Anxiety in Medical Settings: Preoperative Anxiety, ICU Delirium, Cardiac Anxiety, Cancer-Related Distress, and Integrated Management Approaches

Anxiety in medical settings represents one of the most prevalent yet systematically undertreated dimensions of patient suffering. Unlike primary anxiety disorders that emerge in community contexts, medically contextualized anxiety arises at the intersection of physiological threat, procedural uncertainty, loss of autonomy, and existential confrontation with mortality. These forms of distress are not simply psychological epiphenomena of illness — they are biologically active states that alter surgical outcomes, prolong ICU stays, increase cardiac event recurrence, and accelerate cancer progression through well-characterized neuroimmune and neuroendocrine pathways.

The prevalence figures are striking. Preoperative anxiety affects 60–80% of surgical patients, with 10–30% experiencing clinically severe levels. ICU delirium — a syndrome in which anxiety is both a prodromal feature and a sustaining mechanism — occurs in 20–50% of general ICU patients and up to 80% of mechanically ventilated patients. Cardiac anxiety, defined as fear and hypervigilance centered on cardiac sensations, affects 20–40% of patients following acute coronary syndrome (ACS). Cancer-related distress meeting clinical thresholds for intervention is documented in 30–45% of patients across tumor types, with rates exceeding 60% in certain populations such as those with head and neck, pancreatic, or lung cancers.

Despite these figures, screening rates remain inadequate. The landmark study by Zabora and colleagues (2001), which assessed psychological distress in over 4,000 cancer patients across 14 diagnostic groups, found that fewer than 10% of distressed patients were referred for psychosocial care. Similar gaps exist across surgical, cardiac, and critical care contexts. This article provides a detailed clinical review of the neurobiology, epidemiology, diagnosis, and evidence-based management of anxiety across these four major medical domains, with an emphasis on integrated care models that bridge the traditional divide between psychiatric and medical services.

The neurobiology of anxiety in medical settings involves overlapping but distinct mechanisms from primary anxiety disorders. Understanding these pathways is essential for targeted intervention.

The HPA Axis and Cortisol Dysregulation

Medical illness and surgical stress activate the hypothalamic-pituitary-adrenal (HPA) axis, resulting in sustained cortisol elevation. Preoperatively, cortisol levels can increase by 50–100% above baseline, with peak elevations correlating with self-reported anxiety severity. Chronic HPA axis activation, as seen in prolonged ICU stays and advanced cancer, leads to glucocorticoid receptor downregulation in the hippocampus and prefrontal cortex (PFC), impairing negative feedback and producing a self-reinforcing cycle of cortisol excess. This mechanism underlies the cognitive impairment and emotional dysregulation observed in ICU delirium and cancer-related distress.

The Amygdala-Prefrontal Circuit

The basolateral amygdala (BLA) serves as the primary threat detection hub, receiving sensory input and generating fear responses via projections to the central nucleus of the amygdala (CeA), which activates autonomic and endocrine stress responses. In healthy individuals, the ventromedial prefrontal cortex (vmPFC) provides top-down inhibition of amygdalar output. Medical illness disrupts this regulatory circuit through multiple pathways: inflammatory cytokines (particularly IL-6, TNF-α, and IL-1β) cross the blood-brain barrier and directly increase amygdalar excitability while impairing PFC function. Neuroimaging studies in cardiac patients have demonstrated persistent amygdalar hyperactivation months after acute coronary events, even after controlling for depressive comorbidity.

Neuroinflammation and the Cytokine Model

The neuroinflammatory hypothesis has become central to understanding anxiety in medical settings. Surgical trauma triggers a systemic inflammatory response with IL-6 levels peaking 6–12 hours postoperatively and remaining elevated for 48–72 hours. In the ICU, sustained inflammation driven by sepsis, tissue injury, and mechanical ventilation produces a neuroinflammatory cascade involving microglial activation, increased blood-brain barrier permeability, and disruption of cholinergic neurotransmission. The cholinergic deficiency hypothesis of delirium, supported by the observation that anticholinergic drug burden is one of the strongest pharmacologic predictors of ICU delirium, posits that reduced acetylcholine activity in cortical and subcortical circuits impairs attention, arousal regulation, and reality testing — core features of the delirious state.

Autonomic Nervous System Dysregulation

Cardiac anxiety is characterized by a specific pattern of autonomic imbalance: sympathetic hyperactivation with reduced parasympathetic (vagal) tone, measurable via heart rate variability (HRV). Reduced HRV is both a consequence of cardiac anxiety and an independent predictor of adverse cardiac outcomes, creating a pathophysiological feedback loop. The insula, particularly the right anterior insula, plays a critical role in interoceptive processing — the brain's representation of internal bodily states. Hyperactivity of insular cortex in cardiac patients leads to amplified perception of cardiac sensations, which fuels catastrophic interpretation and behavioral avoidance.

