 1 1  :
ga2 2
The estimated sample size desired for this study under the assumption of equal choice prob-
abilities (i.e., g = 50%) is given by
0:5
n 2� ½1:96�
0:5� 0:052
n � 1536:
A minimum of 192 respondents were needed based on the number of available BWS sce-
narios (k = 8) and the formula N = n/k. The anticipated sample size of 200 respondents was
sufficient, and all 200 were included in the final analysis due to a 100% response rate. Prior to
data collection, written consent was obtained from respondents after explaining the study’s
purpose and emphasizing their voluntary participation. Respondents who had dispensed anti-
microbials within the past three months to one year and provided informed consent were
included in the study. However, those who refused consent and five interns in four medicine
sales outlets who had never dispensed antimicrobials over the past three months and lacked
experience were excluded from this study. Respondents were compensated with airtime for
their time spent answering questionnaires.
Experimental design
Best-wort scaling (BWS) is a contemporary approach for conducting choice experiments [24]
that have gained popularity in various fields such as health, conflict handling, environmental
sustainability, and food-related consumer research [25–29]. This method enables the evalua-
tion of the relative importance of attributes on a common scale [30] , which is crucial for this
study. In BWS, participants select their most preferred and disliked items (i.e., specific charac-
teristics) among three or more options. To identify the factors related to antimicrobial pre-
scribing and dispensing practices, we consulted medical sales experts and conducted an
extensive literature review [31]. The initial list of factors was then narrowed to eight plausible
attributes through focus group discussions with eight medicine sales outlet staff. These eight
attributes are presented in Table 1.
At the start of the questionnaire, we collected data about the participants’ demographics
and asked questions about their practices regarding prescribing and dispensing antimicrobial
medications. We then presented respondents with eight sets of BWS scenarios, each with
seven attributes related to antimicrobial dispensing practices. For each set, we asked partici-
pants to indicate which attribute of antimicrobial dispensing practices concerned them the
most and least. Each set of the BWS scenarios was presented separately. The balanced incom-
plete block design [32] was used to allocate the eight plausible attributes in the BWS scenarios.
Fig 1 shows a completed example best-worst scaling scenario.
Ethical considerations. All methods were carried out following relevant guidelines and
regulations. Ethical clearance was obtained from the University of Ghana Ethical and Protocol
PLOS ONE | https://doi.org/10.1371/journal.pone.0288519 July 13, 2023 4 / 15
PLOS ONE Perceived antimicrobial dispensing practices in medicine outlets
Table 1. Items considered in the BWS questionnaire.
Attributes
1 Follow the drug act and do not dispense antimicrobials without a prescription
2 Dispense antibiotics if the patient has a good outcome with previously prescribed antibiotic
3 Dispense antibiotics to poor patients who cannot afford doctor/hospital charges keeping in mind their cost-
effectiveness
4 Evaluate patients and dispense antimicrobials based on symptoms, and not their age or gender
5 Dispense antimicrobials when a patient requests them without a prescription
6 Dispense antimicrobials if the patient has a mild fever
7 Do not dispense antimicrobials and usually refer the patient to get a prescription from a doctor if a patent has a
high fever, generalized malaise, fatigue as symptoms, sinusitis
8 Do not dispense antimicrobials without a prescription for a pediatric or geriatric patient with a severe infection
The eight plausible attributes of antimicrobial dispensing practices were obtained through focus group discussions
with eight medicine sales outlet staff.
https://doi.org/10.1371/journal.pone.0288519.t001
Review Committee (Reference No: ECBAS 034/21-22). Before collecting data, permission was
obtained in the form of written consent from the respondents after explaining the purpose of
the study. Respondents were informed that participation was voluntary and they could choose
whether to participate in the study. Extra precautions were taken due to the COVID-19 out-
break, such as wearing face masks and providing hand-cleansing sanitizers, soaps, and water
before interacting with the study participants. We followed all COVID-19 preventive protocols
during each personal interaction and data collection session.
Empirical strategy
In BWS, items are evaluated based on a random utility framework [33–35], and the frequency
of choices made by participants serve as a metric to compare the relative importance of differ-
ent items. By using a model with an error theory, it is possible to estimate how often one item
might be chosen over another. Here a choice is assumed to have an underlying value, or utility,
to respondents. The maximum difference model is used to estimate these utilities. To formalize
this model, let ℏ with |ℏ|�3 denote the finite set of potentially available choice options in a
Fig 1. A sample completed best-worst scaling scenario. The eight sets of best-worst scaling scenarios generated from
the balanced incomplete block design contain seven attributes each. The sample completed best-worst scaling scenario
represents best-worst scaling scenario number 1 in the balanced incomplete block design and contains plausible
attributes 5, 7, 1, 8, 4, 2, and 3.
https://doi.org/10.1371/journal.pone.0288519.g001
PLOS ONE | https://doi.org/10.1371/journal.pone.0288519 July 13, 2023 5 / 15
PLOS ONE Perceived antimicrobial dispensing practices in medicine outlets
given best-worst scenario and let ψ(ℏ) denote the experimental design, that is, the set of (sub)
sets of antimicrobial dispensing choice alternatives that occur in this study. For any set Y2ψ
(ℏ), Y�ℏ with |Y|�3, letMYðxÞ denote the probability that the alternative x is chosen as best
in set Y, AYðyÞ the probability that the alternative y is chosen as worst in set Y, andMAYðx; yÞ
the probability that alternative x is chosen as best in set Y and the alternative y6¼x is chosen as
worst in Y.
