Journal of Critical Care 40 (2017) 243–250

Contents lists available at ScienceDirect

Journal of Critical Care

journal homepage: www.jccjournal.org

Potential drug-drug interactions in pediatric patients admitted to

intensive care unit of Khyber Teaching Hospital, Peshawar, Pakistan: A
cross-sectional study
Mohammad Ismail, PharmD, PhD a,⁎, Sana Aziz, PharmD a, Sidra Noor, PharmD a, Iqbal Haider, MBBS, FCPS b,
Faryal Shams, PharmD a, Inamul Haq, BPharm, MPH a, Faiza Khadim, PharmD a, Qasim Khan, PharmD a,c,
Fahadullah Khan, PharmD a, Muhammad Asif, PharmD a
a
Department of Pharmacy, University of Peshawar, Khyber Pakhtunkhwa, Pakistan
b
Department of Medicine, Khyber Teaching Hospital, Peshawar. Pakistan
c
Department of Pharmacy, COMSATS Institute of Information Technology, Abbottabad, Pakistan

a r t i c l e i n f o a b s t r a c t

Available online xxxx Purpose: To investigate frequencies, levels, clinical relevance and predictors of potential drug-drug interactions
(pDDIs) in pediatric intensive care unit (PICU).
Keywords: Methods: Case notes of 411 patients were reviewed for pDDIs through Micromedex. Frequencies, levels and
Potential drug-drug interactions clinical relevance of pDDIs were reported. Logistic regression was applied to calculate the odds-ratios for
Drug interactions predictors of pDDIs.
Pediatrics
Results: We recorded pDDIs in 59.4% patients. Major-pDDIs were found in 34.5% patients. Total 990 pDDIs were
Intensive care unit
identified, of which, 37.8% were of moderate-severity and 30.6% of major-severity. Patient's case notes of top-
ten pDDIs showed presence of signs/symptoms such as fever, jaundice, vomiting, anorexia, tachycardia,
drowsiness, & lethargy; and abnormalities in labs such as total leukocytes count, blood urea nitrogen, alanine
aminotransferase, & potassium-level. Odds of exposure to major-pDDIs were significantly higher in patients
aged 6–12 years (p = 0.008); hospital stay of ≥7 days (p = 0.05); and ≥11 prescribed medicines (p b 0.001).
Conclusion: Substantial numbers of patients in PICU are exposed to pDDIs. Major-pDDIs are of particular concern.
Timely identification of pDDIs, preferably with computerized source, is crucial point for their management.
Monitoring of clinically relevant parameters and identification of various predictors are needed to minimize or
prevent the associated negative consequences of pDDIs.
© 2017 Published by Elsevier Inc.

1. Introduction
Hospitalized pediatric patients are often exposed to substantial
polypharmacy [1,2]. The issue is more crucial in case of intensive care set-
tings [3]. Polypharmacy has been identified as a major risk factor for expo-
sure to adverse drug reactions (ADRs) and potential drug-drug
interactions (pDDIs) [3-5]. Drug-drug interactions (DDIs) can result in un-
wanted consequences, varying from ineffective treatment to life threaten-
ing adverse drug events [6-8].
Sufficient published data are available about pDDIs in different clin-
ical specialties including oncology [9], internal medicine [10], geriatrics
Abbreviations: ICU, Intensive Care Unit; PICU, Pediatrics ICU; DDIs, Drug-drug
interactions; pDDIs, potential Drug-Drug Interactions; ADRs, Adverse Drug Reactions; KTH,
Khyber Teaching Hospital; IQR, Interquartile range; PCM, Protein Calorie Malnutrition; DM,
Diabetes Mellitus; DKA, Diabetic Ketoacidosis; CCF, Congestive Cardiac Failure.
⁎ Corresponding author.
E-mail address: ismailrph@upesh.edu.pk (M. Ismail).
[11], cardiology [12], adult intensive care unit (ICU) [13,14] and psychi-
atry [15]. As far as pediatric population is concerned, pDDIs have been
investigated by various studies [2-4,16]. These studies have a number
of limitations and gaps which need to be addressed. Some of these are
not specific to pediatric ICU (PICU) settings, rather these were conduct-
ed either in general pediatric ward or in children hospital [4,16]. There-
fore, their findings cannot be applied to PICU. Some studies have been
conducted specifically in PICU settings [2,3]. One of these studies
worked on adverse drug events [2]. DDIs were not the main target of
this study. Its main implications apply to adverse drug events. DDIs
were given only a brief overview. The study acknowledged this fact in
limitations that more work is needed regarding drug interactions in
PICU [2]. Another study conducted in PICU reported a high exposure
to pDDIs in PICU [3]. The study investigated frequency of pDDIs, their
classes and predictors. However, clinical outcomes were not studied.
The study highlighted need of future research to identify clinical rele-
vance in the form of risk of adverse drug events and actual harm caused

