Inappropriate use of antimicrobials is, to a greater or lesser extent, an undeniable reality in any hospital. However, the degree to which such drugs are inappropriately used is difficult to quantify due to the methodological and practical difficulties inherent in its measurements and the lack of standardized evaluation methods. Various studies in different countries have estimated inappropriate use of antimicrobials in hospitals at levels ranging between 16 and 70%1–4. The consequences of the inappropriate use of antimicrobials include an increased risk of adverse events such as Clostridioides difficile infection, therapeutic failure with potential effects on morbidity and mortality, selection of resistant microorganisms, and increased health costs5–7.

A particularly relevant phenomenon is the increase observed in bacterial resistance to antimicrobials, which has become a global health problem. The World Health Organization has set the antimicrobial stewardship as one of the five strategic goals in the fight against microbial resistance8.

Antimicrobial stewardship programs (ASPs) are being progressively implemented in Spanish hospitals. Their goals include improving clinical outcomes, minimizing adverse events, including resistance, and measuring and improving the appropriate use of antimicrobial agents to ensure the cost-effectiveness of treatments9,10.

Various strategies have been proposed to estimate the quality of antimicrobial use in different settings. One of them consists in establishing quality of use indicators based on consumption11,12. However, the most commonly used method for assessing the prevalence and use of antimicrobials is the performance of cross-sectional studies1,13,14. These are usually carried out in a single day and are an efficient tool when time and resources do not allow for a longitudinal study. A significant sample can provide information on the prescription of antibiotics at different times in the course of treatment. In Spain, the National Epidemiological Surveillance Network (RENAVE) annually records the prevalence of antimicrobial use in hospitals15.

The evaluation of antimicrobial drug prescriptions should consider several items such as indication, spectrum, dosage, duration, efficacy and safety monitoring, and registration. The biggest problem about performing a systematic evaluation of antimicrobial prescription is the difficulties in using a standardized method due to the complexities inherent in the evaluation of each of these items. Another critical point is the fraction of subjectivity in the evaluation that leads to inter-observer variability16–18.

The aim of the PAUSATE study was to evaluate the prevalence and degree of appropriateness of antimicrobial use in Spanish hospitals by means of a pharmacist-led cross-sectional systematic review.

A total of 103 hospitals participated in the study, distributed heterogeneously throughout the Spanish territory. The composition by hospital size is reflected in figure 1, and shows a distribution that is not proportional to the real one, with a greater weight in the study of larger hospitals.

Figure 1.

Distribution and size of participating hospitals.

(0,09MB).

*Source: 2019 Spanish Hospital Catalog. Ministry of Health, Consumer Affairs and Social Welfare.

The prescriptions of 3,568 patients were reviewed, of whom 1,498 (42.0%) received antimicrobials. Of these, 862 (57.5%) were male and median age was 69 years (range: 0-101). A total of 46.6% of these patients were admitted to a medical unit, 31.2% to a surgical unit, 9.9% to a critical care unit, 8.4% to a hemato-oncological unit, 3.1% to a pediatric unit, and 0.8% to other units. Of the patients on antimicrobial therapy, 1,449 (96.7%) received at least one antibiotic, 126 (8.4%) at least one antifungal, and 80 (5.3%) at least one antiviral. The prevalence of the use of the most frequent antimicrobial agents is shown in table 2.

Of the patients who received antimicrobial therapy, 424 (28.3%) did so in combination therapy; 358 (23.9%) received at least two antibiotics and 120 patients (8.0%) were treated with a combination of a β-lactam active against Pseudomonas aeruginosa and an agent active against methicillin-resistant Staphylococcus aureus.

Regarding the appropriateness of treatment, the prescription was considered appropriate in 34% of the cases, suboptimal in 45%, inappropriate in 19% and unevaluable in 2% (Table 3).

The items that most influenced assess a prescription as suboptimal were completeness of medical record entries, type of antimicrobial agent selected, duration of treatment and monitoring of efficacy and safety. In 21% of cases where the prescription was assessed as suboptimal, the only reason was incompleteness of data in the patient's medical record (in 79% of the remaining cases, several items converged). The item that most influenced rating a prescription as inappropriate was the indication of antimicrobial treatment, with the antimicrobial being considered unnecessary in 8% of prescriptions.

