Abstract
Introduction
Infections caused by extended-spectrum β-lactamase-producing Enterobacterales (ESBL-E) and carbapenemase-producing Enterobacterales (CPE) are a major concern in patients with neutropenia with haematological diseases, as they may lead to inadequate empirical therapy and poor outcomes. However, the relationship between intestinal colonisation and subsequent infection in this population remains insufficiently characterised. The SCREEN-IN Study aims to identify factors associated with incident ESBL-E/CPE colonisation and to develop and validate a prediction model for infection caused by these microorganisms in patients with neutropenia.
Methods and analysis
SCREEN-IN is a multicentre prospective observational cohort study in 15 hospitals in Spain. Adults admitted to haematology wards who are expected to develop neutropenia after chemotherapy, conditioning for haematopoietic stem cell transplantation or lymphodepleting therapy before CAR-T cell therapy will be enrolled at treatment initiation. Rectal swabs will be obtained at baseline and weekly during neutropenia, including subsequent neutropenic episodes, to assess dynamic colonisation by ESBL-E and CPE. Isolates from surveillance and clinical samples will undergo local phenotypical testing and central molecular characterisation, including whole-genome sequencing to evaluate the relationship between colonising and infecting microorganisms. Participants will be followed for 90 days after the first rectal swab, and each infectious episode will be followed for 30 days. The two co-primary outcomes, analysed separately, are incident ESBL-E/CPE colonisation among participants who are non-colonised at baseline and the first microbiologically documented infection caused by these organisms. Factors associated with colonisation will be analysed using time-to-event methods. Colonisation will be treated as a time-dependent predictor in the infection model, which will undergo internal validation and evaluation in a temporally separated validation cohort.
Ethics and dissemination
The protocol was approved by the Provincial Research Ethics Committee of Seville on 24 March 2026. Written informed consent will be obtained from all participants. Results will be disseminated through peer-reviewed journals and scientific meetings.
Trial registration number
Keywords: HAEMATOLOGY, Molecular diagnostics, MICROBIOLOGY, Neoplasms
STRENGTHS AND LIMITATIONS OF THIS STUDY.
This is a multicentre prospective cohort study including 15 hospitals in Spain; this approach captures real-world clinical practice and epidemiological heterogeneity in haematology units.
The study uses serial rectal swab screening during neutropenia to assess dynamic colonisation status.
Colonisation status will be analysed as a time-dependent variable in the prediction model for infection.
Whole-genome sequencing will allow assessment of the clonal concordance between surveillance and clinical isolates.
As an observational study, residual confounding and centre-level variability in clinical practice cannot be fully avoided.
Introduction
Infections caused by multidrug-resistant organisms (MDROs) are an important health problem and are associated with higher morbidity, mortality and healthcare costs.1 MDROs are especially relevant in patients with haematological malignancies, who are treated with intensive chemotherapy, haematopoietic stem cell transplantation or CAR-T cell therapy. In this setting, infections are one of the main causes of complications and death. It is estimated that up to 80% of patients with neutropenia develop fever, and 10%–25% suffer bloodstream infections (BSIs). Among MDROs, extended-spectrum β-lactamase-producing Enterobacterales (ESBL-E) and carbapenemase-producing Enterobacterales (CPE) are of particular concern because they can lead to inappropriate empirical treatment and worse outcomes.2 3
Gut colonisation by ESBL-E and CPE is considered an important step before infection. Previous studies in non-neutropenic populations have shown that intestinal colonisation by antibiotic-resistant Enterobacterales significantly increases the risk of later invasive infection.4 5 In patients with neutropenia with haematological diseases, the relationship between colonisation and infection is still not well defined. Reported colonisation rates with CPE range from 6% to 11%, and subsequent BSI has been described in 14%–25% of colonised patients, while colonisation by ESBL-E is usually more frequent.6–8 However, published studies are heterogeneous and often come from single-centre experiences with small sample sizes. In addition, centres differ in their local epidemiology, antibacterial prophylaxis, empirical treatment strategies and infection-control practices.
