The investigators reviewed epidemiological, clinical and laboratory information from medical records and tested serum samples using a qualitative lateral-flow cryptococcal antigen assay. The sample included 63 patients, of whom 77.8% were men. Median age was 38 years, median CD4+ cell count was 33 cells/µL, with an interquartile range of 12.5–85.5, and the median logarithmic viral load was 4.95, with an interquartile range of 4.21–5.53. Among patients with CD4+ counts below 100 cells/µL, antigenemia was detected in four of 47, or 8.5%.

Six antigen-positive patients underwent lumbar puncture for cerebrospinal fluid collection. Cryptococcal meningitis was confirmed in one of them, representing 16.7% of those tested, while another patient had disseminated cryptococcal disease. Isolated cryptococcal antigenemia, without evidence of cryptococcal disease, was identified in five patients, or 7.9% of the full study population. The abstract reports that two patients died during hospitalization without evidence of cryptococcal disease. Two asymptomatic patients received fluconazole as pre-emptive treatment, and one received no antifungal drugs; none developed cryptococcal meningitis during the 12 months after discharge.

These results are limited to the small, single-hospital retrospective sample and the follow-up reported in the abstract. The study did not assess a broader population, and the absence of cryptococcal meningitis during follow-up applies only to the three antigen-positive patients whose discharge treatment was described. The authors concluded that cryptococcal antigen screening should be available for all patients in regions where the burden of cryptococcal disease among people living with HIV is high. The findings support that conclusion for the setting studied, but the abstract does not provide data on screening uptake, diagnostic delays, barriers to care or the effectiveness of screening across other hospitals or regions.

Subject of Research: Prevalence and management of cryptococcal antigenemia in hospitalized patients with advanced HIV in Ceará, Brazil

Article Title: Prevalence and management of cryptococcal antigenemia in hospitalized patients with advanced HIV infection in a tertiary hospital from Fortaleza, Ceará, Brazil between 2020 and 2022

Article References: Damasceno, L. S., Almeida Júnior, A. M. B. D., Brito, J. D. N., Maciel, L. S., Magalhães, L. F., & Mesquita, J. R. L. D. (2026). Prevalence and management of cryptococcal antigenemia in hospitalized patients with advanced HIV infection in a tertiary hospital from Fortaleza, Ceará, Brazil between 2020 and 2022. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14554-2

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14554-2

Keywords: cryptococcal antigenemia, advanced HIV, cryptococcal meningitis, hospital epidemiology, Ceará, Fortaleza, serum screening, in-hospital mortality, retrospective study, Prevalence, management, cryptococcal

The work focused on Hospital Clínic de Barcelona, a predominantly adult tertiary hospital with limited pediatric services. Because notification rules in Spain and Catalonia include both communicable and some noncommunicable conditions, the investigators included every notifiable-disease episode reported at the hospital during the three-month period, rather than selecting diseases in advance. The sample covered 25 conditions grouped into 10 categories, including sexually transmitted infections, mosquito-borne diseases, respiratory infections, hepatitis, zoonoses and asthma. Of the analyzed episodes, 173 involved males, 110 females and one person whose sex was recorded as unknown. Median age was 42 years, with an interquartile range of 25 years. One episode with no information in the clinical record was excluded from subsequent analyses, leaving 283 for assessment by healthcare setting and document type.

Sexually transmitted diseases formed the largest notified group, accounting for 73 episodes, or 25.8%. Most notifications originated in the outpatient clinic, which accounted for 145 episodes, or 51.2%; 96, or 33.9%, came from the emergency room and 42, or 14.8%, from hospital care. Medical doctors reviewed whether sufficient information was available on the reporting date to justify notification under Catalonia’s public health criteria. They classified 210 episodes, representing 74.2%, as adequately suspected and 73, or 25.8%, as inadequate. The investigators also calculated the interval between the first appearance of suspicion in the electronic health record and the actual report.

Among the adequately notified episodes for which relevant records could be identified, 170 were subsequently confirmed as having the reported condition. Thirty-three episodes, equal to 15% of the adequately notified group, had been reported before suspicion was recorded in the electronic health record and were excluded from the delay calculation. Among the remaining 177 episodes, the median delay was zero days and the interquartile range was three days. Sixty-two percent were reported on the same day that the first suspicion appeared, but 10% were reported more than two weeks later. The authors said explanations for reports preceding the electronic record, such as telephone calls or internal messages, remained hypotheses because the original notification source had not been registered.

The leading cause of inadequate reporting was the treatment of a chronic condition as a new reportable event, which explained 34.2% of the 73 inadequate notifications. References to a disease in a patient’s medical history accounted for 26.0%, while disease-related language whose clinical context did not support current suspicion accounted for 23.3%. A further 12.3% lacked supporting information, and 4.1% involved a mismatch between the disease reported and the condition suggested by the available records. Importantly, 83% of inadequate notifications contained at least one term from the disease-specific thesaurus. The results therefore show that the mere presence of a disease word or clinical concept is not enough to establish that an episode warrants notification.

