Delirium in the oncology intensive care unit: Risk, recognition, and recovery strategies

Abstract

Delirium represents a severe neurological complication with exceptionally high prevalence in oncology intensive care units (ICU), affecting up to 88% of terminal cancer patients. This acute brain dysfunction emerges from the convergence of critical illness factors with cancer-specific vulnerabilities, creating unique management challenges. This narrative review synthesizes current evidence on delirium epidemiology, risk factors, diagnostic challenges, management strategies, and long-term cognitive outcomes specific to critically ill cancer patients. Delirium demonstrates a profound prognostic impact, increasing ICU mortality odds tenfold and hospital mortality sixfold. Patients experiencing delirium are nearly four times more likely to have cancer treatment modified or discontinued, directly affecting survival. Risk factors unique to this population include chemotherapy neurotoxicity, central nervous system malignancy, paraneoplastic syndromes, and complex medication regimens. While screening tools show strong performance in general ICU settings, validation in cancer populations remains limited. Evidence strongly supports non-pharmacologic multicomponent interventions over pharmacologic approaches, with early mobilization reducing delirium risk by 47%. Long-term cognitive impairment represents a devastating legacy, creating a double burden when superimposed on chemotherapy-related cognitive dysfunction. Management requires specialized preventive approaches prioritizing brain protection through systematic screening, multicomponent prevention bundles, and integrated palliative care.

Key Words: Delirium; Oncology intensive care unit; Cancer critical care; Cognitive dysfunction; Neurological complications; Palliative care; Intensive care unit outcomes; Chemotherapy neurotoxicity; Post-intensive care syndrome; Cancer survivorship

Core Tip: Delirium affects up to 88% of critically ill cancer patients, increasing intensive care units (ICU) mortality tenfold and making post-discharge treatment discontinuation four times more likely. This review introduces the “double burden” of ICU delirium superimposed on chemotherapy-related cognitive impairment, creating unique survivorship challenges. Cancer-specific vulnerabilities include chemotherapy neurotoxicity, central nervous system malignancy, and chimeric antigen receptor T-associated neurotoxicity. Evidence favors non-pharmacologic prevention, with early mobilization reducing delirium by 47%. Delirium thus extends beyond transient confusion to a key determinant of cancer survivorship, potentially transforming ICU admission into a treatment-limiting event.

INTRODUCTION

Delirium manifests as an acute neurological syndrome characterized by disturbances in attention, awareness, and cognition, affecting between 5% and 90% of increasing intensive care units (ICU) patients, depending on the clinical context[1]. In the oncology ICU, this incidence reaches extraordinary levels, with terminal cancer patients experiencing rates approaching 88%[1]. This heightened vulnerability emerges from the convergence of critical illness factors, including sepsis, inflammation, and mechanical ventilation, with cancer-specific elements such as chemotherapy neurotoxicity, central nervous system (CNS) involvement, paraneoplastic syndromes, and high-dose opioid requirements[1]. The syndrome represents more than transient confusion; it constitutes acute organ failure comparable to renal or respiratory dysfunction, carrying profound implications for mortality and healthcare costs[2]. Understanding delirium's unique manifestations and management requirements in the Onco-ICU is essential for optimizing outcomes in this vulnerable population.

Critically, delirium in the oncology ICU differs fundamentally from general ICU populations through the superimposition of cancer-specific neurotoxic insults upon traditional critical illness precipitants. While general ICU patients face delirium risk from sepsis, mechanical ventilation, and sedation alone, critically ill cancer patients simultaneously contend with chemotherapy-induced cognitive dysfunction, paraneoplastic neurological syndromes, immunotherapy-related neurotoxicity, including chimeric antigen receptor T (CAR-T) cell-associated encephalopathy, tumor-related metabolic derangements, and CNS disease burden. This unique pathophysiological convergence creates distinct diagnostic challenges, alters risk stratification, necessitates specialized screening approaches, and demands management strategies specifically tailored to oncology populations. Despite affecting nearly 40% of oncology ICU admissions and profoundly impacting cancer treatment continuation post-discharge, delirium in this population remains inadequately characterized compared to general critical care settings. This review synthesizes evidence specific to critically ill cancer patients, emphasizing cancer-specific vulnerabilities that necessitate specialized prevention and management approaches distinct from standard ICU delirium protocols.