Now, the best choice model assumes that there is a scale μ such that for all y2Y2ψ(ℏ),
emðyÞ
MYðyÞ ¼X ; ð1ÞemðzÞ
z2Y
where the value μ(y) for an option y is interpreted as the utility for that option. The parallel
model for worst choice assumes that there is a scale v such that for all y2Y2ψ(ℏ),
enðyÞ
AYðyÞ ¼X : ð2ÞenðzÞ
z2Y
Assuming that (1) and (2) are met and that the choice probabilities for sets of two elements
follow a plausible condition for all different pairs, x,y2Y2ψ(ℏ), then
Mfx;ygðxÞ ¼ Afx;ygðyÞ;
that is, we have
e mðyÞ
AYðyÞ ¼X : ð3Þe mðzÞ
z2Y
If the choice probabilities that are considered the best (and worst) satisfy (1) and (3), then
when selecting the best option, the utility of a choice alternative is the opposite of the utility of
that same option when selecting the worst option, and this utility-scale μ is such that for all x,
y2Y2ψ(ℏ), x6¼y,
e½mðxÞ mðyÞ�
MAYðx; yÞ ¼X : ð4Þe½mðpÞ mðqÞ�
fp;qg2Y
Here, a respondent chooses the items they like and dislike the most. This data is analyzed
using the maximum difference model, treating all attributes as generic variables. Statistical
analysis was conducted using John’s Macintosh Project Pro (JMP Pro Version 16.0). The sig-
nificance of the attributes determining antimicrobial dispensing practices was assessed at the
95% confidence intervals (CIs). The relative importance of the attributes is compared through
parameter estimates. A parameter estimate is considered statistically significant when the asso-
ciated 95% confidence interval (CI) is greater than or less than zero. If the 95% CI overlaps, the
parameter estimate is not significantly different. A significant positive/negative parameter esti-
mate indicates a high/low preference for a specific attribute.
Results
Demographic description
This study included 200 staff members from medicine sales outlets, all of whom responded.
Table 2 provides an overview of the population studied. Of the 200 respondents, 106 (77.4%)
were male and aged between 18 and 30 years old, while 47 (74.6%) were female. Most of the
PLOS ONE | https://doi.org/10.1371/journal.pone.0288519 July 13, 2023 6 / 15
PLOS ONE Perceived antimicrobial dispensing practices in medicine outlets
Table 2. Demographic description of the study population.
Variable Gender
Female Male
Age (years) 18–30 47 (74.6%) 106 (77.4%)
31–45 16 (25.4%) 28 (20.4%)
46–60 0 (0.0%) 3 (2.2%)
Educational level completed Secondary School/SHS/SSS or less 14 (22.2%) 20 (14.6%)
Diploma/HND 15 (23.8%) 28 (20.4%)
Undergraduate 30 (47.6%) 85 (62.0%)
Post-graduate degree 4 (6.3%) 4 (2.9%)
Religion Christian 49 (77.8%) 123 (89.8%)
Muslim 14 (22.2%) 11 (8.0%)
Traditional 0 (0.0%) 3 (2.2%)
Marital status Living alone 42 (66.7%) 119 (86.9%)
Living with another (married, partnership) 21 (33.3%) 18 (13.1%)
Number of children None 49 (77.8%) 119 (86.9%)
1–3 13 (20.6%) 13 (9.5%)
4–6 1 (1.6%) 0 (0.0%)
>6 0 (0.0%) 5 (3.6%)
Type of facility you work Drug store 5 (7.9%) 10 (7.3%)
Pharmacy store 58 (92.1%) 127 (92.7%)
Current job role License chemical seller 3 (4.8%) 4 (2.9%)
Pharmacist 30 (47.6%) 65 (47.4%)
Pharmacy technician 11 (17.5%) 46 (33.6%)
Pharmacy/chemical shop assistant 19 (30.2%) 22 (16.1%)
Length of work experience (years) 1–5 50 (79.4%) 116 (84.7%)
>5 13 (20.6%) 21 (15.3%)
Institution you have been involved in antimicrobial training programme Ghana health service 9 (14.3%) 12 (8.8%)
Pharmaceutical companies 7 (11.1%) 22 (16.1%)
Pharmacy council 24 (38.1) 54 (39.4%)
Trained by a professional staff at my facility/manager 23 (36.5%) 49 (35.8%)
Out of the 200 staff members sampled from medicine sales outlets, a significant number were men aged between 18 and 30 years old. These individuals were graduate
pharmacists with 1–5 years of work experience and were employed in pharmacy stores. They also participated in multiple antimicrobial training programs organized by
the pharmacy council.
https://doi.org/10.1371/journal.pone.0288519.t002
respondents, 123 (89.8%), were male Christians who had obtained a first degree 85 (62.0%),
lived alone 119 (86.9%), and had no children 119 (86.9%). Among the female respondents, the
majority were female Christians who had obtained a first degree 30 (47.6%), lived alone 42
(66.7%), and had no children 49 (77.8%). More than half 116 (84.7%) of the male respondents
had 1–5 years of working experience, with 127 (92.7%) working in pharmacy stores as pharma-
cists 65 (47.4%). Among the female respondents, 50 (79.4%) had 1–5 years of working experi-
ence, with 58 (92.1%) working in pharmacy stores and 30 (47.6%) as pharmacists. 54 (39.4%)
male respondents had participated in an antimicrobial dispensing training program organized
by the pharmacy council, and 49 (35.8%) had training from the professional staff at their facil-
ity or manager. In comparison, 24 (38.1%) of the female respondents had participated in an
antimicrobial training program organized by the pharmacy council, and 23 (36.5%) had train-
ing from the professional staff at their facility or manager.