Manual review of six types of electronic health record documents revealed where suspicion-related terms were most likely to appear. When the corresponding document was available, investigators found such terms in 93.0% of hospital discharge notes, 89.0% of unreleased clinical drafts, 84.7% of emergency-room discharge notes, 78.7% of clinical course reports, 63.3% of microbiology reports and 45.4% of laboratory reports. These percentages describe the content of available documents, however, and do not mean that every document type was equally accessible across episodes. Availability differed substantially by healthcare setting and disease group, making the importance of a source depend on where and how a patient was being treated.

The pattern changed markedly across care environments. In emergency-room episodes, emergency-room discharge notes contained suspicion-related information in 85.4% of cases, while unreleased clinical drafts did so in 83.8%. Clinical course reports were less frequent in that setting, and hospital discharge notes and microbiology reports were often unavailable. For hospitalized patients, clinical course reports contained relevant information in every episode in which they were assessed, with additional contributions from hospital discharge notes and unreleased drafts. Outpatient records had a different profile: clinical course reports contained relevant terms in 96.8% of episodes, and laboratory and microbiology reports also contributed, whereas the three other document types were not frequently available. These differences argue against a surveillance tool based on one fixed electronic record source.

Disease group also altered the contribution of each document. Clinical course reports were the most consistently available source, appearing in more than 80% of episodes in every disease group; when available, they contained suspicion-related terms in 81.2% of episodes. Unreleased clinical drafts were present in more than half the episodes of six disease groups and contained relevant terms in 94.8% of those available documents. Emergency-room discharge notes were representative for asthma, exanthematous viral diseases and mumps, and heatstroke, while hospital discharge notes were representative only for respiratory infectious diseases. Microbiology reports were available in at least half the episodes of six groups and contained relevant terms in 81.1% of available reports. Their contribution was particularly prominent for exanthematous viral diseases and mumps, foodborne and waterborne diseases, mosquito-borne diseases, and sexually transmitted infections.

The authors say the findings point toward a future system that selects document sources according to disease and care pathway, maps variations in language to standardized clinical concepts, and interprets those concepts in context rather than treating them as isolated keywords. A system would need to distinguish active suspicion from chronic disease, past history, negation and contextual mentions that do not support notification. The analysis was performed on raw clinical documentation, retaining abbreviations, misspellings, typographical errors, idiomatic expressions and colloquialisms; investigators manually identified and then standardized concepts for classification. A future system would also need expert review in a human-in-the-loop workflow, particularly because the study found overnotification as well as delays.

This work does not demonstrate that natural language processing or large language models can improve disease surveillance. No automated system was tested, and the researchers could not calculate the sensitivity, specificity or diagnostic performance of a future pipeline. The analysis included only episodes that had already been notified, rather than independently identifying cases that escaped reporting, and it did not include prior notification records needed to verify repeated reports. Its three-month summer sample may underrepresent conditions with different seasonal patterns, while the single hospital setting limits generalizability. Future research would need larger and more representative datasets, automated testing on unprocessed text, and direct measurement of accuracy and reporting time. Even with those limitations, the study offers a concrete starting point: the documents most likely to contain early signals, the conditions that generate false alerts, and the clinical pathways any future surveillance tool would have to understand.

Subject of Research: Evaluation of notifiable-disease reporting and clinical documents for future NLP-assisted hospital surveillance

Article Title: Setting the basis for a natural language processing tool to improve the reporting of notifiable diseases: a descriptive study

Article References: Cano-Escobar, C., Barón-Miras, L. E., Gualda-Gea, J. J., Frid, S., & Bertran, M. J. (2026). Setting the basis for a natural language processing tool to improve the reporting of notifiable diseases: a descriptive study. Frontiers in Digital Health, 8, Article 1882377. https://doi.org/10.3389/fdgth.2026.1882377

Image Credits: AI Generated

DOI: 10.3389/fdgth.2026.1882377

Keywords: natural language processing, notifiable diseases, public health surveillance, electronic health records, clinical informatics, disease reporting, hospital epidemiology, text mining, Barcelona, Setting, basis, natural

The research, published in Frontiers in Antibiotics by Tristan Lüdecke and Klaus-Peter Hunfeld of the Institute for Laboratory Medicine, Microbiology and Infection Control at Northwest Medical Center in Frankfurt am Main, evaluated the Bruker IR Biotyper as a first-line surveillance typing tool for Serratia isolates recovered from clinical materials. The team compared the FTIR results against whole genome sequencing coupled with bioinformatic analysis, the current gold standard for determining the phylogenetic relationships of clinical isolates. Their findings indicate that, although with some limitations, the infrared approach may serve as a rapid and cost-effective screening method when an outbreak is suspected.

Serratia marcescens, the type strain of the genus, is a facultative human pathogen of the order Enterobacterales that is encountered primarily as a cause of nosocomial infections, particularly in neonatal intensive care units, with sink drains identified as one of the main reservoirs. The bacterium can colonize a wide range of wet surfaces, from artificial nails to medical equipment, vials, the human gastrointestinal and respiratory tract, and materials such as lotions, antiseptics, and even quaternary ammonium-based disinfectants when handled incorrectly. Multi-year-long outbreaks as well as nationwide clonal spreading of Serratia species have been reported, underlining the significance of this facultative pathogen in healthcare settings.