http://dx.doi.org/10.1016/j.jcrc.2017.04.028
0883-9441/© 2017 Published by Elsevier Inc.
244 M. Ismail et al. / Journal of Critical Care 40 (2017) 243–250
by pDDIs [3]. Therefore, studies are needed to explore pDDIs with par-
ticular focus on their clinical relevance.
There is need of explicit work in countries like Pakistan, because liter-
ature is least reported from these settings. In Pakistan, there is no proper
disease-surveillance system and doctors and other healthcare profes-
sionals are overburdened [17]. Medication-therapy is the most common
method of treatment in this country and average number of medications
prescribed per patient is higher as compared with other parts of the world
[18]. In most of the hospitals, established clinical-pharmacy system does
not exist to monitor and optimize medication usage. It has been pointed
out that irrational use of medicines is a common and crucial problem in
Pakistan [19-21]. On the basis of above mentioned facts, there is need of
specific work about pDDIs in PICU settings in Pakistan. Therefore, this
study aimed to explore frequencies, levels, clinical relevance and predic-
tors of pDDIs in pediatric patients admitted to ICU. The secondary aim
was to develop a list of wide-spread interactions of more clinical impor-
tance along with their potential adverse outcomes.
2. Methods
2.1. Study setting and design
This was a cross-sectional study, carried out in PICU of a tertiary care
hospital, Khyber Teaching Hospital (KTH), Peshawar, Pakistan. KTH is a
1200-bed hospital. It is located on main university road in Peshawar
city, providing healthcare coverage to population residing at university
road and adjacent areas. It has also a referral status and many patients
are received from other districts for tertiary care. As far as pharmacy ser-
vices are concerned, there is no proper clinical pharmacy coverage at
ward level. Patient's profiles are developed in hand written format and re-
cords are maintained manually. Moreover, there is no computerized drug
interaction screening programs available in hospital. ICU in department of
pediatrics is a 12-bed setting with 2 ventilator, 12 specialized nurses, 8
medical officers, 4 consultants and 6 residents. It provides 24/7 coverage.
Critically ill patients are shifted to PICU from pediatric units on call.
2.2. Administrative and ethical approval
Permission from hospital administration was obtained in order to
access patients' case notes. Ethics approval was obtained from Ethical
Committee of the hospital under reference number 02-KTH/EC.
2.3. Inclusion and exclusion criteria
Following were the inclusion criteria:
• Pediatric patients admitted to PICU during a one year period, January
2014 to December 2014.
• Both male and female pediatric patients.
Patients were excluded if their medications profiles were incom-
plete with respect to relevant data needed for study. Moreover, data of
expired patients were not included in this study.
On the basis of above criteria, 411 patients were included in this
study. Data regarding hospital admissions, patients' demographics,
medications therapy and labs tests results were collected retrospective-
ly on a structured data collection form.
2.4. Screening of medications profiles for pDDIs
Different software are available for DDIs screening but we selected
Micromedex Drug-Reax® (Truven Health Analytics, Greenwood Village,
Colorado, USA) [22] because it has been reported to have highest score
in term of sensitivity, specificity and completeness [5,23,24]. Patients'
case notes were reviewed for pDDIs using this software. Interactions
were double confirmed by authors' team. Initially screening was
performed by two members, and then confirmed by other two senior
members of the team. According to description of Micromedex Drug-
Reax®, drug interactions are classified on basis of severity-levels and doc-
umentation-levels as follows [22]:
2.4.1. Severity-levels
• Contraindicated: The drugs are contraindicated for concurrent use.
• Major: The interaction may be life threatening and/or require medical
intervention to minimize or prevent adverse effects.
• Moderate: The interaction may result in exacerbation of the patient
condition and/or require an alteration in therapy.
• Minor: The interaction would have limited clinical effects. Manifesta-
tions may include an increase in the frequency or severity of the side
effects but generally would not require a major alteration in therapy.
2.4.2. Documentation-levels
• Excellent: Controlled studies have clearly established the existence
of the interaction.
• Good: Documentation strongly suggests the interaction exists, but
well-controlled studies are lacking.
• Fair: Available documentation is poor, but pharmacologic consider-
ations lead clinicians to suspect the interaction exists; or documenta-
tion is good for a pharmacologically similar drug.
Frequencies of pDDIs as well as frequencies of severity-levels (con-
traindicated, major, moderate and minor) were identified. Clinical rele-
vance of top-ten interactions was reported by correlating them with
patients' signs, symptoms and lab tests results.
3. Statistical analysis
Quantitative forms of variables including patients' age, hospital stay,
numbers of prescribed medications and number of pDDIs per patient
are presented as median and interquartile range (IQR), whereas, they
are analyzed in categorical forms as well by mentioning their percentages
and frequencies. Categorical variables including gender, frequency of
pDDIs, their severity-levels and documentation-levels are presented in
the form of frequencies and percentages. Logistic regression analysis
was applied in order to identify association of the presence of one or
more pDDIs with patients' age, gender, stay in hospital and number of
medicines prescribed. Moreover, association of the presence of major-
pDDIs with above mentioned variables was also identified. Exposure to
pDDIs of any severity, or, exposures to major-pDDIs were the dependent
variables in the model. Variables that were taken as predictors of pDDIs
included gender, age, stay in hospital, and number of prescribed medica-
tions. For each predictor odds ratios (OR) and 95% confidence intervals
(CI) were determined. Initially, univariate logistic regression analysis
was carried out. Then, for variables with significant univariate p-values,
multivariate analyses were performed. In this study, p-value of 0.05 or
less was considered statistically significant. SPSS-v23 was used for statis-
tical analyses of the data.
4. Results
4.1. Patients' Characteristics
Of the total studied patients (n = 411), 157 (38.2%) were females and
254 (61.8%) were males (Table 1). Majority of the patients were of age
less than two years (n = 315; 76.7%), with hospital stay of four or more
days (n = 243; 59.1%) and with four or more number of prescribed med-
icines (n = 375; 91.2%) (Table 1). Median age was seven months (IQR =
2–24), median stay in hospital was four days (IQR = 3–7) and median
number of medications prescribed was eight (IQR = 5–11). The most fre-
quent diagnoses were Pneumonia (n = 87) followed by Sepsis (n = 84)
and then Meningitis (n = 50).
 M. Ismail et al. / Journal of Critical Care 40 (2017) 243–250 245

