Appropriateness of Antibiotic use for the Management of Acute Diarrhea Among Under-Five Children Treated at Primary Care Centers in Northwest Ethiopia: A Cross-Sectional Study

Background Acute diarrhea management is solely aimed at uid replacement and nutritional support while antibiotics have a very limited role. Antibiotic treatment is recommended only for bloody diarrhea (dysentery), cholera and invasive bacterial diarrhea. This study is launched to assess the appropriateness of antibiotic use for the management of acute diarrhea among under-ve children in Gondar town primary care centers. Methods Institutional based cross-sectional study was conducted in three primary care centers located in Gondar town, Northwest Ethiopia. Children aged from 2–59 months who visited the three primary care centers from September 12, 2015 to September 10, 2016 and received treatment for acute diarrhea were included in the study. We selected 176 cases from Azezo Health Center, 166 from Poly Health Center and 80 from Woleka Health Center. Cases were drawn using systematic random sampling technique. The ndings of the study were summarized using tables and gures; binary logistic analysis was used to identify association between the independent and outcome variables at 95% condence level where p < 0.05 was considered as statistically signicant.

zinc as the mainstay of treatment of acute diarrhea in children. Antibiotic treatment is recommended only for bloody diarrhea (dysentery), cholera and invasive bacterial diarrhea [6,7].
While it is important to use antibiotics appropriately ensure best use of resources, effective therapy, reducing the rising rate of antibiotics resistance pathogens, increasing quality and e ciency of healthcare and minimization of potential side effects, it is common to encounter inappropriate antibiotic use in diarrheal patients, especially in developing countries [8][9][10]. Inappropriate antibiotic use takes many different forms like poly-pharmacy, over-use/misuse of antibiotics and injections, failure to prescribe in accordance with guidelines and inappropriate self-medication [11,12]. Inappropriate use of antibiotics in the treatment of acute diarrhea results in an increased risk of adverse effects, higher rates of antimicrobial resistance by enteric pathogens, inappropriate patient demand, and increase healthcare cost [8][9][10][11][12]. In addition to these, all oral antibiotics are capable of causing, or worsening diarrhea due to their effect on gut micro ora [2,9].
Lack of national policies and strategies related to antibiotic use is associated with increased antibiotic use in acute diarrhea [13]. Factors that in uence appropriateness of antibiotics use in acute acute diarrhea include diagnostic uncertainty, giving something more than just ORS to satisfy the patient, insu cient number of quali ed and competent staff, caregiver preference, source of updating knowledge, experience of the prescriber, age of the child, lack of speci c medicines' information (e.g., clinical guidelines and drug bulletins),lack of continuing medical education and supervision, peer pressure (where senior staff may set inappropriate prescribing practices that are followed by junior staff), lack of diagnostic support services (e.g., laboratory services), poor infrastructure [14][15][16].
Ethiopia is one of developing countries where the healthcare system at a low level of development, factors that predispose to inappropriate antibiotic use are common futures of its most primary healthcare institutions [17]. Despite this fact, there has been limited data on the pattern of antibiotic use in the management of acute diarrhea in pediatric patients treated at primary healthcare facilities in the country.
This study is launched to assess the appropriateness of antibiotic use for the management of acute diarrhea among under-ve children in Gondar town health centers.

Methods
Institutional based cross-sectional study was conducted in three primary care centers (Azezo Health Center, Poly Health Center and Weleka Health Center) located in Gondar town, Northwest Ethiopia.
Children aged from 2-59 months who visited the three primary care centers from September 12, 2015 to September 10, 2016 and received treatment for acute diarrhea were included in the study. Children whose diarrheal illness lasted more than 14 days, children who were referred to hospitals and children who were immunocompromised or had other comorbidities were excluded from the study.
The sample size was determined using single population proportion formula by assuming 95% con dence level, 5% margin of error, estimated proportion of appropriate use of antibiotics (0.5) and 10% no-retrieval rate. Accordingly, the calculated sample size was 422, which is then proportionally distributed over the three primary care centers based on the number of under-ve children acute diarrheal cases in each center. Thus, we selected 176 cases from Azezo Health Center, 166 from Poly Health Center and 80 from Woleka Health Center. Cases were drawn using systematic random sampling technique.
Data was collected by review of patient medical cards and entered into IBM SPSS Statistics 24® software for analysis. The ndings of the study were summarized using tables and gures. Furthermore, binary logistic analysis was used to identify association between the independent and outcome variables at 95% con dence level where p < 0.05 was considered as statistically signi cant.
Ethical clearance was obtained from School of Pharmacy, University of Gondar, Research and Ethics Review Committee. The study followed the principles outlined in the Declaration of Helsinki. The requirement of participants' written informed consent was waived by the ethical review committee of the University of Gondar because this was a retrospective study through document review. Anonymity of study participants was assured by neglecting the personal identi ers such as patients' name and address. The information from patients' cards was kept con dentially, only data collectors and investigator have access to them.

Socio-demographic and clinical characteristics
A total of 422 children presenting with acute diarrhea between the age of 2 and 59 months were enrolled in this study from HCs of Gondar town. However, data of 32 children were excluded from the study due to incompleteness, giving the card retrieval rate of 92.4%.
The mean age of the study participants was 19.1 ± 12.8 months. The majority (60.8%) were males while 80.9% were vaccinated. Height and body temperature records were found only on few patients' medical les, thus these variables were excluded from our analyses. Less than a fth (19.7%) of the study subjects visited health facilities within a day of symptom onset. The majority (85.9%) of the children presented with watery diarrhea (Table 1). Nurses (holding technical level diploma). The practitioner examining a patient took full responsibility of selecting treatment regimens. Antibiotics were prescribed to 66.2% of the children. Trimethoprimsulfamethoxazole (TMP-SMZ) was the predominant antibiotic prescribed (Fig. 1).

