Evaluation of current practice of antimicrobial use and clinical outcome in the management of pneumonia among pediatric patients admitted to Jimma Medical Center, Southwest Ethiopia: a prospective observational study

Abstract

Background:

Inappropriate use of antibiotics is one of the major contributing factors for the development of antibiotic resistance, threatening the effective prevention and treatment of common infections. Appropriate antimicrobial use is the cornerstone for curbing antimicrobial resistance and achieving good clinical and economic outcomes. The issues of appropriate antimicrobial use are of critical importance to the global community.

Objectives:

To evaluate the current practices of antimicrobial utilization and outcomes in the management of Pneumonia among pediatric patients admitted to the hospital.

Method:

A prospective observational study design was conducted on children admitted to the pediatric wards of the hospital. The study was conducted from 3 February 2022 to 3 June 2022, and patients aged <18 years and diagnosed with pneumonia were included. A chart review supplemented by a self-administered questionnaire was used to collect data. Descriptive statistics and binary logistic regressions were performed for data analyses.

Results:

Among the total of 146 patients, 61.6% were male, and the mean age was 40.95 (+47.61) months. Treatment approaches in all were empirical and no de-escalation therapy was made even after culture results. Ceftriaxone was the most commonly prescribed antimicrobial. Patients’ outcomes included a clinical cure rate of 65.1%, an in-hospital mortality rate of 7.5%, and a complication rate of 27.3%. Poor clinical outcome was associated with recent antimicrobial use (AOR = 3.87; 95% CI:1.34–11.16; P = 0.012), antimicrobial changed (AOR = 3.74; 95% CI: 1.522–9.22; P = 0.004), and longer hospital stay (>10 days) (AOR = 6.00; 95% CI: 2.53–14.22; P = 0.029) were all independent predictors of poor clinical outcome in treatment of pneumonia.

Conclusion:

Antimicrobials were prescribed empirically without sufficient evidence of indication such as microbiological and radiologic findings. The practice also is not based on local guidelines and has no multidisciplinary approach in management. These factors likely contributed to higher rates of poor outcomes (more than one-fourth of the patients) when compared with similar studies in other countries. Therefore, these observations need the hospital’s establishment of an antimicrobial stewardship program centered on fundamental strategies in pediatrics, including preauthorization, prospective audits with feedback, and pharmacist-led interventions that are feasible and accepted in the inpatient setting as soon as possible.

HIGHLIGHTS

• Pneumonia is the main factor leading to intensive use of antimicrobials in the hospital.

• In Ethiopia there are indications of inappropriate use of antimicrobial drugs.

• Ensuring that our current cadre of antimicrobials is used more wisely is essential.

• Empirically initiated antimicrobial was completed without evidence of indication.

• More than one-fourth of patients treated for pneumonia experienced poor outcomes.

Introduction

The discovery of antimicrobial agents is one of the most important medical advances in human history to reduce disease and save billions of lives over the years^[1,2]. However, the successful use of antimicrobial agents is compromised by misuse^[3–5], and the development of resistance in the past few years^[5]. Antimicrobial resistance is a global crisis that threatens to reverse a century of medical progress; threatening the effective prevention and treatment of common infections, leading to increased morbidity and mortality, and associated economic costs due to the health care burden^[6,7]. Reports indicate that there will be about 10 million AMR-related deaths every year until 2050, with the majority being in Africa and Asia^[8]. In Ethiopia, reports have indicated that the rates of antimicrobial resistance among commonly isolated pathogens to frequently prescribed antimicrobials are rapidly increasing^[9]. According to a systematic review and meta-analysis of 33 different studies, the pooled prevalence of antimicrobial resistance in the country was 59.7%^[10].

Inappropriate antimicrobial use is a major driver of antimicrobial resistance (AMR). Earlier Studies conducted in pediatric wards in Ethiopian hospitals have shown inappropriate use of antimicrobials^[11–13]. In these studies, one of the main reasons for inappropriate antimicrobial use was the management of pneumonia. Pneumonia is the leading cause of morbidity and mortality in Ethiopia, especially in the pediatric population, and its main factors lead to hospitalization and intensive use of antimicrobials^[14].

In addition to the AMR, Inappropriate and indiscriminate use of antimicrobial lead to undesirable consequences, including severe illnesses, prolonged hospitalization, increased morbidity and mortality, increased Clostridium difficile, and other hospital-acquired infections (HAIs), and increased healthcare expenses^[6,7,15].

The Ministry of Health and its subordinates, in particular, the Ethiopian Food and Drug Administration (EFDA), have worked together to tackle antimicrobial resistance by developing national drug policies^[16], developing standard treatment guidelines for the different healthcare establishments at different times^[17,18], drug formulary, good prescription and dispensing manuals^[13] and different practical guides since the issuance of declaration 661 2009. All of these efforts are being made to tackle AMR through the promotion of rational antimicrobial use and ultimately protect citizens^[19]. However, the effective implementation of standards and most treatment guidelines remains unexplored. Therefore, this study attempted to evaluate antimicrobial use in current practice and clinical outcomes in the management of childhood pneumonia. The findings could help policymakers design appropriate strategies so that antimicrobial utilization could be optimized and patient and economic outcomes are improved. Ultimately, the findings may call for policymakers for effective drug regulation in Ethiopia. This cross-sectional study has been reported in line with the STROCSS guidelines^[20].

