INTRODUCTION

Antibiotic use has significantly increased the life expectancy of the population ¹. However, antibiotic resistance (AR) is a growing problem and one of the greatest challenges for public health today ². Irrational or indiscriminate use of antibiotics is the most important cause for antibiotic resistance, and also results in other adverse consequences, such as toxicity, interactions, infection by Clostridioides (Clostridium) difficile and prolonged hospital stay; which finally increases morbidity, mortality and costs for the health system ^{3 , 4}.

AR rates are particularly high in hospitals ⁵. It has been recognized for several decades that up to 50% of prescribed antibiotics have unnecessary or inappropriate indications ^{6 )} and, additionally, the small number of recently approved antibiotics reduces therapeutic options, forcing the increasingly prudent use of antibiotics. Therefore, to address this problem, the World Health Organization (WHO) promotes the establishment of Antimicrobial stewardship (AMS) programs. The AMS are an effective strategy that ensures the correct use of antibiotics, according to the available scientific evidence, to improve clinical outcomes, minimize adverse events and reduce hospital costs ⁷.

High-income countries have more resources and experience in developing AMS programs, therefore they have data available on antimicrobial use ^{( 8}. On the other hand, in low- and middle-income countries, such as Peru, AMS programs have been implemented in recent years despite the scarcity of trained human resources and the economic limitations of health systems ^{(9, 10)}.

Accordingly, actions have been taken in Peru to curb antimicrobial resistance, through a Multi-sector Plan (2019-2021) ¹¹ and the development of the AMS Technical Standard. However, we still have limited information about the quality and characteristics of antibiotic prescriptions that can serve as a reference for optimizing antimicrobial control programs. For this reason, we present a work of point prevalence in the use of antibiotics, which will provide basic information about the patterns of antibiotic prescription in a third-level hospital in Lima, Peru.

MATERIALS AND METHODS

RESULTS

We identified 358 patients distributed in different hospitalization rooms of the HNDM. The mean age was 49 (standard deviation [SD ]: 25.3) years (<18 years: 25.7%; 18-65 years: 53.9%; >65 years: 34.1%). The distribution by gender was similar (49.2% males, 50.5% females, and one transgender patient 0.3%). Most patients were from adult medical rooms (62.5%), adult surgery (17%), neonatology (5%), high-risk adults (4.7%), pediatric medicine (4.2%) and other rooms (high-risk adults [3.3% ], pediatric ICU [1.1% ], obstetrics-gynecology [1.1% ], pediatric surgery [0.8% ]).

Invasive devices were present in 341 cases (95.2%). The most frequent was the peripheral venous catheter (90.5%), followed by the bladder (16.5%) and the central venous catheter (10.9%), while hemodialysis, peritoneal and other catheters accounted for 7%. Orotracheal intubation was performed in 8.4% of patients, and some surgical procedure was performed in 23.5% of patients during their hospitalization.

Antibiotics were used in 51.7% of the patients; most frequently in pediatric (73.3%), pediatric surgery (66.7%), adult medicine (57.1%), neonatology (50%), obstetrics-gynecology (50%) and adult surgery rooms (42.6%); and less frequently in adult ICUs (29.4%), pediatric ICUs (25%) and high-risk adult hospitalization (8.3%) (Table 1).

The main reason for using antibiotics was community-acquired infections (66.1%), followed by healthcare-associated infections (19%). The most frequent diagnoses for antibiotic use were pneumonia (19.6%), skin and soft tissue infection without non-surgical osteomyelitis (12.1%), lower urinary tract infection (11.1%), intra-abdominal infection (8.9%), clinical sepsis (7.4%), upper urinary tract infection (4.2%). The least frequent diagnoses were acute or chronic exacerbated bronchitis, asymptomatic bacteriuria, and systemic response inflammatory syndrome, which received antibiotics in 2.6%, 2.1%, and 0.5% of cases, respectively (Table 1).

The use of antibiotics was mostly indicated as treatment (87.8%), surgical prophylaxis (7.9%) and for unknown reasons (4.2%). All indications for surgical prophylaxis received multiple doses of the drug on the same day (Table 1).

