1 Introduction

Delirium is a complex, multidimensional neuropsychiatric illness that is common among older hospitalized patients (Wilson et al. 2020). It is characterized by an abrupt onset of disorientation and has been independently associated with a twofold increase in risk of death, a 2.4-fold increase in risk of institutionalization, and a 12.5-fold increase in risk of dementia (Echeverría et al. 2022; Witlox et al. 2020). Delirium can occur in patients with and without pre-existing dementia, significantly impacting the prognosis and management of affected individuals (Inouye et al. 2014). While a plethora of patient and medical factors have been associated with delirium, medication has been considered as both a predisposing and precipitating factor (Inouye et al. 2014). The same medications that are thought to increase the risk of precipitating a delirium, such as antipsychotics and hypnotics, are used as treatment despite evidence that the use of such medications may increase the risk of mortality and prolonged hospital stays (Inouye et al. 2014). A recent meta-analysis reported insufficient evidence for the use of antipsychotics, corticosteroids, or ketamine for the prevention or treatment of delirium. Similarly, insufficient evidence exists for anticholinergics, opioids, benzodiazepines, and H1-antihistamines (Reisinger et al. 2023; Xu et al. 2020; Day et al. 2021; Hatta et al. 2019; Kim et al. 2020; Wang et al. 2021; The American Psychiatric Association 2024). There is a paucity of knowledge on the mechanisms of action, impact of dosage and form, risk associated with drug combinations, and suggested alternatives. There is a distinct lack of comprehensive information on medication risk in international delirium guidelines (American Psychiatric Organisation 2010). The most recently updated National Institute for Health and Care Excellence (NICE) delirium guidelines (2023) do not provide recommendations on certain medications due to a lack of high-quality studies (National Institute for Health and Care Excellence 2023), while the American Psychiatric Association (APA) practice guidelines for the prevention and treatment of delirium (2024) suggest comprehensive medicines reconciliation and the involvement of a clinical pharmacist to manage medication risk (The American Psychiatric Association 2024).

Considering the serious negative health consequences to the patient and resultant cost to the healthcare system, comprehensive guidance is imperative to improve patient and prescribing safety. Especially, the resultant long-term cognitive decline of patients resulting in a higher incidence of dementia is thought to cost the US healthcare system $6.9 billion (Inouye and Ferrucci 2006). Additionally, the costs incurred after hospital discharge resulting from rehospitalization, rehabilitation, home health care, or institutionalization are estimated to be over $100 billion per year (Inouye and Ferrucci 2006). Not to mention the personal cost incurred to the patient due to the functional decline and loss of quality of life (Fong et al. 2009). Especially patients with a pre-existing diagnosis of dementia who experience delirium (delirium superimposed on dementia [DSD]) have a much higher risk of institutionalization and mortality (Han et al. 2022). The overlapping symptom profile of delirium and dementia makes DSD difficult to diagnose and is often mistaken for a sudden worsening of dementia with resultant prolonged hospital stays, increased use of hospital resources, and increased institutionalization leading to loss of independence, distress, diminished ability to engage in meaningful activities and social interactions, as well as an increased caregiver burden (Fong and Inouye 2022; Fick et al. 2005; Fong et al. 2019). A recent systematic review (SR) reports that the cost implications of delirium may be 52% higher when dementia is considered (Kinchin et al. 2021).

The aim of this extensive SR was to collate existing published evidence outside of clinical and national guidelines on medication risks in delirium, mechanisms, treatment, and treatment alternatives in adult patients with and without dementia in an attempt to support prescribing decision-making and patient safety in healthcare practice. Clinical and national guidelines focused on psychiatry, neurology, perioperative, and pediatrics will be looked at in separate SRs (Weidmann et al. 2023, 2024, 2025; Weidmann and Tadic 2024).

2 Methods

An SR methodology was chosen over a scoping review as the aim required a rigorous, bias-minimized synthesis of the existing evidence to provide definitive conclusions on medication risk in relation to the causation, prevention, and treatment of delirium. A scoping review is more suited to the identification of evidence gaps (Joanna Briggs Institute 2024). The research steps are shown in the flowchart (Figure 1).