PLOS ONE | https://doi.org/10.1371/journal.pone.0288519 July 13, 2023 7 / 15
PLOS ONE Perceived antimicrobial dispensing practices in medicine outlets
Maximum difference model results
The results of our model on antimicrobial dispensing practices, including the utility coeffi-
cients (also-called parameter or utility estimates), standard errors (SE), and 95% confidence
intervals (CIs) are presented in Table 3. The statistical significance of the model was measured
at p-value (p< 0.00, 0.01 and 0.05). The likelihood ratio (L-R) Chi-square value of 561.46 with
7 degrees of freedom (DF) and a p< 0.00 tell us that the model as a whole is important at a
pre-specified significance level of 1%, 5% and 10%. To determine the statistical significance of
a utility coefficient, we check if the 95% CIs are greater than or less than zero. If there is an
overlap of the 95% CIs, the utility coefficient associated with the corresponding attribute is not
significantly different. All parameter estimates of attributes related to antimicrobial dispensing
practices have the expected sign at 95% CIs (Table 3). The relative importance of each attribute
is shown in Fig 2 with their corresponding marginal utility estimates, and marginal probability
values (represented by the length of the bars).
It can be inferred that, relatively, respondents follow the drug act and avoid dispensing anti-
microbials without a prescription (marginal utility estimate = 0.96 with a marginal probability
value = 0.26). The results showed stronger preferences for not dispensing antimicrobials and
referring patients to a doctor if they have symptoms like high fever, fatigue, and sinusitis (mar-
ginal utility estimate = 0.87 with a marginal probability value = 0.23). Additionally, respon-
dents were less likely to dispense antimicrobials without a prescription for pediatric or
geriatric patients with severe infections (marginal utility estimate = 0.36 with a marginal prob-
ability value = 0.14). Respondents prioritize evaluating patients based on their symptoms
when dispensing antimicrobials rather than their age or gender (marginal utility estimate = 0.21
with a marginal probability value = 0.12). On the other hand, there was a disutility associated
Table 3. Maximum difference model results of attributes associated with antimicrobial dispensing practices.
Attribute Parameter SE Lower Upper
Estimate 95% 95%
Dispense antimicrobials if a patient has a mild fever -0.39 0.08 -0.55 -0.23
Dispense antimicrobials when patient requests for it without a prescription -0.71 0.07 -0.86 -0.55
Do not dispense antimicrobials and usually refer a patient to get a prescription from a doctor if the patient has a high 0.87 0.07 0.72 1.02
fever, generalized malaise, fatigue as symptoms, sinusitis
Follow the drug act and do not dispense antimicrobials without prescription 0.96 0.07 0.81 1.11
Do not dispense antimicrobials without prescription for a pediatric or geriatric patient with severe infection 0.36 0.08 0.20 0.52
Evaluate patients and dispense antimicrobials based on symptoms, and not their age or gender 0.21 0.08 0.05 0.37
Give antibiotics if the patient has a good outcome with previously prescribed antibiotic -0.06 0.08 -0.22 0.09
Model fits
L-R Chi-Square 561.46
AIC 3915.50
BIC 3946.06
DF 7
P-Value <0.00
*SE, standard error; L-R, likelihood ratio; AIC, Akaike information criteria; BIC, Bayesian information criteria; DF, degree of freedom. The attribute variable following
the drug act and not dispensing antimicrobials without a prescription has the highest utility. Other stronger preferences are observed for not dispensing antimicrobials
and usually referring a patient to get a prescription from a doctor if the patient has complications (like high fever, generalized malaise, fatigue as symptoms, sinusitis),
followed by not dispensing antimicrobials without a prescription if a patient is pediatric or geriatric with a severe infection as well as evaluate patients and dispense
antimicrobials based on symptoms, not their age or gender. However, the attribute variables, such as dispensing antibiotics if a patient has a mild fever and requests an
antimicrobial without a prescription, result in disutility.
https://doi.org/10.1371/journal.pone.0288519.t003
PLOS ONE | https://doi.org/10.1371/journal.pone.0288519 July 13, 2023 8 / 15
PLOS ONE Perceived antimicrobial dispensing practices in medicine outlets
Fig 2. Marginal utility estimates and marginal probability of attributes associated with antimicrobial dispensing
practices. The corresponding estimates of marginal utility and probability values determine the importance attached
to each attribute. The length of the bars represents these values.
https://doi.org/10.1371/journal.pone.0288519.g002
with dispensing antimicrobials if the patient has a mild fever (marginal utility estimate = -0.39
with a marginal probability value = 0.06), as well as dispensing antimicrobials when a patient
requests it without a prescription (marginal utility estimate = -0.71 with marginal probability
value = 0.04). Overall, the survey suggested that medicine sales staffs were careful when dis-
pensing antimicrobials.
Table 4 compares antimicrobial dispensing practices and their associated attributes, includ-
ing utility estimates, marginal probabilities, and odd ratios. Patients with mild fevers have a
higher chance of receiving antimicrobials (odds ratio = 1.37; probability = 0.57) than those who
request them without a prescription (odds ratio = 0.72; probability = 0.42). Overall, medicine
sales outlet staff are more likely to follow the drug act by not dispensing antimicrobials without
a prescription (odds ratio = 3.89; probability = 0.79) and referring patients with high fevers,
malaise, fatigue, and sinusitis to a doctor for a prescription (odds ratio = 3.55; probability = 0.78).
