Antibiotic Utilization Trends in Two State Hospitals of Mongolia from 2013 to 2017

1 School of Pharmacy, Mongolian National University of Medical Sciences, S.Zorig street, Sukhbaatar district, 14210 Ulaanbaatar, Mongolia Norwegian Institute of Public Health and School of Pharmacy, University of Oslo, Postboks 1068 Blindern 0316 OSLO, Norway School of Pharmacy, Curtin University of Technology, GPO Box U1987, Perth 6845, Australia 4 Department of Pharmacy, State ird Hospital of Mongolia, Ard-Ayush avenue, 6th khoroo, Bayangol district, 16000 Ulaanbaatar, Mongolia Department of Pharmacy, State Second Hospital of Mongolia, S.Zorig street, Sukhbaatar district, 210648 Ulaanbaatar, Mongolia 6 School of Public Health, Mongolian National University of Medical Sciences, S.Zorig street, Sukhbaatar district, 14210 Ulaanbaatar, Mongolia


Introduction
Antimicrobial resistance (AMR) is threatening current health care practice. Antibiotic use and especially over-use contributes to increased AMR in society, and WHO recommends every country set up surveillance systems for AMR. Drug utilization research is de ned as "an eclectic collection of descriptive and analytical methods for the quanti cation, understanding and the evaluation of the processes of prescribing, dispensing and consumption of medicines, and for the testing of interventions to enhance the quality of these processes " [1]. Drug utilization research is regarded as essential for rational pharmacotherapy; it investigates what is being used by evaluating the appropriateness of prescribing and use of medicines [2]. Drug utilization studies of antimicrobial use may create awareness and understanding of the volume of use and potential consequences for AMR from the use of antibiotics in the human and animal health sectors. e use of antibiotics varies substantially among countries [3][4][5][6]. Due to different types of health care organizations, variations occur among hospitals within a country, indicating the need for common methodology to facilitate interhospital and longitudinal comparisons. Antibiotic utilization has been monitored frequently at outpatient and inpatient levels in many countries, particularly in high-income countries; however limited data are available from less developed nations. Drug use data have until now not been collected in Mongolia and there is a need to perform such studies, to set a baseline, and when appropriate develop effective interventions towards rational drug prescribing. Recently, the WHO reported antibiotic surveillance of consumption data worldwide, with Mongolia one of the countries with the highest antibiotic consumption (64.4 DDD per 1000 inhabitants per day). However, these data were collected from the records of imported and locally manufactured antibiotics and not at the patient level. ese records of imported and locally produced antibiotics were retrieved from an online database, managed by the Center for Health Development, Ministry of Health, Mongolia [7]. Drug utilization studies should be conducted at the end-user level [1]. However, no data regarding medicine utilization statistics are available for any of the hospitals in Mongolia.
is study aimed to evaluate inpatient antibiotic use in both the State Second Hospital and State ird Hospital in Mongolia, using the WHO developed and standardized ATC/ DDD methodology.

Study Design.
A retrospective longitudinal analysis of antibiotic consumption data from two selected state hospitals of Mongolia was completed for the period from 2013 to 2017.

2.2.
Setting. In Mongolia, there are 13 central hospitals and specialized centers which provide tertiary level care, 5 Regional Diagnostic Treatment Centers, 16 aimag (province) general hospitals, 12 district general hospitals and health centers, 6 provincial general hospitals provide secondary level care, 39 inter-soum (second level subdivision) hospitals, 273 soum general hospitals, 218 family group practices that provide primary level care. In addition, there are 240 private inpatient hospitals and 1226 outpatient hospitals [8]. e study has included data from two public central hospitals that provide tertiary health care for approximately one fi h of the Mongolian population. Antibiotic utilization was monitored longitudinally for five years (2013-2017) at the Second and ird State Central Hospitals of Mongolia. ese two state central hospitals provided health care and comprised of 23 wards covering most branches of medicine with a total of 28,617 inpatients (19.1%) of the 149,648 total number of inpatients admitted to 13 central hospitals and specialized centers in the country (2017). e annual number of in-hospital admissions were 9,126 in the Second Hospital and 19,491 in the ird Hospital in 2017 [9]. Data regarding the bed number and bed days were relatively stable for both hospitals over the selected years (Table 1) [10,11]. Hospital wards and corresponding bed numbers are provided in Table 2.

