We partner with a secure submission system to handle manuscript submissions.
Please note:
Click Proceed to submission system to continue to our partner's website.
Background: The 2025 Measles outbreak has affected over 1900 people across the United States. Close collaborations between public health entities and training institutions can increase public knowledge and trust, promote vaccination, optimize options for testing, and reduce exposures. Methods: The Lexington Fayette County Health Department (LFCHD), Kentucky Department for Public Health (KDPH), and the University of Kentucky (UKY) partnered in February 2025 to develop a comprehensive plan to address measles in Kentucky. Priorities for the partnership were: community and provider education, delineation of roles/responsibilities between the health department and UKY, streamlined pathways for testing and post-exposure management, and centralized communication. Standardized educational sessions were developed and deployed widely to outline testing and post-exposure prophylaxis (PEP) resources. Infection Prevention and Control (IPAC) physicians triaged all testing requests in partnership with the health department using CDC criteria for testing. Measles, Mumps, and Rubella (MMR) vaccination campaigns were conducted, focusing on prevention and post-exposure management. For healthcare workers (HCW) at UKY, Epic® electronic health record (EHR) templates and order sets outlining CDC and KDPH algorithms were developed and distributed widely. Results: From March through October 2025, the standardized educational talk was given to 398 public health and medical professionals (Table 1). Six public health and healthcare systems adapted the public health-academic center partnership and protocol to their centers (Table 2). Joint media outreach and education resulted in 32 stories across Kentucky and Tennessee. Three Epic® EHR templates were developed and deployed widely (Figures 1-3). IPAC received and triaged approximately 50 calls/EMR messages from HCW about measles testing, resulting in 4 tests performed at the KDPH State Lab. One test returned positive and there were zero healthcare worker exposures. Immune globulin for PEP was offered to 2 exposed patients, though they declined to receive it (Table 2). MMR vaccinations increased by 16% and 53% in the 0-6 year and 6–18-year age groups, respectively from 2024 to 2025, during the same time periods. MMR vaccination in persons over 18 years decreased slightly (4.5%) (Figure 4). Conclusions: Combining the infrastructure, knowledge, and resources of a state and local public health systems with an academic medical center’s training and community trust can increase measles awareness, vaccination, and optimize resources during an outbreak.
Background: Respiratory Syncytial Virus (RSV) is a leading cause of infant hospitalizations in the United States and is particularly devastating to infants 6 months and younger. Maternal RSV vaccination given more than two weeks before delivery and early infant nirsevimab administration can provide protection against severe illness in this vulnerable population. Objective: To evaluate uptake of, and barriers to, RSV immunization among infants admitted to the hospital with RSV infection in a large healthcare system. Additionally, clinical outcomes between infants who did and did not receive any RSV protection were compared. Methods: This was a retrospective cohort of infants ? 6 months old hospitalized in a large multi-state healthcare system from November 2024 to January 2025 with a positive RSV test within 7 days prior to or during admission. Patients were identified using an electronic health record report and chart review was performed to obtain data on RSV immunization and clinical outcomes. Infants born outside of our healthcare system where information on RSV immunization was lacking were excluded. Infants were classified as protected if either 1) maternal RSV vaccination was given more than two weeks before delivery or 2) the infant received nirsevimab prior to the onset of RSV infection. All others were classified as unprotected. Clinical outcomes including length of stay, ICU admission, and mechanical ventilation were compared between groups. Identification of barriers to RSV immunization were attempted. Result: Of 152 infants admitted with RSV, 103 were born in our hospitals and included. 17 (16.5%) patients were considered protected (8 with maternal vaccination; 9 infant nirsevimab administrations). Patients in both groups were predominantly male and either Non-Hispanic White or Hispanic. Compared to unprotected patients, protected patients had significantly fewer ICU admissions (0% vs 28%, p = 0.03) and shorter median length of stay (1 vs 2 days, p=0.004). There was no difference between groups in rates of mechanical ventilation or death. Among unprotected patients, chart review most often showed no documentation of any RSV immunization discussions (75.6%), family declined treatment (12.8%), and infants becoming ill before nirsevimab could be administered despite willingness to eventually receive it (11.6%). Post-RSV infection, an additional 12% of unprotected patients received nirsevimab. Conclusion: Unprotected infants were more likely to experience negative clinical outcomes, specifically more ICU admissions and longer hospitalizations. Opportunities exist to improve RSV provider and family education and implement interventions to facilitate discussion and administration of RSV prophylaxis.