Dispensing antimicrobials to pediatric or geriatric patients with severe infections without a pre-
scription has higher odds (odds ratio = 2.12; probability = 0.68), and medicine sales outlet staff
are more likely to evaluate patients based on symptoms, not age or gender, when dispensing
antimicrobials (odds ratio = 1.83; probability = 0.64). Again, dispensing antimicrobials without
a prescription when a patient requests them has lower odds (odds ratio within the range 0.18–
0.39; probability within the range 0.15–0.28) than following the drug act (odds ratio = 5.37;
probability = 0.84), referring pediatric or geriatric patients with severe infections to a doctor for
a prescription (odds ratio = 2.93; probability = 0.74), and evaluating patients based on symp-
toms when dispensing antimicrobials (odds ratio = 2.52; probability = 0.71).
On average, medicine sales outlet staff are more likely to refer patients with high fevers,
malaise, fatigue, and sinusitis to a doctor for a prescription (odds ratio = 4.89; probabil-
ity = 0.83, odds ratio = 1.66; probability = 0.62) compared to dispensing antimicrobials when a
patient requests them without a prescription (odds ratio = 0.20; probability = 0.16) as well as
not dispensing antimicrobials without a prescription for pediatric or geriatric patients with
severe infections (odds ratio = 0.59; probability = 0.37). Following the drug act by not dispens-
ing antimicrobials without a prescription has higher odds (odds ratio = 2.12; probability = 0.68)
than evaluating patients based on symptoms (odds ratio = 0.47; probability = 0.31). Medicine
sales outlet staff are more likely to refer patients with high fevers, malaise, fatigue, and sinusitis
to a doctor for a prescription (odds ratio = 1.93; probability = 0.65) compared to evaluating
patients based on symptoms (odds ratio = 0.51; probability = 0.34).
PLOS ONE | https://doi.org/10.1371/journal.pone.0288519 July 13, 2023 9 / 15
PLOS ONE Perceived antimicrobial dispensing practices in medicine outlets
Table 4. All levels comparison of attributes associated with antimicrobial dispensing practices.
Compared1 Compared2 Utility1 Utility2 Probability1 Probability2 Odds1 Odds2
Dispense antimicrobials if the patient has a mild Dispense antimicrobials when patient requests -0.39 -0.71 0.57 0.42 1.37 0.72
fever for it without a prescription
Do not dispense antimicrobials and usually refer -0.39 0.87 0.21 0.78 0.28 3.55
the patient to get a prescription from a doctor if
the patent has a high fever, generalized malaise,
fatigue as symptoms, sinusitis
Follow the drug act and do not dispense -0.39 0.96 0.20 0.79 0.25 3.89
antimicrobials without prescription
Do not dispense antimicrobials without -0.39 0.36 0.31 0.68 0.46 2.12
prescription for a pediatric or geriatric patient
with severe infection
Evaluate patients and dispense antimicrobials -0.39 0.21 0.35 0.64 0.54 1.83
based on symptoms, and not their age or gender
Dispense antibiotics if the patient has a good -0.39 -0.06 0.41 0.58 0.72 1.38
outcome with previously prescribed antibiotic
Dispense antimicrobials when patient requests Follow the drug act and do not dispense -0.71 0.96 0.15 0.84 0.18 5.37
for it without a prescription antimicrobials without prescription
Do not dispense antimicrobials without -0.71 0.36 0.25 0.74 0.34 2.93
prescription for a pediatric or geriatric patient
with severe infection
Evaluate patients and dispense antimicrobials -0.71 0.21 0.28 0.71 0.39 2.52
based on symptoms, and not their age or gender
Do not dispense antimicrobials and usually refer Dispense antimicrobials when patient requests 0.87 -0.71 0.83 0.16 4.89 0.20
the patient to get a prescription from a doctor if for it without a prescription
the patent has a high fever, generalized malaise,
fatigue as symptoms, sinusitis
Follow the drug act and do not dispense 0.87 0.96 0.47 0.52 0.91 1.09
antimicrobials without prescription
Do not dispense antimicrobials without 0.87 0.36 0.62 0.37 1.66 0.59
prescription for a pediatric or geriatric patient
with severe infection
Give antibiotics if the patient has a good 0.87 -0.06 0.71 0.28 2.56 0.39
outcome with previously prescribed antibiotic
Give antibiotics to poor patients who cannot 0.87 -1.24 0.89 0.10 8.37 0.11
afford doctor/hospital charges keeping in mind
their cost-effectiveness
Follow the drug act and do not dispense Evaluate patients and dispense antimicrobials 0.96 0.21 0.68 0.31 2.12 0.47
antimicrobials without prescription based on symptoms, and not their age or gender
Give antibiotics if the patient has a good 0.96 -0.06 0.73 0.26 2.80 0.35
outcome with previously prescribed antibiotic
Give antibiotics to poor patients who cannot 0.96 -1.24 0.90 0.09 9.18 0.10
afford doctor/hospital charges keeping in mind
their cost-effectiveness
Do not dispense antimicrobials without Follow the drug act and do not dispense 0.36 0.96 0.35 0.64 0.54 1.83
prescription for a pediatric or geriatric patient antimicrobials without prescription
with severe infection
Evaluate patients and dispense antimicrobials 0.36 0.21 0.53 0.46 1.16 0.86
based on symptoms, and not their age or gender
Dispense antibiotics if the patient has a good 0.36 -0.06 0.60 0.39 1.53 0.65
outcome with previously prescribed antibiotic
Give antibiotics to poor patients who cannot 0.36 -1.24 0.83 0.16 5.01 0.19
afford doctor/hospital charges keeping in mind
their cost-effectiveness
(Continued)
PLOS ONE | https://doi.org/10.1371/journal.pone.0288519 July 13, 2023 10 / 15
PLOS ONE Perceived antimicrobial dispensing practices in medicine outlets
Table 4. (Continued)
Compared1 Compared2 Utility1 Utility2 Probability1 Probability2 Odds1 Odds2
