Graf, Erin H.

Abstract Introduction/Objective Infections of prosthetic joints cause significant morbidity and often result in removal of the infected joint. In many cases, an infectious etiology can be difficult to definitively identify. As a result, providers suspecting prosthetic joint infections (PJI) often place large order sets of microbiology tests, including fungal and mycobacterial cultures. Fungal and mycobacterial cultures are labor-intensive for the laboratory taking up to 28 days and 42 days, respectively, to complete. As a potential area for a diagnostic stewardship intervention, we wanted to evaluate how often a fungal or mycobacterial culture produced actionable information in cases of suspected PJI. Methods/Case Report Data reports were retrieved from both the Electronic Health Record (EHR) and the Laboratory Information System (LIS) from May 2022 to May 2023. As sources submitted by the provider often do not specify that the joint is prosthetic, data from LIS was evaluated to identify microbiology culture test sets ordered on joint sources, while data from EHR was examined to isolate patients with implanted joints. Using Microsoft Excel, data from both systems was compared in order to match microbiology culture test sets to the known implanted joints. LIS data from implanted joint sources was assessed to record negative (no growth) and positive (growth) cultures, and to register all organisms isolated. Results (if a Case Study enter NA) Of 794 order sets for cultures collected from prosthetic joint sources, 430 included requests for fungal culture and 79 included requests for mycobacterial culture. Of these 79 requests for mycobacterial culture, 0 organisms were isolated (0%). Of these 430 requests for fungal culture, 7 grew yeast (1.6%) and 7 out of 7 (100%) of these yeasts were also isolated on general (aerobic bacterial) culture. No molds were isolated. Conclusion Discontinuing the routine practice of ordering fungal and mycobacterial culture for prosthetic joint sources would have no negative impact on patient care but would reduce unnecessary work in the microbiology laboratory and potentially avoid additional costs to the patient. Education to the orthopedic provider group as well as best practice alerts in the EHR will be pursued, based on the findings from this study, in order to reduce unnecessary culture orders.

Abstract Metagenomic sequencing of plasma has been advertised by Karius, Inc. as a way to diagnose a variety of infectious syndromes. Due to the lack of robust evidence of clinical utility, our laboratory began actively stewarding Karius testing. Microbiology Directors recommended cancelation of Karius orders when certain criteria were identified. We set out to review Karius test requests in a 52-month period of stewardship, during which we recommended cancellation on 21 of 57 orders (37%). Of Karius tests sent on samples with negative conventional testing, only 3 (7%) had positive results for Karius with plausible explanatory etiologies. Of these three cases, two were empirically covered for the positive finding without improvement and one case was never treated. Twelve (29%) had positive results that were noted by infectious diseases (ID) to reflect insignificant detections. Given the 4-fold higher detection of insignificant Karius results, we set out to systematically analyze the literature for the experience of insignificant detections at other centers. When we compared studies that included healthy controls or had clinical adjudication of positive Karius findings by ID physicians, we found a median of 17.5% of individual patients that had positive insignificant detections of potential pathogenic bacteria or fungi. The most frequently detected species were as likely to be clinically adjudicated to be insignificant as they were to be significant within the same studies. Overall, these findings highlight limited utility of Karius testing and a need for careful stewardship, not only to ensure it is sent on patients who may benefit, but also to ensure results of potential pathogens are interpreted cautiously.