Regarding the number of antibiotics used per patient, 49.7% received one antibiotic, 44.9% two antibiotics and 5.4% three or more antibiotics. Only 57.3% of patient prescriptions followed guidelines based on local or international CPGs for antibiotic indication, while 28.5% did not follow any standardized recommendations. The frequency of antibiotic use in accordance with local or international CPGs was as follows: high-risk adults (100%), neonatology (94.4%), adult medicine (77.7%), pediatric medicine (71.4%), pediatric surgery (66.7%), adult surgery (48.7%), obstetrics-gynecology (25%) and adult ICU (20%) (Figure 1).

Figure 1 Compliance with local or international clinical practice guidelines by hospital room at Hospital Nacional Dos de Mayo, 2018. 

In 141 cases (49.8%), the initial antibiotic prescriber was the specialist physician, while in the remaining 50.2% the resident physician was the prescriber. We also found that 86.8% of the therapies were empirical. Likewise, 37% of the patients had some biological sample collected for microbiological culture (blood [22.7% ], urine [43.3% ], sputum [11.3% ], wound drainage [6.3% ], sterile fluids [7.5% ] and others [8.8% ]). Finally, in 71% of the cases, the culture results were reported to the treating physician (Table 2).

In 64 patients (22.7%), at least one required antibiotic dose was missed, with a median of missed doses of 2.9 (interquartile range: 1-16). Missing a dose was mainly as a consequence of shortage in 48.5% of the cases. Meropenem, vancomycin, and ceftriaxone were the most frequently used antibiotics for community and healthcare-related infections (Table 3).

DISCUSSION

The overall frequency of antibiotic use we found (51.7%) was similar to that of other point prevalence studies ^{13 - 15}. Studies from other parts of the Americas show data similar to our reality. For example, in 2011, a study took place in acute care hospitals in ten U.S. states and reported that nearly 50% of the patients had received some antibiotic ¹³. In 2015, another study that included four Latin American countries (Argentina, Brazil, Chile, and Mexico) described that 47.5% of the patients recruited used antibiotics, of which 90% were for systemic use ¹⁴. Another study on the point prevalence of antibiotic use was carried out in Mexico between August and September 2016, and it described that 51% of 260 patients received at least one antibiotic ¹⁵.

On the other hand, studies on European population present some differences with respect to our reality. In 2016, a study carried out by the national center for nosocomial infection surveillance in Germany found a 26% prevalence of antibiotic use in 64,412 patients from 218 hospitals ¹⁶. Additionally, a study conducted on acute care hospitals in Europe found that one third of the patients received at least one antibiotic (70%, one antibiotic; 24%, two antibiotics; 4.5%, three; and 1.3%, four or more antibiotics) in 2017 ¹⁷. This reinforces the concept that in the last decade, the frequency of antibiotic use (by using point prevalence) is lower in regions of Europe ^{16 , 17} than in critical care units in America ^{13 - 15}.

In our study, 66% of the patients received antibiotic therapy for community-acquired infections; other studies report 76.5% (Mexico, 2016) ¹⁵, 70% (USA, 2011) ¹³, and 60.5% (Latin America, 2015) ¹⁴, which are similar to our results. In Europe, less patients (45%) received antibiotic therapy for community-acquired infections ¹⁷. The studies mentioned, including ours, agree that the most frequent diagnosis for antimicrobial prescription was lower respiratory infections.

The studies we cited show that the most frequently used antibiotics for community acquired infections were penicillins combined with beta-lactamase inhibitors, followed by cephalosporins and fluoroquinolones ^{16 , 17 , 21 , 23}; to prescribe these drugs is related to a rational use of antibiotics and shorter hospital stays ^{( 18}. Some series like ours, mainly from Latin American countries ¹⁴, describe the use of broad-spectrum antibiotics in cases where they were not necessarily needed, such as in hospitalization outside the critical care unit ¹⁹ and community-acquired infections ^{13 , 15}.