Appropriateness of antibiotic use
We determined appropriateness of antibiotic use in acute diarrheal patients using the World Health Organization (WHO) manual for treatment of diarrhea. According to the manual, antimicrobial therapy should only be given for the following cases: 1) in cases of bloody diarrhea; 2) for suspected cholera cases; 3) for proven cases of giardiasis. The guideline provides that cholera should be suspected if there is diarrhea with severe dehydration in children over ve years of age or in cholera endemic areas. It also states that acute diarrhea should not be treated for giardiasis. Since our study is on acute diarrheal cases of under ve children in an area with no cholera incidence in preceding three years, only the rst case for antimicrobial therapy applies for this study. Thus, antibiotic use is designated appropriate if antibiotic was administered to a bloody diarrhea case or if no antibiotic was administered for watery diarrhea case; and considered inappropriate otherwise. Antibiotics prescribed for other coinfections were also taken into consideration.
Accordingly, appropriate antibiotic use was recorded in less than half (47.2%) of the study subjects.
Almost all (98.1%) of children subjected to inappropriate antibiotic use were those with watery diarrhea treated with antibiotics. Out of 253 children who received antidiarrheal antibiotics 202 (79.8%) had acute watery diarrhea which should not be treated with antibiotics (Fig. 2).
Factors associated with appropriateness of antibiotic use

Discussion
The main aim in the treatment of acute diarrhea is treating and prevention of dehydration. Proper nutritional support, uid management and early medical intervention are essential for successful management of acute diarrhea [18]. More than half of patients in our study (53.8%) visited healthcare facilities 72 plus hours after the onset of symptoms. This may be related to lack of health-seeking behavior and low awareness about diarrhea and its risk in the community.
Characteristics of diarrhea, especially low proportion of patients with bloody diarrhea (14.1%) suggests that the majority of cases of acute diarrhea were mild and self-limiting. Despite this, at least one antibiotic was prescribed to 66.2% of children. Widely ranging ndings were reported on similar studies conducted elsewhere; an Indian study reporting antibiotic use lower than our study (43%) [9] while higher antibiotic use (84.9%) was reported on a study from Tanzania [19]. The differences in the prevalence of antibiotic use across different studies may be due to many factors such as difference in the epidemiology of diarrhea between the areas, difference in resources or treatment approaches.
The overall proportion of appropriate use of antibiotics for acute diarrheal treatment in our study was 47.2 %. This proportion is relatively similar to that reported in a hospital based cross-sectional study in Southern Thailand (44%), but less than that of a study conducted in Gambia, West Africa (63.4%) [1,8].
Out of children identi ed with inappropriate antibiotic use 98.1% had watery diarrhea and of all patients who received antidiarrheal antibiotics 79.8% had watery diarrhea which should not be treated with antibiotics. This indicates prescribers' widespread deviation from WHO recommendations particularly, high tendency to prescribe antibiotics for acute watery diarrhea in under-ve children. Prescribers' lack of su cient knowledge, diagnostic uncertainty and easy accessibility of antibiotics may be contributing factors for this wrongful practice.
In our study, TMP-SMZ was the most inappropriately prescribed antibiotic which is consistent with Southern Thailand [1]. However, this result is not supported by other studies conducted in Tanzania and India where cephalosporins, metronidazole and cipro oxacin were frequently inappropriately prescribed antibiotics [9,19,20]. The possible reason may be the fact that TMP-SMZ was more available than other drugs (cephalosporins, cipro oxacin) in our setup.
There was no association between appropriateness of antibiotic use and patients' age, weight, chief compliant as well as health center where they got treatment. In contrast, diagnosis of bloody diarrhea, treatment regimen with no antibiotics and having clinical nurses as prescriber were found to be predictors of appropriate antibiotic use. Similarly, higher level professionals were associated with appropriate antibiotic use as reported by studies conducted in India and Tanzania [9,19].
Recommendations on use of antibiotics in the management of acute watery diarrhea should be followed strictly, since indiscriminate use of antibiotics results in development of resistance. In addition to this, giving unnecessary antibiotics means addition of unnecessary cost to the community. Especially in our setup where available antibiotics are limited, we do not have second line antibiotics for diseases caused by resistant strains of microorganisms. Thus, appropriate use of antibiotics is crucial to maintain effective treatment and prevention of infectious diseases in a community.

Conclusion
The ndings of the study con rm the prevalence of widespread inappropriate antibiotic use on under-ve children presenting with acute diarrhea. The study shows how parameters such as stool characteristic, medication regimens and prescribers' pro le affect the appropriateness of antibiotic use. The ndings can be used by stakeholders as input for promoting appropriate antibiotic use in the healthcare system as well as to deter antimicrobial resistance.

List Of Abbreviations
AOR adjusted odds ratio CI: con dence interval COR: crude odds ratio Ethical clearance was obtained from School of Pharmacy, University of Gondar, Research and Ethics Review Committee. The study followed the principles outlined in the Declaration of Helsinki. The requirement of participants' written informed consent was waived by the ethical review committee of the University of Gondar because this was a retrospective study through document review. Anonymity of study participants was assured by neglecting the personal identi ers such as patients' name and address. The information from patients' cards was kept con dentially, only data collectors and investigator have access to them.

Consent for publication
Not applicable Availability of data and materials The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. Figure 1 Frequency of antibiotic prescription and types of antibiotics prescribed for under-ve children with acute diarrhea