Material and methods

Study setting and period

The study was conducted from 3 February 2022 to 3 June 2022, in the pediatric ward at one of the teaching and referral hospitals located in Ethiopia. The hospital provides services for approximately 15 000 inpatients, 160 000 outpatient attendants, and 11 000 emergency cases per year. The hospital serves more than 20 million people coming to the hospital from the catchment area. It has different wards. The pediatric ward is among the wards that have different units such as level I, level II, neonatal unit, intensive care unit, and oncology unit.

Study design

A hospital-based prospective observational study was conducted on children with pneumonia admitted to the pediatric ward.

Population

All pediatric patients admitted to the pediatric ward of Jimma Medical Center with a diagnosis of pneumonia.

Study population

All pediatric patients admitted to the pediatric ward of the Jimma Medical Center with a diagnosis of pneumonia during the study period and fulfilled inclusion criteria.

Inclusion and exclusion criteria

Inclusion criteria

• All patients aged less than or equal to 18 years admitted to the pediatric wards of Jimma Medical Center with a suspected or proven pneumonia diagnosis by physician decision on duty during the study period were included in the study.

• Patients whose parents gave and signed the written assent.

Exclusion criteria

• Patients who died before they started antimicrobial.

• Patients transferred to other facilities.

• Patient lost follow-up.

Sample size and sampling technique

No sampling technique was applied; instead, all pediatric patients diagnosed with pneumonia who fulfilled illegibility criteria during the study period were recruited in the study. Accordingly, 146 patients were followed and completed data collected.

Study variables

Independent variables

Patient-related factors (age, sex, weight, residence, immunization status, nutritional status, type of admission, reason for admission), Disease-related factors (presence of comorbidities, duration of hospital stay, WHO severity class type of pneumonia, diagnostic method), Antimicrobial factors (antimicrobial regimens, time of change of antimicrobial, number of missed antimicrobial doses, duration of antimicrobial therapy, type of therapy(definitive/empirical), recent antimicrobial exposure, number of antimicrobials)

Dependent variable

Clinical outcome

Data collection tool and procedures

The data collection instrument (data abstraction format) was developed by reviewing the literature on antimicrobial use, antimicrobial resistance, and antimicrobial stewardship program guidelines. The data abstraction format was designed to help extract information on patient socio-demographic and clinical characteristics. A self-administered questionnaire to assess the experience and practice of the practitioners regarding antimicrobial use, diagnosis, and management of pneumonia was developed by reviewing the literature on antimicrobial use, antimicrobial resistance, antimicrobial stewardship programs, and pneumonia management guidelines also used. The questionnaire was filled out by treating physicians in the ward, and information including educational level, specific guidelines used, consideration of antimicrobial drug resistance during prescribing (send culture and sensitivity test), antimicrobial selection preference (broad-spectrum and narrow-spectrum), antimicrobial de-escalation practice (based on culture & sensitivity testing result), intravenous to per oral conversion practice, and practice of consultation with a clinical pharmacist for prescribing antimicrobials in pneumonia management was extracted. The diagnosis was made by treating physicians and was written in medical records as a final diagnosis based on WHO criteria (based on clinical and radiological findings). All treating physicians in charge of the pediatric wards were requested to cooperate in providing information about their respective patients diagnosed with pneumonia. In addition, the data collectors had the opportunity to review patients’ charts to identify patients with suspected/ confirmed pneumonia diagnoses. All the patients were followed starting from the initiation of antimicrobial therapy till a clinical outcome was achieved and antimicrobial therapy was discontinued. The clinical outcomes of the patients were observed immediately after completion of treatment and physicians who were following the patients were asked about their clinical status after antimicrobial therapy. Accordingly, final status was classified as good or poor outcomes which are judged by the treating physician.

Quality control

Training was given to data collectors on the data collection procedure and research objectives. The principal investigator was regularly checking the consistency and completeness of the data and random checks were also being performed. The questionnaire was translated into the local languages Afan Oromo and Amharic and then retranslated back to English to check for consistency of meaning. Finally, a pretest was conducted on 5% of patients whose data were not included in the final analysis before the actual data collection to check the consistency and understandability of the tools.

Data processing and analysis

Data were compiled, cleared, coded, and verified for completeness and accuracy before being included in Epi Data Manager version 4.6.2. Double-entry verification was performed and the data were analyzed using statistical software, SPSS version 25.0. Descriptive statistics (mean, frequency, and cross-tabulation) were used to describe the practice of antimicrobial use. The bivariate logistic regression was performed for clinical outcome and variables with P values less than 0.25 with outcomes (poor outcome or good outcome) were taken into account for further analysis. Then, the independent predictors were identified by multivariate logistic regression analysis. All statistical tests were detailed; a P value of less than or equal to 0.05 was considered statistically significant. The research was reported in line with the STROCSS criteria^[20].