According to the World Health Organization, the rate of nosocomial infections among hospitalized patients in the developed world stands at 7.6 percent, rising to 30 percent in intensive care units. Beyond the risk of increased mortality, these infections prolong hospital stays and drive up antibiotic use and healthcare costs. Although Serratia marcescens accounts for only one to two percent of nosocomial infections, an epidemiological increase in the isolation of Serratia species should prompt further epidemiological workup. The bacterium can carry beta-lactamases such as AmpC, ESBL, and carbapenemases including NDM and KPC, which can render a range of antibiotics ineffective and complicate empirical treatment.

The study was triggered by a six-week period of increased Serratia incidence between February and March 2024 among adult ventilated patients treated in a neurological intensive care unit at a collaborating clinical center. The researchers collected 28 isolates from routine clinical and environmental hygiene samples, including 24 first isolates from respiratory materials such as tracheal secretions, bronchoalveolar lavage, and pharyngeal swabs, a pair of environmental isolates, and two epidemiologically unrelated reference isolates from other institutions. All isolates displayed Serratia-typical antimicrobial susceptibility patterns, including some AmpC beta-lactamase overexpression, but no evidence of multidrug resistance.

Technically, the FTIR workflow was performed with the IR Biotyper using standard manufacturer settings. Isolates were subcultured on Columbia agar with five percent sheep blood, suspended in ethanol and distilled water, and applied to silicon target plates in five technical replicates each. Spectra were acquired in a wave number range of 800 to 1300 inverse centimeters, a region that tends to show the best classification performance. The software applied baseline corrections, smoothing, normalization, second derivative calculation, and vector normalization before hierarchical cluster analysis using the UPGMA algorithm generated dendrograms from which clusters were derived.

The genomic reference work was equally rigorous. DNA was extracted, purified, and sequenced on an Illumina NovaSeq X+ platform with a 2×150 bp paired-end workflow. Core genome multilocus sequence typing was performed using a Serratia marcescens cgMLST scheme comprising 2,692 alleles, with isolates differing by twelve or fewer alleles considered closely related. Pairwise SNP distances within clusters ranged from zero to three in four of the five genomic clusters, confirming clonal relationships, while the two environmental isolates from the same hospital sink differed by 30 SNPs yet were still considered epidemiologically linked because they came from the drain and trap of the same sink.

The species-level results were striking in themselves. While routine biochemical testing and MALDI-TOF mass spectrometry reported Serratia marcescens for all or most isolates, genome-based analysis using Mash Screen, average nucleotide identity, and the type strain genome server revealed a heterogeneous collection spanning Serratia ureilytica, S. bockelmannii, S. sarumanii, S. marcescens, and Serratia nevei, along with a potentially new Serratia species now under investigation. This underscores a known weakness of MALDI-TOF, whose mass spectra are too similar among Serratia species and whose databases do not yet include all recently discovered species.

Performance-wise, the comparison produced a positive predictive value of 0.842 for correctly identifying closely related isolates and a negative predictive value of 1.0 for correctly identifying unrelated ones. FTIR achieved a sensitivity of 1.0, meaning not a single closely related isolate was deemed unrelated, while specificity stood at 0.667, with three unrelated isolates falsely assigned to a cluster. Overall accuracy was 0.88. Concordance with the genomic reference, measured using the Adjusted Rand index and the Adjusted Wallace coefficient, yielded an AR of 0.742, a WGS-to-FTIR AW of 1.0, and an FTIR-to-WGS AW of 0.590, meeting previously suggested criteria for a first-line screening test. The FTIR results also compared favorably with a recent meta-analysis reporting an overall accuracy of 78 percent for Enterobacterales.

The authors caution that their strain collection, the largest set of Serratia isolates comprehensively studied with FTIR to date, came from a single medical center under a suspected outbreak situation, which implies a higher pretest probability of relatedness than a randomly collected set. The small sample size is reflected in wide 95 percent confidence intervals, and FTIR’s reliance on phenotypic traits rather than genetic similarity means wrong clustering events can occur. Strict adherence to standardized culture conditions, incubation times, and sample preparation protocols remains essential. Future research should address larger isolate sets from different locations, single-species sub-analyses, inter-laboratory ring trials, alternative spectral windows, and AI-based raw data analysis. Nevertheless, the study builds a case that infrared fingerprinting could become a complementary analytical addition to the arsenal of clinical microbiology laboratories, enabling the early, focused implementation of infection control countermeasures when every hour counts.