Some predictive tools have been developed to estimate the risk of infection or poor outcome in general patients colonised by MDRO, such as the Giannella Risk Score9 and the INCREMENT-CPE Score,10 respectively. However, these scores were not specifically developed for patients with neutropenia with haematological malignancies. Only a few studies have developed prediction models in this setting, and most of them have important limitations, such as a single-centre design, predicting only BSI or restriction to selected subgroups such as transplant recipients.11–13
There is also increasing interest in the role of the intestinal microbiota in these patients. In patients receiving chemotherapy or stem cell transplantation, repeated antibiotic exposure and mucosal injury often lead to intestinal dysbiosis, which has been associated with infection, graft-versus-host disease and mortality. However, the relationship between the intestinal microbiota, colonisation by multidrug-resistant Enterobacterales and progression to infection has not been comprehensively studied in patients with neutropenia.14 15
The SCREEN-IN Study was designed to address these gaps. This multicentre prospective cohort study aims to identify factors associated with incident ESBL-E and CPE colonisation in patients with neutropenia with haematological diseases and to develop and validate a model that predicts infection caused by these microorganisms. The study also includes an exploratory substudy of the intestinal microbiota in a subgroup of patients.
Study objectives
The SCREEN-IN Study has two co-primary objectives: first, to identify factors associated with incident ESBL-E or CPE colonisation in patients with haematological diseases who develop neutropenia after treatment; and second, to develop and validate a model that estimates the individual risk of infection caused by these microorganisms, with colonisation status evaluated as a principal time-varying predictor. The resulting dynamic risk estimate is intended for use at the onset of febrile neutropenia or another clinically suspected infectious episode, using information available at that time, including the latest colonisation result. The model is intended to identify patients whose predicted risk may justify colonisation-directed broader empirical coverage while also identifying a low-risk group in whom unnecessary escalation might be avoided. Secondary objectives are to describe baseline prevalence and acquisition of colonisation during neutropenia; analyse the frequency of infection in colonised and non-colonised patients; assess the adequacy of antimicrobial treatment and its relationship with outcomes; and explore the role of the intestinal microbiota in a subgroup of patients.
Methods and analysis
Study design, setting and study period
SCREEN-IN is a multicentre prospective observational cohort study to be conducted in 15 hospitals in Spain. All diagnostic procedures and treatments are decided by the treating physicians according to local practice. Adult patients with haematological diseases who are expected to develop neutropenia after chemotherapy, conditioning regimens for haematopoietic stem cell transplantation or lymphodepleting therapy before CAR-T cell therapy will be prospectively included. Patient enrolment is scheduled to begin after site activation and continue until October 2028.
Study population
Eligible participants will be adults aged 18 years or older admitted to haematology wards and scheduled to receive treatment expected to cause neutropenia. Underlying conditions include acute leukaemia, myelodysplastic syndromes, lymphomas, multiple myeloma and other haematological disorders. Patients will be enrolled at the start of treatment expected to cause high-risk neutropenia in the subsequent days, defined as an absolute neutrophil count below 0.5×10⁹/L, or below 1.0×10⁹/L when a decrease below 0.5×10⁹/L is expected within the next 48 hours. Patients will be excluded if they have a life expectancy shorter than 30 days, are unable to understand the study procedures or have previously been included in the cohort.
Screening procedures
All included patients will undergo rectal swab screening for colonisation by ESBL-E and CPE. A rectal swab will be obtained at baseline (start of treatment) and weekly during the neutropenic period or until hospital discharge. Weekly swabbing will resume during subsequent episodes of neutropenia. Patients will be dynamically classified as colonised or non-colonised according to the screening culture results.
Microbiological procedures
Rectal swabs will be processed in the microbiology laboratories of each participating centre using selective chromogenic media for the detection of ESBL-E and CPE. Identification and initial phenotypical characterisation will be performed using routine laboratory methods such as matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry and antimicrobial susceptibility testing. All isolates identified as ESBL-E or CPE from screening or clinical samples will be stored locally and sent to the coordinating centre where confirmatory identification, molecular characterisation of resistance mechanisms and whole-genome sequencing will be performed in order to analyse the genomic concordance between colonising and infecting isolates.
Patients’ follow-up
Patients will be followed for 90 days after inclusion (day 0). Serial rectal swabs will be collected at baseline, weekly during neutropenia and during hospital stay, and at day 90 (±48 hours tolerance window at each time point). Febrile neutropenia episodes will be followed for 30 days postonset, with clinical re-evaluation after 72–96 hours.
Microbiome substudy
An exploratory microbiome substudy will be conducted only in patients included in three hospitals in Seville. In these centres, stool samples will be obtained at baseline and, in patients who become colonised by ESBL-E or CPE, additional samples will be collected at the time colonisation is detected and at days +7 and +14 after detection. In patients who remain non-colonised, a second sample will be obtained either at hospital discharge or at day +30 after inclusion. The intestinal microbiota will be analysed by sequencing the V3-V4 regions of the bacterial 16S rRNA gene and evaluating microbial diversity and composition using standard bioinformatic methods.