Operational definitions

Antimicrobial/antibiotics: are agents that destroy or inhibit the growth of different microorganisms and are used for the treatment of infectious diseases, including pneumonia^[21].

Antimicrobial days: the number of days during which a patient received antimicrobials.

Antimicrobial use practice: antimicrobial prescribing pattern, practice of use of microbiology studies, and practice of intravenous to oral switching.

Recent antimicrobial use: A patient has a previous history of antimicrobial use within 90 days Appropriate therapy: use of at least one antimicrobial to which all isolates were susceptible in vitro observed in culture and sensitivity discs.

Clinical stability/cured: Suspected or diagnosed pneumonia patients who had complete resolution of all signs and symptoms of pneumonia or improvement of all abnormalities on the chest radiograph. This was judged by the treating physician and monitored by vital signs and laboratory studies.

Co-morbidity: diseases associated with another disease, pneumonia in this case.

Length of follow-up: The duration starting from the date of pneumonia diagnosis and the initiation of antimicrobial treatment until a final clinical outcome was achieved (in days).

Pneumonia: children presenting with chest indrawing or cough and difficulty in respiration having age-specific fast breathing (≥50 breaths/min in a child aged 2–11 months, ≥40 breaths/min in a child aged 1–5 years)^[22].

Severe pneumonia: children presenting with Cough or difficulty in breathing with: Oxygen saturation < 90% or central cyanosis, Severe respiratory distress (e.g. grunting, very severe chest indrawing), Signs of pneumonia with a general danger sign (inability to breastfeed or drink, lethargy or reduced level of consciousness, convulsions)^[22].

Confirmed pneumonia: a patient who is given a diagnosis of pneumonia with suggestive radiological findings.

Suspected pneumonia: a patient is given a diagnosis of pneumonia based only on clinical signs and symptoms without radiological findings.

Community-acquired pneumonia: Pneumonia that has been acquired in the community in a patient who has not been hospitalized within 14 days before the onset of symptoms.

Empirical treatment: Antibiotic administration initiated before or without identification of sensitive profile of bacterial pathogens.

Fast breathing: is defined as when the age-specific respiratory rates become ≥60 breaths/minute in neonates and infants aged 30 breaths/minute for >59 months

Parents: include both biological parents and study participants’ guardians.

Pediatrics: are neonates, infants, children, and adolescents who are aged <18 based on the treatment protocol according to the current study.

Reason for Admission: refers to the diagnosis given to the illness of the patient by the physician when the patient is admitted.

Mortality: refers to the patient dead of pneumonia despite treatment during their hospital stay.

Clinical outcomes: are defined in this study as outcomes in pneumonia treatment detected only until discharge. These include good or poor outcomes:

• Good outcome: Patients who have been treated for pneumonia and discharged from the hospital on the grounds of clinical improvement.

• Poor outcome: Patients who have been treated for pneumonia but passed away or complicated or an intensive care unit (ICU) admission.

• Complications: Suspected or diagnosed pneumonia patients who will not improve or admitted to the ICU after getting antimicrobial therapy, which is judged by the treating physician.

• Treating physicians: These are interns, residents, and senior physicians who are practicing at JMC.

Ethical statement

Ethical approval of the research was obtained from the Ethical Review Board (Ref. No. JUIRB016/21). A formal letter was written to the hospital (Ref. No.ፋ/ት/በት133/2014/) and, the hospital medical director and head of the department of the pediatric ward were informed about the purpose of the study to get permission and cooperation. Participants’ parents were informed about the purpose/nature of the study before the data collection and approved the invitation by written informed assent. The participants’ privacy and confidentiality were ensured during data collection.

Results

Socio-demographic data

A total of 161 children admitted to the pediatric ward of the hospital were diagnosed with pneumonia and began antimicrobial treatment for pneumonia during the study period. However, data for 15 patients were not complete as a result of discharges against medical advice and some lost follow-up (Fig. 1). According to this study, the mean age of participants was 40.95 (+47.61) months, with more than half (61.6%) of male children. most of the participants, 65 (44.5%), were in the age range less than 11 months. The socio-demographic data are presented in Table 1.

Figure 1.

Flow diagram showing enrollment of children with pneumonia.

Table 1.

Socio-demographic characteristics of children with pneumonia

Variable	Value	Frequency	%
Sex	Male	90	61.6
	Female	56	38.4
Age category (in months)	≤11	65	44.5
	12-35	22	15.1
	36-71	31	21.2
	≥72	28	19.2
Age, months	(Mean ± SD)	(40.95 ± 47.61)
Weight, kg	1-5.9	29	19.9
	6-10.9	59	40.4
	11-15.9	33	22.6
	≥16	25	17.1
Residence	Rural	129	88.4
	Urban	17	11.6
Type of admission	Direct admission	11	7.5
	Referred from gov’t institution	120	82.2
	Referred from private Institute	15	10.3

Clinical characteristics

More than three-quarters (76%) of patients had comorbidities alongside pneumonia. Malnutrition and cardiovascular disease were the most common comorbidities in the study population, accounting for 52 (35.6%) and 30 (20.5%), respectively. More than half of the participants 89 (61%) had recent antimicrobial use. More than three-quarters of patients 120(82.2%), were diagnosed with community-acquired pneumonia, and 121 (82.9%) of patients had severe pneumonia. The baseline clinical characteristics are presented in Table 2.