Subject of Research: Microbiology and Infectious Disease

Article Title: Evaluation of Fourier transform infrared spectroscopy as a first-line surveillance typing tool for Serratia spp. isolates derived from clinical materials

Article References: Lüdecke, T., & Hunfeld, K.-P. (2026). Evaluation of Fourier transform infrared spectroscopy as a first-line surveillance typing tool for Serratia spp. isolates derived from clinical materials. Frontiers in Antibiotics, 5, Article 1781370. https://doi.org/10.3389/frabi.2026.1781370

Image Credits: AI Generated

DOI: 10.3389/frabi.2026.1781370

Keywords: FTIR spectroscopy, IR Biotyper, Serratia, whole genome sequencing, nosocomial outbreaks, strain typing, clinical microbiology, infection control, cgMLST, hospital epidemiology, MALDI-TOF, surveillance

Pseudomonas aeruginosa is a formidable adversary in hospitals. It thrives in moist environments, infects ventilated patients, and causes ventilator-associated pneumonia, bloodstream infections, catheter-associated urinary tract infections, and surgical site infections, particularly in critically ill or immunocompromised people. Its defenses are layered: the bacterium’s outer membrane is intrinsically impermeable to many drugs, it deploys multidrug efflux pumps that expel antibiotics before they reach their targets, it alters the molecular targets those antibiotics are designed to hit, and it produces β-lactamases, enzymes that chemically dismantle penicillin-class drugs. Together these mechanisms allow the emergence of isolates classified as multidrug-resistant (MDR), extensively drug-resistant (XDR), and difficult-to-treat resistant (DTR) — categories that translate directly into prolonged hospitalization, higher costs, and increased mortality.

The study team, led by Wael A. Alghamdi of King Khalid University together with colleagues at Khamis Mushait General Hospital in the Aseer region, took a deliberately rigorous approach to the data. To avoid inflating resistance rates through repeated sampling of the same patients, the researchers deduplicated isolates at the patient level: for each calendar year, only the first P. aeruginosa isolate per patient was counted, with priority given to the most clinically significant source when several specimens were taken on the same day. Because a single patient could still contribute one isolate in different years, the team ran a sensitivity analysis restricted to 352 patients with no repeated observations, ensuring that temporal trends were not an artifact of non-independent data.

Resistance phenotypes were classified according to established international definitions. MDR isolates were non-susceptible to at least one agent in three or more antimicrobial categories; XDR isolates were non-susceptible to all but one or two categories; and DTR isolates, a clinically grounded classification proposed by Kadri and colleagues, were non-susceptible to all the principal first-line antipseudomonal β-lactams and fluoroquinolones. Bacterial identification and susceptibility testing were performed with the MicroScan automated broth microdilution system throughout the study, with results interpreted against the Clinical and Laboratory Standards Institute breakpoints in force at the time. Overall, 92.5 percent of potential susceptibility results were available, and missing results were left unimputed rather than guessed.

The headline finding is a dramatic arc in resistance prevalence. MDR rose from 51.0 percent of isolates in 2018 to a striking peak of 87.3 percent in 2022, before falling back to 69.6 percent in 2024. XDR climbed from 33.3 percent in 2018 to its highest level of 68.3 percent in 2023, then dropped sharply to 39.1 percent in 2024. DTR, the most clinically restrictive category, stayed lower throughout, ranging from 11.8 percent in 2018 to a peak of 31.7 percent in 2023 before declining to 15.2 percent in 2024. When calendar year was analyzed as a categorical variable, the differences across years were statistically significant for MDR (likelihood-ratio χ² = 33.25, P < 0.001) and XDR (χ² = 20.16, P = 0.003), but not for DTR (χ² = 9.57, P = 0.144).

Crucially, the temporal signal survived statistical scrutiny. In multivariable logistic regression models adjusted for age group, sex, nationality, and specimen source — built on the restricted dataset of 352 unique patients — collection year remained significantly associated with MDR (P < 0.001) and XDR (P = 0.013), but not DTR (P = 0.398). Compared with 2018, the adjusted odds of MDR were roughly six-fold higher in 2022 (adjusted odds ratio 6.12, 95 percent confidence interval 2.25–16.63) and nearly three-fold higher in 2023 (aOR 2.98, 95 percent CI 1.13–7.84), while the adjusted odds of XDR were significantly elevated in 2023 (aOR 3.59, 95 percent CI 1.41–9.11). This means the temporal pattern was not explained simply by shifts in the kinds of patients or specimen types being sampled.

The drug-specific data reveal a subtler story beneath the phenotype curves. Ciprofloxacin non-susceptibility surged from 44.0 percent in 2018 to 97.8 percent in 2024 — a near-total collapse of fluoroquinolone utility against these isolates. In contrast, non-susceptibility to the carbapenems imipenem and meropenem actually declined over the same period, falling from 45.7 percent and 41.2 percent to 30.4 percent and 27.3 percent respectively. This divergence suggests that resistance in P. aeruginosa does not move in one direction uniformly: some drug classes erode steadily while others can recover, presumably reflecting changing prescribing pressures, infection control dynamics, and shifts in circulating strains, though the study’s phenotypic design could not disentangle the mechanisms.

The patient and specimen profile of the cohort adds clinical context. Isolates came more often from male patients (68.5 percent), and nearly half came from people over 60 years of age (49.7 percent), a population with greater healthcare exposure and comorbidity burden. Respiratory tract specimens were the most common source, accounting for 35.1 percent of isolates, followed by urine and genital tract samples (28.3 percent) and wound, pus, and tissue specimens (24.7 percent), with blood and sterile body fluids at 12.0 percent. Interestingly, male sex was associated with higher adjusted odds of all three resistance phenotypes, and specimen source was significantly associated with DTR: isolates from urine and from wounds had markedly lower odds of DTR than bloodstream isolates, a pattern consistent with the severity of infections that reach the bloodstream.