Abstract Background Automated plating instruments have been shown to improve quality in Microbiology laboratories in a number of ways including consistent streaking patterns that provide better isolation of bacterial colonies compared to manual methods. The time per culture set up and automated camera imaging is dependent upon the amount of media set, such that more individual plates slow down turn around time. Conversely, smaller media surface area and the use of non-selective media may result in mixed colonies that require sub-culture for isolation before identification and susceptibilities can be performed. Based upon these findings, we attempted to reduce our media usage per culture with the implementation of automation, with the goal of using as few plates as possible while at the same time not negatively impacting the turn around time of the culture via the need to subculture poorly isolated colonies. We conducted a pre-and post-implementation study evaluating the costs per culture and turn around time delays related to subculturing in a subset. Methods In the pre-implementation period, from April-January 2023, three individual media plates (sheep blood agar, chocolate agar and CNA/MacConkey bi-plate) were routinely set for general aerobic bacterial culture sources including swabs and body fluids. In the post implementation period, from April-January 2024, a single sheep blood/chocolate agar bi-plate was set unless the gram stain showed evidence of mixed organisms (defined as more than one morphology) at which time a secondary selective media (CNA/MacConkey) was manually set. For urine cultures, in the pre-implementation period 2 plates were set (sheep blood and CNA/MacConkey bi-plate) and in the post-implementation period a single chromogenic agar/sheep blood bi-plate was set. Total numbers of cultures in both periods were collected as well as the number of cultures for which additional media were set based on gram stain results. Orders of subculture media in the laboratory information system were compared per positive cultures. Statistical analysis was performed in GraphPad Prism. Results In the pre- versus post-implementation periods, a total of 16,454 vs 17,782 general aerobic cultures and 50,680 vs 53,916 urine cultures were set, respectively. Only 3.6% of general cultures required additional media set based on gram stain. The difference in cost between pre- and post-implementation was approximately a two-thirds reduction in cost for general cultures and a 25% reduction for urine cultures, with a total savings of approximately $34,238 in media avoidance in the post-implementation period. The frequency of positive cultures requiring subculture was not significantly different between the pre- (average 36% of positive growth) and post-implementation (average 22% of positive growth) periods for both culture types and may actually have been reduced due to automation. Conclusions The implementation of reduced media concomitant with streaking via automation resulted in significant cost savings while not negatively impacting isolation, which in turn impacts culture turn around times. The automation may have reduced need for subculture, perhaps due to better isolation, thus subsequent analyses will explore this additional finding.

Abstract Background Several studies have shown that the implementation of automation in microbiology laboratories may allow for increased volume growth without the need for additional staffing. One of the more recently available pieces of microbiology automation, the COPAN Colibri, automates the manual steps of colony spotting and preparation of targets for use on MALDI-TOF instruments for bacterial identification. This same instrument can also prepare a diluted bacterial inoculum for immediate use on automated susceptibility testing platforms. We wanted to evaluate the performance of both of these functions in real-time with clinical cultures, to determine if implementation would be beneficial over current workflows. Methods The COPAN Colibri was operated following the manufacturer’s recommendations with the exception of the identification of gram-positive organisms. The Colibri is FDA cleared for MALDI-TOF target spotting, however not with the use of formic acid, nor is it FDA cleared for susceptibility suspension preparation. Preliminary data suggested that addition of formic acid was necessary for accurate identification. A subset of cultures were digitally read in real-time as part of clinical workflows, selected for Colibri MALDI-TOF target spotting and Vitek2 suspension preparation using the COPAN WebApp software. Targets were run on the MALDI Bioptyper Sirius (Bruker) and suspensions were run on Vitek2 (BioMerieux) using the pre-dilution mode using GN-99 and GN-801 cards for appropriate organisms. All results were compared to standard manual workflows. Results A total of 145 gram-negative and 60 gram-positive organisms were picked for MALDI-target spotting on Colibri. Compared to standard testing, 144 (99%) gram-negative and 57 (95%) gram-positive Colibri spotted isolates matched expected species-level identifications. 97 Gram negative bacilli were run on Vitek 2 generating 95.52% categorical agreement across all drugs but with 13 major errors (MEs) and 2 very major errors (VMEs), which exceeded the acceptable percentage for certain drugs. The total number of failed runs was 13 (13.34% of total GNB isolates run), and all required complete resetting of MALDI-spotting and suspension preparation. Conclusions The Colibri performed well for MALDI-TOF target spotting with the addition of formic acid for gram-positive organisms. It did not perform within the acceptable range for susceptibility suspension preparation as it exceeded the recommended cutoffs for MEs and VMEs for certain drugs. The proportion of failed runs due to instrument errors may offset any advantages in labor savings due to the need for repeat set up.