PAUSATE is the first nationwide study in Spain to simultaneously measure the prevalence and appropriateness of the hospital use of antimicrobials by means of a prevalence survey.

The prevalence data obtained in the PAUSATE study are similar to those in the latest RENAVE survey in 201915. Thus, in the PAUSATE study 42% of the patients were receiving at least one antimicrobial, as compared to 45.8% in the RENAVE survey. It should be considered, however, that both studies differ in the composition of the ATC groups analyzed. Indeed, the RENAVE study, unlike the PAUSATE study, included ATC groups A07 and P01 and excluded groups J04 (except for agents used for non-tuberculous bacteria) and J05, although these groups include agents with a small impact on consumption (the prevalence of each group is less than 2.5%). In the PAUSATE study, 28.3% of patients who received antimicrobials did so in combination therapy, as compared to 27.7% in the RENAVE survey.

Of the 10 most highly consumed antimicrobials in the PAUSATE study, nine were the same as in the RENAVE survey, with virtually identical positions. However, the use of broad-spectrum antimicrobials such as piperacillin-tazobactam, meropenem and linezolid was slightly higher in the PAUSATE study than in the RENAVE survey (5.8%, 4.0% and 2.1% vs. 5.2%, 3.5% and 1.8% respectively). Other antimicrobials such as amoxicillin-clavulanic acid, levofloxacin, cefazolin, and ciprofloxacin were more commonly used in the RENAVE survey than in the PAUSATE study (8.9%, 4.9%, 4.6% and 2.9% vs. 7.2%, 3.3%, 3.0% and 2.4%, respectively), which is consistent with the greater proportion of larger hospitals included in the PAUSATE study. Such hospitals tend to treat cases of greater complexity and usually exhibit higher incidences of nosocomial infections and resistant microorganisms.

Other findings resulting from the higher proportion of larger hospitals in the PAUSATE study include a higher prevalence of antifungals as compared to RENAVE (8.4% vs. 3.7%) and a higher percentage of patients on a combination of a β-lactam active against Pseudomonas aeruginosa with an agent active against methicillin-resistant Staphylococcus aureus (8.0%).

Unlike the RENAVE survey, the PAUSATE study was limited to collecting data on the prevalence and appropriateness of the use of antimicrobials, without gathering information on the indication for antimicrobials (prophylaxis, empirical or targeted therapy, or community or nosocomial infection) or patient-related risk factors (comorbidities, immunosuppression, devices and implants, etc.).

The National Plan against Antibiotic Resistance (PRAN) of the Spanish Agency for Medicines and Health Products describes the levels of consumption of antimicrobials in Spanish hospitals, expressed as the number of Defined Daily Doses per 1,000 inhabitants per day24. Although these values cannot be compared directly with the prevalence data of the PAUSATE study as both the methodology and the units of measurement used are different, it can be seen that four of the five most commonly consumed antibiotics in the PRAN coincide in those reported in the PAUSATE prevalence study.

The international literature reports that between 16% and 70% of antimicrobial treatments are inappropriate. However, most of the data comes from low-quality studies characterized by very heterogeneous designs and methodologies. In Spain, a few studies were published that reveal the high levels of inappropriateness antimicrobial use in the country's hospitals, but they are either single-center or refer to specific areas of hospitalization25–27.

The main problem with assessing the quality of antimicrobial use is that there is no standardized evaluation method that covers all the different items of antimicrobial prescribing and that defines the scale of each of these items and by extension the qualification of the prescription as a whole.

A further critical point in the evaluation of antimicrobial prescribing is inter-observer variability. Subjectivity in the evaluation of antimicrobial prescribing is a consequence of the uncertainties inherent in many cases in the approach to infectious disease and the management of antibiotic therapy.