Background: Blood culture contamination leads to unnecessary clinical interventions and increased healthcare costs. Estimations indicate each contamination incurs over $4,500 in avoidable cost. The 2024 IDSA/ASM guideline advises blood culture collection via peripheral venipuncture and cultures from existing vascular catheters should be reserved for suspected Central line associated bloodstream infection (CLABSI) and paired with a peripheral set. It was observed that routine ordering and collection of blood cultures from central venous catheters (CVC) was common practice at our institution and our blood culture contamination rate was above the recommended 1% goal. A multidisciplinary quality improvement plan was implemented to reduce inappropriate catheter blood culture collection. Methods: A pre-post intervention study was conducted at a large quaternary medical center from January 1, 2023 to December 31, 2024 (pre-intervention) and January 1, 2025 to December 31, 2025 (post-intervention). The intervention included simultaneous updates to 1) provider clinical practice guidelines for blood culture indications, 2) nursing policies on culture collection techniques and 3) nutrition policies regarding surveillance cultures from CVCs prior to initiation of total parenteral nutrition. Education was provided and reinforced at multidisciplinary quality rounds as well as various meetings, huddles and lectures. Blood culture contamination was defined using CLSI criteria and rates and collection sites were compared pre and post intervention. CLABSI was determined using National Healthcare Safety Network (NHSN) definitions. CLABSI rates and blood culture collection sites were compared pre and post-intervention. Results: During the study period, 185,030 sets of blood cultures were collected. The number of blood cultures collected from a CVC significantly decreased (Table 1) and the contamination rates of cultures significantly decreased from both peripheral venipuncture and CVC. By reducing collection from lines, 29 estimated contaminations were avoided, with an associated cost savings of $131,602 and a reduction of 203 antimicrobial days. There was a significant decrease in CLABSI rate in the post-intervention period (Figure 1). There were significantly fewer blood cultures drawn from CVCs in patients with NHSN CLABSI in the post-intervention period (70% vs 50%, p = 0.0005). Conclusions: The simultaneous policy change and multifaceted reinforcement with strong senior leadership support resulted in rapid adoption and sustained change in blood culture ordering and collection practices. There was a significant decrease in the proportion of blood cultures drawn from CVCs, blood culture contamination and CLABSI rate in the post-intervention period leading to significant cost savings, fewer antimicrobial days and improved patient outcomes.
Introduction: The American College of Obstetricians and Gynecologists (ACOG) and Society of Healthcare epidemiology (SHEA) recommend cephalosporins as the standard antibiotic for preoperative prophylaxis for abdominal hysterectomy (AH-includes Laparoscopic and open cases) . However, anaerobic organisms are prevalent in vaginal and intraabdominal flora and can contribute to development of surgical site infections (SSIs). Limited data suggest that including anaerobic coverage may reduce SSI risk. In May 2024, we implemented anaerobic containing regimens as first-line preoperative antibiotic prophylaxis for all AH cases. This study evaluates whether adding anaerobic coverage to the preoperative antibiotic protocol reduces National Healthcare Safety Network (NHSN)-defined SSI after AH surgeries. Methods: We conducted a retrospective cohort study of all patients who underwent inpatient or outpatient from January 2023 through September 2025. Preoperative antibiotic choice and timing as well as patient demographics and comorbidities were obtained from medical records. SSI were determined by NHSN definitions by trained infection preventionists. First line antibiotics with anaerobic coverage were defined as cefazolin plus metronidazole, or vancomycin plus aztreonam plus metronidazole in patients with penicillin allergies. SSI in patients who received first line antibiotics with anaerobic coverage were compared to those who received alternative antibiotic regimens (Around 80% of alternative regimen consisted of only cefazolin). Multivariate logistic regression and overlap weighting methods were applied to evaluate the independent impact of first-line therapy. Results: Among 11,444 hysterectomy procedures, the overall SSI rate was 1% (n=109 cases). Unadjusted SSI rate in patients receiving first-line antibiotics was significantly lower than those receiving alternate prophylaxis (0.7% vs 1.3 %, p =0.001). Individuals who received first-line antibiotics were generally younger, more frequently identified as Caucasian, and exhibited greater rates of diabetes and cancer compared to those who received alternative regimens. Patients who received first line antibiotics also had longer surgeries, a higher frequency of elective and laparoscopic procedures, and higher rates of adherence to preoperative protocols such as chlorhexidine bathing, vaginal preparation, and maintenance of normothermia (see Table 1). After adjusting for baseline differences between groups through multivariate logistic regression and overlap weighting, first-line therapy was associated with a 50% reduction in SSI risk compared to alternative regimens (OR 0.51; 95% CI 0.34–0.76; p=0.001). (Table 2) Conclusion: Compared to alternative regimens for preoperative antibiotic prophylaxis, the routine inclusion of metronidazole to cefazolin was associated with a reduced risk of SSI following AH surgeries even after adjusting for demographic, procedural and clinical differences between groups.