Evaluate patients and dispense antimicrobials Do not dispense antimicrobials and usually refer 0.21 0.87 0.34 0.65 0.51 1.93
based on symptoms, and not their age or gender the patient to get a prescription from a doctor if
the patent has a high fever, generalized malaise,
fatigue as symptoms, sinusitis
Give antibiotics if the patient has a good 0.21 -0.06 0.56 0.43 1.32 0.75
outcome with previously prescribed antibiotic
Dispense antibiotics to poor patients who can 0.21 -1.24 0.81 0.18 4.31 0.23
not afford doctor/hospital charges keeping in
mind their cost-effectiveness
Dispense antibiotics if the patient has a good Dispense antimicrobials when patient requests -0.06 -0.71 0.65 0.34 1.91 0.52
outcome with previously prescribed antibiotic for it without a prescription
Dispense antibiotics to poor patients who -0.06 -1.24 0.76 0.23 3.26 0.30
cannot afford doctor/hospital charges keeping
in mind their cost-effectiveness
Dispense antibiotics to poor patients who Dispense antimicrobials if the patient has a mild -1.24 -0.39 0.29 0.70 0.42 2.35
cannot afford doctor/hospital charges keeping fever
in mind their cost-effectiveness
Dispense antimicrobials when patient requests -1.24 -0.71 0.36 0.63 0.58 1.70
for it without a prescription
The all-level comparison report provides pairwise comparisons for the antimicrobial dispensing attribute variables. The first and second components of the comparison
represent Compared 1 and Compared 2, which are the factor and levels, respectively. Utility 1 and Utility 2 are the estimated utilities for the antimicrobial dispensing
attribute variables specified in the first and second components. Probability 1 is the predicted likelihood that the first component will be preferred over the second
component for the specified antimicrobial dispensing attribute variables in the first component. Similarly, Probability 2 is the predicted likelihood that the second
component will be preferred over the first component for the specified antimicrobial dispensing attribute variables in the second component. Additionally, Odds 1 is
obtained by dividing Probability 1 by Probability 2, while Odds 2 is obtained by dividing Probability 2 by Probability 1.
https://doi.org/10.1371/journal.pone.0288519.t004
Discussion
Antimicrobial drugs are widely used by patients worldwide. However, these drugs are easily
available and can be obtained in developing and middle-income countries without a prescrip-
tion from various medicine sales outlets (i.e., pharmacies and licensed drug stores) because the
pharmaceutical industry needs proper regulation. This study assessed antimicrobial prescrib-
ing and dispensing practices in medicine sales outlets. The study provides quantitative evi-
dence to enhance patient safety and care quality in Ghana and other developing and middle-
income countries where antimicrobial drugs are easily accessible. The study found that most
respondents were males aged 18–30 years. Most medicine sales outlet staff were pharmacists
with 1–5 years of experience working in pharmacy stores. They had participated in several
antimicrobial training programs organized by the pharmacy council and their facility manager
or professional staff. These findings support earlier observations that personnel at the drug
store gained their pharmaceutical knowledge through on-the-job training, conversations with
pharmacists, advertisements, common knowledge about medicines in the community, client
feedback, and presentations by drug representatives at association meetings [1].
A key finding of this study is that staffs at the medicine sales outlets, such as pharmacies
and licensed drug stores, are highly concerned with following the drug act and not dispensing
antimicrobials without a prescription. Other stronger preferences were observed for medicine
sales outlet staff not dispensing antimicrobials and instead referring patients to a doctor if they
have complications like high fever, fatigue, generalized malaise, or sinusitis, followed by not
dispensing antimicrobials without a prescription for pediatric or geriatric patients with severe
infection and evaluate patients based on their symptoms, not age or gender. However,
PLOS ONE | https://doi.org/10.1371/journal.pone.0288519 July 13, 2023 11 / 15
PLOS ONE Perceived antimicrobial dispensing practices in medicine outlets
medicine sellers tend not to dispense antibiotics if a patient only has a mild fever and requests
antibiotics without a prescription.
Our results revealed that medicine sellers prioritize following laws and guidelines to control
the use of antimicrobials and not dispensing them without a prescription. They also prioritize
referring patients to a doctor for a prescription if they exhibit symptoms like high fever,
fatigue, generalized malaise, or sinusitis. However, previous studies have shown conflicting
results on antibiotic prescribing and dispensing rates, with some patients purchasing antibiot-
ics without a prescription and personally requesting specific drugs without seeking advice
from medicine outlet staff [1,2,36,37]. These contrasting findings highlight the need for evi-
dence-based antimicrobial prescribing and dispensing practices at medicine sales outlets in
Ghana and other developing and middle-income nations where antimicrobials are easily
accessible. A national surveillance system is needed to monitor antimicrobial prescribing and
dispensing practices, which would help policymakers to know the pattern of commonly used
antimicrobials in the medicine sales outlets [1] . This would allow for the evidence-based selec-
tion of antimicrobials for treating infectious diseases [37] and contribute significantly to the
fight against antimicrobial resistance [7].