Genetic and Epigenetic Vulnerability

Genetic factors modulate susceptibility to anxiety in medical contexts. The serotonin transporter gene-linked polymorphic region (5-HTTLPR) short allele has been associated with heightened amygdalar reactivity to threat and increased risk of anxiety and depression following medical stressors. Polymorphisms in the COMT gene (Val158Met), which influences prefrontal dopamine catabolism, modulate stress resilience and cognitive flexibility under medical threat. Emerging epigenetic research suggests that prior trauma exposure can methylate the FKBP5 gene promoter region, altering glucocorticoid receptor sensitivity and predisposing individuals to exaggerated HPA axis responses during subsequent medical crises.

Preoperative anxiety is the most common psychological response to impending surgery, and its clinical significance extends far beyond subjective distress. It directly modulates physiological processes that affect anesthetic requirements, intraoperative hemodynamic stability, postoperative pain, wound healing, and recovery trajectory.

Epidemiology and Measurement

Studies consistently report that 60–80% of preoperative patients experience some degree of anxiety, with 10–30% meeting criteria for clinically significant anxiety depending on the surgical procedure and measurement instrument used. The Amsterdam Preoperative Anxiety and Information Scale (APAIS) and the State-Trait Anxiety Inventory (STAI) are the most widely validated screening tools. The Visual Analog Scale for Anxiety (VAS-A), while less comprehensive, offers practical advantages in time-pressured preoperative environments and correlates at r = 0.55–0.75 with the STAI-State subscale.

Predictors of Severe Preoperative Anxiety

Robust predictors include female sex (OR 1.5–2.5 across studies), younger age, prior negative surgical or anesthetic experiences, preexisting anxiety or mood disorders, low perceived control and information deficit, and anticipated pain severity. ASA physical status class itself is inconsistently associated with preoperative anxiety, suggesting that subjective threat appraisal, rather than objective surgical risk, is the primary driver.

Impact on Outcomes

The clinical consequences of unmanaged preoperative anxiety are well documented:

• Increased anesthetic requirements: Anxious patients require 20–30% higher doses of propofol and volatile anesthetics for induction and maintenance, likely mediated by elevated catecholamine levels reducing anesthetic sensitivity.
• Postoperative pain: Preoperative anxiety is one of the strongest psychological predictors of acute postoperative pain intensity, with effect sizes (Cohen's d) in the range of 0.4–0.7 across meta-analytic reviews. This contributes to higher opioid consumption and increased risk of chronic postsurgical pain, which develops in 10–50% of patients depending on procedure type.
• Wound healing: Kiecolt-Glaser and colleagues demonstrated in a seminal series of studies that psychological stress impairs wound healing by 25–40%, mediated by cortisol-driven suppression of proinflammatory cytokines essential for the inflammatory phase of tissue repair.
• Length of stay and recovery: Systematic reviews indicate that preoperative anxiety is associated with longer hospital stays (mean increase of 0.5–1.5 days) and delayed return to functional activities.

Pharmacologic Management

Benzodiazepines (midazolam 1–2 mg IV, lorazepam 0.5–2 mg PO) remain the most commonly used anxiolytics in preoperative settings. They provide rapid onset anxiolysis and anterograde amnesia, though concerns about postoperative cognitive effects in older adults and potential for paradoxical agitation have driven exploration of alternatives. Gabapentin (300–1200 mg) and pregabalin (75–300 mg) administered preoperatively have shown anxiolytic effects comparable to benzodiazepines in several RCTs, with the added benefit of reducing postoperative opioid consumption by 20–35%. A meta-analysis by Mishriky and colleagues (2015) found that preoperative pregabalin reduced VAS-anxiety scores with a standardized mean difference of −0.73 (95% CI: −1.04 to −0.42). Dexmedetomidine, an α2-adrenergic agonist, provides anxiolysis without respiratory depression and is increasingly used for intranasal or intravenous preoperative sedation, though its hypotensive effects require monitoring.

Non-Pharmacologic Interventions

Structured preoperative education and information provision reduces anxiety with effect sizes of d = 0.3–0.5. Music therapy has been examined in over 80 RCTs; a Cochrane review by Bradt, Dileo, and Shim (2013) found that perioperative music reduces anxiety (SMD = −0.69, 95% CI: −0.89 to −0.49) and reduces postoperative pain and analgesic requirements. Guided imagery and brief cognitive-behavioral interventions show promise but are limited by implementation challenges in acute surgical settings. Virtual reality (VR) distraction is an emerging intervention with early RCT data showing 30–50% reductions in preoperative anxiety scores, though the evidence base is not yet mature.