What drove the rise and fall? The authors are appropriately cautious. An antimicrobial stewardship program was introduced at the study institution during the surveillance period, and the decline in MDR and XDR prevalence after 2022 and 2023 may be consistent with such interventions taking hold — but the retrospective, observational design makes it impossible to attribute the downturn to any specific measure. The fluctuations may equally reflect changes in antimicrobial prescribing practices, infection prevention measures, and evolving healthcare utilization, including the massive disruption that the COVID-19 pandemic years imposed on hospitals worldwide. Continuous longitudinal surveillance, the authors argue, is precisely what allows institutions to detect these swings and evaluate stewardship efforts over time.

The study also carries important limitations. It was retrospective and laboratory-based, so detailed clinical information — antimicrobial exposure, comorbidities, treatments, and outcomes — was unavailable, and infection could not be reliably distinguished from colonization for every isolate. No molecular characterization was performed, so the phenotypic trends cannot be pinned on clonal expansion or the spread of specific resistance genes. Annual isolate numbers varied considerably, from as few as 41 in 2023 to 75 in 2021, which makes year-to-year percentages sensitive to small changes. Susceptibility was interpreted using CLSI breakpoints in effect at the time of testing, so revisions during the seven-year window may have introduced some variability in longitudinal classification. Residual confounding by unmeasured clinical factors also remains possible.

Even with those caveats, the numbers land with force. A national multicenter surveillance study from Saudi Arabia recently reported that 25.8 percent of clinical P. aeruginosa isolates in the country’s southern region were DTR and 4.5 percent were pandrug-resistant, while regional analyses across the Arabian Gulf have documented rising resistance to carbapenems, fluoroquinolones, and aztreonam. The new single-center data, drawn from a secondary care setting that is often underrepresented in national surveillance reports, reinforce the message that resistance epidemiology is highly heterogeneous across institutions and regions — and that local data are essential for choosing empirical therapy wisely.

Saudi Arabia has responded at the policy level with its National Action Plan for Antimicrobial Resistance (2022–2025), which aims to strengthen surveillance systems, stewardship, and infection prevention nationwide. Studies like this one are the ground truth against which such plans will be judged. The study’s inclusion of the DTR classification alongside MDR and XDR is particularly valuable, because DTR captures the loss of susceptibility to the principal first-line antipseudomonal agents — the drugs clinicians actually reach for first — and has been linked to worse outcomes in Gram-negative infections.

The overall picture is neither pure doom nor easy reassurance. More than two-thirds of isolates in 2024 were still multidrug-resistant, nearly double the 2018 level, and fluoroquinolone resistance has become almost universal. Yet the significant declines in XDR prevalence in 2024, and the recovery in carbapenem susceptibility, show that resistance is not a one-way ratchet: it responds, sometimes relatively quickly, to the ecological pressures a hospital applies. For a pathogen that the global burden studies place among the deadliest sources of antimicrobial-resistance-associated deaths, that dynamism is the most actionable finding of all — a reminder that surveillance is not a bureaucratic exercise but a live instrument for steering therapy, and that stewardship decisions made today will be visible in the resistance curves of tomorrow.

Subject of Research: Longitudinal trends in antimicrobial resistance phenotypes of Pseudomonas aeruginosa in a Saudi hospital, 2018–2024

Article Title: Longitudinal trends in antimicrobial resistance phenotypes in Pseudomonas aeruginosa: a seven-year study

Article References: Alghamdi, W. A., Orayj, K. M., Ibrahim, A. R. N., Abdelsalam, M. A., & Alshehri, A. M. (2026). Longitudinal trends in antimicrobial resistance phenotypes in Pseudomonas aeruginosa: a seven-year study. Frontiers in Antibiotics, 5, Article 1912287. https://doi.org/10.3389/frabi.2026.1912287

Image Credits: AI Generated

DOI: 10.3389/frabi.2026.1912287

Keywords: antimicrobial resistance, Pseudomonas aeruginosa, multidrug-resistant, extensively drug-resistant, difficult-to-treat resistance, surveillance, antimicrobial stewardship, Saudi Arabia, healthcare-associated infections, carbapenems, fluoroquinolones, hospital epidemiology

Difficult-to-treat resistance is a recently formalized category, and its technical definition matters for understanding why these infections are so alarming. According to the criteria widely adopted from United States Centers for Disease Control and Prevention guidance and reflected in the Clinical and Laboratory Standards Institute’s M100 performance standards, a Gram-negative pathogen is labeled DTR when it is nonsusceptible to all agents in the standard first-line categories: all penicillins and beta-lactam and beta-lactamase inhibitor combinations, all cephalosporins, all carbapenems, the fluoroquinolones, and at least one of the aminoglycosides. In practical terms, a DTR Klebsiella isolate leaves clinicians with almost nothing from the traditional antibiotic armamentarium, forcing reliance on newer beta-lactam and beta-lactamase inhibitor combinations, tetracycline derivatives such as eravacycline, or the last-resort polymyxin colistin, each of which carries its own limitations in efficacy, toxicity, and availability.