3 Results

The literature search resulted in 3867 publications. After removal of duplicates and independent screening, 106 publications were included in this study: n = 35, original research studies; n = 28, systematic and narrative reviews; n = 44, case studies; and n = 0, qualitative studies (Table 1). Details of the inclusion criteria (Table S1), search strategy (Table S2), PRISMA chart (Figure S1), and full reference list of publications included (Table S3) can be found in the Supporting Information.

A variety of research designs, were included, totaling 3,563,243 delirium patients across all 106 studies (n = 3,563,189, original research studies; n = 41, case reports; and n = 13, case series) of which 15,546 patients had dementia or mild cognitive impairment.

3.2 Medications at Risk of Causing a Delirium

A total of n = 158 individual medications were identified across 20 different drug classes (Figure 2 and Figure S2). The most frequently mentioned medication classes opioid analgesics (n = 24), antipsychotics (n = 23), anxiolytics/hypnotics (n = 19), antidepressants (n = 18), corticosteroids (n = 16), cardiovascular (n = 16), cholinesterase inhibitors (n = 13), and antibiotics (n = 11) are described in this Section 3. Less frequently mentioned medication detail are shown in Table S4.

3.2.1 Opioid Analgesic

A total of n = 22 papers (20.7%) suggest that opioid analgesics are a substantive risk factor for delirium (Tables 3 and 4) (see references [S4, S5, S8, S10, S11, S13, S17–S19, S24, S35, S41, S45, S46, S53, S64, S78, S79, S91, S93, S98, S101] in the Supporting Information). Mechanisms such as an altered blood–brain barrier integrity, facilitating the entry of pro-inflammatory cytokines and toxins, as well as the apoptosis of human neurons and microglial cells, through the upregulation of Bax and downregulation of the anti-apoptotic protein Bcl-2 are considered to be responsible for the opioid-associated cognitive impairment (Cunha-Oliveira et al. 2008). An equivalent dose of > 54 mg/day was reported to be a factor significantly associated with persistent delirium [S79], while the odds ratio for morphine daily doses of 7.2–18.6 mg (OR = 9.20) was associated with a significant increase in the risk of developing delirium among intensive care (ICU) patients [S98]. The incidence of delirium was 4.8% (n = 83) for morphine in an observational study of 259 opioid-naive oncology inpatients [S91]. However, there was no detectable significant difference in the incidence of delirium between morphine, oxycodone, or tapentadol hydrochloride [S91]. The administration of intravenous morphine equivalent dose was associated with a 2.4% increased risk of delirium, as was epidural ropivacaine hydrochloride use during anesthesia [S91]. While it is most difficult to manage delirium/hallucinations or somnolence induced by transdermal fentanyl, characterized by its long half-life of 6–7 h, oxycodone has the shortest half-life (t_1/2 = 3 h), resulting in a more predictable dose–response relationship [S10, S11]. It is noteworthy that dose levels of > 80 mg methadone were associated with the precipitation of a delirium. This risk doubled in doses above 150 mg and doubled again in doses above 200 mg/day [S101].

TABLE 3. Summary of reported drug-specific mechanism, dosing, and symptom information for the most frequently mentioned drugs and drug classes. Only drugs for which enough information was available were include.