EPIDEMIOLOGICAL BURDEN AND PROGNOSTIC IMPLICATIONS

Incidence patterns

The prevalence of delirium in critically ill cancer patients consistently demonstrates alarming rates. Research examining 135 cancer ICU patients revealed a 39.3% overall incidence[1], while a larger cohort of 915 patients showed 40.5% affected[3]. Mechanical ventilation dramatically escalates risk, with 64.6% of ventilated cancer patients developing delirium[4], approaching or exceeding rates in general ICU populations[1].

Mortality associations

Delirium independently predicts death with striking consistency. Research in advanced cancer patients demonstrated that delirium increased ICU mortality odds more than tenfold with an odds ratio (OR) = 10.75 and 95% confidence interval (CI): 5.91-19.55, P < 0.001, and hospital mortality nearly sixfold with an OR = 5.84 and 95%CI: 4.03-8.46, P < 0.001[5]. This mortality risk extends years beyond discharge, with ICU delirium survivors showing reduced four-year survival probability, particularly among patients over 55[6]. Timing and duration further influence prognosis. Late-onset delirium developing after day three and lasting three or more days carries the worst outcomes, with significantly higher ICU mortality showing an OR = 4.45 with 95%CI: 1.92-10.30, P < 0.001, and hospital mortality with an OR = 2.91 and 95%CI: 1.37-6.19, P = 0.005, compared to early, brief episodes[7].

Healthcare utilization impact

Delirium substantially prolongs ICU and hospital stays[8]. Cancer patients with coronavirus disease 2019 who developed delirium had median ICU stays of 19 days vs 8 days for non-delirious patients[8]. Extended mechanical ventilation duration further increases complications and deconditioning risk[9].

Cancer treatment disruption

Critically, delirium profoundly affects subsequent cancer therapy. Patients experiencing ICU delirium are nearly four times more likely to have cancer treatment modified or discontinued post-discharge with an adjusted OR = 3.80 and 95%CI: 2.72-5.35[10]. This effect, partially mediated by new functional dependence, subsequently increases one-year mortality 2.68-fold[10]. Thus, delirium directly determines cancer survivorship by rendering patients unable to tolerate further oncologic therapy.

PATHOPHYSIOLOGICAL RISK FACTOR CONVERGENCE

Baseline predisposing factors

Advanced age consistently emerges as a primary non-modifiable risk factor, with significantly higher incidence in patients over 60[1]. Each additional year confers incremental risk with an OR = 1.01 per year, P = 0.038[3]. Poor baseline performance status strongly predicts delirium risk[1], while pre-existing cognitive impairment, though often excluded from studies, likely contributes substantially in clinical practice[8]. Chronic conditions, including hypertension and chronic obstructive pulmonary disease, increase vulnerability through inflammation, vascular compromise, and hypoxia susceptibility[11]. Advanced cancer induces systemic catabolism and cachexia, identified as high-intensity risk factors for delirium development[12]. Recent evidence suggests that specific cancer types carry differential delirium risk. Research examining over 10000 hospitalized elderly cancer patients found that pancreatic cancer, leukemia, and oropharyngeal cancer demonstrated significantly higher delirium risk compared to gastric cancer, with hazard ratios of 1.26, 1.24, and 1.30, respectively[13]. Patients with acute promyelocytic leukemia treated with intensive chemotherapy protocols including all-trans retinoic acid face unique challenges, with delirium potentially attributed to both treatment effects and complications, including severe infections and CNS bleeding[14].

Acute precipitating factors

Higher Simplified Acute Physiology Score II, Sequential Organ Failure Assessment, and Mortality Probability Model II scores at admission correlate with increased delirium risk[15]. Sepsis drives neuroinflammation and represents a major independent risk factor[5]. The systemic inflammatory response disrupts blood-brain barrier integrity and neurotransmitter metabolism[12]. Mechanical ventilation increases delirium odds 2.67-fold after adjustment for confounders[15], through both illness severity and intervention-related factors, including sedation and immobility. Benzodiazepines particularly increase risk[1], while opioids represent lower-intensity but significant contributors[12].