Table 1 101 of major-severity, 147 of fair scientific-evidence and 127 of good

General characteristics of study subjects. scientific-evidence.
Characteristic Patients: n (%)

Gender
4.4. Predictors of pDDIs
Male 254 (61.8)
Female 157 (38.2) Table 2 shows exposure to pDDIs, both all types of pDDIs and major-
Age
pDDIs, stratified with respect to patient's characteristics i.e., gender, age,
≤1 month 78 (19) stay in hospital, and number of prescribed medicines. In males, pDDIs
1 month−2 years 237 (57.7) was more frequent (35.3%) as compared with females (24.1%). This
2 years–6 years 54 (13.1) table further shows that as number of prescribed medicines increases
6 years–12 years 36 (8.8)
frequency of pDDIs also increased. Similarly, males (20.2%) have high
12 years–18 years 6 (1.5)
Median 7 frequency of major-pDDIs as compared with females (14.4%). A corre-
IQR 2–24 sponding increase was observed in exposure to major-pDDIs with in-
creased number of prescribed medicines.
Hospital stay (days)
≤3 168 (40.9) Results of univariate logistic regression analysis for exposure to one
4–6 134 (32.6) or more pDDIs are displayed in Table 3. In univariate logistic regression
≥7 109 (26.5) analysis, association of the presence of pDDIs with patient's gender was
Median 4 not significant (p = 0.2). However, significant association was observed
IQR 3–7
for the presence of pDDIs with patient's age of 1 month-2 years (p =
Prescribed medications per patient 0.02), 2–6 years (p = 0.001), and 6–12 years (p b 0.001); hospital stay
≤3 36 (8.8) of ≥7 (p b 0.001); 4–6 prescribed medicines (p = 0.01), 7–10 prescribed
4–6 132 (32.1)
medicines (p b 0.001), and ≥11 prescribed medicines (p b 0.001).
7–10 132 (32.1)
≥11 111 (27) Results of univariate logistic regression analysis based on exposure
Median 8 to major-pDDIs were as follows (Table 3). Association of major-pDDIs
IQR 5–11 with patient's gender was not significant (p = 0.3). Whereas, associa-
tion was significant with patients' age of 1 month-2 years (p = 0.03),
Diagnosis n
Pneumonia 87
2–6 years (p = 0.02), 6–12 years (p = 0.001); hospital stay of ≥7 (p b
Sepsis 84 0.001); 7–10 prescribed medicines (p = 0.004), and ≥ 11 prescribed
Meningitis 50 medicines (p b 0.001).
Bronchiolitis 29 Results of multivariate logistic regression analysis were displayed in
PCM 19
Table 4. As per Nagelkerke R2, regression model explains 38% variation
Myocarditis 17
Hypoxic-Ischemic Encephalopathy 16 for presence of pDDIs of any severity. Significant association was
Asthma 14 found for the presence of pDDIs with patient's age of 2–6 years (p =
DM 13 0.008), 6–12 years (p = 0.005); 4–6 prescribed medicines (p = 0.01),
Pulmonary Tuberculosis 12
7–10 prescribed medicines (p b 0.001), and ≥11 prescribed medicines
Post Measles Pneumonia 12
DKA 12
(p b 0.001). While, considering odds-ratios for predictors of pDDIs, it
Gastroenteritis 11 was higher in patient's age 6–12 years and ≥11 prescribed medicines.
Epidermal Nervous Syndrome 11 According to Nagelkerke R2, for presence of major-pDDIs regression
Diarrhea 11 model explains 37% of variation. Significant association was observed
CCF 11
for the presence of major-pDDIs with patients' age 6–12 years (p =
Seizure 10
Fits 10 0.008); hospital stay of ≥7 days (p = 0.05); 7–10 prescribed medicines
Measles 10 (p = 0.004), and ≥11 prescribed medicines (p b 0.001) (Table 4). Simi-
Miscellaneous 334 larly, odds of exposure to major-pDDIs were also higher for these
IQR = Interquartile range; PCM = Protein Calorie Malnutrition; DM = Diabetes predictors.
Mellitus; DKA = Diabetic Ketoacidosis; CCF = Congestive Cardiac Failure.
4.5. Wide-spread interacting drug combinations
4.2. Frequencies of pDDIs
In this study, most commonly detected contraindicated, major or
Results about frequencies of pDDIs are presented in Fig. 1 which moderate-pDDIs (n = 456) resulted in 46.0% of all pDDIs i.e., 456 out
shows that of total 411 patients, 244 were exposed to pDDIs (overall- of 990 (Table 5). Potential adverse outcomes of these interactions
frequency = 59.4%). Similarly, frequencies of moderate-pDDIs and included hepatotoxicity, QT interval prolongation and arrhythmias,
major-pDDIs were 152 (37%) and 142 (34.5%) respectively. Lowest fre- ototoxicity, nephrotoxicity, postural hypotension, digoxin toxicity,
quencies were recorded for contraindicated-pDDIs and minor-pDDIs. In alteration of drug absorption, tendon rupture, neuromuscular weak-
28.7% cases, one or two pDDIs were detected. However, a limited por- ness, phenobarbital toxicity and reduction in therapeutic effectiveness.
tion (17.8%) of the study patients was identified with five or more
than five pDDIs. 4.6. Clinical relevance