Drug class	Drug [ATC code]	Mechanism of action	CYP p450	Dose	Symptom	Comments	Supporting Information references
Opioid analgesic	General information	Mitochondrial dysfunction, oxidative stress, and upregulation of apoptogenic proteins, which can produce permanent white matter lesions	—	Higher opioid doses increase the risk Cumulative daily doses of opioids above 90 mg	Hallucination	Similar risk of delirium in opioid-naive patients among morphine, oxycodone, and tapentadol oral opioids Delirium is the most common potential opioid-related ADE	[S11, S18, S35, S41, S79, S91, S93, S101]
	Buprenorphine [N02AE01; N07BC01]	Dopamine-releasing effect	CYP3A4 and CYP2D6 inhibitor	—	Inactivity and lethargy	■ Transdermal application increases risk ■ IV opioids carry the highest risk of mortality	[S45]
	Fentanyl [N01AH01; N02AB03]	Possible muscarinic receptor binding	CYP3A4 inhibitor	Dose-dependent toxicity	Hallucination somnolence	■ Transdermal application increases risk ■ IV opioids carry the highest risk of mortality	[S10, S18, S19, S46]
	Morphine [N02AA01]	Inhibition of GABA receptors allows dopaminergic neurons to fire more vigorously	CYP2D6 inducer	> 54 mg/day PO associated with persistent delirium Each daily 10-mg IV morphine-equivalent dose was associated with a 2.4% increased risk for delirium the next day	—	■ Extended release tablets increases risk ■ IV opioids carry the highest risk of mortality	[S98, S46, S91, S24]
	Oxycodone [N02AA05]	Reduction of GABAergic inhibition of dopamine neurons, leading to an increase in dopamine production and release	CYP2D6	—	Delusion	■ Oxycodone is associated with a lower risk for delirium compared to other opioids	[S5, S53, S64, S91]
	Pethidine [N02AB02]	Inhibition of the dopamine and norepinephrine transporter	CYP2B6 substrate	—	—	■ Pethidine was associated with a higher risk of causing delirium than oxycodone	[S4, S11, S13, S17, S19, S78]
	Methadone [N07BC02]	Affects the release of acetylcholine, norepinephrine, substance P, and dopamine	CYP3A4 substrate CYP2B6 substrate CYP2D6 inhibitor	> 80 mg; risk doubled above 150 mg and doubled again above 200 mg/day	—	■ Previous methadone-related delirium predicted recurrence	[S101]
	Tramadol [N02AZ03]	Inhibits norepinephrine, serotonin, and dopamine	CYP2D6 substrate	—	Confusion psychotic symptoms, violent behavior delusion	■ Tramadol IM induces delirium ■ Only use in older adults after careful consideration	[S8, S17, S78]
Antipsychotics	General information	High anticholinergic burden	—	—	—	■ Antipsychotics are a highly possible risk factor for delirium. ■ Moderate degree of QTc prolongation ■ Meta-analysis found that antipsychotic use was not associated with changes in the duration or severity of delirium ■ Only use unless patient is at risk of harm to self and others [Beers/SToppFall] ■ All antipsychotic use for delirium and insomnia is therefore off-label ■ Dementia may predispose patients to ACH drug-induced delirium	[S12, S46, S57, S72, S84]
	Haloperidol [N05AD01]	D2 receptor antagonist	CYP3A4 substrate CYP2D6 inhibitor	IV haloperidol Haloperidol (treatment)—use 1 week or less	Delirium, neuroleptic malignant syndrome	■ Conflicting evidence of causal risk for delirium ■ IV haloperidol increased risk of QTc prolongation (medium–high risk) ■ Haloperidol + omeprazole—avoid ■ Co-administration with carbamazepine leads to a significant change in the pharmacokinetics of haloperidol ■ Haloperidol may negatively affect survival compared to placebo	[S2, S4, S13, S17, S18, S31, S60, S65, S79, S82, S84, S89, S95]
	Olanzapine [N05AH03]	D2 receptor antagonist	CYP1A2 substrate	PO, IM, and IV olanzapine as safe and effective treatments for critically ill patients with delirium	Impaired consciousness and attention, agitation visual hallucination	■ Reports are rare ■ Fluvoxamine increases the plasma concentrations of olanzapine ■ Smoking increases the olanzapine clearance by 40% ■ In females, olanzapine clearance is decreased by 30% ■ Moderate risk for QTc prolongation	[S13, S17, S60, S62, S82, S84, S85, S94, S95]