Even though 88% of the patients who used antibiotics (for treatment of an infectious pathology), about a third of them did not follow an indication based on any CPG. Similar findings are shown in two American studies, such as the one from Ohio University Hospital, where 30% of the antibiotic indications were considered unnecessary ²⁰. A Mexican study (2016) ^{15 )} found that antibiotic use was not justified in 21% of the patients. Similarly, in Latin America (Argentina, Brazil, Chile, Mexico) it was only 64%. In contrast, in Australian hospitals (2014) ²¹, the percentage of inappropriate prescriptions was about a quarter of the total prescriptions and non-guideline use was observed in 27% of the patients. Another multicenter study (2015) ¹⁴ reported a similar percentage of adherence to CPG (77.4%), and finally a study in Singapore ²³ showed that the use of antibiotics without justification was infrequent (0.4%). The causes described for the unjustified use of antibiotics were prolonged therapy times, non-infectious processes, and treatment of colonizing microorganism findings ²⁰.

Antibiotic prophylaxis in surgery represented 8% of the total prescriptions in our study, which highlights its irrational use, as it exceeded the number of days for antibiotic treatment. So, it is important to point out that multiple doses with a duration of more than 24 hours were prescribed in almost all the indications. In other countries, such as Australia, this behavior is observed in 40% of the prescriptions ²¹ and in European countries between 54% and 56% ^{16 , 17}. In contrast, a study carried out in 13 hospitals in Singapore showed that only 8.4% of the prescriptions for surgical prophylaxis lasted more than one day ²³. These data reveal that there is much to be done to raise awareness among professionals in surgical specialties regarding the duration of antibiotic prophylaxis, since this causes an increase in antibiotic resistance.

Regarding microbiological isolation, our study highlights that only 37% of cases had a microbiological sample taken to find the possible causal agent; most of these samples (77.3%) consisted of blood, urine and sputum. In a retrospective study carried out at the University Medical Center of the University of Amsterdam, Netherlands, with 733 hospital beds, about 7,500 blood cultures are carried out by year, which contrasts with the figures we obtained ²⁴. In a study of hospitals in Singapore (2015-2016) ²³, 45.5% showed some positive microbiological culture.

Due to the limited number of beds, the pediatric and adult ICUs represented a small percentage (1.1% and 4.7%, respectively) of the total number of patients evaluated. However, there was a lower percentage of antibiotic use in those units (25% in pediatric ICU and 29.4% in adult ICU), which suggests a greater intervention of the AMS of the HNDM in the mentioned rooms. The latest consensus promoted by the Joint Programming Initiative on Antimicrobial Resistance (2019) standardizes suggestions for optimal follow-up and monitoring of antimicrobial control programs, and emphasizes the item referring to the measurement of drug use, where it is specified that the defined daily dose (DDD) or days of therapy (DOT) are the best parameters for evaluation ²⁵.

The most frequently used antibiotics were ceftriaxone (22.2%), meropenem (12.8%), vancomycin (10.8%), clindamycin (8%), ceftazidime (6.7%) and ciprofloxacin (6%). It is important to note the remarkably similar percentages of the use of meropenem and vancomycin for both nosocomial and community infections, as well as the percentage of ceftriaxone use for nosocomial infections. This reflects the unwise use of antibiotics in our reality.

This study did not evaluate antibiotic consumption in a DDD, patient outcome, and duration of antibiotic therapy, due to the design of the cross-sectional study that sought to describe the characteristics of antibiotic use in hospitalized patients at a given time. Therefore, these variables could be considered in future studies.

In conclusion, this study describes a high frequency of irrational use of antibiotics in hospitalized patients. We found that 15-25% of the cases had an unjustified use of cephalosporins, clindamycin and amikacin, in addition 13% used carbapenems for community acquired infections. In about 90% of patients who received antibiotics, the prescription was empirical and not directed. Our hospital requires urgent measures to strengthen the institutional AMS, in order to decrease the rates of unnecessary and inadequate use of antibiotics, as well as to improve prescriptions according to CPGs, grant timely access to antibiotics and lower the figures of bacterial resistance.