Data Collection.
In Mongolia, the pharmaceutical procurement sector is 100% privatized. Medicines are distributed through organizations such as drug wholesalers and retail drug outlets (community pharmacies and revolving drug funds). Medicines are usually supplied to state central hospitals through purchasing by tender from private pharmaceutical companies, wholesalers, as well as distributors.
In this study, reports submitted by the Pharmacy Departments of each hospital to the Statistics Office and hospital purchases from pharmaceutical wholesalers were used to capture the total hospital antibiotic utilization.
As specified in the Law of Mongolia on Procurement of Goods, Works and Services with State and Local Funds [12], medicines and medical devices are procured through an open-tender [12]. e evaluation is based on criteria set by the Drug erapeutic Coordination Committee of each hospital and the tender with the lowest price is evaluated as "the best" and deemed high value for comparison and estimation to procuring the medical equipment, tools and drugs. A successful wholesaler or tenderer will supply and deliver goods, including medicines to the Hospital Pharmacy Department. Hence, all wards of the two state hospitals receive medicines from pharmacy departments and at the end of each month the Pharmacy Department compiles a report on medicines given to patients. ese public pharmacies are called Pharmacy Departments and they are located in the hospitals and they provide medicines only to inpatients during their hospitalization period in Mongolia, [13] hence only data regarding medicine utilization of patients admitted to hospitals were available from the public hospitals. Medicines given while in the hospital are counted as inpatient medication, if patients are to continue at home they will get a prescription and this will be regarded as used as an outpatient. Outpatients seek medication outside a hospital, o en from retail pharmacies located nearby or within the hospital. However, in Mongolia, no data are available regarding the utilization of antibiotics for outpatients. erefore, the study included records of inpatients who were dispensed antibiotics. Antibiotic data were identi ed to be all antibacterials for systemic use, i.e., group J01 of the Anatomical erapeutic and Chemical (ATC) classi cation and for classes and subclasses of this group [14]. Data collected included unique chemical substance name, generic name, unit strength, pack size, quantity of packs, route of administration, and manufacturer. DDDs were calculated for each product according to the ATC/DDD methodology, ATC/DDD Index 2014 [15]. DDD per 1000 inhabitants per year is only recommended for international comparison.

Data Management and Analysis.
All data were recorded in and analyzed by Microso Excel 2016. Data validation was completed by means of ESAC checklist [16] for possible biases, including population and medicine coverage.
As the WHO recommends, if medicine utilization is assessed among inpatients, it should be expressed as number of de ned daily doses used either per admission (DDD/ admissions, bed days (DDD/100 bed days) and/or per 1000 inhabitants (DID/1000) [2]. In this study, the population prescribed antibiotics were admitted patients staying at least overnight (24 hours). e de nition of a bed day may di er between hospitals or countries. In this study the common de nition for the two hospitals selected was: a bed day is a day during which a person is con ned to a bed for a day surgery procedure and/or in which the patient stays overnight in a hospital. Moreover, DDD/admission is another indicator that can be used to show antibiotic use according to hospital activity. e day of discharge was also included in the analysis. Both indicators are reported to be important to compare and benchmark in and between hospitals [17,18].
Census data regarding bed days and hospital admissions were obtained from the Statistical O ce of the two state hospitals. e statistical analysis was performed using Student's t-test for parametric data. A value of ≤0.05 was considered to be statistically signi cant.

Total Antibiotic Consumption.
Total antibiotic consumption di ered between the hospitals. e consumption in the State Second Hospital was stable over the ve years, while the annual consumption in the State ird Hospital varied considerately between years. Overall, the total antibiotic consumption has uctuated for the State ird Hospital and decreased in the State Second Hospital, from 2013 till 2017 ( Figure 1).
Data regarding DDDs per 100 bed days were calculated for both hospitals in the selected years. First and third generation cephalosporins were the highest consumed antibiotics in the State Second Hospital in 2013. In the State Second Hospital, the 1 st generation cephalosporins were the highest with 16.6 DDD/100 bed days in 2013 and it decreased to 5.0 in 2017. Moreover, the use of 3 rd generation cephalosporins has increased by 13% (DDD/100 bed days 11.1 versus 13.0) in the State Second Hospital. e consumption of metronidazole (DDD/100 bed days 0.25 versus 0.5) and aminoglycosides (DDD/100 bed days 0.9 versus 1.8) have almost doubled in the study period.
On the other hand, signi cantly higher DDD/100 bed days for systemic antibiotic classes were reported from the State ird Hospital. DDDs/100 bed days for systemic (J01) antibiotics consumed at the State ird Hospital showed to be decreasing in the study period (2013: 205.3 DDD/100 bed days, 2017: 138.7 DDD/100 bed days). However, a contrary result occurred for macrolides, 3 rd generation cephalosporins, aminoglycosides, and combination of penicillins were observed. Increase in consumption of 3 rd generation cephalosporins was the highest (DDD/100 bed days 4.2 versus 15.7 DDD/100 bed days) for macrolides were 13.7 in 2013 and they have increased rapidly to 24.7 in 2017 (Table 3).

