The study, led by Camilla Genovese of the Department of Infectious Diseases at Ospedale Luigi Sacco in Milan, together with Jean-François Timsit and Michael Thy of Bichat-Claude Bernard University Hospital and INSERM research units in Paris, set out to map how intensivists and nuclear medicine physicians around the world actually use 2-deoxy-2-[18F] fluoro-D-glucose PET/CT, or [18F]-FDG-PET/CT, in critically ill patients. The researchers distributed an electronic questionnaire worldwide through two of the most influential societies in the field: the European Society of Intensive Care Medicine, known as ESICM, and the European Association of Nuclear Medicine, known as EANM. The instrument collected demographic information, hospital characteristics, the availability and physical location of nuclear medicine services, the clinical indications for which PET/CT is requested in intensive care, the frequency of use, details of workflow organisation, and the obstacles that clinicians perceive when considering the test.

Forty-one respondents from 16 countries completed the survey. Although the sample is modest, its composition is telling. Most participants, 63 percent, worked in university hospitals, and the respondents divided almost neatly into two professional worlds: 71 percent specialised in intensive care medicine while 29 percent specialised in nuclear imaging. That split mirrors the very fault line the study set out to explore. PET/CT in the critically ill sits at the intersection of two specialties, and the survey suggests that neither has yet built the shared routines that would make the examination routine. A nuclear medicine department was available in 78 percent of the centres that responded, which sounds reassuring until one examines the geography: only 60 percent of those departments were co-located with the intensive care unit. In a patient whose blood pressure is maintained by infusions of vasopressors and whose oxygenation depends on a mechanical ventilator, distance is not a trivial detail. Every metre of corridor is a corridor of risk.

The raw numbers on utilisation are striking. Among centres that do use the technique, 43 percent performed only one to five scans per year, and a quarter of respondents performed just one or two scans per month. In an environment where diagnostic uncertainty is a daily companion, those figures suggest that a technology capable of resolving that uncertainty is being left largely on the shelf. The survey also asked what the scans were actually being used for. Infection topped the list, serving as the primary diagnostic indication in 74 percent of cases, followed by suspected malignancy at 57 percent and fever of unknown origin at 49 percent. The pattern makes physiological sense. Critically ill patients frequently harbour deep-seated infections, ventilator-associated pneumonia, catheter-related bloodstream infection, intra-abdominal sepsis, that conventional imaging fails to localise. Because activated immune cells and many tumour cells consume glucose avidly, the fluorinated glucose analogue accumulates precisely where the disease hides, and the paired CT supplies the anatomical coordinates clinicians need to intervene.

How, then, does the examination actually proceed in practice? The survey provides a snapshot of workflow that is as revealing as it is fragmented. Radiotracer administration took place mainly within the nuclear medicine department, in 77 percent of sites, meaning that the unstable patient had to be physically transported for the injection even before the imaging itself. Protocols for the dietary preparation that PET/CT demands, typically fasting to suppress endogenous insulin and ensure that glucose competition does not blunt tracer uptake, existed in only half of the participating centres. The absence of such standardisation matters at a technical level. Blood glucose levels that are too high reduce FDG uptake in target tissues and can degrade diagnostic sensitivity, which in a septic patient on parenteral nutrition is a genuine challenge. That only 50 percent of sites had formal preparation protocols suggests that much of the current practice is improvised rather than codified, and that image quality, and therefore diagnostic yield, may vary considerably from one centre to another.