	Quetiapine [N05AH04]	Anticholinergic hypothesis	CYP3A4 substrate	PO formulation reaches a peak plasma concentration 1.5 h after administration and has a half-life of 6 h Require dose adjustment in patients with hepatic dysfunction. Renal dose adjustment is not necessary	Temporal disorientation, visual hallucination	■ Adjunct for ventilated patients in ICU settings	[S6, S18, S60, S84, S85, S95]
	Risperidone [N05AX08]	Has a higher affinity for 5-HT2A receptors compared to D2, and reduces dopaminergic and serotonergic activity	CYP2D6 substrate	Risperidone reaches peak serum concentrations in 1–2 h	Extreme excitement hallucination, neuroleptic malignant syndrome	■ Long-acting injection increases risk ■ CYP2D6 inhibitors like fluoxetine and paroxetine increase the plasma concentrations of risperidone ■ Risperidone can cause delirium compared to placebo (95% CI, 0.09–0.86; p = 0.02) ■ Compared to placebo, patients had more extrapyramidal effects (95% CI, 0.09–1.37; p = 0.03)	[S2, S4, S13, S17, S60, S66, S71, S84, S95]
	Lorazepam [N05AD01]	Potent D2 blocker	CYP3A4 substrate CYP2D6 inhibitor CYP1A2 substrate	2–30 mg/day	Agitation, perceptual disturbance, sleep–wake cycle disruption	■ Treatment of delirium for max. 1 week or less ■ Duration of action may be increased substantially in the setting of systemic ■ Inflammatory states (e.g., sepsis)	[S13, S18, S31, S46, S60, S79, S98, S104]
Anxiolytics/Hypnotics	General information	—	—	Daily equivalent dose of 5 mg or more statistically significantly increased the risk as did cumulative daily doses of benzodiazepines above 2 mg	Delusion	■ Long-acting agents trended toward a stronger association compared to short-acting benzodiazepines ■ Benzodiazepine withdrawal	[S3, S5, S17, S33, S53, S105]
	Midazolam [N05CD08]	Binds to GABA receptors, thereby inhibiting dopamine release	CYP3A4 substrate	Daily dose of 5 mg to a patient who is both coma- and delirium-free will increase the odds by 4%	Delirium	■ Delirium-inducing drug commonly used in palliative medicine	[S13, S46, S104]
	Zopiclone [N05CF01]	It being highly plasma-protein bound (92.5%), its high drug–drug interaction risk by means of hepatic cytochrome P450-3A4	CYP3A4	Single dose of 7.5 mg zopiclone	Disorientation agitation, hallucination, speech disturbance	■ Zolpidem plasma concentration is gender based	[S11, S47]
Antidepressant	General information	Anticholinergic action and tetracyclic antidepressant are α-2 adrenergic receptor antagonists	—	—	—	■ Delirium inducing drugs commonly used in palliative medicine ■ Antidepressants had a disproportional association with delusion as a symptom	[S17, S28, S41, S46, S64]
	Duloxetine [N06AX21]	Causes dopaminergic excess	CYP2D6 inhibitor	N/A	Visual hallucination, sleep alteration, delusion	■ N/A	[S5, S63, S92]
	Amitriptyline [N06AA09]	Anticholinergic	CYP2D6 substrate	Overdose 1000 mg/day	Confusion	■ All tricyclic antidepressant drugs exert an anticholinergic effect, with amitriptyline having the strongest and nortriptyline the weakest	[S31, S46, S77]
	Bupropion [N06AX12]	Dopamine reuptake inhibition Possible inhibition of CYP2D6 may increase the concentration of hydroxybupropion and cause further dopaminergic excess and subsequent delirium	CYP2D6 inhibitor		Disorientation auditory and visual hallucination, persecutory delusion	■ Lower rates of sexual dysfunction than escitalopram, fluoxetine, paroxetine, and sertraline ■ Its use is not commonly associated with the side effect of dopamine excess ■ Bupropion-related delirium is rare	[S63, S105]
Corticosteroid	General information	Steroid induced sleep disruption	CYP3A4 inducers (weak)	High-dose pulse steroids cumulative daily doses of corticosteroids above 15 mg	Acute hyperactive delirium with psychosis, delusion, confusion	■ Intra-articular administration	[S13, S51, S58, S68, S72, S77, S78]
Cardiovascular	General information	—	—	—	Cognitive impairment, simple acute confusion, visual hallucination	■ CV drugs with central nervous system (CNS) bioavailability carry the highest risk (disopyramide, quinidine, digoxin, clonidine, methyldopa, propranolol, reserpine) ■ Delirium was more common with calcium channel blockers than with renin–angiotensin system agents (∼40% higher) but less common than with beta-blockers (∼20% lower) ■ Visual hallucinations have been reported with angiotensin-converting enzyme (ACE) inhibitors in older patients, particularly those with a history of dementia or mild cognitive impairment	[S4, S13, S36, S38, S49, S67]
