Introduction
On July, 10 2024, the U.S. Food and Drug Administration (FDA) announced a nationwide shortage of BD BACTEC blood culture media bottles (BCx) due to supplier issues. 1 In response, the FDA recommended conservation strategies to prioritize BCx use based on clinical need, ensuring quality and safety in patient care. This shortage was anticipated to impact diagnosis, care, and antimicrobial stewardship, particularly in emergency departments (ED), where usage rates are high.
At the Mount Sinai Health System, the largest hospital network in New York City, we promptly informed clinicians of the shortage on July 11, 2024 and launched our first BCx stewardship initiative within two weeks using an electronic clinical decision-support system (CDSS) (Supplementary Figure S1). The CDSS was initially implemented in the ED - a high-pressure environment where rapid decisions are critical and where the intervention’s impact was expected to be greatest. Although later expanded to inpatient settings, this study, therefore, focuses on the ED.
Adapted from Fabre et al., this CDSS-guided BCx ordering categorized the pretest probability of positive BCx for non-neutropenic patients into three groups, low (<10%), intermediate (10–50%), and high (>50%). Reference Fabre, Sharara, Salinas, Carroll, Desai and Cosgrove2 Neutropenic and suspected sepsis cases were exempt. For suspected sepsis cases, the Epic CDS tool directed providers to follow the institutional sepsis pathway, which includes obtaining two BCx sets for adult patients. Importantly, CDSS did not give automated recommendations or enforce restrictions. Clinicians reviewed the risk table and were required to select a matching risk category to proceed with the blood culture order, documenting their assessment (Supplementary Figure S1).
We assessed the impact of the shortage and intervention by examining BCx volume by examining BCX volume, positivity rate, length of stay (LOS), and in-hospital mortality rates in ED from May 1 to September 14, 2024.
Methods
Data collection
We obtained data from six EDs within the Mount Sinai Health System in New York, NY, which include Mount Sinai Beth Israel, Mount Sinai Brooklyn, Mount Sinai Hospital, Mount Sinai Morningside, Mount Sinai Queens, and Mount Sinai West. All BCx collections in these EDs from May 1 to September 14, 2024, were identified. Data on BCx results, classified as positive or negative for microbial growth, and presumptive contaminants, were retrieved to calculate the diagnostic yield of BCx. Contaminants were defined in consistence with the criteria by the Clinical and Laboratory Standards Institute. 3 LOS and in-hospital mortality data from the same period was collected. In the two weeks following the initial broadcast, which included emailing the CDSS figure to clinicians and announcing it during the ED weekly operations meeting attended by both clinicians and nurses, we implemented the CDSS for BCx stewardship. To account for adaptation to the new system, data from this two-week period were excluded from the analysis.
Statistical analysis
The primary outcome was the change in BCx volume, while secondary outcomes included BCx diagnostic yield, LOS, and in-hospital mortality. A two-week buffer period (July 11–25, 2024) allows for clinician adaptation, with data analyzed for the pre-intervention (May 1–July 10) and post-intervention (July 26–September 14) periods.
An Interrupted Time Series was performed using Autoregressive Integrated Moving Average (ARIMA), while Mann–Whitney U and Fisher’s Exact tests were used to compare LOS and mortality rates before and after the intervention. Statistical analyses were performed using Python version 3.11.5. The study received approval from the Institutional Review Board at the Icahn School of Medicine at Mount Sinai (approval no. 19-00951).
Results
Between May 1 and September 14, 16,820 BCx were collected (54–218 per day). Following a 2-week period of clinician notification regarding the bottle shortages and appropriate indications on July 11, 2024, with the implementation of CDSS, the median daily BCx utilization per 1,000 ED visit decreased from 141.5 (IQR: 127.6–155.3) to 77.9 (IQR: 68.3–86.3). Daily positive culture rates increased from 6.2% (IQR: 4.7%–7.6%) to 8.8% (IQR:6.1%–11.5%). The ARIMA model results indicated that, on average, the CDSS implementation was associated with a significant decrease of 61.8 BCx usages per 1,000 ED visit (95% CI, –71.8 to –51.8; P < 0.001). Figure 1 illustrates trends in daily BCx usages and positive rates. Analysis of positive cultures, including possible contamination, can be found in Supplementary Table S2.
(a) Daily blood culture media bottles utilizations. (b) Daily positive rate of blood cultures.