Emerging biomarkers and pathophysiological insights

Recent investigations into delirium biomarkers have identified promising candidates for early detection and risk stratification. Systematic reviews demonstrate that inflammatory markers, including C-reactive protein, tumor necrosis factor-alpha (TNF-α), and interleukin-6 (IL-6), show significant elevation among delirious patients with a pooled odds ratio of 1.88[16]. S100β and neurofilament light chain have emerged as the most consistent markers of CNS injury, supporting the role of astrocytic and axonal damage in delirium pathogenesis[17]. Studies examining metabolic ratios found that the neutrophil-to-lymphocyte ratio, platelet-to-lymphocyte ratio, and lymphocyte-to-monocyte ratio may serve as accessible screening biomarkers, particularly in ischemic stroke patients admitted to the ICU[18].

Glial injury markers: Glial injury markers provide critical insights into delirium pathogenesis in cancer populations. S100β, a calcium-binding protein released by activated astrocytes, demonstrates consistent elevation in delirious patients and correlates with blood-brain barrier disruption severity[19,20]. Meta-analysis of postoperative delirium demonstrates that blood-based S100β levels show a significant positive association with delirium prevalence, with a standardized mean difference of 0.40 and 95%CI: 0.11-0.69[19]. Glial fibrillary acidic protein (GFAP), an intermediate filament protein specific to astrocytes, serves as a sensitive marker of astrocytic activation and injury[20]. Both markers reflect astrocytic stress responses and blood-brain barrier (BBB) compromise, with S100β elevations occurring during active delirium episodes while GFAP indicates more persistent glial activation[20]. In oncology patients, these markers hold particular relevance given exposure to neurotoxic chemotherapeutic agents that directly compromise BBB integrity and induce astrocytic stress[21]. The combination of S100β and GFAP may enable early identification of patients at the highest risk for delirium development, particularly following neurotoxic cancer treatments including methotrexate, cytarabine, and platinum-based compounds.

The neuroinflammatory cascade involves complex interactions between peripheral inflammation and CNS dysfunction. Pathologic processes, including hypoxia, infections, and malignancies, produce oxidative stress while compromising the body’s redox systems, potentially causing cerebral damage through this oxidative stress pathway[22]. Furthermore, cholinergic deficiency exhibits synergistic effects with neuroinflammation, with reduced acetylcholine levels enhancing secretion of inflammatory markers, thereby exacerbating the inflammatory response through nuclear factor-kappa B signaling pathways[23].

Oncology-specific vulnerabilities

Primary or metastatic brain disease independently doubles delirium odds with an adjusted odds ratio of 2.25[15]. Postoperative hematoma after neurosurgery particularly predicts occurrence[24]. Though rare, paraneoplastic conditions like anti-N-methyl-d-aspartate receptor encephalitis or limbic encephalitis can present as delirium, sometimes revealing occult malignancy[25,26]. Chemotherapy exposure independently increases risk[1], with agents like methotrexate, cytarabine, and ifosfamide causing direct neurotoxicity[27]. Chronic “chemo-brain” creates baseline vulnerability[28]. Immunotherapy, particularly checkpoint inhibitors, can cause neurological immune-related adverse events presenting as delirium[29]. Pre-admission CNS-active medication use, especially three or more classes, dramatically increases risk with an odds ratio of 11.15[30].

CAR-T cell therapy and immune effector cell-associated neurotoxicity

The advent of CAR-T therapy has introduced a novel neurotoxicity spectrum distinct from traditional delirium. Immune effector cell-associated neurotoxicity syndrome (ICANS) occurs in 20%-70% of CAR-T recipients, with pooled incidences of 26.9% for all-grade and 10.5% for high-grade ICANS[31]. Common manifestations include encephalopathy affecting 21% of patients, aphasia in 10%, and delirium specifically in 6%, typically occurring within the first week following infusion[32]. Risk factors for ICANS development include high disease burden, older age, and specific CAR-T products, with anti-CD19 therapies showing significantly higher incidences than other targets[31].