Background: Electronic surveillance for hospital-onset sepsis using CDC’s Adult Sepsis Event definition could provide an efficient and objective method to identify a broad array of serious healthcare-associated infections, many of which are missed through current reporting processes. We developed risk adjustment models of varying complexity to support facility-level comparison of hospital-onset sepsis rates, evaluated trade-offs between model performance and feasibility, and quantified residual inter-facility variation that may reflect gaps in care. Methods: We conducted a retrospective study of adults hospitalized for <3 days within 113 community hospitals between 2022-2023. Hospital-onset Adult Sepsis Events (HO-ASEs) occurring on day 4 or later were identified using updated CDC surveillance criteria. We used logistic regression to develop three risk adjustment models of increasing complexity using covariates from administrative and electronic health record data: basic model (hospital and aggregate patient descriptors), intermediate model (replacing aggregate patient descriptors with patient-level descriptors), and maximal model (adding detailed physiologic and clinical data from hospital days 1-3; Figure 1). We evaluated model performance using Area Under Receiver Operating Curve (AUROC), calculated hospital-level Standardized Infection Ratios (SIRs) for each model and assessed concordance in hospital rankings using Kendall’s tau coefficient (τ). Results: The cohort included 1,557,252 hospitalizations of <3 days, of which 24,169 (1.6%) met HO-ASE criteria and 8,500 (35.2%) died in-hospital. The basic model had limited discrimination (AUROC 0.589, 95% CI, 0.585-0.593). Adding patient-level characteristics to form the intermediate model markedly improved performance (AUROC 0.839, 95% CI 0.836-0.841), with further inclusion of detailed clinical data in the maximal model yielding modest additional improvement (AUROC 0.850, 95% CI 0.848-0.853). Concordance between hospital rankings derived from the crude or basic risk-adjusted HO-ASE rates versus rankings derived from the intermediate or maximal models was moderate (τ 0.40-0.51) whereas concordance between rankings derived from the intermediate vs. maximal models was high (τ 0.86). There was a wide distribution of HO-ASE SIRs across facilities even after risk adjustment using the maximal model (Figure 2), with high signal-to-noise ratios and good calibration. Conclusions: Risk adjustment models incorporating hospital characteristics and patient-level data perform well and might explain substantial variability in HO-ASE rates between facilities. The persistence of residual variability after highly detailed adjustment may reflect differences in care processes, suggesting that risk-adjusted HO-ASE comparisons can help identify gaps and opportunities in the prevention of severe healthcare-associated infections. Our findings support the use of HO-ASE as an electronic, scalable, risk-adjusted metric for facility-level benchmarking to inform quality improvement initiatives.