Evaluation of outcomes
The study has two co-primary outcomes, which will be analysed separately and do not constitute a composite outcome. The first is incident ESBL-E or CPE colonisation, defined as the first positive rectal swab during follow-up among participants who are non-colonised at baseline; baseline colonisation prevalence will be reported separately. The second is the first clinically relevant, microbiologically documented infection caused by ESBL-E or CPE within 90 days after inclusion. Febrile neutropenia episodes will be classified as microbiologically documented infection, clinically documented infection or febrile neutropenia without documented infection according to predefined criteria (see online supplemental table 1). Secondary outcomes include 90-day all-cause mortality after study inclusion and 30-day all-cause and 30-day infection-attributable mortality after an infectious episode; 30-day postinfection recurrence; appropriateness of empirical and targeted antimicrobial treatment; length of hospital stay; and antimicrobial consumption (measured as days of treatment) during follow-up. The occurrence of other infectious complications such as fungal infections or Clostridioides difficile infection will also be recorded. Because the analyses are primarily estimative and address two distinct prespecified objectives, no single confirmatory hypothesis test or family wise type I error claim is planned; effect estimates will be presented with 95% CIs and multiplicity will be considered when interpreting secondary analyses. In patients in whom both colonising and infecting isolates are available, genomic analysis will be performed to evaluate the clonal concordance between the microorganisms detected in surveillance cultures and those causing infection. This analysis will allow assessment of whether invasive infection originates from previously detected colonising isolates.
Other variables and data collection
Baseline and prospective data from electronic medical records, microbiology laboratory databases and routine clinical documentation will be entered into a dedicated electronic case report form (eCRF) by designated clinical investigators or trained research staff at each site. All sites will use a common data dictionary, operational manual and predefined outcome definitions, supported by study-specific training before enrolment. To ensure data quality and integrity, the eCRF incorporates automated validation checks at the point of entry, complemented by central monitoring, systematic queries and resolution of data discrepancies. Periodic completeness and consistency reports will be reviewed by the coordinating team. Because predictors are recorded prospectively prior to outcome occurrence, predictor assessment is inherently masked to future infection events. However, as colonisation results and clinical events form part of routine care, local investigators cannot be masked to them. Ambiguous outcome classifications will undergo central review using the prespecified definitions, and the statistical team will analyse a pseudonymised data set with centre identities coded.
Variables were selected based on clinical relevance and through a targeted, non-systematic review of the literature on colonisation and infection by resistant Enterobacterales in haematological populations. For analytical purposes, variables will be grouped into demographic characteristics, underlying haematological disease and treatment, clinical risk factors, microbiological variables, infection episodes and clinical outcomes (table 1).
Table 1. Variables collected during the study period summarised by blocks.
| Block | Key variables (minimum set) | Timepoint |
|---|---|---|
| Study setting/enrolment | Centre; patient ID; informed consent date; eligibility (inclusion/exclusion yes/no) | Baseline (inclusion) |
| Baseline demographics and comorbidity | Age; sex; weight/height (BMI); residence/long-term care; transfer from other hospital; Charlson; Charlson age-adjusted | Baseline |
| Underlying haematological disease and therapy | Haematological diagnosis (single); disease status (debut/PR/CR/relapse/refractory); planned/intended treatment category (chemotherapy/HSCT/CAR-T/other) | Baseline/updated at index date if needed |
| Rectal swab screening (colonisation) | Swab date; ANC at swab; result (pos/neg); organism(s); ESBL (yes/no, type if available); CPE (yes/no, type if available) | Weekly during neutropenia |
| Index window for risk factors (−90 days) | Index date (first positive swab or last negative); hospitalisations (number, cumulative days); prior febrile neutropenia episodes (number); ICU stay; invasive procedures; devices (CVC/PICC/urinary catheter etc.) Systemic antibiotics (classes+total days); fluoroquinolone prophylaxis (yes/no, total days); systemic corticosteroids (yes/no, total days, prednisone-equivalent); antifungal prophylaxis/treatment (yes/no, agents, days) |
Retrospective −90 days from index date |
| Advanced therapies (if applicable) | HSCT: type, donor, source, engraftment delay, GVHD (type/grade/treatment), CMV reactivation; CAR-T: product, lymphodepletion/bridge, CRS/ICANS grade, ICU, immunosuppressive treatment | Retrospective −90 days from index date |