Mechanistic pathways in ICANS: The pathophysiology of ICANS involves multiple interconnected neurotoxic processes distinct from traditional delirium mechanisms[33,34]. Cytokine-driven neuroinflammation represents the primary initiating event, with massive elevations in IL-6, IL-1, and interferon-gamma (IFN-γ) triggering downstream neurotoxic cascades[33,35]. Following CAR-T cell activation and tumor antigen recognition, activated T cells release IFN-γ, TNF-α, and granulocyte-macrophage colony-stimulating factor, which further activate myeloid cells, including monocytes and macrophages, to produce additional IL-1, IL-6, and TNF-α, establishing a self-amplifying inflammatory loop[34]. These pro-inflammatory cytokines compromise blood-brain barrier integrity through endothelial dysfunction, leading to vasogenic edema and increased vascular permeability that allows peripheral inflammatory mediators and immune cells direct CNS access[33,34,36]. Endothelial activation produces microvascular injury characterized by tight junction disruption mediated through Ras/mitogen-activated protein kinase, phosphatidylinositol 3-kinase/protein kinase B, and Janus kinase/signal transducer and activator of transcription 3 pathways, with pericyte dysfunction and increased transcellular trafficking[34,36]. This BBB compromise facilitates excitotoxicity through altered glutamatergic signaling, with excessive glutamate accumulation causing neuronal overstimulation and calcium-mediated cytotoxicity[37]. Simultaneously, neurotransmitter dysregulation affects both GABAergic and dopaminergic pathways[38]. Reduced GABAergic inhibition contributes to seizure risk observed in severe ICANS, while dopamine pathway disruption manifests as the motor and cognitive features characteristic of ICANS-associated delirium[33,36]. Additional mechanisms include dysfunction of astrocytes affected by IL-1, IL-6, IFN-γ, and TNF-α, leading to further neuroinflammation and neuronal damage[36]. This mechanistic understanding explains why ICANS management differs fundamentally from traditional delirium, requiring immunosuppression with corticosteroids rather than antipsychotics, targeting the underlying inflammatory cascade rather than symptomatic neurotransmitter imbalance[33,35].

Management of ICANS follows distinct protocols from traditional delirium, with corticosteroids representing first-line therapy rather than antipsychotics, while tocilizumab addresses concurrent cytokine release syndrome[39]. The Cornell Assessment for Pediatric Delirium has been adapted for ICANS detection in younger patients, with changes in baseline scores serving as early indicators requiring closer monitoring[40]. Prolonged neurologic symptoms following ICANS resolution affect over 25% of patients, occurring more frequently after high-grade ICANS, though pre-treatment characteristics fail to predict this complication[41]. The pathophysiological convergence of delirium in the oncology ICU is presented in Figure 1.

Figure 1 A schematic illustration showing the multifactorial pathogenesis of delirium in critically ill cancer patients. CRP: C-reactive protein; TNF: Tumor necrosis factor; IL: Interleukin; BBB: Blood-brain barrier; CNS: Central nervous system; ICU: Intensive care units; OR: Odds ratio; CAR-T: Chimeric antigen receptor T-cell; ICANS: Immune effector cell-associated neurotoxicity syndrome; irAEs: Immune-related adverse events; IFN: Interferon; GFAP: Glial fibrillary acidic protein; NMDA: N-methyl-D-aspartate.

CLINICAL RECOGNITION CHALLENGES

Subtype heterogeneity

Delirium manifests across three psychomotor subtypes with distinct prevalence and prognosis[2]. Hyperactive delirium, representing 22.7% of cases, presents the classic picture with agitation and restlessness, being easily recognized but the least common[42]. Hypoactive delirium, comprising 50.3% of cases, is characterized by lethargy and withdrawal, frequently missed due to overlap with cancer-related fatigue, and associated with higher six-month mortality of 32.0% vs 8.7%, P = 0.04[2,42]. Mixed delirium, accounting for 27.7% of cases, shows fluctuating features associated with the longest duration, hospital stays, and highest mortality[42].