Table 2.

Baseline clinical characteristics of children with pneumonia

Variable	Value	Number (%)
Temperature	Febrile	74(50.7)
	Afebrile	72(49.3)
Respiratory rate	Fast	95(65.1)
	Not fast	51(34.9)
Oxygen saturation	>90%	74(50.7)
	<90%	72(49.3)
Vaccination status	Fully vaccinated	82(56.2)
	Incomplete vaccinated	33(22.6)
	Not vaccinated	31(21.2)
Nutritional status	Normal	86(58.9)
	Malnourished	60(41.1)
Type of pneumonia	Community-acquired	120(82.2)
	Hospital-acquired	11(7.5)
	Aspiration-pneumonia	15(10.3)
Severity of pneumonia	Severe pneumonia	121(82.9)
	Non-severe pneumonia	25(17.1)
Diagnostic method	CBC	82(56.2)
	Chest radiography	3(2)
	CBC + Chest radiography	48(32.9)
	CBC + Chest radiography + Chest CT	13(8.9)
Culture done	No	99(67.8)
	Yes	47(32.2)
Comorbid condition	Yes	111(76)
	No	35(24)
Comorbid condition	Malnutrition	52(35.6)
	Cardiovascular disorder	30(20.5)
	Respiratory and airway diseases	13(8.9)
	Gastrointestinal condition	14(9.5)
	Neuromuscular disorder	4(2.7)
	Hematologic condition	4(2.8)
Pre-admit antibiotic use	Yes	89(61)
	No	57(39)
Hospital stay duration	= <5	64(43.8)
	6-9 days	39(26.7)
	= >10 days	43(29.5)

CBC, complete blood count; CT, computed tomography.

Practice of microbiology studies

Among the 146 children diagnosed with pneumonia and hospitalized and given antimicrobials, microbiologic testing was performed only for 47 (32.19%) children. Of the cultured microbiology samples 16 (34.0%) were collected before start of empiric antimicrobial therapy. The most common samples taken were pleural fluid 24 (16.4%), followed by blood 11 (97.5%). The yield of microbiological culture was low in this study. Only 13 bacterial species (28.2%) were designated probable pathogens and were isolated out of 47 cultures (Fig. 2).

Figure 2.

Identified organisms during the management of children with pneumonia.

Fifteen treating physicians who did not order microbiologic studies wrote down reasons why they didn’t send them, and more than half said that they did not order them because “it is usual practice to treat childhood pneumonia based on clinical presentation” (Table 3).

Table 3.

Treating physicians’ reasons for not considering microbiological studies for the management of pneumonia

Reason given by physicians	N (%)
The usual practice is to treat children pneumonia is based only on clinical (signs and symptoms)	9(60.1)
Patient already started antimicrobial	2(13.3)
Culture result always negative and only radiologic information is sufficient to treat pneumonia in pediatrics	2(13.3)
It takes time waiting lab. result	2(13.3)

Antimicrobial prescribing pattern and susceptibility

All treatment approaches (100%) were found to be Empirical and the practice of intravenous to oral switching was limited (0%). The initiated antimicrobial was seen to continue in many cases during the desired duration of therapy. The mean and standard deviation of antimicrobial therapy duration was 11(± 6) days. About 30% of patients received antimicrobials therapy for more than 10 days. In the hospital, it can be observed that about 7 types of antimicrobials were used for the treatment of different types of pneumonia. Ceftriaxone was the most commonly prescribed empiric antimicrobial agent for pneumonia (Fig. 3). Regarding polypharmacy, the average number of antimicrobial prescriptions per patient was 1.68, these meant that almost all of the patients were exposed to at least 2 antimicrobial agents regardless of the type of pneumonia. It was assessed whether there was a change in the initial antimicrobial regimens or not during treatment; it was found that there were 64(43.8%) first-time changes, and the most common reason associated with the changes was a poor response to the initial antimicrobials (Fig. 4). Of a total of 47 microbiology tests done, only 13(28.2%) showed bacterial growth. Most of the detected organisms were found to be resistant to the commonly prescribed Beta-lactam antimicrobials for the treatment of pneumonia based on the susceptibility data collected from the 13 culture-positive results (Table 4). The appropriateness or adequacy of the initial antimicrobial regimens is described in light of the identified microorganisms (Table 5). The most common prescribing physicians were residents 12(63.2%), medical interns 5(26.3%), and specialists 2(10.5%). The guidelines/protocol used for prescribing antimicrobials for the management of children with pneumonia were treating physicians asked and the most common guideline to prescribe antimicrobials was reference eBooks for 13(68.4%), international guidelines (IDSA/BTS) 4(21.1%), and WHO guideline 2(10.5%).

Figure 3.

The most commonly used empiric antimicrobials (percent) for the management of children with pneumonia.

Figure 4.

Reasons for change in initial antimicrobial regimen for the management of children with pneumonia.

Table 4.