This study further revealed that staff members at medicine outlets evaluate patients and dis-
pense antimicrobials based on symptoms rather than age or gender. This result is in line with
acceptable regulations [14] because pharmacists are authorized to prescribe certain drugs,
such as Amoxicillin, Flucloxacillin, Norfloxacin + Tinidazole, Ciprofloxacin, Doxycycline, Tet-
racycline, Erythromycin, and Ampicillin without a doctor’s prescription while keeping accu-
rate records of transactions [13]. However, the practice of prescribing and dispensing
antimicrobials by unlicensed individuals in medicine sales outlets violates the Ghana National
Action Plan on Antimicrobial Resistance. It can lead to the misuse of antimicrobials, resulting
in community-wide antimicrobial resistance, unsuccessful treatments, and high healthcare
costs for patients and the state [7,37].
Our results revealed that medicine sales outlet staff hesitated to provide antibiotics to
patients with mild fevers and did not give out antimicrobials without a prescription. A previ-
ous study revealed that only 46% of pharmacists reported always giving antibiotics with a pre-
scription. However, the most frequently prescribed antibiotics without a prescription were
Amoxicillin, Metronidazole, and Cephalexin, accounting for 51% of cases [38]. It is essential to
implement regulations for dispensing antimicrobial drugs, offer ongoing education, and pro-
vide training to medicine outlet staff, including pharmacists and pharmacy assistants, to
address the issue of antimicrobial resistance. These can be achieved through organizations like
the Pharmacy Council [15], pharmaceutical companies, the Ghana Health Service (GHS), and
the Ministry of Health (MOH).
According to our findings, dispensing antimicrobials to patients with mild fevers is better
than providing them to patients who request them without a prescription [3,7]. If a patient has
a high fever, general malaise, fatigue, or sinusitis, referring them to a doctor for a prescription
instead of dispensing antimicrobials is best. It is important to adhere to the drug act [14] and
not give out antimicrobials without a prescription, especially for pediatric or geriatric patients
with severe infections. Patients should be evaluated based on symptoms rather than age or gen-
der when dispensing antimicrobials. Dispensing antimicrobials without a prescription should
be avoided, and patients should be referred to a doctor for a prescription if necessary, as cov-
ered in the Health Professions Regulatory Act, 2013 (Act 857) [13].
The authors of this study acknowledge certain limitations. Firstly, since the research was
conducted solely in Accra, it may not accurately represent how antimicrobials are typically dis-
pensed in other parts of Ghana and Sub-Saharan Africa. However, the easy drug access in
Accra may reflect dispensing practices in other areas. Secondly, the study only investigates
PLOS ONE | https://doi.org/10.1371/journal.pone.0288519 July 13, 2023 12 / 15
PLOS ONE Perceived antimicrobial dispensing practices in medicine outlets
specific factors related to antimicrobial dispensing practices and does not cover other impor-
tant attributes such as the type of antimicrobials dispensed, record maintained on antimicrobi-
als dispensed, laboratory/diagnostic test of microbes before prescription of antimicrobials,
patients demand antibiotics without prescription or when they do not fully understand their
condition, utilization of already prescribed antibiotics for the same infection or illness, the pro-
portion of the nonprescription sale of antimicrobials for patients with particular symptoms,
among others are not covered. This may lead to biased utility estimates. Future studies should
include these questions to address excluded attribute biases. Additionally, the study relied on
provider reports, and their integrity could not be verified. However, future studies should
employ observational study or anonymously approach medicine sales outlets as "patients" to
validate the results.
Although there are limitations, this study provides important data on the dispensing of
antimicrobials in medicine sales outlets. This information is crucial for policymakers to
improve patient safety and care quality in the pharmaceutical industry in Sub-Saharan Africa,
particularly in Ghana, where the misuse of antimicrobials has become a pressing issue. This
misuse can lead to the development of antimicrobial resistance, resulting in ineffective treat-
ment of infectious diseases and high healthcare costs for patients and the state. This study is
the first to utilize the best-worst scaling experiment approach to assess antimicrobial dispens-
ing practices.
Supporting information
S1 Data.
(XLSX)
Acknowledgments
The authors extend their heartfelt thanks to all the informants who generously shared their
knowledge and experiences. We are also grateful to the medicine sales outlet proprietors who
allowed to include their establishments for this study.
Author Contributions
Conceptualization: Eric Nyarko, Kwabena Doku-Amponsah.
Data curation: Francisca Mawulawoe Akoto.
Formal analysis: Eric Nyarko.
Investigation: Eric Nyarko, Francisca Mawulawoe Akoto, Kwabena Doku-Amponsah.
Methodology: Eric Nyarko.
Project administration: Eric Nyarko, Francisca Mawulawoe Akoto.
Software: Eric Nyarko.
Supervision: Eric Nyarko, Kwabena Doku-Amponsah.
Validation: Eric Nyarko, Kwabena Doku-Amponsah.
Visualization: Eric Nyarko.
Writing – original draft: Eric Nyarko, Francisca Mawulawoe Akoto.
Writing – review & editing: Eric Nyarko, Kwabena Doku-Amponsah.