Screening tool performance

While the confusion assessment method (CAM)-ICU and intensive care delirium screening checklist demonstrate strong psychometric properties in general ICU populations, with CAM-ICU showing specificity of 95%-96% and sensitivity of 80%-84%[43], validation in Onco-ICU settings remains limited. Concerning data from medical oncology wards showed CAM-ICU sensitivity as low as 18% despite 99% specificity[44], suggesting potential limitations in cancer populations. The challenge intensifies when establishing baseline cognitive status, crucial for accurate diagnosis[26]. Many cancer patients enter the ICU with abnormal baselines from chemotherapy effects, CNS disease, or chronic opioid use, complicating acute change detection.

Differential diagnostic complexity

The Onco-ICU differential includes chemotherapy-induced encephalopathy vs acute delirium[26], paraneoplastic encephalitis[26], structural CNS lesions including metastases and hemorrhage, non-convulsive status epilepticus, cancer-specific metabolic emergencies such as hypercalcemia and tumor lysis syndrome, and withdrawal syndromes.

EVIDENCE-BASED MANAGEMENT FRAMEWORK

Non-pharmacologic multicomponent interventions

Prevention through simultaneous risk factor modification represents the most effective approach[45]. The ABCDEF bundle provides a comprehensive framework addressing modifiable factors through pain assessment and management, spontaneous awakening and breathing trials, sedation minimization, delirium screening and prevention, early mobilization, and family engagement[46]. Early mobilization reduces delirium risk by 47% with an OR = 0.53 and 95%CI: 0.34-0.83, and shortens duration by 1.8 days[47]. Benefits include increased ventilator-free days and improved functional outcomes[48]. Implementation requires consideration of thrombocytopenia, neutropenia, and bone metastases in cancer patients[49]. Multicomponent sleep hygiene bundles effectively reduce delirium through noise and light reduction, care clustering, and circadian rhythm support[50,51]. Melatonin shows promise in medical patients, though ICU evidence remains inconclusive[52]. Family involvement reduces delirium incidence by 54% with a risk ratio of 0.46 and 95%CI: 0.31-0.69, and shortens duration[53]. Direct participation in care exceeds passive visitation benefits[53,54].

Pharmacologic symptom management

No Food and Drug Administration-approved delirium treatments exist, and medications don’t reduce duration or improve mortality[46]. Pharmacotherapy addresses severe symptoms threatening safety or care delivery[55]. Haloperidol remains first-line for agitation despite a lack of efficacy for delirium itself[56]. Recent multicenter trials definitively demonstrate this limitation, with 1000 patients randomized to haloperidol vs placebo showing no difference in days alive and out of hospital at 90 days[57]. Cancer patients face amplified risks from myelosuppression and electrolyte abnormalities predisposing to QTc prolongation[55]. When sedation is necessary, avoid benzodiazepines and prefer propofol or dexmedetomidine[45]. Dexmedetomidine is associated with lower delirium incidence vs benzodiazepines[45].

Interdisciplinary integration

Delirium signals severe illness requiring goals-of-care reassessment[5]. Early palliative care consultation addresses complex symptoms and facilitates crucial discussions[58]. With delirium present in 81% of consulted ICU cancer patients[58], automatic palliative referral ensures comprehensive support[59]. Table 1 provides an overview of predisposing and precipitating risk factors, prognostic impact, and preventive strategies for delirium in oncology ICU patients.

Table 1 Risk factors, outcomes, and management strategies for delirium in oncology intensive care unit patients.