ICU delirium occupies a unique position in the spectrum of medical anxiety: it is simultaneously a consequence of physiological stress, a manifestation of acute brain dysfunction, and a condition in which anxiety features prominently as both prodrome and sustaining factor. The clinical significance of ICU delirium has been transformed by landmark research demonstrating its association with increased mortality, prolonged hospitalization, and long-term cognitive impairment.

Epidemiology

ICU delirium affects 20–50% of non-ventilated ICU patients and 60–80% of mechanically ventilated patients. The BRAIN-ICU study (Pandharipande et al., 2013), published in the New England Journal of Medicine, followed 821 ICU patients and found that 74% developed delirium during their ICU stay. Longer delirium duration was independently associated with worse global cognition and executive function at 3 and 12 months post-discharge, with cognitive deficits comparable to mild Alzheimer's disease or moderate traumatic brain injury.

Diagnostic Challenges

ICU delirium is profoundly underdiagnosed, with detection rates as low as 25–35% by clinical impression alone without structured assessment tools. The Confusion Assessment Method for the ICU (CAM-ICU) and the Intensive Care Delirium Screening Checklist (ICDSC) are the recommended instruments. The CAM-ICU has sensitivity of 80–95% and specificity of 89–100% when performed by trained nurses, though sensitivity drops substantially when applied by untrained staff. A critical diagnostic pitfall is the failure to recognize hypoactive delirium, which accounts for 40–70% of ICU delirium cases and presents with reduced alertness, psychomotor slowing, and withdrawal rather than the easily recognized agitation of hyperactive delirium. Hypoactive delirium carries a worse prognosis — patients have higher mortality rates and are more likely to be discharged to long-term care facilities — yet it is the subtype most frequently missed.

Pathophysiology: Beyond Cholinergic Deficiency

The neurobiology of ICU delirium is multifactorial:

• Cholinergic deficiency: Reduced acetylcholine synthesis due to hypoxia, metabolic derangements, and anticholinergic medications impairs cortical arousal and attentional processing.
• Dopaminergic excess: Relative dopamine excess, particularly in mesolimbic circuits, contributes to hallucinations, psychomotor agitation, and impaired reality testing. This forms the rationale for antipsychotic use, though clinical trial evidence has challenged this approach.
• Neuroinflammation: Systemic inflammation drives microglial activation, disrupts the blood-brain barrier, and produces direct neuronal injury. CSF studies in delirious ICU patients show elevated S100B (a marker of astrocytic damage) and increased IL-6 and IL-8 concentrations.
• Sleep architecture disruption: ICU patients lose normal circadian organization of sleep, with fragmentation, loss of slow-wave sleep, and reduction of REM sleep. This impairs glymphatic clearance of neurotoxic metabolites and exacerbates cognitive dysfunction.

Prevention and Treatment: The ABCDEF Bundle

The ABCDEF bundle represents the current standard of care for ICU delirium prevention and management:

• A — Assess, prevent, and manage pain
• B — Both spontaneous awakening and breathing trials
• C — Choice of analgesia and sedation (favoring light sedation protocols and avoiding benzodiazepines)
• D — Delirium assessment and management
• E — Early mobility and exercise
• F — Family engagement and empowerment

Implementation of the ABCDEF bundle has been associated with reductions in delirium incidence of 25–50%, decreased ventilator days, and lower ICU mortality. The MIND-USA trial (Girard et al., 2018), an NICHD-funded RCT comparing haloperidol, ziprasidone, and placebo for ICU delirium treatment, found that neither antipsychotic improved delirium duration, coma-free days, or survival compared to placebo, fundamentally challenging the routine use of antipsychotics for delirium treatment. Dexmedetomidine has shown more promise: the MENDS and SEDCOM trials demonstrated that dexmedetomidine-based sedation reduces delirium prevalence compared to benzodiazepine-based sedation (prevalence 54% vs. 77%, p < 0.001 in SEDCOM). The more recent MENDS2 trial (Hughes et al., 2021) compared dexmedetomidine to propofol sedation in sepsis-associated delirium and found no significant difference in days alive without delirium or coma, suggesting the benefit of dexmedetomidine may be driven primarily by avoidance of benzodiazepines rather than intrinsic anti-delirium properties.

Cardiac anxiety describes a syndrome of heightened fear, body vigilance, and avoidance behavior centered on perceived cardiac threat. It emerges most characteristically following acute coronary syndrome (ACS), cardiac surgery, or implantable cardioverter-defibrillator (ICD) placement, though it also occurs in patients with benign cardiac presentations such as noncardiac chest pain or supraventricular tachycardia.