The list of identified pathogens and their susceptibility patterns to selected antimicrobial in the management of children with pneumonia

	Coagulase-negative streptococcus (n = 1)	Streptococcus agalatiae (n = 1)	K. pneumonia (n = 2)	Pseudomonas Spp. (n = 2)	S. aureus (n = 2)	LF-Enterobacteriaceae (n = 1)	Salmonella group A (N = 1)	Proteus mirabilis (n = 1)	Streptococcus pneumonia (n = 2)
Amoxicillin	-	-	R	R	-	-	R	R	R
Ampicillin	-	S	R	R	-	-	R	R	R
C. penicillin	-	S	-	-	-	-	-	-	-
Gentamycin	-	-	R	S	R	S	S	-	-
Azithromycin	-	-	-	-	R	-	-	-	-
Ceftriaxone	-	-	R	R	R	-	R	-	R
Ceftazidime	-	-	-	-	-	-	R	R	R
Vancomycin	R	-	-	-	S	-	-	-	S
CAF	S	-	S	S	S	-	S	S	S
Oxacillin	R	-	-	-	S	-	-	-	R
Erythromycin	-	R	-	-	R	-	-	-	R
Meropenem	-	-	S	-	S	S	-	-	-
TTC	-	R	-	R	R	S	-	-	-
Cefazoline			R		R				R
Cotrimoxazole	-	-	-	R	S	S	S	-	-
P/T	-	-	S	R	-	S	-	S	S
Cefotaxime	-	-	-		-	-	S	-	-
Imipenem	-	-	S	S	-	-	S	-	S

CAF, chloramphenicol; P/T, piperacillin/tazobactam; TTC, tetracycline.

Table 5.

Organisms isolated, susceptibility data, and initial antimicrobial therapy to demonstrate adequacy and appropriateness

Identified organism	Susceptible to	Initial antimicrobial regimen given	Appropriateness/adequacy (yes/no)
Streptococcus pneumonia	Vancomycin, imipenem	Ceftriaxone	No
	Piperacillin/tazobactam, chloramphenicol	Ampicillin and gentamycin	No
Klebsiella pneumonia	Meropenem, chloramphenicol	Ceftriaxone	No
	Piperacillin/tazobactam, imipenem,	Ceftriaxone	No
Pseudomonas spp.	Ceftazidime, cefuroxime, imipenem, P/T, CAF, gentamycin	Ceftriaxone and gentamycin	No/yes
	Meropenem, chloramphenicol, P/T	Ceftazidime and vancomycin	-
Staphylococcus aureus	Oxacillin, gentamycin, CAF, TTC, ciprofloxacin, clindamycin, cotrimoxazole, Erythromycin	Ampicillin and gentamycin	Yes
	Oxacillin, penicillin, gentamycin, ciprofloxacin	Crystalline penicillin	Yes
LF-Enterobacteriaceae	Cotrimoxazole, P/T, TTC, meropenem	Ceftriaxone	No
Coagulase-negative staphylococcus	Chloramphenicol	Ceftriaxone and vancomycin	-/No
Streptococcus agalatiae	Ampicillin, penicillin	Ampicillin and gentamycin	Yes/-
Salmonella group A	Cotrimoxazole, gentamycin, cefotaxime, imipenem, CAF	Ceftriaxone and gentamycin	No/yes
Proteus miribilis	P/T, CAF	Ceftriaxone	No

CAF, chloramphenicol; P/T, piperacillin/tazobactam; TTC, tetracycline.

Clinical outcome

All the patients were followed starting from the initiation of antimicrobial therapy till a clinical outcome was achieved and antimicrobial therapy was discontinued. The patient’s clinical outcome was observed immediately after the end of treatment, and the physicians who followed the patient were asked about their clinical status following antimicrobial treatment. The observed outcomes were good outcomes, with 95 cases (65.1%), and poor outcomes, with 51 cases (34.9%). The results, especially the clinical cured, have not been confirmed by an independent physician and may be biased toward respondents (Table 6).

Table 6.

The type of pneumonia and clinical outcomes of children with pneumonia

	Outcome status
Type of pneumonia	Good outcome	Poor outcome	Total
Community-acquired	82	36	118
Hospital-acquired	6	5	11
Aspiration-pneumonia	7	10	17
Total	95	51	146

Predictors of poor outcome in children with pneumonia

Initially, the clinical outcome was categorized into good outcomes (clinical cured and improved patients) and poor outcomes (death and complicated patients) because a logistic regression model requires the dependent variable to be expressed dichotomously. A Multivariate logistic regression analysis was used to evaluate independent predictors of poor outcomes. Accordingly, it was discovered that patients who had the following three characteristics were found to be more likely to have a poor outcome: recent antimicrobial use [P = 0.012, AOR 3.87(1.34–11.16)], patients with hospital stay above 10 days [P = 0.029, AOR 3.49(1.13–10.72)], patients whose antimicrobial regimen changed [P = 0.004, AOR 3.74(1.52–9.22.4)], patients with comorbidities [P = 0.012, AOR 5.00(1.42–17.59)] (Table 7).

Table 7.