When the researchers asked respondents directly what stood in the way, a coherent picture of systemic friction emerged. The barriers most commonly reported included the lack of standardised protocols, the prolonged duration of the procedure, limited imaging slots, ambiguous clinical indications, insufficient staff expertise and the logistical difficulties of transporting unstable intensive care patients. Each barrier deserves unpacking. A PET/CT examination is not a five-minute chest radiograph; it involves tracer production or delivery, injection, a waiting period of roughly an hour for uptake, and then acquisition on a scanner that is often shared with oncology services whose demand for slots is enormous. Staff expertise is a second constraint: positioning, monitoring and safely transporting a ventilated, sedated patient require a coordinated team of intensivists, nurses, radiographers and nuclear medicine physicians, and many centres simply lack experience in this choreography. Ambiguous indications compound the problem. Because no formal guidelines exist to direct the use of [18F]-FDG-PET/CT in the intensive care unit, clinicians have no authoritative framework telling them when the scan will genuinely change management and when it will merely delay it.

That absence of guidance is perhaps the survey’s most consequential finding. The authors emphasise that routine use of the technique in intensive care remains sporadic and that there are, at present, no formal guidelines governing its application. Medicine has repeatedly shown that imaging technologies with real diagnostic power underperform when their indications are left to individual judgement. The result is a vicious circle: because guidelines are lacking, use is rare; because use is rare, centres never develop the protocols, training and logistics that would make scans safe and efficient; and because the infrastructure never matures, the evidence base needed to write guidelines remains thin. The survey’s data, collected from both sides of the specialty divide, provide the empirical foundation needed to break that circle.

The study’s authors are explicit about the trajectory they hope the findings will set. The data, they write, highlight an unmet need for consensus guidelines and streamlined workflows to optimise the diagnostic value of [18F]-FDG-PET/CT in critical-care patients, and the results will inform forthcoming joint recommendations to be issued by ESICM and EANM. Such recommendations, grounded in the real-world practice patterns the survey documented, could specify which patients are most likely to benefit, how dietary preparation and glucose control should be managed, how transport of ventilated patients should be organised and staffed, and how scanner time might be reserved for intensive care indications. The stated aim is ultimately clinical rather than technical: to enhance patient outcomes through evidence-based imaging protocols. In intensive care, where every day of undiagnosed infection carries a measurable mortality cost, a scan that localises the source of sepsis and redirects therapy is not a luxury but a potential lifeline.

The survey is not without limitations, and the authors are careful about what it can and cannot claim. It is a self-reported questionnaire study with 41 respondents, and it captures practice rather than outcome: no one can yet say from these data whether PET/CT in the intensive care unit improves survival, shortens ventilation or reduces antibiotic exposure. The study involved no direct interaction with human subjects and relied entirely on secondary data, an ethical design that facilitated its broad international reach. What it delivers instead is a baseline, a map of where the world currently stands. It reveals a technology of proven diagnostic sensitivity for the very conditions that complicate critical illness, used only a handful of times a year in most centres, hampered by distance, duration, cost, protocol gaps and uncertainty about when it helps. Closing that gap between capability and practice is now a defined task, and the joint guidelines that ESICM and EANM are preparing will be the first serious attempt to define it. If they succeed, the hidden infections and tumours that PET/CT can illuminate may finally be found, not in the rarest of cases, but wherever a critically ill patient needs them found.

Subject of Research: Use of FDG PET/CT imaging in critically ill intensive care patients

Article Title: Practices and barriers in PET/CT use for critically ill patients: an international survey

Article References: Genovese, C., Timsit, J.-F., & Thy, M. (2026). Practices and barriers in PET/CT use for critically ill patients: an international survey. European Journal of Nuclear Medicine and Molecular Imaging. https://doi.org/10.1007/s00259-026-08172-y

Image Credits: AI Generated

DOI: 10.1007/s00259-026-08172-y

Keywords: PET/CT, FDG PET, intensive care, critical care imaging, nuclear medicine, occult infection, fever of unknown origin, international survey, ESICM, EANM, diagnostic imaging, clinical guidelines

Outside critical care, a growing body of evidence has linked clinician gender to differences in how care is delivered. Studies have found that patients treated by female physicians show better chronic disease control and, in hospital medicine, lower mortality and readmission rates. Consensus recommendations, meanwhile, are only as robust as the panel that drafts them, and research on such panels has shown that seniority, professional role, and gender can all shape how much an individual contributes to the final document. Against this backdrop, researchers led by Dr. Nastaran Sharifian and assisted by medical student Aoife Bates at University of Galway, working with international collaborators, set out to determine whether a panelist’s gender influenced how they rated proposed recommendations for the management of very old patients in the intensive care unit.