Background: Current US healthcare-associated infection (HAI) surveillance misses many serious nosocomial infections. Electronic identification of hospital-onset sepsis using CDC Adult Sepsis Event (ASE) criteria could expand the breadth and efficiency of HAI surveillance, but little is known about the preventability of these events. Methods: We identified hospital-onset sepsis events on hospital day ?4 using updated ASE criteria amongst adults admitted to 9 hospitals in Massachusetts, 2021-2023. We randomly selected 224 sepsis events to adjudicate infection probability and source (50% academic, 50% community hospitals) and then analyzed 100 confirmed infections for preventability using a 6-point Likert scale. Two physicians reviewed cases for infection source, following joint review and discussion of the first 25 cases with two senior physicians to standardize approach. All preventability assessments were classified by consensus amongst three physicians. Results: Amongst 224 hospital-onset sepsis events, 100 (45%) had definite or probable infection, 82 (37%) had possible infection, and 42 (19%) had no infection. Most sepsis events were due to respiratory (47%), bloodstream (16%), abdominal (13%) and urinary (9%) infections (Figure 1). Only 11 events were CMS-reportable NHSN HAIs (mostly central-line associated bloodstream infections and catheter-associated urinary tract infections). Crude inpatient mortality for hospital-onset sepsis cases was 35%. Amongst 100 cases assessed for preventability, 30% were judged potentially preventable (Figure 2), most commonly due to lapses in managing established infections (41%) including delays in source control, microbiologic testing, antimicrobial administration, imaging, and subspecialty consultation. Gaps in nosocomial respiratory infection prevention measures were also common (36%) including insufficient oral care, lack of dysphagia management, and failure to mask (Figure 3). Lapses in traditional device-related HAI prevention bundles were uncommon (7%). Some hospital-onset sepsis events may have been preventable with better general medical care (17%), including earlier recognition and management of non-infectious contributors to physiologic deterioration (e.g., earlier diuresis to optimize respiratory reserve among patients at risk for aspiration, proactive bowel management to prevent stercoral colitis and downstream infectious complications). Conclusions: Hospital-onset sepsis events identify serious, often fatal infections, many of which are potentially preventable. Opportunities extend beyond current HAI prevention bundles and include the need for more timely recognition and management of established infections to prevent progression to sepsis, introducing measures to prevent nosocomial respiratory infections, and improving general inpatient medical care. Incorporating hospital-onset sepsis surveillance into infection prevention and quality improvement programs could help identify actionable steps to improve the safety and quality of inpatient care.
Background: Microbiome disruption is linked to risk of mortality and infection. Measuring the impact of patient-level antibiotic exposure to microbiome disruption is needed to inform stewardship and microbiome restoration therapy efforts. The National Healthcare Safety Network (NHSN) standard, days of therapy (DOT), widely used to capture antibiotic use, does not account for spectrum of activity. Other metrics better account for spectrum, including antibiotic spectrum coverage (ASC), antibiotic spectrum index (ASI), and anaerobic activity index (AAI) – and with NHSN antibiotic groups (Clostridioides difficile infection [CDI] high risk, and broad-spectrum hospital-onset infections [BSHO]). We compared DOT with spectrum-weighted metrics to determine which best track patient-level microbiome disruption. Method: We retrospectively analyzed peri-rectal swab metagenomic data and antibiotic use for acute and long-term care facility patients. Shannon diversity index, pathogen abundance, and butyrate-producing bacterial abundance were calculated for each participant. Antibiotic exposure within 30 and 90 days prior to sampling were summarized six ways: crude duration – DOT for all antibiotics, NHSN-CDI antibiotics, and NHSN-BSHO antibiotics (all unweighted) and three weighted DOT values (sum of weights per day for each ASC, ASI, and AAI) across all days in the exposure window, accounting for antibiotic spectrum. Spearman correlations were calculated among all six antibiotic metrics and between the metrics and microbiome disruption features. Result: One hundred participants were included with median 30-day and 90-day DOT of 16 (IQR:6-31) and 27 (IQR: 13-64), respectively. Weighted metrics (ASC, ASI, AAI) demonstrated strong correlations with each other (Figure 1, darker orange). Crude duration metrics (DOT, CDI, BSHO) were only moderately correlated with weighted metrics (Figure 1, lighter orange). Shannon diversity was correlated with all exposure metrics except the NHSN CDI high-risk group, and AAI had the strongest association (30-day r=-0.41, p Conclusion: Patient-specific weighted DOT of antibiotic exposure, in particular the AAI weights, correlates with gut microbiome disruption well. Interestingly, the NHSN high-CDI-risk antibiotic group did not. Patient-level 30-day exposure had similar observed correlations as 90-day exposure windows and may be a practical window to measure microbiome antibiotic effects. Further study is needed to test causality of identified associations, the influence of absolute vs relative abundance measures, and best ways to incorporate these measures into stewardship efforts and identification of microbiota therapy candidates.