| Febrile neutropenia episodes (clinical) | Episode dates (start/end fever); ANC at onset; qSOFA; SOFA (onset and worst); clinical syndrome; definitive focus | Each FN episode |
| Febrile neutropenia episodes (microbiology) | Samples obtained; pathogen(s); resistance phenotype (ESBL/CPE if relevant); concordance with colonisation (phenotypical) | Each FN episode |
| Management and biomarkers | Empirical regimen(s) + dates; targeted regimen(s) + dates; CRP/PCT at onset±at switch (if available); empirical adequacy; time-to-active therapy (if documented infection) | Each FN episode |
| Outcomes (short-term) | 30-day all-cause mortality; infection-attributable mortality; recurrence during same neutropenia (relapse/reinfection/new focus) | 30 days after each episode |
| Follow-up (day +90) | 90-day mortality (all-cause/attributable); total re-admissions+duration; Clostridioides difficile infection (episodes, treatment, recurrences); invasive fungal infection (type, focus, treatment); ICU admissions | Day +90 from inclusion |
ID, identification
BMI, body mass index
PR, partial remission
CR, complete remission
HSCT, Haematopoietic stem cell transplantation
CAR-T, Chimeric antigen receptor T-cell therapy
ANC, Absolute neutrophil count
ICU, Intensive care unit
CVC, Central venous catheter
PICC, Peripherally inserted central catheter
GVHD, Graft-versus-host disease
CMV, Cytomegalovirus
CRS, Cytokine release syndrome
ICANS, Immune effector cell-associated neurotoxicity syndrome
qSOFA, Quick Sequential Organ Failure Assessment
CRP, C-reactive protein
PCT, Procalcitonin
Cytomegalovirus; ANC, Absolute neutrophil count; BMI, Body mass index; CAR-T, Chimeric antigen receptor T-cell therapy; CMV, Cytomegalovirus; CMV, Cytomegalovirus; CMV, CPE, carbapenemase-producing Enterobacterales; CR, Complete remission; CRP, C-reactive protein; CRS, Cytokine release syndrome; CVC, Central venous catheter; ESBL-E, extended-spectrum β-lactamase-producing Enterobacterales; GVHD, Graft-versus-host disease; GVHD, Graft-versus-host disease; GVHD, Graft-versus-host disease; HSCT, Haematopoietic stem cell transplantation; ICANS, Immune effector cell-associated neurotoxicity syndrome; ICU, Intensive care unit; ID, Identification; PCT, Procalcitonin; PICC, Peripherally inserted central catheter; PR, Partial remission; qSOFA, Quick Sequential Organ Failure Assessment.
Sample size and statistical analysis
To address the investigation of predictors for colonisation and for infection, the commonly used rule of at least 10 outcome events per predictor variable in multivariable models was applied.16 The lower number of outcomes would be that for infections caused by ESBL and/or CPE; according to previous unpublished data from our site, we estimated a 15% rate of infection so we would need 335 patients to have 50 infections and therefore include up to five predictors in the model. In addition, to validate the model, a cohort of 200 patients will also be included. Therefore, the total planned sample size of the study is 535 patients.
Descriptive analyses will summarise baseline characteristics, colonisation rates and infection outcomes. Continuous variables will be described using means and SD or medians and IQRs, as appropriate; categorical variables will be summarised using frequencies and percentages. Acquisition of colonisation will be analysed among participants non-colonised at baseline. Cumulative incidence functions will treat death before colonisation or infection as a competing event. Risk factors for colonisation will be analysed using time-to-event methods, with Cox proportional hazards models used to estimate HRs. For the infection outcome, colonisation status will be treated as a time-dependent predictor.
The first 335 participants will form the model-development cohort and the subsequent 200 the temporal validation cohort. Bootstrap resampling will be used for internal validation. The model will then be evaluated without refitting in the validation cohort. Performance will be assessed using measures of discrimination, calibration, overall prediction error, clinical utility via decision-curve analysis, and centre-level heterogeneity will be explored using internal-external cross-validation.17 18 The added prognostic value of serial colonisation screening will be explored by comparing model performance before and after its inclusion. The model will estimate absolute risk using only information available at the onset of each suspected infection. Rather than applying a single cut-off, low-risk thresholds will prioritise sensitivity and negative predictive value, while high-risk thresholds will consider specificity, positive predictive value and the consequences of incorrect classification. Candidate predictors and their coding will be prespecified before outcome modelling based on previous evidence, clinical relevance, availability at the intended time of prediction and consistent measurement across participating centres. The candidate set will be limited according to the anticipated number of outcome events. No data-driven preselection based on univariable values of p or automated stepwise procedures will be used.