The biology of Klebsiella pneumoniae makes it a particularly formidable opponent in cancer care. The organism is a Gram-negative bacillus equipped with an outer membrane rich in porins and a polysaccharide capsule that shields it from complement-mediated killing and phagocytosis. Resistance in DTR strains typically arises through the acquisition of beta-lactamase genes, including extended-spectrum beta-lactamases and carbapenemases such as those of the KPC, NDM, OXA-48-like, and VIM families, which hydrolyze the drugs that would otherwise destroy the cell wall. These genes frequently sit on plasmids that can transfer horizontally between bacterial species, meaning that resistance can spread within a hospital’s microbial ecosystem even when individual bacterial cells do not. For patients whose immune defenses are dismantled by chemotherapy, neutropenia, invasive devices, and repeated hospitalizations, such an organism finds an environment with few natural barriers.

Cancer patients represent a uniquely vulnerable population for these infections, a point the Varanasi team emphasizes by situating their investigation squarely in supportive care rather than pure microbiology. Intensive chemotherapy regimens produce prolonged periods of severe neutropenia, during which the body’s primary cellular defense against bacteria is largely absent. Mucosal barriers lining the gastrointestinal tract are damaged by cytotoxic drugs, allowing intestinal colonizing bacteria to translocate into the bloodstream. Central venous catheters, urinary catheters, and ventilatory support provide direct conduits for microorganisms to bypass skin defenses. Prolonged and repeated hospital admissions increase exposure to the hospital flora, and broad-spectrum prophylactic or empiric antibiotics, often necessary in this population, apply the selective pressure that favors resistant strains over susceptible competitors. The result is a patient population in which multidrug-resistant Gram-negative infections are both more common and more lethal than in the general hospitalized population.

Recent evidence underscores how grim the prognosis has become even in high-resource settings. A retrospective cohort study published in The Lancet Infectious Diseases in 2026 by Walker and colleagues examined survival trends among patients with difficult-to-treat, antibiotic-resistant Gram-negative infections in the United States during the era of next-generation antibiotics, and its inclusion in the Varanasi paper’s reference list signals the authors’ awareness that mortality in this syndrome remains substantial despite the arrival of new drugs. Severe infections caused by difficult-to-treat Gram-negative bacteria, as reviewed by Dettori, Bassetti, and colleagues in Current Opinion in Critical Care, frequently present as bloodstream infections, ventilator-associated pneumonia, or complicated intra-abdominal infections, and delays in appropriate empiric therapy are consistently associated with worse outcomes. In oncology patients, the combination of host immunosuppression and near-total antibiotic resistance creates a clinical scenario in which every hour of ineffective therapy matters.

Accurate susceptibility testing is therefore not a laboratory formality but a life-or-death variable, and this technical dimension occupies a significant place in the Varanasi group’s broader research program. Colistin, the polymyxin antibiotic often reserved for DTR organisms, is notoriously difficult to test reliably. Its large, amphipathic molecule aggregates in aqueous solution and does not diffuse well in agar, which is why traditional disk diffusion performs poorly. The reference standard, broth microdilution, is technically demanding and impractical for many laboratories in resource-limited settings. In response, the Clinical and Laboratory Standards Institute has endorsed broth disk elution as an alternative, a method in which colistin is eluted from a disk into broth before inoculation. A 2025 study by Chandankhede, Karyakarte, and colleagues compared broth disk elution with broth microdilution for carbapenem-resistant Enterobacterales and Pseudomonas aeruginosa, and a prospective study by CV, Priya, Chauhan, and Dhingra specifically evaluated the diagnostic performance of the two methods for carbapenem-resistant Gram-negative bacilli from cancer patients in India. These methodological comparisons matter because an erroneous report of susceptibility, or of resistance, directly changes whether a critically ill cancer patient receives an effective regimen.

The Varanasi investigation itself was structured as a formal healthcare-associated outbreak investigation, a discipline with well-established epidemiological methods. Outbreak investigation in a hospital setting begins with case ascertainment and the construction of a case definition, followed by descriptive epidemiology that maps cases by person, place, and time to identify patterns of transmission. When multiple patients harbor isolates with identical or closely related resistance profiles, the possibility of a common source or of clonal spread via healthcare workers’ hands, contaminated equipment, or the shared hospital environment must be systematically evaluated. The authors acknowledge support for the infection prevention and control assessment component of their investigation from Professor Purva Mathur of the Jai Prakash Narayan Apex Trauma Centre at AIIMS New Delhi and from Dr. A. S. Valan, a public health specialist at the Centers for Disease Control and Prevention India Country Office in New Delhi, indicating that the assessment was aligned with international standards, including the World Health Organization’s 2023 assessment tool on infection prevention and control minimum requirements for tertiary health care facilities.