PLOS ONE | https://doi.org/10.1371/journal.pone.0288519 July 13, 2023 13 / 15
PLOS ONE Perceived antimicrobial dispensing practices in medicine outlets
References
1. Ahiabu MA, Magnussen P, Bygbjerg IC, Tersbøl BP. Treatment practices of households and antibiotic
dispensing in medicine outlets in developing countries: The case of Ghana. Research in Social and
Administrative Pharmacy. 2018; 14: 1180–1188. https://doi.org/10.1016/j.sapharm.2018.01.013 PMID:
29428578
2. Belachew SA, Hall L, Selvey LA. Non-prescription dispensing of antibiotic agents among community
drug retail outlets in Sub-Saharan African countries: a systematic review and meta-analysis. Antimicro-
bial Resistance & Infection Control. 2021; 10(1): 1–15. https://doi.org/10.1186/s13756-020-00880-w
PMID: 33446266
3. WHO. Antimicrobial resistance: global report on surveillance: World Health Organization; 2014.
4. Torres NF, Chibi B, Middleton LE, Solomon VP, Mashamba-Thompson TP. Evidence of factors influ-
encing self-medication with antibiotics in low and middle-income countries: a systematic scoping review.
Public health. 2019; 168: 92–101. https://doi.org/10.1016/j.puhe.2018.11.018 PMID: 30716570
5. Ocan M, Obuku EA, Bwanga F, Akena D, Richard S, Ogwal-Okeng J, et al. Household antimicrobial self-
medication: a systematic review and meta-analysis of the burden, risk factors and outcomes in develop-
ing countries. BMC public health. 2015; 15(1): 1–11. https://doi.org/10.1186/s12889-015-2109-3 PMID:
26231758
6. Auta A, Hadi MA, Oga E, Adewuyi EO, Abdu-Aguye SN, Adeloye D, et al. Global access to antibiotics
without prescription in community pharmacies: a systematic review and meta-analysis. Journal of Infec-
tion. 2019; 78(1): 8–18. https://doi.org/10.1016/j.jinf.2018.07.001 PMID: 29981773
7. MOH. Ghana National Action Plan on Antimicrobial Resistance. [Online].; 2022 [cited 2022 September
22. Available from: https://www.moh.gov.gh/wp-content/uploads/2018/04/NAP_FINAL_PDF_A4_19.
03.2018-SIGNED-1.pdf.
8. Nsimba SED. Assessing the performance, practices and roles of drug sellers/dispensers and mothers’/
guardians’ behaviour for common childhood conditions in Kibaha district, Tanzania. Tropical doctor.
2007; 37(4): 197–201. https://doi.org/10.1258/004947507782333099 PMID: 17988473
9. Indalo AA. Antibiotic sale behaviour in Nairobi: a contributing factor to antimicrobial drug resistance.
East African medical journal. 1997; 74(3): 171–173. PMID: 9185415
10. Mbonye AK, Buregyeya E, Rutebemberwa E, Clarke SE, Lal S, Hansen KS, et al. Prescription for antibi-
otics at drug shops and strategies to improve quality of care and patient safety: a cross-sectional survey
in the private sector in Uganda. BMJ open. 2016; 6(3): e010632. https://doi.org/10.1136/bmjopen-
2015-010632 PMID: 26980439
11. Anong DN, Akoachere JFK. Prescribing patterns and associated factors of antibiotic prescription in pri-
mary health care facilities of Kumbo East and Kumbo West Health Districts, North West Cameroon.
PloS one. 2018; 13(3): e0193353. https://doi.org/10.1371/journal.pone.0193353 PMID: 29505584
12. Alrasheedy AA, Alsalloum MA, Almuqbil FA, Almuzaini MA, Aba Alkhayl BS, Albishri AS, et al. The
impact of law enforcement on dispensing antibiotics without prescription: a multi-methods study from
Saudi Arabia. Expert review of anti-infective therapy. 2020; 18(1): 87–97. https://doi.org/10.1080/
14787210.2020.1705156 PMID: 31834825
13. Afari-Asiedu S, Kinsman J, Boamah-Kaali E, Abdulai MA, Gyapong M, Sankoh O, et al. Afari-Asiedu S.,
Kinsman J., Boamah-Kaali E., Abdulai M. A., Gyapong M., Sankoh O.,.To sell or not to sell; the differ-
ences between regulatory and community demands regarding access to antibiotics in rural Ghana.
Journal of pharmaceutical policy and practice. 2018; 11(1): 1–10.
14. GBRR. E-Registry of Business Regualtion in Ghana. [Online].; 2023 [cited 2023 April 26. Available
from: https://www.bcp.gov.gh/new/reg_details.php?id=NQ==#:~:text=AN%20ACT%20to%20revise%
20the,to%20provide%20for%20related%20matters.
15. PCG. PHARMACY COUNCIL, GHANA. [Online].; 2022 [cited 2022 October 5. Available from: https://
pcghana.org/.
16. Prah J, Kizzie-Hayford J, Walker E, Ampofo-Asiama A. Antibiotic prescription pattern in a Ghanaian pri-
mary health care facility. Pan African Medical Journal, 28(1). 2017.
17. Asante KP, Boamah EA, Abdulai MA, Buabeng KO, Mahama E, Dzabeng F, et al. Knowledge of antibi-
otic resistance and antibiotic prescription practices among prescribers in the Brong Ahafo Region of
Ghana; a cross-sectional study. BMC health services research. 2017; 17(1): 1–9.
18. Amponsah OKO, Nagaraja SB, Ayisi-Boateng NK, Nair D, Muradyan K, Asense PS, et al. High Levels
of Outpatient Antibiotic Prescription at a District Hospital in Ghana: Results of a Cross Sectional Study.