Table 6 shows clinical findings (relevant signs, symptoms, and labo-

4.3. Levels of pDDIs
A total of 301 interacting-drugs pairs were detected that were pre-
sented in total 990 pDDIs (Fig. 2). The identified pDDIs were classified
on the basis of severity type and scientific-evidence. Of 990 interactions,
37.8% were of moderate-severity and 30.6% of major-severity; whereas,
45.8% and 43.5% were of fair and good scientific-evidence, respectively.
Of 301 interacting drug combinations, 127 were of moderate-severity,
ratory test results) and monitoring/management guidelines for top-ten
most frequent interactions. Patients with the interaction, ceftriaxone
+ calcium containing preparations, presented with fever, sepsis,
nephrolithiasis, hydronephrosis, breathing difficulty, and raised levels
of serum creatinine, BUN, TLC, and neutrophils. Patients with the inter-
action pyrazinamide and rifampin presented with jaundice, vomiting,
anorexia, fever, & lethargy; and raised ALT, AST and ALP. In patients
with digoxin and furosemide interaction, tachycardia, tachypnea,
246 M. Ismail et al. / Journal of Critical Care 40 (2017) 243–250

Fig. 1. Frequencies of potential drug-drug interactions. - pDDIs = Potential drug-drug interactions. - Overall-frequency is the presence pDDIs irrespective of severity type. As total patients
were 411, therefore overall-frequency was 59.4%. - Severity-wise frequencies do not add up to 244 (59.4%) because many study subjects were exposed to interactions of different severity-
levels. - Number of pDDIs per patients: Median = 1 (Interquartile range = 0–3).

Fig. 2. Levels of the identified potential drug-drug interactions. - pDDIs = Potential drug-drug interactions. - In this study, 301 interacting-drugs pairs were detected that were presented in
total 990 pDDIs.
 M. Ismail et al. / Journal of Critical Care 40 (2017) 243–250 247

Table 2 Table 4
Patients' characteristics stratified according to exposure to potential drug-drug Multivariate logistic regression analysis based on exposure to all pDDIs and pDDIs of ma-
interactions. jor-severity.

Variables All types of Only major Variables All pDDIs Major-severity

interactions interactions
OR 95% CI p-value OR 95% CI p-value
n % n %
Age
Gender ≤1 month Reference Reference
Male 145 35.3 83 20.2 1 month−2 1.6 0.9–3 0.1 1.7 0.9–3.5 0.1
Female 99 24.1 59 14.4 years
Total 244 59.4 142 34.5 2 years–6 years 3.3 1.4–8 0.008 2.2 0.9–5.5 0.08
6 years– 5 1.6–14.2 0.005 3.8 1.4–10.3 0.008
Age 12 years
≤1 month 34 8.3 17 4.1 12 years– 3.1 0.5–19 0.2 2.2 0.2–23.3 0.5
1 month −2 years 139 33.8 82 20 18 years
2 years– 6 years 39 9.5 22 5.4
6 years – 12 years 29 7.1 20 4.9 Hospital stay (days)
12 years – 3 0.7 1 0.2 ≤3 Reference Reference
18 years 4–6 0.6 0.4–1.1 0.1 0.7 0.4–1.2 0.2
Total 244 59.4 142 34.5 ≥7 1.4 0.7–2.7 0.3 1.9 1–3.4 0.05

Hospital stay (days) Prescribed medicines

≤3 91 22.1 46 11.2 ≤3 Reference Reference
4–6 70 17.0 35 8.5 4–6 3.9 1.4–11.2 0.01 5.2 0.7–41 0.1
≥7 83 20.2 61 14.8 7–10 15.5 5.6–44.7 b0.001 19.4 2.5–148.3 0.004
Total 244 59.4 142 34.5 ≥11 56 17–183.3 b0.001 65.5 8.4–511 b0.001

Prescribed medicines pDDIs = potential drug-drug interactions; OR = odds ratio; CI = confidence interval; p =
≤3 5 1.2 1 0.2 p-value
4–6 48 11.7 17 4.1
7–10 91 22.1 47 11.4
≥11 100 24.3 77 18.7
Total 244 59.4 142 34.5 5. Discussion