Epidemiology

Following ACS, 20–40% of patients develop clinically significant anxiety, with 10–25% meeting criteria for a specific anxiety disorder (most commonly GAD, panic disorder, or adjustment disorder with anxiety). A distinct subset of 12–25% of post-ACS patients develop post-traumatic stress disorder (PTSD), a finding established by the landmark research of Edmondson and colleagues, whose 2012 meta-analysis of 24 studies involving 2,383 patients found a pooled PTSD prevalence of 12% (95% CI: 9–16%) following ACS. Cardiac anxiety is measured by the Cardiac Anxiety Questionnaire (CAQ), which captures three dimensions: heart-focused attention, avoidance of physical activity, and fear and worry about cardiac sensations.

Diagnostic Complexity

A major clinical challenge is the phenomenological overlap between cardiac anxiety symptoms and cardiac disease symptoms. Palpitations, chest tightness, dyspnea, diaphoresis, and dizziness are core features of both panic attacks and cardiac ischemia. This overlap creates a bidirectional diagnostic trap:

• False negatives: Genuine cardiac events may be dismissed as anxiety, particularly in women and younger patients. Studies show that women presenting to emergency departments with chest pain and concurrent anxiety are less likely to receive cardiac catheterization, even after adjusting for clinical presentation variables.
• False positives: Noncardiac chest pain accounts for 50–70% of emergency department chest pain presentations, and a substantial subset of these patients develop chronic cardiac anxiety without underlying structural heart disease. Up to 30% continue to have recurrent presentations and functional impairment years later.

Prognostic Impact

Cardiac anxiety is not benign from a cardiovascular standpoint. The INTERHEART study, a case-control study across 52 countries involving over 25,000 participants, identified psychosocial factors (including anxiety and stress) as contributing to a population-attributable risk of 32.5% for first myocardial infarction — a risk comparable to smoking. Mechanistically, chronic anxiety promotes cardiovascular damage through multiple pathways: sustained sympathetic activation increases heart rate, blood pressure, and myocardial oxygen demand; reduced vagal tone lowers the threshold for ventricular arrhythmias; platelet activation and endothelial dysfunction promote atherosclerosis progression; and HPA axis activation produces metabolic syndrome features (insulin resistance, visceral adiposity, dyslipidemia).

In a landmark prospective study, Roest and colleagues (2010) conducted a meta-analysis of 20 studies with 249,846 participants and found that anxiety was associated with a 26% increased risk of incident coronary heart disease (HR = 1.26, 95% CI: 1.15–1.38) and a 48% increased risk of cardiac mortality (HR = 1.48, 95% CI: 1.14–1.92) independent of depressive symptoms and traditional cardiovascular risk factors.

Treatment Evidence

Cognitive-behavioral therapy (CBT) is the best-studied psychological intervention for cardiac anxiety. A meta-analysis by Whalley et al. (2011) for the Cochrane Collaboration found that psychological interventions (predominantly CBT) for coronary heart disease patients reduced anxiety symptoms (SMD = −0.25, 95% CI: −0.48 to −0.03), though effects on cardiac mortality were not statistically significant. The ENRICHD trial (Berkman et al., 2003), while primarily targeting post-MI depression, demonstrated that CBT improved psychosocial outcomes but did not significantly reduce all-cause mortality or recurrent MI at a median follow-up of 29 months, leading to important debates about whether psychological interventions can alter hard cardiac endpoints or whether they primarily improve quality of life.

SSRIs are considered first-line pharmacotherapy for cardiac anxiety. The SADHART trial (Glassman et al., 2002) established the cardiac safety of sertraline in post-ACS patients, demonstrating no adverse effects on left ventricular ejection fraction, ventricular arrhythmias, or QTc interval. Sertraline showed modest but significant improvement in depression scores compared to placebo, with NNT for response of approximately 7–10. The CREATE trial (Lespérance et al., 2007) found citalopram superior to placebo for depression in coronary patients, with response rates of 52.8% vs. 40.4%. Benzodiazepines should be used cautiously in cardiac populations: while they reduce acute anxiety, observational data suggest potential increased mortality risk in post-MI patients, though confounding by indication limits causal inference.

Cardiac rehabilitation (CR) with integrated psychological support is recommended by the American Heart Association and European Society of Cardiology. CR programs that include systematic anxiety screening and psychological components show anxiety reduction effect sizes (d) of 0.30–0.50, improvement in exercise capacity, and in some studies, reductions in all-cause mortality of 20–30% over 2–5 years.