International Journal of Environmental Research and Public Health. 2022; 19(16): 10286. https://doi.
org/10.3390/ijerph191610286 PMID: 36011917
PLOS ONE | https://doi.org/10.1371/journal.pone.0288519 July 13, 2023 14 / 15
PLOS ONE Perceived antimicrobial dispensing practices in medicine outlets
19. Sneddon J, Thompson W, Kpobi LN, Ade DA, Sefah IA, Afriyie D, et al. Exploring the Use of Antibiotics
for Dental Patients in a Middle-Income Country: Interviews with Clinicians in Two Ghanaian Hospitals.
Antibiotics. 2022; 11(8): 1081. https://doi.org/10.3390/antibiotics11081081 PMID: 36009950
20. Baba SM, Salifu AT. Medicines prescribing pattern in northern Ghana: does it comply with WHO recom-
mendations for prescribing indicators? African Journal of Pharmacy and Pharmacology. 2019; 13(6):
70–75.
21. Bekoe SO, Ahiabu MA, Orman E, Tersbøl BP, Adosraku RK, Hansen M, et al. Exposure of consumers
to substandard antibiotics from selected authorised and unauthorised medicine sales outlets in Ghana.
Tropical Medicine & International Health. 2020; 25(8): 962–975.
22. GSS. Ghana Statistical Service. [Online].; 2022 [cited 2023 April 26. Available from: https://statsghana.
gov.gh/gssmain/fileUpload/pressrelease/2021%20PHC%20General%20Report%20Vol%203A_
Population%20of%20Regions%20and%20Districts_181121.pdf.
23. Hensher DA, Rose JM, Greene WH. Applied Choice Analysis: A Primer Cambridge: Cambridge Univer-
sity Press; 2005.
24. Finn A, Louviere JJ. Determining the appropriate response to evidence of public concern: the case of
food safety. Journal of Public Policy & Marketing. 1992; 11(2): 12–25.
25. Dobischok S, Metcalfe R, Matzinger E, Palis H, Marchand K, Harrison S, et al. Dobischok S., Metcalfe
R., Matzinger E, PaliMeasuring the preferences of injectable opioid agonist treatment (iOAT) clients:
Development of a Person-Centered Scale (Best-Worst Scaling). International Journal of Drug Policy.
2023; 112: 103948.
26. Xiong X, Dalziel K, Huang L, Rivero-Arias O. Test-Retest Reliability of EQ-5D-Y-3L Best-Worst Scaling
Choices of Adolescents and Adults. Value in Health. 2023; 26(1): 50–54. https://doi.org/10.1016/j.jval.
2022.07.007 PMID: 35970707
27. Daly TMJ, Anne L, Soutar GN, Rasmi S. Conflict-handling style measurement: A best-worst scaling
application. International Journal of Conflict Management. 2010; 21(3): 281–308.
28. Van Schoubroeck S, Chacon L, Reynolds AM, Lavoine N, Hakovirta M, Gonzalez R, et al. Environmen-
tal sustainability perception toward obvious recovered waste content in paper-based packaging: An
online and in-person survey best-worst scaling experiment. Resources, Conservation and Recycling.
2023; 188: 106682.
29. Rolfe J, Rajapaksa D, De Valck J, Star M. Rolfe J., RajaWill greenhouse concerns impact meat con-
sumption? Best-worst scaling analysis of Australian consumers. Food Quality and Preference, 104,
104755. 2023.
30. Najafzadeh M, Lynd LD, Davis JC, Bryan S, Anis A, Marra M, et al. Barriers to integrating personalized
medicine into clinical practice: a best–worst scaling choice experiment. Genetics in medicine. 2012; 14:
520–526. https://doi.org/10.1038/gim.2011.26 PMID: 22241089
31. Wittenberg E, Bharel M, Bridges JF, Ward Z, Weinreb L. Using best-worst scaling to understand patient
priorities: A case example of papanicolaou tests for homeless women. The Annals of Family Medicine.
2016; 14: 359–364. https://doi.org/10.1370/afm.1937 PMID: 27401425
32. Auger P, Devinney TM, Louviere JJ. Using best–worst scaling methodology to investigate consumer
ethical beliefs across countries. Journal of business ethics. 2007; 70: 299–326.
33. Thurstone LL. A law of comparative judgment. Psychological Review. 1927; 34(2): 273–286.
34. McFadden D. Conditional logit analysis of qualitative choice behavior Zarembka P, editor. New York:
Academic Press: Frontiers in Econometrics, 105–42; 1974.
35. Flynn TN, Marley AA. Best-worst scaling: theory and methods. In Stephane H, Andrew D. Handbook of
choice modelling.: Edward Elgar Publishing; 2014. p. 178–201.
36. Tagoe DNA, Attah C. A Study of Antibiotic Use and Abuse in Ghana: a case study of the Cape Coast
Metropolis. The Internet Journal of Health. 2010; 11: 1–5.
37. Yevutsey SK, Buabeng KO, Aikins M, Anto BP, Biritwum RB, Frimodt-Møller N, et al. Situational analysis
of antibiotic use and resistance in Ghana: policy and regulation. BMC public health. 2017; 17(1): 1–7.
38. Kumar KS, Saranya S, Rani NV. Community pharmacists’ knowledge, attitude, and nonprescription dis-
pensing practices of antibiotics: An explorative study in a selected city of South India. Journal of
Research in Pharmacy Practice. 2022; 11(2): 51. https://doi.org/10.4103/jrpp.jrpp_48_21 PMID:
36798105
PLOS ONE | https://doi.org/10.1371/journal.pone.0288519 July 13, 2023 15 / 15