The underlying material came from a recent international Delphi process convened by the European Society of Intensive Care Medicine. A steering group of 28 specialists drafted 59 candidate statements covering goals of care, ICU admission, treatment decisions, and post-discharge management. A further 82 experts then rated their agreement across two anonymous rounds, and consensus was ultimately reached on 48 of the statements. The full panel was predominantly medical, with roughly 85 percent of members being doctors, drawn mainly from intensive care, though geriatric and emergency medicine, clinical pharmacy, and allied health professions were also represented, reflecting the multidisciplinary reality of caring for very old patients.

The secondary analysis, published online on August 8, 2026, in the Journal of Intensive Medicine, focused on the 128 panelists who reported both their gender and clinical background, comprising 53 female and 75 male respondents. The investigators concentrated on the 26 statements that reached only moderate consensus, where opinion had diverged most during the Delphi process, and compared agreement scores on a scale from zero to 100. Importantly, what was measured throughout was how strongly panelists endorsed a proposed recommendation, an expression of opinion rather than a record of what any clinician actually does at the bedside.

The dominant finding was agreement. Across the great majority of the proposed recommendations, there was no significant difference between female and male respondents, and in analyses adjusted for provider category and age, independent effects of gender were limited. Interactions between gender and specialty were almost all non-significant, indicating a pattern that held consistently across intensivists, emergency physicians, and geriatricians. Where differences did emerge, they clustered in a single domain: multidisciplinary input. Female respondents recorded higher agreement on five statements concerning non-physician involvement, including medication review by a pharmacist, geriatrician, or other specialist, scoring 91.9 compared with 81.0 for male respondents, a statistically significant gap. Female panelists also rated clinical pharmacist involvement more highly, at 90.4 versus 81.8, as well as dietitian input to optimize nutrition, at 93.0 versus 85.6, speech and language therapy advice, at 88.9 versus 80.6, and transfer from the ICU to a geriatric ward where available, at 86.9 versus 78.7.

Male respondents recorded higher agreement on just one statement: that decisions to limit life-sustaining therapy should be based on expected functional outcome, scoring 87.8 compared with 81.2 for female respondents. The gender gaps widened further when the researchers examined who agreed emphatically rather than simply agreed. Two-thirds of female respondents strongly endorsed dietitian involvement, against 37.8 percent of male respondents, and for clinical pharmacist involvement the figures were 54.7 percent and 27 percent respectively. Other domains showed no separation at all, including ethics consultation, quality of life as a long-term outcome, and hospital readmissions.

The authors are careful to frame the work as an exploratory secondary analysis intended to generate hypotheses rather than settle questions. Twenty-six statements were compared without correction for multiple testing, so while six significant results exceed what chance alone would predict, most individual differences would not withstand a conservative statistical correction. The comparison groups are also small, with roughly 51 female and 71 male respondents rating each statement, meaning the estimates carry wide margins of uncertainty. Furthermore, the recommendations showing gender differences are those concerning non-physician disciplines, and nursing and allied health professions are predominantly female. Because the analysis adjusted for medical specialty rather than professional role, it cannot fully separate a genuine gender effect from the simple tendency of practitioners to endorse their own discipline. These are ratings of proposed recommendations, not observations of care actually delivered to patients.

What the analysis does suggest is that stated support for recommendations on collaborative care varied modestly with who was asked, while the broad architecture of the recommendations commanded consensus regardless of the respondent’s gender. For those convening guideline panels, the finding underscores the need for broad representation, since the composition of an expert group may subtly shape which aspects of care receive the strongest endorsement. Whether these attitudinal differences translate into measurable differences in patient outcomes remains an open question that will require direct observational and interventional research.