Cancer-related distress encompasses a spectrum from expected emotional responses to clinical syndromes of debilitating anxiety that impair treatment adherence, quality of life, and potentially survival. The National Comprehensive Cancer Network (NCCN) defines cancer-related distress as a "multifactorial unpleasant emotional experience of a psychological, social, and/or spiritual nature that may interfere with the ability to cope effectively with cancer, its physical symptoms, and its treatment." This definition deliberately destigmatizes psychological suffering by framing it as an integral dimension of oncologic care.

Epidemiology

The Zabora et al. (2001) study, using the Brief Symptom Inventory (BSI) in 4,496 cancer patients across 14 cancer diagnoses, found an overall prevalence of clinically significant distress of 35.1%, with the highest rates in lung cancer (43.4%), brain cancer (42.7%), and Hodgkin's disease (37.8%), and the lowest in gynecologic cancers (29.6%). More recent data suggest these figures may be conservative, with some studies reporting distress rates of 40–50% in patients with advanced disease. Specific anxiety disorders are prevalent: adjustment disorder with anxiety occurs in 10–20% of cancer patients, GAD in 5–15%, panic disorder in 3–10%, and PTSD in 5–19% depending on cancer type and treatment intensity.

Fear of Cancer Recurrence (FCR)

Fear of cancer recurrence is the most prevalent unmet psychological need among cancer survivors. A meta-analysis by Simard et al. (2013) found that 49% of cancer survivors report moderate-to-high FCR, with 7–15% experiencing clinically severe levels that impair functioning. FCR is characterized by intrusive thoughts about recurrence, body scanning and hypervigilance to physical symptoms, excessive health care utilization, and difficulty making future plans. Unlike acute perioperative or treatment-related anxiety, FCR often intensifies after active treatment ends, when the protective structure of medical monitoring is withdrawn and the patient confronts uncertainty in the absence of active intervention.

Neurobiological Considerations

Cancer-related anxiety involves distinctive neurobiological features beyond the generic stress response:

• Chemotherapy-induced neurotoxicity: Chemotherapeutic agents, particularly platinum-based compounds and taxanes, cause direct neurotoxicity affecting hippocampal neurogenesis, white matter integrity, and frontal-subcortical circuit function. This contributes to "chemobrain" (cancer-related cognitive impairment), which compounds anxiety by reducing patients' cognitive capacity for emotion regulation.
• Immune-brain interactions: Tumor-derived and treatment-induced inflammatory cytokines (IL-6, TNF-α, IL-1β) produce "sickness behavior" that overlaps substantially with anxiety and depression symptomatology. In a study by Bower et al. (2011), persistent elevation of inflammatory markers was associated with fatigue and mood disturbance up to 5 years post-treatment in breast cancer survivors.
• Endocrine disruption: Hormonal therapies (tamoxifen, aromatase inhibitors, androgen deprivation therapy) directly alter CNS neurotransmitter systems. Tamoxifen crosses the blood-brain barrier and modulates serotonergic neurotransmission; aromatase inhibitors reduce brain estrogen, which normally facilitates serotonergic and GABAergic function. Androgen deprivation therapy in prostate cancer increases anxiety and depression rates by 40–50% relative to non-hormonal treatment.

Screening and Assessment

The NCCN mandates distress screening for all cancer patients, recommending the Distress Thermometer (DT) as a single-item screening tool (0–10 scale; cutoff ≥ 4 identifies clinically significant distress with sensitivity of 77% and specificity of 68%). The Hospital Anxiety and Depression Scale (HADS) is widely used, with a cutoff of ≥ 8 on the anxiety subscale providing sensitivity of 75–85% and specificity of 70–80% for anxiety disorders in cancer populations. For FCR specifically, the Fear of Cancer Recurrence Inventory (FCRI) is the most validated measure.

Treatment Evidence

CBT is the most extensively studied psychological intervention. A meta-analysis by Osborn et al. (2006) found significant effects of CBT on anxiety (SMD = −0.42, 95% CI: −0.56 to −0.28) and depression (SMD = −0.36) in cancer patients. For FCR specifically, the ConquerFear program (Butow et al., 2017), a manualized CBT intervention incorporating metacognitive therapy, attention training, and values-based living, demonstrated sustained reduction in FCR severity at 6-month follow-up (between-group effect size d = 0.57) in a randomized trial.

Mindfulness-based stress reduction (MBSR) and mindfulness-based cognitive therapy (MBCT) have accumulated substantial evidence in oncology. A meta-analysis by Haller et al. (2017) including 29 RCTs found moderate effect sizes for anxiety reduction (Hedges' g = −0.51, 95% CI: −0.70 to −0.33). The MBSR(BC) program, adapted for breast cancer, showed sustained anxiety reduction at 6 weeks post-intervention with NNT of approximately 4–6 for clinically meaningful improvement.