This study presents the frequency and nature of pDDIs in PICU pa-
drowsiness, vomiting and abnormal potassium-levels were recorded. tients. Comparison with studies conducted in adult intensive care set-
Clinical features indicating poor response were observed in patients tings shows reports with a higher prevalence in the range of 70% to
with interactions such as artemether + phenytoin, artemether + phe- 91% [7,27,28] as well as with a lower prevalence of 53% [29]. This incon-
nobarbital, lumefantrine + dexamethasone, artemether + dexametha- sistency may be due to one or more of the following reasons: variability
sone, lumefantrine + phenobarbital, and lumefantrine + phenytoin. In in study designs, nature of study subjects, pattern of drugs prescribing/
patients with the interaction of clarithromycin and levofloxacin, abnor- utilization, and types of DDIs screening tools. Moreover, 59.4% preva-
malities in heart rate and potassium levels were observed. lence in the present work is higher than large number of other studies
conducted in settings other than ICU such as 46% in oncology [9],
52.8% in internal medicine [10] and 25.8% in pediatrics [4]. Furthermore,
as in most of the hospitals in Pakistan there is neither clinical pharmacy
Table 3
department nor DDIs screening programs [10]; irrational use of medi-
Univariate logistic regression analysis based on exposure to all pDDIs and pDDIs of major-
severity. cines, lack of facilities, poor access to available health facilities, burden
of health care professionals and drug related problems including
Variables All pDDIs Major-severity
pDDIs remains a considerable issue [20]; therefore, present results war-
OR 95% CI p OR 95% CI p rant proper consideration of pDDIs in PICU settings. In order to manage
Gender pDDIs, different evidence based strategies are recommended such as
Male Reference Reference screening of medications profiles for pDDIs by the use of computerized
Female 1.3 0.9–1.9 0.2 1.2 0.8–1.9 0.3 screening programs [30] procedure for structured assessment of pDDIs
Age [31], clinical pharmacist participation in evaluating patient case notes
≤1 month Reference Reference for pDDIs [32], assessment of pertinent laboratory tests of clinical rele-
1 month −2 1.8 1.1–3.1 0.02 1.9 1–3.5 0.03 vance of interactions [33], and improving knowledge of health care pro-
years
fessionals about pDDIs [34].
2 years– 3.4 1.6–7.1 0.001 2.5 1.1–5.3 0.02
6 years In terms of severity-levels of pDDIs, in this study, moderate-pDDIs
6 years – 5.4 2.1–14 b0.001 4.5 2–10.5 0.001 were mostly observed followed by major-pDDIs and concerning scien-
12 years tific-evidence, fair type were more common followed by good type of
12 years – 1.3 0.2–6.8 0.8 0.7 0.07–6.6 0.8
scientific-evidence. These findings are consistent with other studies,
18 years
carried out in other specialties or wards [4,35]. A study conducted in pe-
Hospital stay (days) diatric ward reported total 260 pDDIs in medication profiles of 400 pa-
≤3 Reference Reference tients, of these pDDIs, moderate-pDDIs severity (41.5%) and good
4–6 0.9 0.6–1.5 0.7 0.9 0.6–1.6 0.8
≥7 2.7 1.6–4.6 b0.001 3.4 2–5.6 b0.001
scientific-evidences (76.9%) were more common [4]. In another study,
medical profiles of 407 study subjects were screened for interactions.
Prescribed medicines Of total identified pDDIs, most were of moderate severity and good sci-
≤3 Reference Reference
4–6 3.5 1.3–9.7 0.01 5.2 0.7–40.3 0.1
entific-evidences [35]. Severity of pDDIs is one of the main factors to be
7–10 13.8 4.9–37.9 b0.001 19.4 2.6–145.8 0.004 considered for proper evaluation and management of pDDIs. Therefore,
≥11 56.4 18.1–174.7 b0.001 79.3 10.4–602.5 b0.001 it is important for health care providers to properly identify and classify
pDDIs = potential drug-drug interactions; OR = odds ratio; CI = confidence interval; p = pDDIs. It is crucial for clinical management of pDDIs, minimizing their
p-value risk and designing prophylactic measures for prevention.
248 M. Ismail et al. / Journal of Critical Care 40 (2017) 243–250

Table 5
Most frequently identified interactions, their levels and potential adverse outcomes.

Interactions Frequency Severity Evidence Potential adverse outcomes

Ceftriaxone + Calcium containing preparations 45 Contraindicated Good Formation of ceftriaxone-calcium precipitates

Pyrazinamide + Rifampin 37 Major Good Severe hepatic injury
Digoxin + Furosemide 20 Moderate Good Digoxin toxicity
Artemether + Phenytoin 20 Contraindicated Fair Decreased artemether exposure and loss of efficacy
Artemether + Phenobarbital 19 Contraindicated Fair Decreased artemether exposure, loss of efficacy
Lumefantrine + Dexamethasone 18 Contraindicated Fair Decreased lumefantrine exposure, loss of efficacy
Artemether + Dexamethasone 18 Contraindicated Fair Decreased artemether exposure, loss of efficacy
Lumefantrine + Phenobarbital 17 Contraindicated Fair Decreased lumefantrine exposure, loss of efficacy
Clarithromycin + Levofloxacin 17 Major Fair Increased risk of QT interval prolongation
Lumefantrine + Phenytoin 17 Contraindicated Fair Decreased lumefantrine exposure and loss of efficacy
Dexamethasone + Rifampin 16 Moderate Good Decreased dexamethasone effectiveness
Amikacin + Furosemide 14 Major Fair Increased ototoxicity and/or nephrotoxicity
Captopril + Furosemide 14 Moderate Good Postural hypotension
Clarithromycin + Dexamethasone 14 Major Fair Decrease clarithromycin exposure and increased dexamethasone exposure
Levofloxacin + Zinc 14 Moderate Fair Decreased levofloxacin effectiveness
Chelated iron +Zinc 14 Moderate Excellent Decreased gastrointestinal absorption of iron and/or zinc
Phenytoin + Valproic acid 12 Moderate Good Altered valproate levels or phenytoin levels
Ciprofloxacin + Zinc 11 Moderate Fair Decreased ciprofloxacin effectiveness
Clarithromycin + Rifampin 11 Moderate Excellent Decreased serum concentration of clarithromycin
Dexamethasone + Levofloxacin 11 Moderate Excellent Increased risk for tendon rupture
Artemether + Rifampin 11 Contraindicated Excellent Decreased artemether exposure and loss of efficacy
Lumefantrine + Rifampin 11 Contraindicated Excellent Decreased lumefantrine exposure and loss of efficacy
Dexamethasone + Phenobarbital 10 Moderate Good Decreased dexamethasone effectiveness
Ciprofloxacin + Metronidazole 9 Major Fair Increased risk of QT-interval prolongation and arrhythmias
Furosemide + Ibuprofen 8 Moderate Good Decreased diuretic and antihypertensive efficacy
Amikacin + Antacid 8 Moderate Fair Neuromuscular weakness
Clarithromycin + Phenobarbital 7 Major Fair Reduced clarithromycin exposure and loss of efficacy
Furosemide + Gentamicin 7 Major Good Ototoxicity and/or nephrotoxicity
Ciprofloxacin + Calcium containing preparations 7 Moderate Good Decreased ciprofloxacin efficacy
Phenobarbital + Valproic acid 7 Moderate Good Phenobarbital toxicity or decreased valproic acid effectiveness
Digoxin + Rifampin 7 Moderate Excellent Decreased digoxin levels

