Antimicrobial use in pediatric oncology and hematology in Germany and Austria, 2020/2021: a cross-sectional, multi-center point-prevalence study with a multi-step qualitative adjudication process

Abstract

Background: Due to the high risk of severe infection among pediatric hematology and oncology patients, antimicrobial use is particularly high. With our study, we quantitatively and qualitatively evaluated, based on institutional standards and national guidelines, antimicrobial usage by employing a point-prevalence survey with a multi-step, expert panel approach. We analyzed reasons for inappropriate antimicrobial usage.

Methods: This cross-sectional study was conducted at 30 pediatric hematology and oncology centers in 2020 and 2021. Centers affiliated to the German Society for Pediatric Oncology and Hematology were invited to join, and an existing institutional standard was a prerequisite to participate. We included hematologic/oncologic inpatients under 19 years old, who had a systemic antimicrobial treatment on the day of the point prevalence survey. In addition to a one-day, point-prevalence survey, external experts individually assessed the appropriateness of each therapy. This step was followed by an expert panel adjudication based upon the participating centers' institutional standards, as well as upon national guidelines. We analyzed antimicrobial prevalence rate, along with the rate of appropriate, inappropriate, and indeterminate antimicrobial therapies with regard to institutional and national guidelines. We compared the results of academic and non-academic centers, and performed a multinomial logistic regression using center- and patient-related data to identify variables that predict inappropriate therapy.

Findings: At the time of the study, a total of 342 patients were hospitalized at 30 hospitals, of whom 320 were included for the calculation of the antimicrobial prevalence rate. The overall antimicrobial prevalence rate was 44.4% (142/320; range 11.1-78.6%) with a median antimicrobial prevalence rate per center of 44.5% (95% confidence interval [CI] 35.9-49.9). Antimicrobial prevalence rate was significantly higher (p < 0.001) at academic centers (median 50.0%; 95% CI 41.2-55.2) compared to non-academic centers (median 20.0%; 95% CI 11.0-32.4). After expert panel adjudication, 33.8% (48/142) of all therapies were labelled inappropriate based upon institutional standards, with a higher rate (47.9% [68/142]) when national guidelines were taken into consideration. The most frequent reasons for inappropriate therapy were incorrect dosage (26.2% [37/141]) and (de-)escalation/spectrum-related errors (20.6% [29/141]). Multinomial, logistic regression yielded the number of antimicrobial drugs (odds ratio, OR, 3.13, 95% CI 1.76-5.54, p < 0.001), the diagnosis febrile neutropenia (OR 0.18, 95% CI 0.06-0.51, p = 0.0015), and an existing pediatric antimicrobial stewardship program (OR 0.35, 95% CI 0.15-0.84, p = 0.019) as predictors of inappropriate therapy. Our analysis revealed no evidence of a difference between academic and non-academic centers regarding appropriate usage.

Interpretation: Our study revealed there to be high levels of antimicrobial usage at German and Austrian pediatric oncology and hematology centers with a significant higher number at academic centers. Incorrect dosing was shown to be the most frequent reason for inappropriate usage. Diagnosis of febrile neutropenia and antimicrobial stewardship programs were associated with a lower likelihood of inappropriate therapy. These findings suggest the importance of febrile neutropenia guidelines and guidelines compliance, as well as the need for regular antibiotic stewardship counselling at pediatric oncology and hematology centers.

Funding: European Society of Clinical Microbiology and Infectious Diseases, Deutsche Gesellschaft für Pädiatrische Infektiologie, Deutsche Gesellschaft für Krankenhaushygiene, Stiftung Kreissparkasse Saarbrücken.

Keywords: Antimicrobial resistance; Antimicrobial stewardship; Cancer; Expert panel; Pediatric hematology; Pediatric oncology; Point-prevalence survey.

Fig. 1

Reasons for individual patient exclusion.

Fig. 2

Rate of appropriate, indeterminate, and inappropriate therapies; individual adjudication regarding institutional standards, and expert panel adjudication regarding institutional standards and national guidelines, respectively. Number of adjudications before expert panel: 426 (3 experts × 142 therapies); Number of adjudications after expert panel: 142 (3 experts agreed on adjudication of 142 therapies).

Fig. 3

Rate of appropriate, inappropriate, and indeterminable therapies according to individual adjudication, expert panel adjudications with institutional standards as reference, and expert panel adjudication with national guideline as reference. Each letter code ID indicates one center. ∗y-axis runs between 0 and 3 since 3 independent adjudications are shown which all sum up to 100% (i.e., 1).

Fig. 4

Reasons for inappropriate therapy after expert panel process. More than one reason for inappropriate therapy could be named: 111 reasons for inappropriate therapy regarding institutional standard; 141 reasons for inappropriate therapy regarding national guideline; ∗defined as more than 20% deviation from the corresponding recommendation of the institutional standard or the national guideline; ^#including no de-escalation, early escalation, delayed escalation, broad-spectrum therapy; ^§including prolonged therapy, missing indication.