In order to minimize these aspects in the PAUSATE study, the AFinf group designed a method based on an evaluation form with specific and objective records related to local guidelines or, in their absence, to national or international reference guidelines. The form was put together by a group of pharmacists with long experience in the care of patients with infectious diseases who, in many cases, were members of their hospitals’ antibiotic stewardship teams.

The PAUSATE study showed that 45% of antimicrobial prescriptions were suboptimal and 19% were inappropriate.

In the analysis by items, the one with the greatest room for improvement was completeness of medical record entries, with entries being incomplete in 34% of prescriptions. This item was not considered crucial enough to assess the whole prescription as inappropriate, but it was the reason why one out of every five prescriptions were assessed as suboptimal rather than appropriate.

Documenting the clinical management plan, as well as reflecting the antimicrobial treatment selected together with its indication and the duration of treatment are recommendations included in stewardship programs21,28. Explicitly stating an antibiotic plan in the patients’ medical record helps physicians reflect on the decisions made and facilitates the review of antibiotic treatment.

The choice of antimicrobial agent was suboptimal in 23% and inappropriate in 3% of the prescriptions evaluated. A suboptimal assessment indicated non-adherence to reference protocols or guidelines or the use of equally effective but less safe, ecological, or inexpensive therapeutic alternatives. An inappropriate assessment refers to a larger issue of ineffectiveness or safety of the prescribed agent.

Other items with less favorable results were the duration of treatment (excessive in 19% of cases and too short in 2%) and the monitoring of efficacy and safety (suboptimal in 17% of cases and inappropriate in 4%).

The most concerning result of the study was the considerable percentage of prescriptions classified as inappropriate (19%), i.e., those where, in the evaluator's opinion, the patient did not require an antimicrobial, or its use did not guarantee the patient's cure or caused unacceptable harm. This data reinforces the need to implement measures to optimize the use of antimicrobials.

This study presents with several limitations. Firstly, the voluntary participation of the centers, apart from providing a heterogeneous distribution that is not representative of reality, with a disproportionately high participation of larger hospitals, can lead to a selection bias as it could be argued that the centers participating in the study were those most keen on optimizing the use of antimicrobials. Another selection bias is that the study sample only included patients who received at least one antimicrobial, excluding those who did not receive an antimicrobial although they required it (default antimicrobial indications were not analyzed).

Another limitation of the study is seasonality. The use of antimicrobials varies between different times of the year. The PAUSATE study was carried out in the month of April and overlapped with the RENAVE survey, which collected its data in May. Another factor that could distort the data obtained on antimicrobial use is the presence of patients admitted with SARS-CoV-2 pneumonia. During the data collection period, the percentage of hospital beds occupied by patients with SARS-CoV-2 infection ranged between 6.4% and 8.2%29.

The fact that the study was conducted by hospital pharmacists could have constituted an observer bias. Evaluation of antimicrobial prescriptions is assumed to be subject to nuances dependent on the observer's judgment. The influence of the observer's academic background in his evaluation of antimicrobial prescriptions was not studied. A forthcoming study, to be published as an extension of the present one, will measure the concordance of assessments by pharmacists and physicians using the AFinf method.

In short, the PAUSATE study provides updated data on the prevalence of the use of antimicrobials in Spanish hospitals. The data obtained are consistent with those previously published in the RENAVE survey and those reported by the PRAN Plan. Moreover, the study is the first of its kind to evaluate appropriateness of antimicrobial use, contemplating all the items in antimicrobial prescribing. An understanding of the prevalence and appropriateness of antimicrobial use is the first step for designing and undertaking the measures required to improve and measuring the impact of implementing antimicrobials within the framework of antimicrobial stewardship plans.

No funding was received for this research. The authors and investigators, and their collaborators, received no remuneration for participating in this study.

The authors would like to thank the researchers and other professionals involved in the collection of the data. Their names are contained in a specific section.

The authors have declared no conflict of interest with respect to this research.

PAUSATE is a nationwide study aimed at measuring the prevalence and appropriateness of antimicrobials use in hospitals.

The results of the study should subsequently be used to design actions geared towards improving antimicrobial stewardship programs.