Seriously ill patients usually receive multidrug therapy to get
pharmaco-therapeutic cure. There is positive relationship between
number of prescribed drugs and risk of pDDIs [36]. In this respect, ICU
patients are more likely to be at risk to pDDIs. Significant associations
of pDDIs with increased age and number of prescribed medicines,
found in this study, are in accordance with previous studies [4,10,27,
29]. We have calculated odds of exposure to major-pDDIs separately.
Significant results for association of major-pDDIs with age, longer stay
in hospital and increased numbers of prescribed medicines are consis-
tent with other studies which have reported odds of exposure for
pDDIs [4,9,10]. Based on these odds-ratios patient's in PICU are at con-
siderable risk to pDDIs, particularly to major-pDDIs. Due to sensitive na-
ture of pediatric patients in PICU, therapy of all patients, particularly
patients with longer hospital stay and poly-pharmacy must be screened
for pDDIs, preferably through computerized DDIs screening tools. Such
screening is necessary before prescribing/administering the drugs, in
order to identify pDDIs and manage or prevent them accordingly.
List of most frequently detected major, moderate and contraindi-
cated pDDIs is of prime importance for physicians and clinical pharma-
cists. It can be used for selective/targeted screening of DDIs. Patients
with these interactions may be given more attention and their therapy
should be carefully monitored for all potential adverse outcomes.
Clinical relevance of interactions shows potential effects of interac-
tions on clinical features and laboratory tests results, which further
highlights the importance of medication profiles screening for DDIs,
also enlightens by other studies too [37,38]. Monitoring parameters
for interactions and management guidelines will be helpful for physi-
cians and clinical pharmacists to evaluate and manage DDIs in PICU
setting.
Following are the potential limitation of this study. This study has
been conducted in a single setting. Although a similar pattern is ex-
pected in other settings but multicenter data may explore the issue
in more depth. Another limitation is the retrospective nature of the
study, which does not allow us to study interactions in more details
with respect to clinical relevance. Therefore, further studies should
be carried out to consider in depth magnitude of interactions in
clinical area of PICU.
6. Conclusion
Substantial numbers of patients in PICU are exposed to pDDIs.
Major-pDDIs are of particular concern. Timely identification of pDDIs,
preferably with some computerized source, is a crucial point for their
management. Monitoring of all clinically relevant signs, symptoms
and labs; and proper identification of various predictors such as
polypharmacy, longer hospital stay and specific age groups are needed
to minimize or prevent the associated negative consequences of pDDIs.
Conflict of interest
All authors declare that they have no conflict of interest.
Funding
No funding had been taken for this study from any source.
Authors' contributions
Mohammad Ismail (MI); design and develop concept and meth-
odology of research, literature review, manuscript writing, statistical
analysis, overall supervision.
Sana Aziz (SA); literature review, data collection, data compilation
and DDIs screening.
Sidra Noor (SN); data compilation, manuscript writing, literature
review.
Iqbal Haider (IH1); design and develop concept and methodology of
research, data collection, statistical analysis, overall supervision.
 M. Ismail et al. / Journal of Critical Care 40 (2017) 243–250 249

Table 6
Clinical relevance, monitoring and management guidelines of top-ten interactions.

Interactions⁎ Signs and symptoms⁎ Laboratory investigations⁎ Monitoring/management guidelines [22,25,26]

Ceftriaxone + Calcium containing
preparations (45)
Pyrazinamide + Rifampin (37)
Digoxin + Furosemide (20)
Artemether + Phenytoin (20)
Artemether + Phenobarbital (19)
Lumefantrine + Dexamethasone (18)
Artemether + Dexamethasone (18)
Lumefantrine + Phenobarbital (17)
Clarithromycin + Levofloxacin (17)
Lumefantrine + Phenytoin (17)
Fever (31) Serum Creatinine: normal (13), high (4) Do not mix or administer ceftriaxone concurrently
Cough (15) BUN: normal (8), high (15) with calcium-containing IV solutions in the same IV
Sepsis (3) TLC: low (1), normal (15), high (10) administration line, including continuous calcium-containing
Breathing difficulty (19) Neutrophils: normal (4), high (7) infusions such as parenteral nutrition via a Y-site.
Nephrolithiasis (1) Lymphocytes: low (5), normal (5), high (1)
Hydronephrosis (1)
Fever (27) Serum bilirubin: abnormal (1) Monitoring of liver function tests at baseline and at 2, 4, 6,
Jaundice (11) ALT: normal (9), high (7) and 8 weeks of treatment. Patient education about
Vomiting (8) AST: high (2) reporting symptoms of liver injury.
Lethargy (6) ALP: low (3), normal (5), high (3)
Anorexia (2)
Weight loss (2)
Abdominal pain (1)
Anemia (1)
Tachycardia (8) Potassium level: low (1), Potassium and magnesium monitoring. Patient education
Tachypnea (8) normal (12), high (2) about decrease intake of dietary potassium
Drowsiness (5) and/or potassium supplements.
Vomiting (3)
Abdominal pain (2)
Lethargy (2)
Distress (1)
Diarrhea (1)
Fever (17) TLC: low (1), normal (10), high (5) Monitoring efficacy of artemether by checking signs
Vomiting (6) Neutrophils: normal (3), high (4) and symptoms of malaria and malarial parasite in blood.
Pale (3) Lymphocytes: low (3), normal (4) Alternate anti-malarial which have no interacting
Anorexia (1) capability may be given.
Fever (13) TLC: low (3), normal (8), high (3) Monitoring efficacy of artemether by checking signs
Pale (4) Neutrophils: normal (1), high (3) and symptoms of malaria and malarial parasite in blood.
Anorexia (3) Lymphocytes: low (2), normal (2) Alternate anti-malarial which have no interacting
Vomiting (2) capability may be given.
Anemia (1)
Lethargy (1)
Fever (15) TLC: low (1), normal (9), high (3) Monitoring efficacy of artemether by checking signs
Pale (6) Neutrophils: normal (1), high (4) and symptoms of malaria and malarial parasite in blood.
Vomiting (5) Lymphocytes: low (3), normal (2) Alternate anti-malarial which have no interacting
Anorexia (2) capability may be given.
Lethargy (2)
Fever (15) TLC: low (1), normal (9), high (3) Monitoring efficacy of artemether by checking signs
Pale (6) Neutrophils: normal (1), high (4) and symptoms of malaria and malarial parasite in blood.
Vomiting (5) Lymphocytes: low (3), normal (2) Alternate anti-malarial which have no interacting
Lethargy (4) capability may be given.
Anorexia (2)
Fever (11) TLC: low (3), normal (6), high (3) Monitoring efficacy of artemether by checking signs
Vomiting (5) Neutrophils: normal (1), high (3) and symptoms of malaria and malarial parasite in blood.
Pale (4) Lymphocytes: low (2), normal (2) Alternate anti-malarial which have no interacting
Lethargy (3) capability may be given.
Anorexia (2)
Anemia (1)
Tachycardia (7) Potassium level: low (2), Monitoring of signs and symptoms of QT interval
Bradycardia (1) normal (5), high (1) prolongation, especially in high risk patients.
Fever (14) TLC: low (1), normal (9), high (4) Monitoring efficacy of artemether by checking signs
Vomiting (6) Neutrophils: normal (2), high (3) and symptoms of malaria and malarial parasite in blood.
Pale (5) Lymphocytes: low (2), normal (3) Alternate anti-malarial which have no interacting
Lethargy (3) capability may be given.
Anorexia (1)