Pharmacotherapy for cancer-related anxiety follows general psychopharmacologic principles with important oncologic considerations. SSRIs and SNRIs are first-line, though drug interactions with tamoxifen must be carefully managed: paroxetine and fluoxetine are strong CYP2D6 inhibitors that reduce conversion of tamoxifen to its active metabolite endoxifen by 50–65%, potentially compromising therapeutic efficacy. Sertraline, citalopram, escitalopram, and venlafaxine have minimal CYP2D6 inhibition and are preferred in patients on tamoxifen. Benzodiazepines are appropriate for acute procedural anxiety (e.g., MRI, chemotherapy administration) but carry risks of falls, cognitive impairment, and dependence in long-term use. Mirtazapine deserves special mention in oncology: its antihistaminic properties provide anxiolysis, appetite stimulation, and sedation — addressing three common cancer-related complaints simultaneously.

Anxiety in medical settings rarely occurs in isolation. Understanding common comorbidity patterns is essential for accurate diagnosis and effective treatment planning.

Anxiety-Depression Comorbidity

The overlap between anxiety and depression in medically ill populations is substantial: 50–70% of patients meeting criteria for one condition also meet criteria for the other. In cardiac populations, the correlation between HADS anxiety and depression subscales typically exceeds r = 0.60. This comorbidity pattern is clinically significant because the combination of anxiety and depression confers worse prognosis than either condition alone: post-ACS patients with both anxiety and depression have approximately double the risk of major adverse cardiac events compared to patients with either condition alone (HR approximately 2.0–2.5).

Anxiety and Pain

Anxiety and pain share neurobiological substrates, particularly in the anterior cingulate cortex (ACC), insula, and periaqueductal gray (PAG). In postoperative contexts, anxiety amplifies pain perception through descending facilitation pathways from the rostral ventromedial medulla (RVM). Meta-analytic data indicate that preoperative anxiety explains 10–15% of variance in postoperative pain intensity, making it one of the strongest psychological predictors of acute surgical pain. The fear-avoidance model, originally developed for chronic musculoskeletal pain, applies directly to cardiac anxiety: catastrophic interpretation of cardiac sensations leads to activity avoidance, physical deconditioning, reduced exercise tolerance, and paradoxically increased vulnerability to both pain and cardiac events.

Anxiety, Delirium, and Post-ICU Syndrome

The relationship between anxiety and delirium is bidirectional and clinically important. Pre-ICU anxiety disorders increase delirium risk (OR 1.5–2.5), while ICU delirium increases the risk of post-ICU anxiety and PTSD. Post-intensive care syndrome (PICS), defined by the Society of Critical Care Medicine, encompasses cognitive, physical, and mental health impairments persisting months to years after ICU discharge. The mental health component of PICS includes anxiety (prevalence 25–46%), depression (17–43%), and PTSD (10–39%) at 3–12 months post-discharge. Delusional memories from the ICU stay — particularly persecutory delusions experienced during delirium — are among the strongest predictors of post-ICU PTSD, with one study reporting an odds ratio of 5.4 for PTSD in patients with delusional ICU memories compared to those without.

Substance Use

Alcohol and substance use disorders complicate anxiety management in all four medical domains. In cardiac populations, alcohol use disorder has a prevalence of 8–15%, exceeding general population rates, and is associated with medication non-adherence and increased cardiac risk. In cancer patients, the prevalence of alcohol use disorder ranges from 5–20% depending on tumor site, with highest rates in head and neck cancers. In the ICU, acute alcohol withdrawal is a major precipitant of delirium and must be actively screened for using the CIWA-Ar protocol. Benzodiazepine anxiolytics must be used with extreme caution in patients with substance use vulnerability.

Identifying prognostic factors enables clinical stratification and resource allocation. Evidence across the four domains reveals both shared and domain-specific predictors.

Shared Predictors of Poor Outcome

• Prior psychiatric history: Pre-existing anxiety or depressive disorders consistently predict more severe and persistent anxiety across all medical settings (OR 2.0–4.0 for clinically significant anxiety).
• Neuroticism (trait anxiety): High scores on trait anxiety measures predict worse acute anxiety responses and slower recovery trajectories. The STAI-Trait subscale is a strong predictor of preoperative anxiety severity (r = 0.50–0.70).
• Low social support: Both perceived and structural social support deficits predict worse anxiety outcomes. In the ENRICHD trial, low perceived social support independently predicted mortality after MI (HR = 1.44).
• Catastrophic cognitive style: Tendency toward catastrophic interpretation of bodily sensations and medical information predicts more severe cardiac anxiety, FCR, and postoperative pain-anxiety cycles.
• Avoidant coping: Behavioral avoidance and experiential avoidance (emotional suppression) predict chronicity across all four domains, while active problem-solving and emotional processing predict better outcomes.