Univariate and multivariate logistic regression analysis of factors associated with poor outcome among children with pneumonia who received antimicrobial therapy

Variables	Clinical outcome
	Good	Poor	COR (95% CI)	P value	AOR (95% CI)	P value
Sex
Male	57(63.3)	33(36.7)	1.222(0.603–2.476)	0.577
Female	38(67.9)	18(32.1)		1
Age category (in months)
≤11	40(61.5)	25(38.5)		1		1
12–35	13(59.1)	9(40.9)	1.108(0.413–2.969)	0.839	0.920(0.311–02.724)	0.880
36–71	21(67.7)	10(32.3)	0.762(0.309–1.881)	0.302	0.595(0.222–1.594)	0.555
≥72	24(85.7)	4(14.3)	0.267(0.083–0.860)	0.027	0.337(0.097–1.175)	0.088
Weight (Kg)
1–5.9	19(65.5)	10(35.5)	1.353(0.424–4.323)	0.610
6–10.9	35(59.3)	24(40.7)	1.763(0.638–4.870)	0.274
11–15.9	23(60.7)	10(30.3)	1.118(0.355–3.517)	0.849
≥16	18(72)	7(28)	1	1
Residence
Rural	85(65.9)	44(34.1)	0.739(0.263–2.076)	0.567
Urban	10(58.8)	7(41.2)	1	1
Admission type
Direct admission	6(54.50)	5(45.5)	1	1
Transferred from gov’t institution	79(65.8)	41(34.2)	0.623(0.179–2.164)	0.456
Transferred from private institution	10(66.7)	5(33.3)	0.600(0.121–2.973)	0.532
Temperature
Febrile	39(52.7)	35(47.3)	3.141(1.53–6.446)	0.003	3.878(0.413–30.250)	0.233
Afebrile	56(77.7)	16(22.3)		1		1
Respiratory Rate
Fast	55(57.9)	40(42.1)	2.645(1.210–5.779)	0.015	1.503(0.633–3.564)	0.356
Not fast	40(78.4)	11(21.6)		1		1
Oxygen saturation
<90%	52(70.3)	22(29.7)		1		1
>90%	43(59.7)	29(40.3)	1.594(0.803–3.165)	0.183	1.149(0.506–2.609)	0.741
Vaccination status
Full vaccinated	47(57.3)	35(42.7)		1
Partially vaccinated	22(66.7)	11(33.3)	1.429(0.540–3.783)	0.472
Not vaccinated	26(83.9)	5(16.1)	1.072(0.374–3.070)	0.897
Nutritional status
Normal	66(70.2)	28(29.8)		1
Malnourished	29(55.8)	23(44.2)	1.869(0.925–3.770)	0.081	3.247(0.785–13.245)	0.104
Type of pneumonia
Community acquired	82(69.5)	36(30.5)		1		1
Hospital-acquired	6(54.5)	5(45.5)	1.898(0.544–6.624)	0.315	1.745(0.050–11.054)	0.830
Aspiration-pneumonia	7(41.2)	10(58.8)	3.254(1.147–9.228)	0.027	2.051(1.045–94.525)	0.713
Severity of pneumonia
Severe pneumonia	80(66.1)	41(33.9)	1.769(0.318–1.861)	0.560
Non-severe pneumonia	15(60)	10(40)		1
Diagnostic method
CBC	54(65.9)	28(34.1)		1		1
Chest radiography	3(100)	0	.000(0.000)	0.999	0.000(.000)	0.999
CBC + Chest radiography	32(66.7)	16(33.3)	0.964(0.454–2.050)	0.925	1.087(0.431–2.742)	0.860
CBC + Chest radiography + Chest CT	6(46.2)	7(53.8)	2.250(0.690–7.338)	0.179	1.126(0.211–6.001)	0.890
Culture done
No	60(60.6)	39(39.4)		1		1
Yes	35(74.5)	12(25.5)	1.896(0.878–4.093)	0.103	1.264(0.529–3.017)	0.598
Co-morbidities
Yes	64(57.7)	47(27.8)	5.691(1.881–17.22)	0.002	5.004(1.423–17.594)	0.012
No	17(4.7)	21(55.3)		1		1
Pre-admit antibiotic use
Yes	44(49.4)	45(50.6)	8.693(3.387–22.31)	0.000	3.870(1.342–11.16)	0.012
No	51(89.5)	6(10.5)		1		1
Antimicrobial regimen changed
Yes	29(45.5)	35(54.7)	4.978(2.387–10.38)	0.000	3.746(1.522–9.222)	0 .004
No	66(80.5)	16(19.5)		1		1
Duration of hospital stay
≤5	51(79.9)	13(20.3)		1		1
6–9 days	27(69.2)	12(30.8)	2.7444(1.700–4.34)	0.002	1.585(1.728–12.233)	0.047
≥10 days	17(39.5)	26(60.5)	6.000(2.532–14.22)	0.000	3.492(1.138–10.72)	0 .029

AOR, adjusted odds ratio; CBC, complete blood count; CT, computed tomography; CI, confidence interval; COR, crude odds ratio.