Class-Speci c Trends.
e average consumption of systemic antibiotics was approximately twice in the State  Overall, the use of 1 st generation cephalosporins was similar in both hospitals (50.7 vs 49.9 DDD/100 bed days); however, comparatively more 3 rd generation cephalosporins were used for inpatient care in the State Second Hospital  consists of four centers, 11 wards providing hospital care specializing in Cardiology, Neurology, Diabetes, and Consultancy services. In addition, the ird Hospital provides medical care through 25 departments. e State Second Hospital has about 215 beds, 9 wards, in addition to providing ambulatory and inpatient hospital care for all patients in Mongolia, the hospital provides emergency and planned surgeries of residents in Sukhbaatar district, Ulaanbaatar city. Both hospitals should conduct microbiology testing upon administration; however, due to lack of nancial resources, routine testing is not performed.
A previous study conducted in two tertiary hospitals in Mongolia found out that 19% of infections microbiological diagnostics were done and antibiotics were given without sensitivity testing in 92% of cases [24]. e level of antibiotic use was 84.0 DDD/100 bed days [DDD/100 bed days of 29.3 (State Second Hospital) and 138.7 (State ird Hospital)] and it is still exceeding that from other countries, including Croatia (32.9 DDD/100 bed days) [25], Norway (47.5 DDD/100 bed days) [26], and Slovenia (50.4 DDD/100 bed days) [27]. Oral antibiotics ranged from 60.3% to 55.8% and parenteral antibiotics were almost equal throughout the years, ranging from 39.7% to 44.2% in both hospitals combined. However, the individual analysis in this study indicated the State ird Hospital used more oral antibiotics (66.4%) than parenteral medications (33.6%). As for the State Second Hospital, parenteral forms were more utilized starting from 71% in 2013, increasing to 82% in 2017. Di erences between consumption of parenteral versus oral forms of antibiotics can be related with hospital specialties, as well as di erent levels of knowledge and attitude among prescribers practicing in each hospital [28]. Other ndings reported the proportion of parenteral DDD being higher (around 70%) than Tajikistan (31.1%) [29], but lower than China (98%) [30].
In general, the ndings indicated a decreasing but uctuating pattern of antibiotic consumption in both hospitals for all medicines. Of the seven most frequent used nitroxoline is the one that is much of interest compared to use elsewhere in the world. e main indication is for urinary tract infections cephalosporins and the macrolides represented around 70% of the antibiotic consumption in a ve years period (2013-2017).
In the State ird Hospital, above mentioned three antibiotic groups represented only 31%, while nitroxoline, uoroquinolones, and penicillins with extended spectrum represented 52% of total DDD/admissions in the same period. e seven most frequently used antibiotics represented around 75% of all DDDs in both hospitals, in the whole period 2013-2017; these were; amoxicillin (J01CA04), cefazolin, cefotaxime, ce riaxone, clarithromycin, cipro oxacin, and nitroxoline. However, this was not true when looking at the separate years, in 2015 and 2016, these substances represented approximately 50% (Table 4).