Category	Specific factors	Risk/impact	Prevention/management strategy
Baseline predisposing factors	Age > 60 years	Higher incidence, OR = 1.01 per year (P = 0.038)	Enhanced screening, proactive monitoring
	Poor performance status	Strong predictor	Optimize functional status pre-admission
	Pre-existing cognitive impairment	Substantial contributor	Establish baseline cognitive assessment
	Hypertension, COPD	Increased vulnerability	Manage comorbidities, optimize oxygenation
	Advanced cancer cachexia	High-intensity risk	Nutritional support, symptom management
Cancer-specific factors	Pancreatic cancer	HR = 1.26 vs gastric cancer	Disease-specific risk stratification
	Leukemia	HR = 1.24 vs gastric cancer	Heightened vigilance during treatment
	Oropharyngeal cancer	HR = 1.30 vs gastric cancer	Enhanced monitoring protocols
	CNS malignancy	OR = 2.25	Neurological assessment, imaging
	Paraneoplastic syndromes	Variable	Consider the autoimmune encephalitis workup
Acute precipitating factors	High SAPS II, SOFA scores	Increased risk	Early intervention, severity mitigation
	Sepsis	Major independent risk	Prompt antimicrobial therapy
	Mechanical ventilation	OR = 2.67	Minimize duration, spontaneous breathing trials
	Benzodiazepine use	Particularly high risk	Avoid or minimize, prefer alternatives
Treatment-related	Chemotherapy (methotrexate, cytarabine, ifosfamide)	Direct neurotoxicity	Dose adjustment, nephroprotection
	Immunotherapy checkpoint inhibitors	Neurological immune-related adverse events	Early recognition, corticosteroids
	CAR-T cell therapy	ICANS in 20%-70%	Distinct protocol: Corticosteroids, tocilizumab
	CNS-active medications (≥ 3 classes)	OR = 11.15	Medication reconciliation, deprescribing
Prognostic impact	ICU delirium presence	ICU mortality OR = 10.75, hospital mortality OR = 5.84	Multicomponent prevention bundle
	Late-onset (> day 3), prolonged (≥ 3 days)	ICU mortality OR = 4.45, hospital mortality OR = 2.91	Early detection and intervention
	Delirium in COVID-19 cancer patients	Median ICU stay 19 days vs 8 days	Prevention prioritization
	Post-ICU delirium	Cancer treatment modification OR = 3.80	Rehabilitation, survivorship support
Prevention strategies	Early mobilization	47% risk reduction (OR = 0.53)	Structured mobility protocols
	Family involvement	54% incidence reduction (RR = 0.46)	Liberal visitation, care participation
	Sleep hygiene bundle	Effective reduction	Noise/Light reduction, circadian support
	ABCDEF bundle	Comprehensive risk reduction	Systematic implementation

OR: Odds ratio; COPD: Chronic obstructive pulmonary disease; HR: Hazard ratio; CNS: Central nervous system; SAPS: Simplified Acute Physiology Score; SOFA: Sequential Organ Failure Assessment; CAR-T: Chimeric antigen receptor T-cell; ICANS: Immune effector cell-associated neurotoxicity syndrome; ICU: Intensive care units; COVID-19: Coronavirus disease 2019; RR: Risk ratio.

ECONOMIC BURDEN AND HEALTHCARE RESOURCE UTILIZATION

The economic impact of delirium extends far beyond immediate hospitalization costs, representing a substantial burden on healthcare systems globally. Recent systematic reviews demonstrate that delirium results in adjusted increased costs ranging from $1532 to $22269 per patient, varying by included cost categories, country, and hospital department[60]. The patient-level 30-day cumulative cost of ICU delirium attributable to increased resource utilization reaches $17838, with professional services, dialysis, and bed costs accounting for the largest percentage of incremental expenses[61]. Notably, these costs would be approximately 20% higher if not for delirium-associated early mortality, suggesting the true economic burden is partially masked by increased death rates[61]. In cancer populations specifically, the economic consequences are particularly severe. Delirium in acute respiratory failure patients, a common ICU presentation in cancer patients, is associated with hospitalization costs of $15395 vs $9393 in non-delirious patients, representing a 49% increase after adjustment for confounders[62]. Long-term financial implications compound these acute costs, with one-year healthcare costs following delirium substantially exceeding those of matched controls[63].