That WHO tool provides a structured framework for evaluating whether a tertiary facility meets minimum IPC standards across core components such as governance and dedicated IPC staffing, guidelines and protocols, surveillance of healthcare-associated infections, microbiology laboratory capacity, environmental cleaning and sterilization, and the built environment and equipment. In cancer hospitals, the stakes of each component are amplified. High-touch surfaces around immunosuppressed patients, shared bathroom facilities, water systems, and medical devices all represent potential reservoirs for Klebsiella, which can survive on inanimate surfaces for extended periods. Hand hygiene compliance, cohorting of colonized or infected patients, rigorous terminal cleaning of rooms, and antimicrobial stewardship programs that limit unnecessary broad-spectrum exposure form the interlocking defenses that keep resistant organisms contained. An outbreak of DTR Klebsiella in such a setting is therefore as much a systems failure as a microbiological event, and the investigation’s value lies in identifying which links in the chain of transmission broke down.

The broader significance of the study extends well beyond a single institution in northern India. Carbapenem-resistant Klebsiella pneumoniae has been designated a critical-priority pathogen on the World Health Organization’s bacterial priority pathogens list, and the global spread of carbapenemase genes, particularly those encoding NDM-type enzymes that are endemic across the Indian subcontinent, means that cancer centers everywhere face converging threats. The Varanasi paper’s title phrase, beyond the tumor, captures a conceptual shift that supportive oncology has been undergoing: survival in cancer medicine increasingly depends not only on the effectiveness of antineoplastic therapy but on the invisible war being waged against resistant bacteria in the wards, laboratories, and plumbing of the hospitals where treatment occurs. As next-generation antibiotics arrive, their durability will depend on how well healthcare systems prevent the transmission events that this outbreak investigation documents, because every new drug deployed against a resistant cluster applies the selective pressure that can eventually render it ineffective too.

For clinicians and hospital administrators, the message of the study is operational as much as scientific. Rapid detection of DTR phenotypes through reliable susceptibility methods, prompt notification of infection control teams when clusters emerge, disciplined application of contact precautions, and sustained investment in laboratory capacity and IPC infrastructure are the concrete measures that determine whether a cluster remains a cluster or becomes a sustained institutional epidemic. For patients and families, the study is a reminder that modern cancer care is a team effort in which microbiologists, infection preventionists, and oncologists share responsibility for outcomes, and that the fight against antimicrobial resistance is inseparable from the fight against cancer itself.

Subject of Research: Healthcare-associated outbreak of difficult-to-treat Klebsiella pneumoniae infection among cancer patients

Article Title: Beyond the tumor: cluster of difficult-to-treat (DTR) Klebsiella pneumoniae infection in cancer patients—healthcare-associated outbreak investigations

Article References: Bajpai, V., Mishra, A., John, N., Sarode, R., Chaurasia, A., Bharti, S., Kapoor, A., & Mishra, B. K. (2026). Beyond the tumor: cluster of difficult-to-treat (DTR) Klebsiella pneumoniae infection in cancer patients—healthcare-associated outbreak investigations. Supportive Care in Cancer, 34(10), Article 1055. https://doi.org/10.1007/s00520-026-11277-7

Image Credits: AI Generated

DOI: 10.1007/s00520-026-11277-7

Keywords: Klebsiella pneumoniae, difficult-to-treat resistance, antimicrobial resistance, cancer patients, healthcare-associated infection, outbreak investigation, infection prevention and control, carbapenem-resistant Enterobacterales, colistin susceptibility testing, supportive oncology, immunocompromised hosts, hospital epidemiology

The research, led by Zihuan Li and Tian Wang with corresponding author Yisui Cen, was structured as a before-and-after analysis spanning nearly eight years of hospital operations. The investigators divided the study period into two distinct phases. The pre-intervention period, designated Period 1, ran from 2018 through 2022 and captured routine surveillance practices as they existed before the enhancement. The post-intervention period, Period 2, covered 2023 through 2025, after the hospital deployed its enhanced active surveillance program. During the intervention phase, a multidisciplinary infection prevention and control team conducted daily active surveillance using a novel intelligent early warning system, and provided real-time feedback to clinical staff on infection diagnosis, antimicrobial use, isolation practices, and environmental hygiene.

The scale of the dataset lends the findings considerable weight. In total, the researchers identified 2,600 HAI cases during the five-year pre-intervention period and 1,697 cases during the three-year post-intervention period. Rather than simply comparing crude averages between the two eras, the team employed interrupted time-series analysis, a statistical framework widely regarded as one of the strongest quasi-experimental designs for evaluating the impact of a policy change when a randomized trial is not feasible. This method models the underlying trend of an outcome over time, then tests whether the intervention produced an immediate level shift at the moment of implementation, a change in the slope of the trend, or both.

The headline result concerns timeliness of reporting. Before the intervention, 32.85 percent of HAI cases were reported late, a figure that fell to 28.23 percent afterward, a statistically significant decrease with a p-value of 0.001. The interrupted time-series analysis added granularity: during Period 2, the delayed reporting rate declined by an average of 1.240 percentage points per quarter, with a 95 percent confidence interval stretching from minus 2.408 to minus 0.078 percent, and the downward trend itself was significant at p equals 0.033. In practical terms, this means the improvement was not a one-time blip but a sustained, quarter-by-quarter erosion of the delayed reporting problem, exactly the pattern one would hope to see from a surveillance system that learns and embeds itself into daily workflow.