Discussion
To the best our knowledge, this is the rst study to report the antibiotic consumption in two selected state hospitals of Mongolia, using the WHO methodology.
e changes of antibiotic consumption over a ve-year period were quanti ed in order to inform decision makers, medical professionals, and the community.
Antibiotic consumption is the major driver leading to antibiotic resistance. e study has indicated that the consumption has decreased in both hospitals in the study period. Considering the Government's e ort to promote appropriate use of antibiotics and reduce antibiotic resistance, antibiotic use was decreased by almost 30% over the period of study [State Second Hospital: 47.8 DDD/ 100 bed days (2013) vs 29.3 DDD/100 bed days (2013); State ird Hospital: 205.3 DDD/ 100 bed days (2013) vs 138.2 DDD/100 bed days (2017). In addition, this result may be accounted for in part by the greater emphasis on development of recently produced documents, implemented activities [19,20] and State issued regulations [21][22][23]. Di erences in the consumption levels of antimicrobial agents in the two hospitals were observed in this study and it can be related with various factors, including hospital specialties, patient population, and adherence to clinical guidelines and protocols. e State ird Hospital Hospital), which is comparable with those reported from Iran (5.8 DDD/100 bed days) [35] and Israel (5.9 DDD/100 bed days) [36]. Moreover, the use of quinolones (6.8 DDD/100 bed days; State Second Hospital: 3.66 DDD/100 bed days, State ird Hospital: 10.12 DDD/100 bed days) was lower when compared with results from Iran (9.3 DDD/100 bed days) [35] and Israel (10.8 DDD/100 bed days) [36].
Carbapenems (0.026 DDD/admission in State Second Hospital, 0.002 DDD/admission in State ird Hospital) and glycopeptides (0.006 DDD/admission in State Second Hospital and 0.005 DDD/admission in State ird Hospital), o en required as last-line treatment for multiresistant bacteria, were used at low levels in two hospitals. is is likely because their use requires pre-authorization due to high cost. Nevertheless, data from the State Second Hospital showed that carbapenem consumption increased by 32.9% from 0.019 DDD/admission in 2013 to 0.026 DDD/admission in 2017. is may be in response to the rising prevalence of Extended-Spectrum Beta-Lactamase (ESBL)-producing bacteria, identi ed in surveillance studies [37]. Also, a ve-time rise in glycopeptide, i.e., vancomycin use (from 0.001 DDD/admission in 2013 to 0.006 DDD/admission in 2017) was observed in the State Second Hospital. Similarly, the consumption of vancomycin was increased in the State ird Hospital dramatically from 0.001 DDD/admission in 2013 and 0.005 DDD/admission in 2017.
Interestingly, the consumption of beta-lactamase sensitive penicillins, including benzyl penicillin (J01CE01) has declined dramatically from 11.4 DDD/100 bed days in 2013, consequently dropping to 0.2 DDD/bed days in 2014 to no consumption in 2017. Hospital reports indicated that the resistance to these agents and ine ective treatment were the main reasons to stop procuring these antimicrobials.
(UTI) which is reported to be one of the prevalent diseases in Mongolia [8]. A recent literature review on e cacy and tolerance of nitroxoline in the treatment of uncomplicated UTI demonstrated equivalent e cacy of nitroxoline with the controls tested (cotrimoxazole, nor oxacin) and justi ed the use of nitroxoline as one of the rst line antibiotics for the treatment of uncomplicated UTI [31]. Uncomplicated UTI is more likely to be prevalent in the ambulatory setting, on the other hand patients preference for inpatient services and the perceived poor quality of outpatient services have been reported previously [13]. Hence the prescribing practice of nitroxoline at these hospitals should be further investigated and justi ed.
Existing inappropriate use [32], access and availability of essential medicines [33], lack of knowledge among prescribers and dispensers [28], and hospital specialties could be the main contributors to huge di erences between hospitals.
Particularly, uctuations seemed to have occurred dramatically during 2015-2017.
is can be somehow related with Health Minister's Order on increasing the Government Budget for medicines and medical devices to public hospitals. Moreover, uctuations may also have occurred due to outbreaks of infectious diseases and increasing trend in the development of antimicrobial resistance. e local production of some antibiotics (ce riaxone, cefotaxime), in addition to commercial promotions may also lead to the increased use of cephalosporins. Availability of newer antibiotics on the market, by the pharmaceutical industry, as well as insu cient knowledge of prescribers [29] may also result in the uctuating results. On the other hand, the procurement and supply of pharmaceuticals and medical devices are still highly dependent on foreign currency [34]. Procurement is also heavily impacted by nancial instability, irregular access to and unavailability of funds. Recently approved guidelines and protocols as well as their compliance and treatment outcomes are yet to be assessed and monitored. Together, the quantity of pharmaceuticals, including antibiotics procured and consumed at public hospitals may have varied.
In addition, the consumption of macrolides was 6.6 DDD/100 bed days ( analysis and revised the manuscript. BG conceived of the study, and participated in its design and coordination and revised the manuscript. All authors read and approved the final manuscript.
In addition, aggressive marketing of newer agents and a lack of comprehensive antibiotic-control programs might also have had an effect on consumption rates. Inappropriate use of antimicrobials and increasing number of resistant antimicrobial agents were found to be prevalent in Mongolia [42,43].

Limitation.
e reports from the Pharmacy Departments in two selected hospitals allowed for precise measurement of total amounts of drugs dispensed at the hospital, including judgement of the total dose, dosage, and volume of each individual prescription. However, due to the unavailability of data on antibiotics dispensed for outpatients which can be obtained with or without prescription from private pharmacies located within hospitals, the extent of antibiotic consumption can be systematically underestimated.
e OTC sale of antibiotics has been prevalent in Mongolia and it has been addressed by the Government [23,44,45].

Conclusion
e first hospital-based study of antibiotic consumption in Mongolia has demonstrated a measurably higher rate of antibiotic consumption in Mongolia than that in other countries. is is a critical first step in planning approaches to limit the emergence of antibiotic resistance through regular data gathering and analysis and the guidance of antibiotic prescribing. Data Availability e datasets used and analyzed during the current study are available from the corresponding author on reasonable request.

Conflicts of Interest
e authors declare that they have no conflicts of interest.