Background: With increasing utilization of oral antibiotics in serious infections, complex outpatient antimicrobial therapy (COpAT) programs are vital for improving patient safety and outcomes. We discuss the Michigan Medicine (MM) COpAT program, including processes, interventions, and evaluate its impact. Methods: Adult patients discharged from MM hospital between Jan-Oct 2025 were eligible for COpAT program enrollment if prescribed antibiotics for ≥14 days, including oral antibiotics requiring lab monitoring, long-acting lipoglycopeptides, or intravenous (IV) antibiotics administered at a hemodialysis center, subacute nursing facility, or home. Eligible diagnoses include osteomyelitis, prosthetic joint infection, septic joint, spinal hardware infection and infectious tenosynovitis. Established in October 2024, the COpAT program consists of an infectious diseases (ID) pharmacist with ID physician oversight. The pharmacist conducts a telephone-encounter within 2 business days of discharge to review antibiotic plan, lab monitoring, appropriate antibiotic administration, duration and adherence, and assess for adverse effects. The outpatient ID physician is notified of treatment concerns and lab monitoring results. Another telephone-encounter occurs 2 weeks post-discharge to re-evaluate antibiotic therapy. COpAT patient outcomes and process measures were compared to a historical cohort of patients discharged from MM on oral or IV antibiotics for an osteoarticular infection from May – August 2023. Chi-squared tests were used for categorical variables; the Mann-Whitney U test was used to compare median length of stay (LOS). Results: From January-October 2025, 163 patients with osteoarticular infections were enrolled in COpAT post-discharge. There were 124 patients in the historical cohort. Among COpAT patients, 79.8% (130/163) were discharged on oral antibiotics compared to 33.9% (42/124) in the historical cohort (p<0.0001, Figure 1). COpAT patients had a shorter median LOS of 6.1 days [IQR 4.2-8.7] versus 8.2 days [IQR 5.3-12.1, p<0.0001], and numerical reduction of 30-day readmission rate of 12% (19/158) versus 16.9% (22/132) in the historical cohort (p=0.258). In a subset of 85 patients enrolled from 1/1– 5/31/25, 96.5% had a COpAT telephone-encounter within 2 business days and 638 pharmacist interventions occurred with an average 7.4 interventions/patient (Figure 2). Conclusion: COpAT patients had significantly increased oral antibiotic utilization and shorter LOS, and numerically fewer readmissions compared to a historic cohort. The increase in oral and long-acting antibiotics, rapid post-discharge follow-up and multiple COpAT pharmacist interventions likely contributed to improved outcomes and safer transitions of care. COpAT programs should be instituted to improve patient outcomes and ensure safe use of oral and IV antibiotics in serious infections.
Background: Previously published studies, including our own, have demonstrated that empiric antibiotic therapies with activity against methicillin-resistant Staphylococcus aureus (MRSA) and resistant gram-negative (RGN) bacteria are often unnecessarily prescribed for patients hospitalized with diabetic foot infections (DFI) and lower extremity osteomyelitis (OM). We implemented a multidisciplinary intervention based on local and national data to optimize antibiotic prescribing for these patients, including development of a “diabetic foot algorithm,” stakeholder education, and creation of an infectious diseases bone and joint consult service. We evaluated changes in antibiotic use and clinical outcomes after implementation of this intervention. Methods: This was a “before and after study” of all patients hospitalized with DFI and/or lower extremity OM in our hospital system in 2021 and 2025. Patients were included if they had an International Classification of Disease, Tenth Revision (ICD-10) diagnosis code of M86, E10.621, E11.621, or E08.621. Patients were excluded if antibiotics were for a different indication or if they were less than 18 years of age. Empiric antibiotic therapy included antibiotics started by the admitting team. Definitive antibiotic therapy included the final antibiotic course either completed during admission or prescribed at the time of discharge. The intervention began in October 2023. The primary outcomes were changes in use of empiric and definitive antibiotic therapies with activity against MRSA and RGN bacteria. The secondary outcomes were changes in amputation rates and hospital length of stay (LOS). Results: There were 259 and 242 unique patients hospitalized with DFI or lower extremity OM who met inclusion criteria in 2021 and 2025, respectively. Comparing 2021 to 2025, the percentage of patients receiving empiric therapies with activity against MRSA decreased from 87% to 70% and against RGN bacteria decreased from 84% to 68% (p<0.001) (Figure 1). There was also a trend in the decreased use of definitive therapies with these spectra of activity, while the prevalence of resistant organisms in cultures remained similar. Inpatient days of antibiotic therapy with activity against MRSA and RGN bacteria also decreased 16% and 11%, respectively. There was no difference in clinical outcomes with 44% of patients in both 2021 and 2025 undergoing inpatient amputation surgeries (p=0.997). There was a significant decrease in hospital LOS (8 vs. 6 days, p=0.002). Conclusions: A multidisciplinary, multifaceted intervention was associated with significant decreases in use of antibiotic therapies with activity against MRSA and RGN bacteria with no noted adverse effects on clinical outcomes.