Missing values will be handled using multiple imputation. Outcome status will not be imputed; participants lost to follow-up will be censored at the last date known to be event-free. Complete-case and alternative-assumption sensitivity analyses will be performed when appropriate.
A detailed, version-controlled statistical analysis plan, specifying predictor coding, handling of missing data and full model specifications, will be finalised prior to database lock and made publicly available.
Timeline
Patient recruitment began in September 2026 and is expected to continue until May 2028, with the final patient follow-up anticipated to be completed in August 2028. The study follows a predefined sequence of procedures from patient inclusion to the end of follow-up. After enrolment, patients undergo baseline assessment and rectal screening for colonisation. During the neutropenic period, rectal swabs are repeated weekly until recovery from neutropenia or hospital discharge. Patients are prospectively monitored for the occurrence of febrile neutropenia or clinically relevant infection. When an infectious episode occurs, specific clinical and microbiological data are collected, and the episode is followed for 30 days to assess outcomes. All participants remain under observation for a total follow-up period of 90 days from inclusion (figure 1).
Figure 1. Timeline of the SCREEN-IN project.
Patient and public involvement
Patients and/or the public were not involved in the design, conduct, reporting or dissemination plans of this research.
Ethics and dissemination
The study will be conducted in accordance with the principles of the Declaration of Helsinki (World Medical Association, 2024 revision), as well as applicable Spanish regulations governing observational research, including Law 14/2007 of 3 July on Biomedical Research and Royal Decree 957/2020 of 3 November regulating observational studies with medicines for human use. The protocol was approved by the Provincial Research Ethics Committee (administración.eecc.hvm.sspa@juntadeandalucia.es) of Seville on 24 March 2026 (SICEIA-2026-000228). All participants will provide written informed consent before inclusion. Data will be collected and managed in compliance with current data-protection regulations, including the General Data Protection Regulation (EU) 2016/679 and Organic Law 3/2018 of 5 December on Personal Data Protection and Guarantee of Digital Rights, ensuring confidentiality and pseudonymisation of participant information.
As this is an observational study that does not involve experimental treatments or modification of routine clinical care, no additional risks for participants are anticipated beyond those related to standard medical practice.
This study protocol was prepared to enable the reporting of results in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology statement (see online supplemental table 2). Moreover, development, validation, updating and reporting of the prediction model will follow the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis + Artificial Intelligence (TRIPOD+AI) statement, which supersedes the 2015 TRIPOD checklist and applies to both regression and machine-learning approaches (see online supplemental table 3).
The results of the study will be disseminated through publication in peer-reviewed scientific journals and presentation at national and international conferences. Findings may also contribute to improving risk stratification strategies and guiding empirical antimicrobial treatment in patients with neutropenia with haematological diseases.
The study was prospectively registered at ClinicalTrials.gov on 30 January 2026 (identifier: NCT07396571).
Discussion
Patients with neutropenia with haematological diseases remain one of the most challenging groups in daily infectious diseases practice. When fever develops, antibiotic treatment usually needs to be started immediately, often before microbiological confirmation is available. In this setting, screening for ESBL-E and CPE may help to guide empirical treatment decisions. Recent 10th European Conference on Infections in Leukaemia (ECIL-10) guidelines recommend broader-spectrum empirical therapy, including new β-lactams and carbapenems for selected patients, such as those who are critically ill or haemodynamically unstable, those with previous colonisation or infection due to resistant Gram-negative bacteria and those managed in settings with a high prevalence of resistance.19 However, the guideline provides limited discussion on how this recommendation should be applied in routine practice, particularly in relation to individual absolute risk. In our view, empirical treatment should not be uniformly broadened for all colonised patients but rather tailored to specific clinical profiles using risk-based tools. In selected patients, alternative strategies, including combination regimens with aminoglycosides for suspected ESBL-E or CPE infection, may also represent a reasonable option. SCREEN-IN was conceived to address this practical question: whether colonisation screening can improve estimation of infection risk and support more individualised empirical treatment decisions in real-world haematology wards.