Meta-analyses of economic evaluations reveal that multicomponent non-pharmacological interventions demonstrate cost-effectiveness, with a pooled incremental net benefit of $8014 per patient[64]. The Hospital Elder Life Program, when implemented systematically, saved $831 per intervention patient for acute hospital costs and $9446 per patient per year in long-term nursing home costs, with annual savings reaching $7.4 million across six hospital units after accounting for program costs[65]. These findings underscore that despite the substantial investment required for comprehensive delirium prevention programs, the return on investment through reduced complications and shortened stays justifies their implementation.

LONG-TERM COGNITIVE OUTCOMES AND SURVIVORSHIP

Post-intensive care syndrome

ICU delirium independently predicts lasting cognitive impairment, with duration directly correlating with worse three-month and twelve-month cognitive performance[66]. Delirium increases new dementia diagnosis odds 5.4-fold and is associated with worse objective cognitive testing[67]. Psychiatric sequelae, including depression, anxiety, and substance disorders, frequently emerge[68].

The "double burden" in cancer survivors

Cancer patients often enter the ICU with chemotherapy-related cognitive impairment. Superimposed critical illness delirium creates compounded dysfunction, potentially more severe than in previously healthy patients. This cognitive burden impairs treatment comprehension and decision-making capacity, informed consent provision, medication adherence, and side effect recognition and reporting. This cascade undermines the ICU admission's purpose of enabling continued cancer therapy by creating cognitive barriers to treatment continuation[10], ultimately affecting cancer survival.

Cognitive rehabilitation and recovery strategies

Emerging evidence supports cognitive rehabilitation interventions for ICU survivors experiencing persistent deficits. Studies demonstrate that survivors typically experience stark and dramatic cognitive decline rather than gradual deterioration, with patients reporting being substantially different after critical illness[69]. While cognitive impairment often remains severe at hospital discharge, modest improvements occur over time before stabilizing, with most patients demonstrating persistent difficulties at remote timepoints[69]. Promising rehabilitation strategies include computerized cognitive training programs, which have shown efficacy in populations with chemotherapy-related cognitive impairment and are now being tested in critical illness survivors[70]. The Early Cognitive Intervention in Delirium study examines whether cognitive training initiated within 24 hours improves four-month global cognition in delirious elderly patients, including those with pre-existing dementia[71]. Additional approaches under investigation include remote neurofeedback training platforms and combined interventions incorporating physical exercise with brain training to optimize recovery trajectories[71]. These multimodal rehabilitation programs recognize that cognitive recovery requires addressing both neurological and physical deconditioning simultaneously, particularly important for cancer patients facing the dual challenges of critical illness recovery and ongoing oncologic treatment. Prevention and Management of Delirium in the ICU provides additional comprehensive strategies for addressing this complex syndrome[72].

CLINICAL RECOMMENDATIONS

Based on synthesized evidence, key recommendations include implementing universal enhanced screening with twice-daily CAM-ICU coupled with robust baseline cognitive assessment using collateral history. Standardized prevention requires adopting the ABCDEF bundle, emphasizing benzodiazepine avoidance, sleep optimization, early mobilization, and family engagement. Early palliative integration should occur automatically for new delirium diagnoses, particularly in advanced disease. Recovery planning must screen survivors for persistent impairments and provide rehabilitation referrals and survivorship program integration.

FUTURE RESEARCH PRIORITIES

Critical knowledge gaps require investigation, including large-scale Onco-ICU validation of CAM-ICU and intensive care delirium screening checklist against psychiatric assessment gold standards. Biomarker development should identify markers differentiating delirium etiologies to guide targeted therapy. Tailored intervention studies must test multicomponent bundles adapted for cancer-specific challenges. Survivorship trajectory research requires longitudinal studies tracking cognitive recovery and cancer treatment impacts.

CONCLUSION

Delirium in the Onco-ICU represents a devastating complication transcending transient confusion to determine short-term mortality, long-term survival, and cancer treatment feasibility. The unique convergence of critical illness and cancer-specific vulnerabilities demands specialized preventive approaches prioritizing brain protection through evidence-based, non-pharmacologic interventions. By implementing systematic screening, multicomponent prevention bundles, and integrated palliative care while advancing targeted research, clinicians can begin mitigating this silent epidemic's impact on critically ill cancer patients.