Background: Admission to a room previously occupied by a patient with Clostridioides difficile infection (CDI) has been identified as a risk for CDI. However, previous studies have not included molecular typing to definitively link healthcare-associated CDI (HA-CDI) cases to prior room occupants or to residual spore contamination on surfaces. Methods: In an acute care hospital, we conducted a 1-year cohort study to determine the proportion of HA-CDI cases linked to prior room occupants with CDI or to room surfaces remaining contaminated after post-discharge cleaning and disinfection. Cultures were collected from post-discharge CDI and non-CDI rooms. Whole genome sequencing was used to determine relatedness of isolates. We calculated the percentage of HA-CDI cases infected with isolates genomically related to prior room or ward-level exposures. Results: Of 5,746 patients admitted, 22 had community-associated CDI and 55 were diagnosed with HA-CDI. C. difficile was recovered from 79 of 327 (21%) post-discharge rooms, including 14 of 36 (39%) CDI rooms and 83 of 287 (29%) non-CDI rooms. Of 1,773 patients with room-level exposure to prior CDI patients or contaminated surfaces, 21 (1%) developed HA-CDI, but none were infected with genomically-related isolates (Figure 1.A). Of 49 patients developing HA-CDI after ward-level exposures, 3 were infected with isolates genomically related to prior CDI patients or contaminated surfaces on the ward (1.B). Conclusion: Despite frequent exposure to rooms previously occupied by patients with CDI or contaminated with C. difficile, no HA-CDI cases were linked to prior room exposures based on whole genome sequencing.
Background: The safety of midlines (MLs) versus peripherally inserted central catheters (PICCs) in the context of outpatient parenteral antimicrobial therapy (OPAT) is still debated. The Infectious Diseases Society of America guidelines currently offer a weak recommendation to consider MLs for antibiotic durations of <14 days. The purpose of this study was to quantify the ML complication rate compared to PICCs for OPAT. Methods: We performed a single-center, retrospective cohort study of MLs and PICCs placed by the bedside vascular access team for OPAT as recommended by the infectious diseases consult service at a 670-bed urban academic medical center. Between July 2019 and June 2022, 1351 catheters were placed in 1252 unique patients for OPAT. The primary exposure was the placement of a ML versus PICC, informed by selection guidelines from the Infusion Nurses Society. The primary outcome was any bloodstream infection (BSI) or superficial or deep venous thromboembolism (VTE) while the catheter was present. We estimated the hazard ratio (HR) of developing the outcome among MLs compared to PICCs using Cox proportional hazards models, accounting for multiple episodes of OPAT-related catheter placement over time in the same patient and adjusting for confounding by planned antibiotic duration of ≥14 days and the use of vancomycin. We tested whether planned antibiotic duration modified the association between catheter type and the outcome using an interaction term. Results Demographic and clinical characteristics are in Table 1. The ML complication rate was 6.1 per 1000 line-days (26 [3 BSI, 23 VTE] of 322 MLs) versus 1.2 per 1000 line-days for PICCs (40 [11 BSI, 30 VTE] of 1029 PICCs), with a higher complication rate for MLs (adjusted HR [aHR] 2.97, 95% confidence interval [CI] 1.32-6.70, p=0.01). Adjusted survival curves are given in Figure 1. The aHR was 0.97 (95% CI 0.30-3.11, p=0.96) among catheters with a planned antibiotic duration of ≤14 days and 4.11 (95% CI 2.08-8.16, p<0.01) for <14 days (p-interaction=0.03). Conclusion In this study of patients receiving OPAT, complications were uncommon but significantly more frequent for MLs than PICCs. This association appeared to differ by planned antibiotic durations of <14 days than ≤14 days. Our study calls into question the safety of MLs for OPAT, especially for prolonged durations of therapy.