BUN = blood urea nitrogen, TLC = total leukocytes count, ALT = alanine aminotransferase, AST = aspartate aminotransferase, ALP = alkaline phosphatase.
⁎ Frequencies are given in parenthesis and calculated among patients with respective interaction.

Faryal Shams (FS); data collection, data compilation and DDIs
screening.
Inamul Haq (IH2); design and develop concept and methodology of
research, data collection.
Faiza Khadim (FK1); data collection, data compilation and DDIs
screening.
Qasim Khan (QK); design and develop concept and methodology of
research, data compilation.
Fahadullah Khan (FK2); data collection, data compilation and DDIs
screening.
Muhammad Asif (MA); data collection, data compilation and DDIs
screening.
All authors read and approved final version of manuscript.
Acknowledgements
We are very thankful for the cooperation of staff and administration
of the hospital.
References
[1] Feudtner C, Dai D, Hexem KR, Luan X, Metjian TA. Prevalence of polypharmacy expo-
sure among hospitalized children in the United States. Arch Pediatr Adolesc Med
2012;166:9–16.
[2] Silva DC, Araujo OR, Arduini RG, Alonso CF, Shibata AR, Troster EJ. Adverse drug
events in a paediatric intensive care unit: a prospective cohort. BMJ Open 2013;
3:e001868.
[3] Dai D, Feinstein JA, Morrison W, Zuppa AF, Feudtner C. Epidemiology of
polypharmacy and potential drug–drug interactions among pediatric
250 M. Ismail et al. / Journal of Critical Care 40 (2017) 243–250
patients in ICUs of US Children's hospitals. Pediatr Crit Care Med 2016;17
e218–e228.
[4] Ismail M, Iqbal Z, Khan MI, Javaid A, Arsalan H, Farhadullah Khan F, et al. Frequency,
levels and predictors of potential drug-drug interactions in a pediatrics ward of a
teaching hospital in Pakistan. Trop J Pharm Res 2013;12:401–6.
[5] Reis AMM, Cassiani SHDB. Evaluation of three brands of drug interaction software
for use in intensive care units. Pharm World Sci 2010;32:822–8.
[6] Rodrigues AT, Stahlschmidt R, Granja S, Falcão ALE, Moriel P, Mazzola PG. Clinical
relevancy and risks of potential drug–drug interactions in intensive therapy. Saudi
Pharm J 2015;23:366–70.
[7] Reis AM, Cassiani SH. Prevalence of potential drug interactions in patients in an in-
tensive care unit of a university hospital in Brazil. Clinics (Sao Paulo) 2011;66:9–15.
[8] Krahenbuhl-Melcher A, Schlienger R, Lampert M, Haschke M, Drewe J, Krahenbuhl S.
Drug-related problems in hospitals: a review of the recent literature. Drug Saf 2007;
30:379–407.
[9] van Leeuwen RW, Brundel DH, Neef C, van Gelder T, Mathijssen RH, Burger DM, et al.
Prevalence of potential drug-drug interactions in cancer patients treated with oral
anticancer drugs. Br J Cancer 2013;108:1071–8.
[10] Ismail M, Iqbal Z, Khattak MB, Khan MI, Arsalan H, Javaid A, et al. Potential drug-drug
interactions in internal medicine wards in hospital setting in Pakistan. Int J Clin
Pharmacol 2013;35:455–62.
[11] Hoffmann W, van den Berg N, Thyrian JR, Fiss T. Frequency and determinants of po-
tential drug-drug interactions in an elderly population receiving regular home visits
by GPs-results of the home medication review in the AGnES-studies.
Pharmacoepidemiol Drug Saf 2011;20:1311–8.
[12] Ismail M, Iqbal Z, Khattak MB, Khan MI, Javaid A, Khan TM. Potential drug-drug in-
teractions in cardiology ward of a teaching hospital. HealthMED 2012;6:1618–24.
[13] Askari M, Eslami S, Louws M, Wierenga PC, Dongelmans DA, Kuiper RA, et al. Fre-
quency and nature of drug-drug interactions in the intensive care unit.
Pharmacoepidemiol Drug Saf 2013;22:430–7.
[14] Ismail M, Khan F, Noor S, Haider I, I-u Haq, Ali Z, et al. Potential drug–drug interac-
tions in medical intensive care unit of a tertiary care hospital in Pakistan. Int J Clin
Pharmacol 2016;38:1052–6.
[15] Ismail M, Iqbal Z, Khattak MB, Javaid A, Khan MI, Khan TM, et al. Potential drug-drug
interactions in psychiatric ward of a tertiary care hospital: prevalence, levels and as-
