An antimicrobial stewardship program implemented in an intensive care unit successfully decreased patient antibiotic treatment days but did not lead to narrower antibacterial spectrum selection, according to a retrospective before-and-after investigation published in Frontiers in Pharmacology. The study, conducted by researchers at The First Affiliated Hospital with Nanjing Medical University, analyzed two hundred critically ill adult patients suffering from systemic infections across consecutive baseline and intervention years. By deploying an innovative dual-dimensional monitoring framework that paired conventional utilization measures with spectrum-weighted metrics, known as days of antibiotic spectrum coverage and antibiotic spectrum coverage use, the research team demonstrated that volume reductions in cumulative antimicrobial administration largely coincided with shorter hospital stays rather than genuine de-escalation toward narrower-spectrum agents.
Patients admitted to intensive care units face substantial risks of severe bacterial infections due to high acute illness acuity, frequent invasive clinical procedures, prolonged hospitalization, and compromised host immunity. Consequently, intensive care units represent critical environments where broad-spectrum antimicrobials are heavily utilized, making critically ill populations primary targets for structured antimicrobial stewardship interventions. A cornerstone strategy within stewardship programs is antimicrobial de-escalation, a clinical practice defined as systematically narrowing the breadth of antibacterial coverage or discontinuing redundant empiric agents once clinical stability is attained or definitive microbiological identification and susceptibility findings emerge.
Historically, hospital stewardship teams have relied almost exclusively on standard volume-oriented surveillance indicators, primarily Days of Therapy, Defined Daily Doses, and Antimicrobial Use Density. Days of Therapy quantifies the overall duration of exposure by counting each unique antibacterial agent administered on a given calendar day, irrespective of dosage strength. In contrast, Antimicrobial Use Density measures the gross volume of drug consumption by normalizing accumulated Defined Daily Doses to patient observation time, expressed per one hundred bed-days or days of stay. Although these classical metrics provide valuable benchmarks for tracking institutional consumption trends, they possess inherent structural limitations because they do not reflect differences in antibacterial spectrum breadth.
To overcome these qualitative surveillance gaps, clinical researchers introduced composite spectrum-based indicators that integrate coverage breadth into utilization measurements. Days of antibiotic spectrum coverage incorporates antibiotic spectrum scores directly into exposure assessment by multiplying the assigned breadth score of each administered drug by its corresponding days of therapy. Expanding upon this foundation, antibiotic spectrum coverage use calculates spectrum exposure weighted by antimicrobial use density, offering an adaptable alternative when granular patient-level administration logs are unavailable. Furthermore, derived spectrum-intensity ratios, specifically days of antibiotic spectrum coverage divided by days of therapy and antibiotic spectrum coverage use divided by antimicrobial use density, capture the average spectrum breadth delivered per treatment day and per consumption unit, establishing duration-independent indicators of treatment intensity.
The investigation was designed as a single-centre retrospective cohort study conducted within the intensive care unit of The First Affiliated Hospital with Nanjing Medical University, with ethical review approval granted by the hospital ethics committee. The investigators screened adult patients hospitalized with documented infectious diseases between January 1, 2024, and December 31, 2025, defining 2024 as the baseline phase of routine practice and 2025 as the formal stewardship intervention period. Eligible individuals were aged eighteen years or older, had infections confirmed through clinical, laboratory, radiographic, and microbiological criteria requiring systemic antibacterial therapy, and received active treatment for at least twenty-four continuous hours. From initial pools of 610 eligible baseline patients and 552 intervention patients, the researchers applied computer-generated random number sequences to assemble two equally balanced analytical groups of one hundred patients each.
Patient baseline demographic, clinical, and laboratory profiles were thoroughly balanced between the two cohorts. Median patient age was 69.00 years in the baseline cohort and 66.50 years in the intervention cohort, with male representation standing at 62.00 percent and 59.00 percent, respectively. Median body mass index was 23.20 kilograms per square meter versus 24.00 kilograms per square meter. Severe illness markers showed close parity, including invasive mechanical ventilation requirements among 93.00 percent of baseline subjects and 91.00 percent of intervention subjects. Median Charlson Comorbidity Index scores were identical at 7 in both cohorts, while median Acute Physiology and Chronic Health Evaluation II scores were comparable at 15 versus 13.5. Renal and hepatic laboratory markers, including estimated glomerular filtration rates and liver enzyme levels, exhibited no statistically meaningful baseline divergence.
The clinical distribution of infection sites and identified microbial species also remained consistent between study phases. Pulmonary infections represented the predominant clinical diagnosis in both groups, accounting for 92.00 percent of baseline admissions and 84.00 percent of intervention admissions, followed by urinary tract, bloodstream, and intra-abdominal sources.