The study matters because, despite widespread use of rectal-swab surveillance in many centres, comparable data are scarce. Much of the available evidence comes from single-centre experiences, often in institutions with highly integrated electronic systems, automated alerts and decision-support tools.11–13 Those environments do not represent the reality of many public hospitals, where microbiology and prescribing workflows are less automated and where implementing advanced informatics solutions is not straightforward. Furthermore, models available differ from SCREEN-IN in their target populations, predicted outcomes, prediction times and predictor structures, and none incorporates serial colonisation status as a time-varying predictor. Recalibration of previous models would not overcome these fundamental differences, while more extensive revision would effectively constitute the development of a different model. This is particularly relevant in haematology, where therapeutic options are limited and the consequences of both undertreatment and overtreatment are serious. Inadequate empirical therapy in a patient with neutropenia can have an immediate impact on complications and mortality.20 21 On the other hand, systematic escalation to last-line drugs for all colonised patients is not sustainable and carries a clear ecological cost. The key question is not whether colonisation is frequent, but which colonised patients are most likely to develop infection, when that risk peaks and whether surveillance information can distinguish patients who genuinely need broader empirical coverage from those who do not.
SCREEN-IN is designed to provide answers that clinicians can actually use. A risk-based approach is attractive in this context, offering a way to rationalise empirical therapy without relying on ‘one size fits all’ rules, which tend to either miss a minority of high-risk cases or expose a majority to unnecessary broad-spectrum antibiotics. This is also one of the areas where haematology clinicians repeatedly ask for clearer guidance, because even the best international recommendations often have to remain general.
Our study has several limitations. As an observational study, residual confounding is unavoidable. Antibiotic exposure, severity of illness, duration of neutropenia and the likelihood of being screened (or rescreened) are intertwined, and not all these factors can be fully captured in routine data. In addition, the impact of screening depends heavily on local practice: turnaround time, laboratory methods, how results are displayed in the electronic record and how clinicians incorporate those results into prescribing. These centre-level differences can influence both measured outcomes and the apparent effect of screening-informed decisions. From a clinical perspective, the implications are clear. If screening can help identify patients at particularly high risk of ESBL-E/CPE infection, it provides a rational basis for early escalation in those individuals, potentially improving adequacy at the moment it matters most. Conversely, if there are subgroups in whom colonisation does not translate into a high likelihood of infection, that is equally valuable, because it supports more conservative empirical regimens and avoids unnecessary exposure to the newest agents. Either way, the central goal is to make empirical therapy in febrile neutropenia less empirical than it currently is.
The inclusion of 15 hospitals with differing epidemiology, laboratory workflows and prescribing practices should improve transportability across comparable haematology settings. Nevertheless, the cohort is restricted to adults admitted to Spanish haematology wards. Applicability to paediatric patients, non-haematological causes of neutropenia, low-resource or healthcare systems with markedly different resistance ecology will require independent external validation.
Even a well-validated prediction model cannot be assumed to improve care merely because it is made available to clinicians. If SCREEN-IN demonstrates adequate predictive performance and clinical utility, a subsequent clinical impact study will be required to test whether a prespecified implementation strategy changes empirical prescribing and improves patient outcomes. Because use of the model would influence team-level and ward-level behaviour and contamination between individual patients would be likely, a cluster-randomised design comparing model-guided implementation with usual care would be an appropriate preferred approach.22
The microbiome substudy may also provide additional insights into the mechanisms linking intestinal colonisation and infection in patients with neutropenia. We expect that specific patterns of intestinal dysbiosis, such as reduced microbial diversity or the predominance of particular bacterial taxa, may be associated with an increased risk of colonisation and subsequent infection. In the future, this knowledge could open the door to strategies aimed at modulating the intestinal microbiota, such as optimisation of antibiotic exposure, microbiota-preserving approaches or microbiome-based interventions.
In summary, SCREEN-IN addresses a common clinical dilemma in haematology with a pragmatic approach. The study is intended to generate evidence that is applicable under routine conditions and to inform decisions in a population where both the risk of severe infection and the limitations of available antimicrobial options are particularly pronounced.
Study status
The current protocol version is 1.0, dated 9 January 2026. The study protocol was submitted to the Research Ethics Committee in March 2026 and received a favourable opinion on 24 March 2026 (SICEIA-2026-000228). Patient recruitment began in September 2026, following the enrolment of the first participant.