Discussion

Appropriate antimicrobial utilization is critical for antimicrobial resistance containment and for good clinical and economic outcomes^[7]. However, inappropriate uses of antimicrobials have a devastating effect on patients as well as the general population. The present study is one of the few studies from Ethiopia, to assess the practice of antimicrobials utilization to treat pneumonia in hospitalized children. Childhood pneumonia was selected because it is one of the most common infectious diseases in the study area.

Microbiological studies were performed on only 47(32.19%) patients. This result was lower than reports from Italy (98%) and Indonesia (43.5%) which had microbiological cultures obtained during hospitalization, respectively^[23,24]. The huge gap seen in the current study setting might be as a result of weak laboratory infrastructure and the laboratory not being equipped with rapid diagnostic kits.

The findings showed that most children were treated without evidence of microbiological data, and this is in agreement with the attending physician’s opinions on the use of microbiological studies; that more than three-quarters 15(78.9%) of the physicians hesitate to send culture and sensitivity tests. Recommendations from published guidelines for managing children with pneumonia recommend starting treatment with broad-spectrum antibiotics, which should be accompanied by proper de-escalation based on culture results. This may be due to the poor attention of physicians to the microbiological data. As most physicians have stated, it is sufficient to treat children with pneumonia based on clinical information. Also, loss of trust in achieving a positive culture result.

In the current study, all treatment approaches were found to be empirically initiated. Empiric antimicrobial therapy is generally categorized as appropriate (adequate) or inappropriate (inadequate) based on microbiological culture and susceptibility results. An empirical treatment regimen is considered appropriate; if the identified microorganism is sensitive to at least one of the prescribed antimicrobial agents^[2]. Studies have shown that appropriate empiric antimicrobial therapy is associated with decreased mortality in patients with many different types of infection. However, the lack of cultural and susceptibility data has devastating consequences for both patient and economic perspectives^[8]. However, in this study, only 5 out of 13 (38.4%) patients who received empiric antimicrobial regimens were seen to have culture results with appropriate coverage. This finding was higher than reports from Japan and Indonesia, which were (17%) and (15.7%)^[25,26]. This might be due to the higher prevalence of pneumonia or the difficulty of differentiating between bacterial and viral infections in our study.

In this study, out of the 208 prescriptions, 129(88.4%) were third-generation cephalosporin. In addition, empirically, about 6.4% of patients received vancomycin. The result also showed that most treating physicians preferred broad-spectrum antimicrobials over narrow-spectrum on treatment initiation 14(73.6%). This finding was in contrast with a report from Nepal in which ampicillin was the most common (70%) and broad-spectrum antibiotics such as linezolid, vancomycin, and meropenem were used in less than 1% of patients^[24]. However, the result was in line with reports from Nekemte Referral Hospital (60%) that use third-generation cephalosporins as the first-line treatment of children with pneumonia, respectively^[27]. It was seen that many factors influenced antimicrobial selection. On the one hand, this was due to the lack of a standardized hospital-specific treatment guideline, which allowed prescribing physicians to use whatever antimicrobial agent they wanted. On the other hand, it is due to the frequent stockouts of most of the antimicrobials in the hospital, and this led the treating physician to prescribe the available antimicrobial agents.

In the current study, many clinically stable and unstable patients (complicated cases) treated with broad-spectrum antimicrobial combinations completed a full-course initiated treatment. Antimicrobial de-escalation is a strategy for the rational use of antimicrobials to balance empirical use and reduce the development of resistance. To that effect, antimicrobial use practices in the current study are increasing at an alarming rate and require coordinated immediate intervention. Therefore, unless continuous delivery of antimicrobials is an issue; proper strategies and balances should be maintained in the selection of empirical antimicrobials to minimize their overuse and/or unnecessary use.

The present study evaluated patient eligibility for IV to PO conversion using predefined criteria adopted from the SHEA guidelines^[17]. Therefore, it was found that 32 patients were eligible for conversion. However, only three patients were transferred and the rest continued to receive IV antimicrobials. This may be due to treating physicians’ limited attitude toward the benefits of switching from IV to PO. This was reflected in their response to the question about the very limited practice of converting IV to PO. About 37% of treating physicians reported switching at discharge and IV administration antimicrobials in inpatients is a common practice (26.3%). A similar was demonstrated in a study in TASH, in which 82(41.8%) of the treating physicians said that administration of IV antimicrobials for hospitalized patients is a usual practice in this hospital and conversion is only made at the time of discharge, and about 10(5.1%) of the treating physicians also said IV is more effective than PO antimicrobials^[28].

The current study also evaluated the guidelines used by treating physicians to prescribe antimicrobials for managing children with pneumonia. Most physicians’ responses showed that they were using international guidelines and eBooks to treat children with pneumonia. No use of local guidelines was observed in this study. The reason behind this might be that the local standard treatment guidelines were not prepared taking into consideration tertiary care hospitals. It is also known that the guidelines were developed without actual local antibiogram data. This suggests that clinicians relied primarily on eBooks and international guidelines. Different guidelines and books have been published at different times, and their recommendations are often not the same^[7]. This concept is also demonstrated in the guide to good prescribing, which was prepared by WHO^[7]. This has diversified the use of antimicrobials in treating patients diagnosed with similar pneumonia. The referenced eBooks and international guidelines are primarily based on national antimicrobial resistance patterns and are mostly for educational purposes. In addition, these eBooks and guidelines recommend having facility-specific guidelines based on the drug susceptibility of your facility^[2]. Several studies and published guidelines describe how to use of different strategies has been found helpful for the more appropriate and cost-effective use of antimicrobials in hospitals. Such strategies include education, guidelines, and clinical pathways, antimicrobial order forms, de-escalation of therapy, intravenous-to-oral switch therapy, and dose optimization, common infection-based interventions such as community-acquired pneumonia^[7,23].