The stewardship program deployed during the 2025 intervention period incorporated three structured clinical components executed through dedicated multidisciplinary collaboration. First, stewardship teams implemented enhanced evaluations of antimicrobial de-escalation opportunities, advising attending clinicians to narrow coverage or discontinue broad-spectrum empirical agents upon laboratory culture confirmation or measurable clinical stabilization. Second, the hospital instituted standardized classification and restriction controls aligned with Chinese national stewardship classifications, categorizing drugs into special-use antimicrobials requiring formal authorization and non-special-use antimicrobials.
Initial unadjusted comparisons demonstrated significant reductions across raw cumulative antibiotic exposure metrics following stewardship implementation. Median patient-level Defined Daily Doses dropped from 26.25 to 17.90, days of therapy decreased from 24.00 days to 14.00 days, and days of antibiotic spectrum coverage fell from 219.50 to 135.00, all yielding statistically significant differences. However, the median length of hospital stay also decreased substantially from 15.00 days in the baseline group to 11.00 days in the intervention cohort. Correlation analyses revealed that cumulative Defined Daily Doses, days of therapy, and days of antibiotic spectrum coverage correlated strongly with individual patient length of stay, exhibiting Spearman correlation coefficients between 0.74 and 0.79, indicating that gross cumulative consumption declines were heavily influenced by shortened patient observation periods.
To rigorously account for observation time and potential confounding from clinical disease severity, the investigators performed multivariable regression modeling on complete patient cases using generalized linear models. When evaluating normalized treatment-day rates per patient-day using negative binomial models adjusted for age, sex, body mass index, APACHE II score, Charlson Comorbidity Index, mechanical ventilation, and renal function, the stewardship intervention maintained a statistically significant association with a lower antibiotic treatment-day rate, producing an adjusted incidence rate ratio of 0.834 with a 95 percent confidence interval from 0.736 to 0.944. Conversely, multivariable Gamma regression models for Defined Daily Dose intensity and days of antibiotic spectrum coverage rates showed that their reductions were no longer statistically significant after full adjustment, confirming that treatment exposure contraction was primarily characterized by fewer active therapy days.
Stratifying antimicrobial consumption by administrative restriction tiers revealed marked divergence between drug classes that explained overall patterns. Within the non-special-use stratum, days of antibiotic spectrum coverage fell significantly from 114.50 to 63.50, accompanied by meaningful drops in antimicrobial use density from 99.87 to 87.24 and antibiotic spectrum coverage use from 923.93 to 761.25. In stark contrast, metrics within the special-use stratum showed no statistically significant decreases, with antimicrobial use density and antibiotic spectrum coverage use trending upward by 21.4 percent and 37.9 percent, respectively. This compensatory upward drift in restricted, broad-spectrum agents partially offset reductions achieved in the non-special-use category, illustrating how targeted reductions in standard agents can occur alongside persistent reliance on high-tier broad-spectrum therapies in critically ill cohorts.
Most crucially, duration-independent spectrum intensity ratios demonstrated remarkable stability throughout the entire investigation. The overall ratio of days of antibiotic spectrum coverage to days of therapy remained virtually unchanged, shifting from 9.15 to 9.22, while the ratio of antibiotic spectrum coverage use to antimicrobial use density moved from 9.55 to 9.68, with neither difference approaching statistical significance. Similar preservation was observed across both restriction tiers, proving that the average spectrum breadth administered per day of therapy was fully maintained. The authors underscored that this discrepancy between lowered treatment duration and unchanged spectrum breadth illustrates a pattern of treatment-day reduction without spectrum de-escalation, demonstrating that conventional volume reductions do not inherently signify narrowing spectrum coverage and highlighting the essential complementary role of spectrum-based metrics in hospital stewardship.
Subject of Research: Antimicrobial stewardship evaluation using spectrum-based DASC and AUSC metrics in critically ill patients
Article Title: Application of DASC and AUSC metrics in evaluating antimicrobial stewardship interventions in critically ill patients
Article References: Wang, S., Liu, Y., Wang, H., Yu, Y., Luo, C., & Hu, J. (2026). Application of DASC and AUSC metrics in evaluating antimicrobial stewardship interventions in critically ill patients. Frontiers in Pharmacology, 17, Article 1993215. https://doi.org/10.3389/fphar.2026.1993215
Image Credits: AI Generated
DOI: 10.3389/fphar.2026.1993215
Keywords: antimicrobial stewardship, intensive care unit, days of antibiotic spectrum coverage, antibiotic spectrum coverage use, days of therapy, antimicrobial use density, defined daily doses, antimicrobial de-escalation, critically ill patients, bacterial infections, pharmacology, special-use antimicrobials, hospital length of stay, multivariable regression, treatment duration, clinical pharmacy, Acinetobacter baumannii, pathogen coverage
News Source: Ophelia Keating. (November 16, 2026). ICU Stewardship Program Shortens Antibiotic Treatment Days Without Narrowing Spectrum Breadth. Scienmag.