As very old patients continue to make up a growing share of intensive care admissions worldwide, understanding how expert consensus forms, and where it frays, becomes increasingly consequential. This study offers reassurance that on the fundamentals of caring for patients over 80 in the ICU, female and male specialists across specialties are largely reading from the same page, while highlighting that the value placed on pharmacists, dietitians, therapists, and geriatric wards is where perspectives most often diverge.

The Delphi technique used to generate the underlying recommendations deserves some explanation, because its design shapes what the secondary analysis could detect. In a Delphi process, experts rate statements anonymously and iteratively, with feedback between rounds, so that consensus emerges without the pressure of face-to-face debate or hierarchy. Anonymity is intended to equalize contributions, yet prior research on expert panels suggests that demographic characteristics, including gender, can still influence how individuals rate proposals. That makes the near-universal agreement observed here across female and male panelists noteworthy, even if the divergence on multidisciplinary statements was the exception rather than the rule.

The specific statements that divided respondents map closely onto well-documented vulnerabilities of the very old intensive care population. Polypharmacy is common among patients over 80, many of whom arrive in intensive care already taking multiple medications that may be inappropriate, duplicative, or poorly tolerated in the setting of acute organ dysfunction. Structured medication review by a pharmacist or geriatrician is one of the most consistently recommended interventions in geriatric care, and malnutrition and swallowing difficulties are similarly prevalent in this age group, which helps explain why dietitian and speech and language therapy input featured among the candidate statements at all. The placement of these services within intensive care varies widely between institutions, so attitudes toward them may reflect professional experience with how well, or how poorly, such support is currently delivered.

The single statement on which male respondents agreed more strongly, that decisions to limit life-sustaining therapy should rest on expected functional outcome, touches on one of the most contested questions in geriatric critical care. Functional outcome as a criterion for withholding or withdrawing treatment raises difficult issues about how recovery is predicted, whose definition of acceptable function prevails, and how such judgments weigh against patient values and family perspectives. Prior studies of end-of-life decision-making in intensive care have reported varied findings on clinician gender and attitudes toward treatment limitation, and the modest difference observed here, on a single statement with a small gap, should be interpreted with corresponding caution.

The statistical structure of the analysis also merits emphasis for readers weighing its conclusions. When many comparisons are made simultaneously, a handful will reach conventional significance thresholds by chance alone. The researchers note that six significant results out of twenty-six comparisons exceed what random variation would be expected to produce, which lends the overall pattern some credibility, but the individual estimates remain imprecise given the modest number of respondents per statement. The adjustment for provider category and age, while useful, could not account for every characteristic that might shape a panelist’s response, such as years of experience, country of practice, or the intensity of their exposure to allied health services.

Looking forward, the most productive extension of this work would move from stated opinions to observed practice. Linking panelist characteristics or, more feasibly, clinician characteristics in routine practice to objective measures such as medication review completion, dietitian referral rates, rehabilitation uptake, and ultimately survival and functional recovery would test whether the attitudinal signal detected here has any clinical footprint. Studies of physician gender and patient outcomes suggest such links are possible, though the mechanisms remain debated. Until then, the practical takeaway for guideline developers is procedural rather than clinical: assembling panels that genuinely represent the multidisciplinary teams who care for very old patients may be the surest way to ensure that recommendations about collaborative care reflect the full range of expertise that such care depends upon.

Subject of Research: The influence of clinician gender on agreement with intensive care recommendations for patients aged 80 and over

Article Title: Clinician gender and agreement with intensive care unit recommendations for patients aged 80 and over

Article References: Clinician gender and agreement with intensive care unit recommendations for patients aged 80 and over. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: intensive care, clinician gender, Delphi consensus, geriatrics, ESICM, very old patients, multidisciplinary care, critical care, Journal of Intensive Medicine, agreement ratings, frailty, guideline panels