Perhaps the most striking finding was the jump in pre-discharge reporting, the proportion of infections identified and documented before the patient leaves the hospital. This rate rose from 74.54 percent to 89.69 percent, a highly significant improvement with p less than 0.001. The time-series model pinpointed an immediate increase of 13.453 percentage points at the intervention point itself, with a 95 percent confidence interval of 3.143 to 23.763 percent and a p-value of 0.012. That instantaneous level shift is telling. It suggests the intelligent early warning system began surfacing infections that would previously have slipped through the cracks almost as soon as it went live, likely because automated alerts flag suspicious clinical patterns, laboratory results, and antimicrobial prescriptions in real time rather than relying on clinicians to notice and report them at the end of a busy shift.

Speed mattered in the details as well. The median time of delayed reporting, among cases that were still reported late, dropped from 11 days, with an interquartile range of 7 to 18 days, to 8 days, with an interquartile range of 6 to 12 days, a difference significant at p less than 0.001. Even when the system could not prevent a delay entirely, it compressed the lag. The median interval from infection onset to discharge shortened from 11 days to 9 days, and the median time from reporting to discharge fell from 9 days to 7 days, both with p-values below 0.001. Shorter intervals between onset, detection, and discharge carry real clinical meaning: earlier recognition allows earlier isolation, earlier targeted therapy, and earlier discharge planning, all of which reduce the window in which a pathogen can spread to other patients on the ward.

What makes this study technically interesting is the architecture of the intervention itself. Active surveillance, in contrast to passive reporting, means that trained personnel actively hunt for infections rather than waiting for them to be reported. The Chinese team layered an intelligent early warning system on top of this human effort, presumably mining electronic health record data to generate alerts, although the published abstract does not detail the underlying algorithm. Crucially, the system did not operate in isolation. A multidisciplinary infection prevention and control team reviewed the alerts daily and closed the loop with real-time feedback covering four domains: whether the infection diagnosis was correct, whether antimicrobial use was appropriate, whether isolation practices were being followed, and whether environmental hygiene standards were being met. This combination of machine detection and human verification addresses a well-known weakness of purely automated surveillance, namely alert fatigue and false positives that erode clinician trust.

The findings arrive at a moment when health systems worldwide are under intensifying pressure to curb HAIs. Antimicrobial resistance has turned once-routine hospital infections into therapeutic nightmares, and regulatory bodies increasingly tie reimbursement and accreditation to infection metrics. Surveillance is the foundation of every infection prevention program because you cannot control what you cannot see, yet underreporting and delayed reporting remain endemic problems in hospitals everywhere. Traditional surveillance depends on infection control practitioners manually reviewing charts, a labor-intensive process that scales poorly and lags behind clinical reality. The Guangzhou results suggest that intelligent early warning systems, coupled with disciplined human follow-up, can close much of that gap, raising the pre-discharge reporting rate by roughly 15 percentage points and shaving days off critical time intervals.

Several caveats deserve honest acknowledgment. The study was conducted at a single tertiary hospital in China, and its retrospective design means the researchers could not control for concurrent changes in case mix, staffing, or the lingering effects of the COVID-19 pandemic era on hospital operations during the early period. Interrupted time-series analysis is robust, but it cannot fully exclude secular trends or co-interventions that happened to coincide with the 2023 implementation. The abstract also does not report whether the surveillance improvements translated into lower overall HAI incidence or reduced mortality, outcomes that matter most to patients. What the data do establish is a chain of process improvements: faster reporting, more complete reporting before discharge, and shorter times from infection onset and from report to discharge. Whether those process gains compound into harder clinical endpoints is the natural next question for future research.

Even with those limitations, the study offers a template that other hospitals can adapt. The core recipe, an automated early warning layer feeding a multidisciplinary team that delivers immediate, specific feedback on diagnosis, antimicrobials, isolation, and environmental hygiene, does not require exotic technology, only integration of existing electronic health record data with a well-resourced infection control workforce. The authors received no external funding for the work, and the study was approved by the hospital’s ethics committee with informed consent waived owing to the de-identified retrospective database. As hospitals everywhere grapple with resistant organisms and stretched infection control staff, the message from Guangzhou is clear: surveillance that watches actively, warns intelligently, and feeds back instantly can turn the slow, leaky pipeline of infection reporting into something approaching real time, and that shift alone may be worth days of hospital stay for every affected patient.

Subject of Research: Effectiveness of enhanced active surveillance with an intelligent early warning system for healthcare-associated infections

Article Title: Effectiveness analysis of enhanced active surveillance for healthcare-associated infections

Article References: Li, Z., Wang, T., Wu, H., Lin, G., Fan, C., Liu, B., Liu, D., & Cen, Y. (2026). Effectiveness analysis of enhanced active surveillance for healthcare-associated infections. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14566-y

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14566-y

Keywords: healthcare-associated infections, active surveillance, infection prevention and control, intelligent early warning system, interrupted time-series analysis, delayed reporting rate, pre-discharge reporting, hospital epidemiology, antimicrobial stewardship, patient safety, BMC Infectious Diseases, tertiary hospital