Supplementary material
Acknowledgements
The authors thank the investigators from the 14 participating hospitals in Spain: Tania Arámbula Amaya (Hospital Universitario Virgen de Valme, Sevilla), Manuela Aguilar Guisado (Hospital Universitario Virgen del Rocío, Sevilla), Rubén Lobato Cano (Hospital Universitario de Jerez, Jerez de la Frontera, Cádiz), José Ángel Raposo (Hospital Universitario Puerta del Mar, Cádiz), Ignacio Márquez Gómez (Hospital Regional Universitario de Málaga, Málaga), María Teresa Pérez Rodríguez (Hospital Universitario de Vigo, Pontevedra), Alejandro Avendaño (Hospital Universitario de Salamanca, Salamanca), Juan Carlos Ramos (Hospital Universitario La Paz, Madrid), José Luis del Pozo (Clínica Universidad de Navarra, Pamplona), Pedro González Sierra (Hospital Universitario Virgen de las Nieves, Granada), Alexandra M Aceituno Cano (Hospital Universitario Torrecárdenas, Almería), Susana Ramírez García (Hospital Universitario Juan Ramón Jiménez, Huelva), Adaia Albasanz Puig (Hospital Universitario Vall d’Hebron, Barcelona) and Esther Serrano Conde (Hospital Universitario Ramón y Cajal, Madrid). The authors also thank all the professionals involved in patient care and study procedures at the participating centres, including the haematology, infectious diseases and microbiology teams, as well as laboratory technicians and nursing staff involved in patient screening, sample collection and clinical follow-up, for their contributions.
Footnotes
Funding: The study was funded by the Instituto de Salud Carlos III, AES 2025 (grant PI25/01224) and co-funded by the European Union. The investigators also receive funds for research from the Instituto de Salud Carlos III, Subdirección General de Redes y Centros de Investigación Cooperativa, Ministerio de Ciencia, Innovación y Universidades, CIBERINFEC.
Prepublication history and additional supplemental material for this paper are available online. To view these files, please visit the journal online (https://doi.org/10.1136/bmjopen-2026-121248).
Provenance and peer review: Not commissioned; externally peer reviewed.
Patient consent for publication: Not applicable.
Patient and public involvement: Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
References
- 1.Antimicrobial Resistance Collaborators Global burden of bacterial antimicrobial resistance in 2019: a systematic analysis. Lancet. 2022;399:629–55. doi: 10.1016/S0140-6736(21)02724-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Gustinetti G, Mikulska M. Bloodstream infections in neutropenic cancer patients: A practical update. Virulence. 2016;7:280–97. doi: 10.1080/21505594.2016.1156821. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Willems RPJ, van Dijk K, Vehreschild M, et al. Incidence of infection with multidrug-resistant Gram-negative bacteria and vancomycin-resistant enterococci in carriers: a systematic review and meta-regression analysis. Lancet Infect Dis. 2023;23:719–31. doi: 10.1016/S1473-3099(22)00811-8. [DOI] [PubMed] [Google Scholar]
- 4.Pérez-Galera S, Bravo-Ferrer JM, Paniagua M, et al. Risk factors for infections caused by carbapenem-resistant Enterobacterales: an international matched case-control-control study (EURECA) EClinicalMedicine. 2023;57:101871. doi: 10.1016/j.eclinm.2023.101871. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Cano A, Gutiérrez-Gutiérrez B, Machuca I, et al. Risks of Infection and Mortality Among Patients Colonized With Klebsiella pneumoniae Carbapenemase-Producing K. pneumoniae: Validation of Scores and Proposal for Management. Clin Infect Dis. 2018;66:1204–10. doi: 10.1093/cid/cix991. [DOI] [PubMed] [Google Scholar]
- 6.Cattaneo C, Di Blasi R, Skert C, et al. Bloodstream infections in haematological cancer patients colonized by multidrug-resistant bacteria. Ann Hematol. 2018;97:1717–26. doi: 10.1007/s00277-018-3341-6. [DOI] [PubMed] [Google Scholar]
- 7.Chen X, Wen X, Jiang Z, et al. Prevalence and factors associated with carbapenem-resistant Enterobacterales (CRE) infection among hematological malignancies patients with CRE intestinal colonization. Ann Clin Microbiol Antimicrob. 2023;22:3. doi: 10.1186/s12941-023-00554-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Gallardo-Pizarro A, Lopera C, Peyrony O, et al. Rectal colonization by multidrug-resistant Gram-negative bacteria and subsequent bacteraemia in haematological patients. Clin Microbiol Infect. 2025;31:1579–83. doi: 10.1016/j.cmi.2025.05.033. [DOI] [PubMed] [Google Scholar]