Domain-Specific Prognostic Factors

Preoperative anxiety: Good prognostic indicators include adequate preoperative information, perceived surgical team competence, and prior positive surgical experiences. Emergency surgery, perceived inadequate analgesia, and uncontrolled perioperative nausea predict poor psychological recovery.

ICU delirium: The strongest predictor of delirium resolution is treatment of the underlying precipitant (infection, metabolic derangement, medication-related cause). Duration of delirium is the primary predictor of long-term cognitive outcome: each additional day of delirium in the BRAIN-ICU study was associated with significantly worse cognitive scores at 12 months. Baseline cognitive reserve (estimated by pre-ICU education level and cognitive function) moderates long-term cognitive outcomes.

Cardiac anxiety: Perception of disease severity (rather than objective cardiac function) is the strongest predictor of persistent cardiac anxiety. Patients with preserved ejection fraction can have severe cardiac anxiety, while some patients with significant ventricular dysfunction develop minimal anxiety. Return to exercise and sexual activity within 3–6 months post-ACS are positive prognostic indicators, while persistent activity avoidance predicts chronicity.

Cancer-related distress: Disease stage, recurrence history, and treatment toxicity are medical predictors. Younger age is consistently associated with higher distress and FCR, potentially because cancer diagnosis violates developmental expectations. Active treatment phase generally shows higher distress than surveillance, but the transition from treatment to surveillance is itself a high-risk period. Having a sense of meaning, purpose, or spiritual engagement is one of the strongest protective factors against existential distress in advanced cancer.

The evidence strongly supports integrated care models that embed psychological services within medical settings rather than relying on external referral pathways, which typically result in attrition rates of 50–80% between referral and first appointment.

Collaborative Care Models

The collaborative care model (CoCM), originally developed and tested in the IMPACT trial for depression in primary care, has been adapted for medical settings with strong outcomes. CoCM features three core components: (1) a care manager who provides systematic screening, patient education, and brief interventions; (2) a consulting psychiatrist who provides case-based supervision; and (3) a measurement-based treatment-to-target approach with systematic outcome monitoring. In the SMaRT Oncology trials (Sharpe et al., 2014), collaborative care for depression in cancer patients produced response rates of 62% vs. 17% for usual care (NNT = 2.2), with sustained benefits at 24 weeks and 48 weeks, and significant improvement in anxiety symptoms as a secondary outcome. This represents one of the strongest treatment effects in the consultation-liaison psychiatry literature.

Stepped Care

Stepped care models allocate intervention intensity to clinical severity: low-distress patients receive psychoeducation and self-management support; moderate distress triggers brief structured interventions (4–8 sessions of CBT, supportive counseling, or pharmacotherapy); high distress prompts referral to specialist psychiatric or psychological services. Implementation of stepped care in cancer settings has been associated with 30–40% improvements in distress detection rates and 25–35% reductions in time to first psychological intervention.

The Perioperative Surgical Home and Enhanced Recovery After Surgery (ERAS)

ERAS protocols represent a systematic approach to perioperative care that now increasingly incorporates psychological preparation. ERAS bundles typically include preoperative education, carbohydrate loading, multimodal analgesia, early mobilization, and increasingly, anxiety screening and management. Implementation of ERAS protocols has been associated with 30–50% reductions in postoperative complications and 1–3 day reductions in length of stay, though the specific contribution of anxiety management components is difficult to isolate from the bundle.

ICU Diary Programs

ICU diary programs, in which staff and family members maintain a factual record of the patient's ICU stay, address the memory fragmentation and delusional recall that fuel post-ICU anxiety and PTSD. The RACHEL trial (Jones et al., 2010) found that ICU diaries reduced new-onset PTSD from 13% to 5% at 3-month follow-up, though subsequent replication has shown more modest effects. Current evidence supports diaries as a low-cost, low-risk intervention that should be offered routinely, though they are not sufficient as a standalone intervention for patients with severe post-ICU psychological distress.

Technology-Assisted Interventions

Digital health platforms offer scalability advantages for managing anxiety across medical settings. Internet-delivered CBT (iCBT) for cancer distress has been tested in multiple RCTs with effect sizes comparable to face-to-face CBT (d = 0.30–0.60). Mobile applications for cardiac rehabilitation that include anxiety monitoring and coping skill modules show promising engagement rates of 60–75% over 12 weeks, though long-term adherence data remain limited. Virtual reality applications for preoperative anxiety reduction, chemotherapy-related distress, and ICU rehabilitation are active areas of research, with early evidence suggesting 30–50% anxiety reduction in preoperative VR studies.