sociation with risk factors. Trop J Pharm Res 2012;11:289–96.
[16] Feinstein J, Dai D, Zhong W, Freedman J, Feudtner C. Potential drug− drug interac-
tions in infant, child, and adolescent patients in children's hospitals. Pediatrics 2015;
135:e99-108.
[17] World Health Organization. Report of the health system review mission – Pakistan.
Islamabad: World Health Organization; 2007.
[18] Creese A, Gasman N. Mariko M. The World Medicines Situation: World Health Orga-
nization; 2004.
[19] Khan FA, Hoda MQ. Drug related critical incidents. Anaesthesia 2005;60:48–52.
[20] Nishtar S. Pharmaceuticals–strategic considerations in health reforms in Pakistan. J
Pak Med Assoc 2006;56:S100–11.
[21] Babar ZU, Jamshed S. Social pharmacy strengthening clinical pharmacy: why phar-
maceutical policy research is needed in Pakistan? Pharm World Sci 2008;30:617–9.
[22] Micromedex Drug-Reax®, Greenwood Village, CO: Truven Health Analytics., http://
www.micromedexsolutions.com/home/dispatch; [accessed 10 November, 2016].
[23] Patel RI, Beckett RD. Evaluation of resources for analyzing drug interactions. J Med
Libr Assoc 2016;104:290.
[24] Kheshti R, Aalipour M, Namazi S. A comparison of five common drug–drug interac-
tion software programs regarding accuracy and comprehensiveness. J Res Pharm
Pract 2016;5:257.
[25] Tatro DS. Drug Interaction Facts: Facts and Comparisons. Wolters Kluwer Health: St.
Louis, Missouri; 2009.
[26] Stockley's Baxter K, Interactions Drug. 9th Editon. London, Chicago: Pharmaceutical
Press; 2010.
[27] Abideen S, Vivekanandan K, Mishra P. Assessment of prevalence of potential drug-
drug interactions in medical intensive care unit of a tertiary care hospital in india.
Asian J Pharm Clin Res 2015;8:125–30.
[28] Hasanloei V, Amin M, Hamdolah S, Aysa H. Drug–drug interactions prevalence in in-
tensive care unit patients of a university hospital in Iran. Bull Env Pharmacol Life Sci
2014;3:87–91.
[29] Mohamed SMAES, Gad ZM, El-Nimr NA, Razek AAHA. Prevalence and pattern of po-
tential drug-drug interactions in the critical care units of a tertiary hospital in Alex-
andria. Egypt Adv in Pharmacoepidemiol Drug Saf 2014;2013.
[30] Moura CS, Prado NM, Belo NO, Acurcio FA. Evaluation of drug–drug interaction
screening software combined with pharmacist intervention. Int J Clin Pharmacol
2012;34:547–52.
[31] van Roon EN, Flikweert S, le Comte M, Langendijk PN, Kwee-Zuiderwijk WJ, Smits P,
et al. Clinical relevance of drug-drug interactions. Drug Saf 2005;28:1131–9.
[32] Roblek T, Deticek A, Leskovar B, Suskovic S, Horvat M, Belic A, et al. Clinical-pharma-
cist intervention reduces clinically relevant drug–drug interactions in patients with
heart failure: a randomized, double-blind, controlled trial. Int J Cardiol 2016;203:
647–52.
[33] Geerts AF, De Koning FH, De Smet PA, Van Solinge WW, Egberts TC. Laboratory tests
in the clinical risk management of potential drug-drug interactions. Drug Saf 2009;
32:1189–97.
[34] van Leeuwen RW, Jansman FG, van den Bemt PM, de Man F, Piran F, Vincenten I,
et al. Drug–drug interactions in patients treated for cancer: a prospective study on
clinical interventions. Ann Oncol 2015:mdv029.
[35] Riechelmann RP, Tannock IF, Wang L, Saad ED, Taback NA, Krzyzanowska MK. Po-
tential drug interactions and duplicate prescriptions among cancer patients. J Natl
Cancer Inst 2007;99:592–600.
[36] Johnell K, Klarin I. The relationship between number of drugs and potential drug-
drug interactions in the elderly: a study of over 600,000 elderly patients from the
Swedish prescribed drug register. Drug Saf 2007;30:911–8.
[37] Zwart-van Rijkom JE, Uijtendaal EV, Ten Berg MJ, Van Solinge WW, Egberts AC. Fre-
quency and nature of drug–drug interactions in a Dutch university hospital. Br J Clin
Pharmacol 2009;68:187–93.
[38] Cruciol-Souza JM, Thomson JC. A pharmacoepidemiologic study of drug interactions
in a Brazilian teaching hospital. Clinics 2006;61:515–20.