- 9.Giannella M, Trecarichi EM, De Rosa FG, et al. Risk factors for carbapenem-resistant Klebsiella pneumoniae bloodstream infection among rectal carriers: a prospective observational multicentre study. Clin Microbiol Infect. 2014;20:1357–62. doi: 10.1111/1469-0691.12747. [DOI] [PubMed] [Google Scholar]
- 10.Gutiérrez-Gutiérrez B, Salamanca E, de Cueto M, et al. Effect of appropriate combination therapy on mortality of patients with bloodstream infections due to carbapenemase-producing Enterobacteriaceae (INCREMENT): a retrospective cohort study. Lancet Infect Dis. 2017;17:726–34. doi: 10.1016/S1473-3099(17)30228-1. [DOI] [PubMed] [Google Scholar]
- 11.Liu J, Zhang H, Feng D, et al. Development of a Risk Prediction Model of Subsequent Bloodstream Infection After Carbapenem-Resistant Enterobacteriaceae Isolated from Perianal Swabs in Hematological Patients. Infect Drug Resist. 2023;16:1297–312. doi: 10.2147/IDR.S400939. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Zhang Y-Q, Wu W-Q, Xu J, et al. A clinical predictive model for pre-transplantation Klebsiella pneumoniae colonization and relevance for clinical outcomes in patients receiving allogeneic hematopoietic stem cell transplantation. Microbiol Spectr. 2024;12:e0203923. doi: 10.1128/spectrum.02039-23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Gallardo-Pizarro A, Teijón-Lumbreras C, Monzo-Gallo P, et al. Development and Validation of a Machine Learning Model for the Prediction of Bloodstream Infections in Patients with Hematological Malignancies and Febrile Neutropenia. Antibiotics (Basel) 2024;14:13. doi: 10.3390/antibiotics14010013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Gavriilaki M, Sakellari I, Anagnostopoulos A, et al. The Impact of Antibiotic-Mediated Modification of the Intestinal Microbiome on Outcomes of Allogeneic Hematopoietic Cell Transplantation: Systematic Review and Meta-Analysis. Biol Blood Marrow Transplant. 2020;26:1738–46. doi: 10.1016/j.bbmt.2020.05.011. [DOI] [PubMed] [Google Scholar]
- 15.Smith M, Dai A, Ghilardi G, et al. Gut microbiome correlates of response and toxicity following anti-CD19 CAR T cell therapy. Nat Med. 2022;28:713–23. doi: 10.1038/s41591-022-01702-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Vittinghoff E, McCulloch CE. Relaxing the rule of ten events per variable in logistic and Cox regression. Am J Epidemiol. 2007;165:710–8. doi: 10.1093/aje/kwk052. [DOI] [PubMed] [Google Scholar]
- 17.Labarère J, Renaud B, Fine MJ. How to derive and validate clinical prediction models for use in intensive care medicine. Intensive Care Med. 2014;40:513–27. doi: 10.1007/s00134-014-3227-6. [DOI] [PubMed] [Google Scholar]
- 18.Steyerberg EW. Clinical prediction models: a practical approach to development, validation, and updating. New York: Springer; 2009. [Google Scholar]
- 19.Averbuch D, Vanbiervliet Y, Baccelli F, et al. Empirical and targeted antimicrobial therapy in patients with febrile neutropenia and haematological malignancy or after haematopoietic cell transplantation: recommendations from the 10th European Conference on Infections in Leukaemia. Lancet Infect Dis. 2026;26:e232–47. doi: 10.1016/S1473-3099(25)00619-X. [DOI] [PubMed] [Google Scholar]
- 20.Martinez-Nadal G, Puerta-Alcalde P, Gudiol C, et al. Inappropriate Empirical Antibiotic Treatment in High-risk Neutropenic Patients With Bacteremia in the Era of Multidrug Resistance. Clin Infect Dis. 2020;70:1068–74. doi: 10.1093/cid/ciz319. [DOI] [PubMed] [Google Scholar]
- 21.Tang Y, Wu X, Cheng Q, et al. Inappropriate initial antimicrobial therapy for hematological malignancies patients with Gram-negative bloodstream infections. Infection. 2020;48:109–16. doi: 10.1007/s15010-019-01370-x. [DOI] [PubMed] [Google Scholar]
- 22.Giraudeau B, Weijer C, Eldridge SM, et al. Why and when should we cluster randomize? J Epidemiol Popul Health. 2024;72:202197. doi: 10.1016/j.jeph.2024.202197. [DOI] [PubMed] [Google Scholar]