Background: Community-acquired pneumonia (CAP), a heterogeneous clinical entity with a wide range of inciting pathogens, remains a significant cause of morbidity and mortality worldwide. The recent COVID-19 pandemic and subsequent shifts in clinical and diagnostic practices necessitate an updated understanding of the contemporary epidemiologic and etiologic landscape for non-COVID CAP. Method: This retrospective cohort study included adult patients hospitalized with CAP between January 1, 2017, and December 31, 2024, across 68 hospitals participating in the Michigan Hospital Medicine Safety consortium. Eligible patients included hospitalized adults with a pneumonia discharge code, symptoms and imaging consistent with pneumonia, and receipt of antibiotic treatment on hospital day 1 or 2. Patients admitted to intensive care, those with concomitant infections, positive SARS-CoV-2 tests, or severe immunosuppression were excluded. Diagnostic testing methods and the pathogens detected were reviewed and summarized by year. Result: A total of 62,340 patients were included. Mean age was 69.0 ±15.9 years, and 50.3% (31,359) were male. 94.0% (58,613) received a chest X-ray and 53.1% (33,084) received a CT, with use of CTs increasing over time. Microbiological evaluation was performed in 92.8% (57,820) of patients, with a median 4 [IQR 2-6] etiologic tests per patient. Blood cultures were the most common microbiologic test (74.7%; 46,566) with use slightly decreasing over the study period. In contrast, the use of molecular respiratory pathogen assays increased substantially from 28.2% in 2017 to 77.1% in 2024 (overall performed in 55.6%, 34,655). Over the entire study period, one or more pathogens were detected in 15.9% (9,915) of patients: 10.5% (6,515) had a bacterial pathogen, 6.3% (3,908) had a viral pathogen, and 0.8% (508) had both. The most common pathogens were Streptococcus pneumoniae (2.6%), influenza virus (2.6%), Pseudomonas aeruginosa (1.6%), and human rhinovirus (1.4%). Conclusion: The 2019 CAP guideline recommendations to limit traditional bacterial tests to severe CAP patients appear to have had minimal impact on diagnostic practices. Despite a substantial post-pandemic increase in molecular diagnostic use, pathogen detection rates remained low, revealing an etiology in only a minority of patients. These findings underscore the need for a reassessment of the diagnostic approach to optimize resource allocation and ensure high-value testing in hospitalized CAP patients.
Background: Blood culture (BCx) algorithms can safely and effectively reduce unnecessary BCxs. We compared the effect of implementing a BCx algorithm across multiple inpatient units within our hospital system, focusing on blood culture event (BCE) rates and secondary adverse events. Method: The same BCx algorithm (Figure 1) was implemented in multiple inpatient units across three hospitals in our health system (Figure 2). BCE rates were compared between pre- and post-intervention periods with interrupted time series (ITS) analyses. Differences in secondary outcomes, including incidence of central line-associated bloodstream infections (CLABSI) and Clostridioides difficile infection (CDI), were compared using χ tests. Result: The rate of BCE decreased after implementation of the BCx algorithm across all units (Figure 3). An immediate decrease in BCE rate was noted in most units upon implementation of the algorithm, except in the cardiac ICU, cardiac floors and neurocritical care unit. No unit demonstrated a significant post-intervention decrease in the BCE rate, with the surgical ICU demonstrating a significant regression to the mean with time. Antibiotic use decreased across all analyzed units (Figure 4). Readmission rates decreased significantly on the cardiac floor, neurocritical care unit and cardiothoracic ICU in hospital 1 and also in the critical care unit in hospital 2, while it increased significantly in the cardiothoracic stepdown unit. CLABSI rates decreased significantly in the cardiac ICU and the solid oncology floor, while CDI rates decreased significantly in the neurocritical care unit. No significant increase was seen in CLABSI and CDI rates or antibiotic use. Conclusion: This study demonstrated that the BCx algorithm was effective in decreasing BCEs across all units and hospitals without any clear harms. However, the effect of BCx algorithms is partially dependent on the nature of the unit where it is implemented. The decrease in BCEs was limited to the time period directly after implementation of the algorithm. No units demonstrated ongoing decreases in BCE rates, with the surgical ICU in particular demonstrating a steady increase of BCE rates as time progressed, although rates of BCE remained below pre-intervention levels. Consistent decreases were seen in antibiotic use post-implementation, although causality between the two cannot be determined. Further research is warranted on how to implement a more durable change and on the use of such algorithms in more specific populations, including immunocompromised hosts.