Despite significant advances, the evidence base for managing anxiety in medical settings has notable limitations and evolving frontiers.

Psychedelic-Assisted Therapy in Cancer Distress

Two landmark randomized, double-blind, crossover trials — one at Johns Hopkins (Griffiths et al., 2016) and one at NYU (Ross et al., 2016) — demonstrated that single-dose psilocybin-assisted psychotherapy produced rapid, large, and sustained reductions in anxiety and depression in patients with life-threatening cancer diagnoses. The NYU trial reported response rates for anxiety of 83% at 7 weeks (vs. 14% for placebo) and sustained responses at 6.5 months in approximately 60–80% of participants. Effect sizes were large (d = 1.3–1.5). These findings are being pursued in larger phase II/III trials, and the FDA has granted breakthrough therapy designation for psilocybin for treatment-resistant depression. The primary limitations are small sample sizes (N = 29 and N = 51 in the original trials), self-selected populations, challenges in blinding, and inability to disentangle pharmacological from expectancy effects.

Biomarker-Guided Treatment

Research is exploring whether inflammatory biomarkers (CRP, IL-6), cortisol profiles, or HRV measurements can guide treatment selection. For example, patients with high inflammation may respond preferentially to anti-inflammatory augmentation strategies, while those with low HRV may benefit most from vagal nerve stimulation or HRV biofeedback. This precision psychiatry approach remains in early stages, with no validated predictive algorithms for clinical use.

Ketamine and Esketamine

Subanesthetic ketamine is being investigated for rapid anxiolysis in palliative care settings, with case series and small RCTs suggesting significant anxiety reduction within hours. The NMDA receptor antagonism and rapid neuroplasticity-promoting effects of ketamine make it theoretically attractive for acute medical anxiety, though evidence is limited by small sample sizes and short follow-up periods.

Limitations of Current Evidence

Several cross-cutting limitations constrain the evidence base:

• Heterogeneity of anxiety constructs: Studies use inconsistent definitions, measures, and thresholds for clinically significant anxiety, making cross-study comparisons difficult.
• Underrepresentation of minority populations: Most RCTs have been conducted in predominantly White, English-speaking populations. Cultural factors profoundly influence illness appraisal, help-seeking behavior, and treatment preferences.
• Short follow-up: Most intervention studies have follow-up periods of ≤ 6 months, inadequate for understanding chronic trajectories of medical anxiety.
• Lack of head-to-head trials: There are very few direct comparisons of CBT vs. pharmacotherapy vs. combined treatment in medical anxiety populations, forcing reliance on indirect comparisons across trials.
• Implementation gap: Even effective interventions show poor real-world uptake. The gap between evidence and practice is arguably the largest barrier to improving outcomes for anxiety in medical settings.

Anxiety in medical settings is biologically active, clinically consequential, and treatable. Effective management requires systematic screening, accurate diagnostic formulation, and integration of evidence-based interventions within existing medical workflows.

Key Clinical Recommendations

• Screen routinely: Use validated instruments (APAIS for preoperative, CAM-ICU for delirium, CAQ for cardiac anxiety, DT for cancer distress) at standardized time points.
• Treat preoperative anxiety proactively: Combine structured education with pharmacologic anxiolysis (gabapentinoids or short-acting benzodiazepines) and non-pharmacologic approaches (music, guided imagery) tailored to patient preference.
• Implement the ABCDEF bundle in ICU care: Prioritize delirium prevention through light sedation, benzodiazepine avoidance, early mobilization, and sleep promotion. Do not rely on antipsychotics for delirium treatment.
• Recognize cardiac anxiety as a modifiable cardiac risk factor: Integrate anxiety screening into cardiac rehabilitation, offer CBT or structured psychological support, and prescribe SSRIs (preferably sertraline or escitalopram) when indicated.
• Assess cancer distress as a sixth vital sign: Follow NCCN guidelines for universal distress screening, implement stepped care pathways, and attend to FCR in survivorship.
• Address comorbidity systematically: Evaluate for concurrent depression, pain, substance use, and cognitive impairment, as these conditions interact to amplify anxiety severity and treatment resistance.
• Adopt integrated care models: Collaborative care and stepped care models dramatically outperform usual care for anxiety and depression in medical populations.

The evidence is clear: anxiety is not an inevitable and untreatable accompaniment of medical illness. It is a clinically distinct entity with identifiable neurobiological mechanisms, validated assessment tools, and effective interventions. The primary barrier to better outcomes is not knowledge — it is implementation.