AbstractCentral venous catheters are among the most used medical devices in hospitals today. Despite advances in modern medicine, catheter infections remain prevalent, causing significant morbidity and mortality worldwide. Here, SteriGel is reported, which is a multifunctional hydrogel engineered to prevent and treat central line‐associated bloodstream infections (CLABSI). The mechanical properties of SteriGel are optimized to ensure appropriate gelation kinetics, bio‐adhesiveness, stretchability, and recoverability to promote durability upon application and to provide persistent protection against infection. In vitro assays demonstrated that SteriGel exhibits long‐term antimicrobial efficacy and has bactericidal effects against highly resistant patient‐derived pathogens known to be frequently associated with CLABSI. SteriGel outperformed Biopatch, which is a clinically used device for CLABSI, in ex vivo cadaver studies that simulate clinical scenarios. Furthermore, SteriGel has biocompatible, pro‐healing, and anti‐inflammatory properties in vitro and in a rat subcutaneous injection model, suggesting a potential synergistic effect in the prevention and treatment of CLABSI. SteriGel is a multifunctional adherent biomaterial with potent antimicrobial effects for sustained sterility while promoting healing of the catheter incision site to protect against infection.

Abstract Objectives This review summarizes the current and potential uses of artificial intelligence (AI) in the current state of clinical microbiology with a focus on replacement of labor-intensive tasks. Methods A search was conducted on PubMed using the key terms clinical microbiology and artificial intelligence. Studies were reviewed for relevance to clinical microbiology, current diagnostic techniques, and potential advantages of AI in routine microbiology workflows. Results Numerous studies highlight potential labor, as well as diagnostic accuracy, benefits to the implementation of AI for slide-based and macroscopic digital image analyses. These range from Gram stain interpretation to categorization and quantitation of culture growth. Conclusions Artificial intelligence applications in clinical microbiology significantly enhance diagnostic accuracy and efficiency, offering promising solutions to labor-intensive tasks and staffing shortages. More research efforts and US Food and Drug Administration clearance are still required to fully incorporate these AI applications into routine clinical laboratory practices.

AbstractPersistent or recurrent bleeding from microvessels inaccessible for direct endovascular intervention is a major problem in medicine today. Here, an innovative catheter‐directed liquid embolic (P‐LE) is bioengineered for rapid microvessel embolization to treat small vessel hemorrhage. Tested in rodent, porcine, and canine animal models under normal and coagulopathic conditions, P‐LE outperformed clinically used embolic materials in both survival and non‐survival experiments, effectively occluding vessels as small as 40 microns with no signs of recanalization. P‐LE occlusion is independent of the coagulation cascade, and its resistance to displacement is ≈ 8 times greater than systolic blood pressure. P‐LE is also found to be biocompatible and x‐ray visible and does not require polymerization or a chemical reaction to embolize. To simulate the clinical scenario, acute microvascular hemorrhage is created in the pig kidney, liver, or stomach; these are successfully treated with P‐LE achieving immediate hemostasis. Furthermore, P‐LE is found to be bactericidal to highly resistant patient‐derived bacteria, suggesting that P‐LE may also protect against infectious complications that may occur following embolization procedures. P‐LE is safe, easy to use, and effective in treating ‐microvessel hemorrhage.

Benign prostatic hyperplasia and prostate cancer are often associated with lower urinary tract symptoms, which can severely affect patient quality of life. To address this challenge, we developed and optimized an injectable compound, prostate ablation and drug delivery agent (PADA), for percutaneous prostate tissue ablation and concurrently delivered therapeutic agents. PADA is an ionic liquid composed of choline and geranic acid mixed with anticancer therapeutics and a contrast agent. The PADA formulation was optimized for mechanical properties compatible with hand injection, diffusion capability, cytotoxicity against prostate cells, and visibility of an x-ray contrast agent. PADA also exhibited antibacterial properties against highly resistant clinically isolated bacteria in vitro. Ultrasound-guided injection, dispersion of PADA in the tissue, and tissue ablation were tested ex vivo in healthy porcine, canine, and human prostates and in freshly resected human tumors. In vivo testing was conducted in a murine subcutaneous tumor model and in the canine prostate. In all models, PADA decreased the number of viable cells in the region of dispersion and supported the delivery of nivolumab throughout a portion of the tissue. In canine survival experiments, there were no adverse events and no impact on urination. The injection approach was easy to perform under ultrasound guidance and produced a localized effect with a favorable safety profile. These findings suggest that PADA is a promising therapeutic prostate ablation strategy to treat lower urinary tract symptoms.