Additionally, we evaluated the BCx contamination rate. Daily contamination culture rates increase from 1.9% (IQR: 1.2–2.5%) to 2.7% (IQR: 1.2–3.6%). However, ARIMA model analysis showed a 0.8% (95% CI, –0.4 to 2.0; P = 0.20) increase, which was not statistically. Figure 2 illustrates the daily trends in BCx contamination.
Daily contamination rate of blood cultures.
There was no significant difference in LOS (1.36 vs. 1.49 days; P = 0.18) or in-hospital mortality rate (2.91% vs. 2.88%; P = 0.90) during this time-period.
Discussion
This study shows that CDSS-guided BCx ordering can assist clinicians in making more selective decisions during the nationwide shortage, reducing BCx volume while increasing positivity rates without compromising patient outcomes.
In ED, BCx is frequently collected from patients with suspected infections, even in cases with a low likelihood of bacteremia (eg, non-severe pneumonia, cellulitis), due to concerns about missing severe infections, Reference Long and Koyfman4 This practice has associated with low positive BCx yield globally. Reference Long and Koyfman4–Reference Nannan Panday, Wang, van de Ven, Hekker, Alam and Nanayakkara6 Previous studies suggest that stewardship strategies, including evidence-based provider education, and feedback, can help clinicians refine their criteria for ordering BCx and reduce unnecessary collections. Reference Schinkel, Boerman and Bennis7–Reference Wang, Zhou and Shay9
The shortage provided a unique opportunity to implement and reinforce BCx stewardship practices. In response, the FDA recommended conservation strategies to preserve supplies for high-risk patients. 1 These strategies included minimizing contamination, reducing repeat cultures, prioritizing use for higher-risk patients, and collecting BCx only when medically necessary. Reference Ryder, Van Schooneveld, Diekema and Fabre10 In alignment with these recommendations, CDSS has been implemented in our EDs to guide ordering clinicians to adhere to recommended conservation strategies.
While a recent inpatient study using order-set guidance and a hard stop on repeat cultures saw a 49% reduction in BCx use, Reference Humphries, Wright and Banerjee11 our approach focused on clinician-guided stewardship rather than rigid restrictions. We believe this strategy promotes sustainable, evidence-based practice beyond crisis scenarios and offers a scalable model for routine ED workflows. To the best of our knowledge, our study is the first to report a significant increase in BCx yield following the implementation of a clinical decision-support tool during the shortage, along with evidence that patient outcomes—in terms of LOS and in-hospital mortality—were unaffected by the substantial reduction in BCx collections. Further prospective studies are warranted to evaluate BCx restriction on outcomes and its use post-shortage.
This study has several limitations. First, all six hospitals where we collected data were within a single health system, limiting generalizability. Second, the retrospective design also restricts causal conclusions. Additionally, we lacked data on clinician adherence or perceptions to the decision-support tool. Finally, patients suspected of sepsis were exempt from the tool, allowing clinicians discretion. In practice, this meant that clinicians could still order blood cultures for patients they suspected of being septic, even if those patients did not meet the CDSS criteria. It is important to note that overall blood culture utilization still improved despite this flexibility.
Conclusions
Providing guidance on appropriate indications and implementing decision-support in response to the BD BACTEC BCx shortage led to reduced BCx volumes and increased positivity rates in emergency departments, with no adverse impact on patient length of stay or mortality. This approach offers a framework for effective conservation strategies during supply shortages.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/ice.2025.83.
Author contributions
KT, BC, AS designed and conceptualized the study. KT, WO, BC, and AS performed data collection and analysis. KT, WO, BC, and AS wrote the first draft of the manuscript. All authors interpreted the data analyses, critically revised the manuscript for important intellectual content, and read and approved the final version. All authors gave final approval for the version to be published. KT is the first author, and BC and AS provided supervision for the study as co-senior authors.
Funding
This work was supported by the National Institutes of Health (NIH) grants K08DK131286 awarded to A.S., R01DK108803, U01HG007278, U01HG009610, and U01DK116100 awarded to G.N.N.
Competing interests
G.N.N. is a founder of Renalytix, Pensieve, Verici and provides consultancy services to AstraZeneca, Reata, Renalytix, Siemens Healthineer and Variant Bio, serves a scientific advisory board member for Renalytix and Pensieve. He also has equity in Renalytix, Pensieve and Verici. L.C. is a consultant for Vifor Pharma INC and has received honorarium from Fresenius Medical Care. All other authors report no potential conflicts.