In the current study, good or poor clinical outcome was the primary endpoint. Therefore, the observed finding is in-hospital mortality 11(7.5 %). In addition, despite the fact that two or more combinations of antimicrobials were given for more than 10 days 40(27.3%), patients were complicated (no improvement). This finding is higher than both results from the Nekemte Referral Hospital, Ethiopia, in which poor treatment was observed in 30.6% of pediatric patients^[27] and from TASH, in which 7.7% of patients had poor treatment outcomes^[29]. The findings of the present study suggest that the quality of care of patients with infectious diseases was minimal in the study area as compared to the previous study settings. The absence of a sufficient number of infectious diseases specialists, hospital-specific antimicrobial treatment protocols, continuous supply of antimicrobials, and better microbiological laboratory services indicates poor quality care. This could also be due to a high proportion of comorbidities, such as cardiovascular disease and malnutrition, which can lead to a poor prognosis.

Identifying predictors for poor outcome was a critical strategy during infectious disease management to provide specialized care based on the number of risk factors associated with the patients. Hence, this study tried to assess the possible predictors of poor outcomes in patients with pneumonia. Accordingly, around three predictors were identified, and these risk factors are in line with different studies done across the globe.

In this study, a history of recent antimicrobial use (P = 0.012) was one of the predictors of poor clinical outcomes. Patients who were exposed to antimicrobials before admission had 3.8 times worse outcomes than those who were not. This is because patients coming to the hospital are referred from different lower health facilities and other hospitals according to the country’s referral structure or system. This can lead to colonization and subsequent infection by resistant bacteria^[12]. As a result, clinicians caring for these patients should take their recent antimicrobial experience into account and balance empiric antimicrobial regimens for suspected infection until microbiological studies are completed.

In this study, initial antimicrobial regimen changes were associated with an increased incidence of poor outcomes compared to those who completed their entire course of therapy with the initial antimicrobial regimen. This may also be due to the patient’s antimicrobial changed being affected by resistant organisms to prescribed antimicrobials. Therefore, these factors that patients who had antimicrobial changes could have contributed to the poor outcomes as the difference was significant on multivariate logistic regression (P = 0.004).

In the current study, another factor found to have a statistically significant association with poor outcomes was hospital stay duration. Hospital stay duration above ten (P = 0.029) was a negative predictor of clinical outcome in the management of pneumonia. Patients with increased hospital days above 10 days were about 3.4 times more likely to have a poor treatment outcome compared with patients who stayed ≤5 days (AOR = 3.49, 95% CI:2.2–143.1). Similar results were found in a TASH study, where patients who stayed 8 days were approximately 14.3 times more likely to have a poor treatment outcome than patients who stayed 3 days (AOR = 14.3, 95% CI: 1.35–151.1)^[30].

Strengths and limitations of study

The strength of this study is its prospective study design which allowed better control of data quality and consecutive patient enrollment. However, our study is not without limitations and it should be interpreted with caution. This study was conducted at a single tertiary hospital, and the follow-up was also limited to the patient’s hospital stay. The findings; particularly stability/clinical improvement were not confirmed by an independent physician and might have interviewee bias

Conclusion

Empirical initiated antimicrobial was seen completed without sufficient evidence of indication and microbiological and radiographic findings. More than one-fourth of the patients treated for pneumonia experienced poor outcomes, implicating the need for more attention during treatment. The observations indicate a need for the hospital to establish an antimicrobial stewardship program focused on essential strategies in pediatrics. This program should include preauthorization, prospective audits with feedback, and pharmacist-led interventions that are both feasible and well-accepted in the inpatient setting, and it should be implemented as soon as possible.

Ethical approval

Ethical approval of the research was obtained from the Jimma University Ethical Review Board (Ref. No. JUIRB016/21).

Consent

Written informed consent from the child’s parent or guardian was obtained for publication and any accompanying images after providing information regarding the purpose/nature of the study.

Sources of funding

None.

Author contributions

A.W.G. and B.T.D.: conception and design, reviewing and editing of manuscript; A.W.G.: material preparation, data collection, formal analysis, the first draft of the manuscript. Both authors read and approved the final manuscript.

Conflicts of interest disclosure

None.

Research registration unique identifying number (UIN)

Unique identifier number: NCT06830915 (https://clinicaltrials.gov/study/NCT06830915).

Guarantor

Abduba Wariyo Guyo.

Provenance and peer review

Not commissioned, externally peer-reviewed.

Data availability statement

Data can be made available upon reasonable request from author.