	Digoxin [C01AA04]	Hypokalemia and hypomagnesemia due to thiazide and loop diuretics may increase digoxin toxicity.	Not metab. by CYP	—	Visual hallucination, bilateral ballism	—	[S19, S42, S67]
	Clonidine [C02AC01]	Presynaptic alpha2-receptor agonist inhibiting the release of norepinephrine.	CYP2D6 substrate	Risk is increased when medications are used in high or toxic doses	Confusion, lapses in memory, poor concentration	■ Clonidine can cause disruption in the release of various catecholamines and/or serotonin (5-HT) in cortical and subcortical regions	[S13, S22, S67]
	Reserpine [C02AA02]	Reserpine depletes stores of 5-HT, dopamine, and norepinephrine in the CNS	Metab. by the CYP system to a major degree but neither induces nor inhibits CYP activity	Risk is increased when medications are used in high or toxic doses	Confusion, lapses in memory, poor concentration	■ Relative lack of hepatotoxicity	[S67]
Antibiotics	General information	—	—	—	Delusion	■ No evidence that any particular antibiotic or class of antibiotics has a consistent or predictable adverse effect on cognition ■ Quinolone and Macrolide carry the greatest risk ■ Fluoroquinolones are a rare cause of seizures and delirium	[S4, S5, S11, S13]
	Tigecycline [J01AA12]	Glutamate receptor activation A substrate for p-glycoprotein (P-gp)	Not metab. via CYP	Infusion with a loading dose (mg not specified)	Delirium	■ Delirium as a side effect of tigecycline is very rare ■ As it does not metabolize via CYP, pharmacokinetic drug interactions are uncommon	[S66, S106]
	Ciprofloxacin [J01MA02; S01AE03; S02AA15; S03AA07]	N/A	CYP1A2 inhibitor	N/A	Mania, Insomnia, psychosis	■ The number of reports of delirium is greater for quinolone and macrolide antibiotics ■ Switchen to a cephalosporine	[S1, S11, S13, S46]
	Moxifloxacin [J01MA14/ S01AE07]	N/A	Not metab. via CYP	N/A	Confusion, attention alteration, temporal and spatial disorientation	■ The number of reports of delirium is greater for quinolone and macrolide antibiotics ■ Switchen to a cephalosporine	[S5, S80]
	Gatifloxacin [J01MA16; S01AE06]	N/A	Negligible inhibitors of CYP1A2 and CYP2C9	400 mg	Delusion, concern, agitation	■ Delirium as a side effect of gatifloxacin is rare	[S87]
	Scopolamine [A04AD01; N05CM05; S01FA02]	Competitive inhibition of antimuscarinic receptors	CYP3A4	11.7 (3–27) h after the application of the patch. The mean dosage of scopolamine was 2.25 mg	Insomnia, restlessness, disorientation, visual hallucination, repetitive behavior, gait disturbance, dysarthria, delusion	■ Transdermal application increases the risk	[S4, S13, S88]

TABLE 4. Risk associated with opioid analgesics in older patients with and without dementia.

Drug	Opioid characteristics and delirium-specific risk	Recommendations in dementia
Similar risk of delirium in opioid-naive patients among morphine, oxycodone, and tapentadol oral opioids.
Buprenorphine patches	Patients potentially experience fewer side effects compared to other opioids.	■ Could be a good choice in dementia but reduced elimination due to its lipophilic character can result in increased drug plasma concentration in older patients.
Codeine	Considerable variation in response and adverse effects can increase the risk of falls.	■ Poorly tolerated in dementia.
Dextropropoxyphene	Has a propensity for neural and cardiac toxicity.	■ Best avoided in dementia.
Fentanyl patches	Their half-life of 6–17 h makes treating side effects more difficult. Even after patch is removed, it will take at least 24 h for a significant drop in plasma concentration.	■ Useful in chronic pain and palliative care, although it should not be used to initiate opioid analgesia in frail older people. ■ The incidence of delirium after the commencement of fentanyl injection was significantly lower. ■ Transdermal fentanyl can be switched to slow-release morphine (260 mg/day), allowing pain control, but the risk of hallucinations and somnolence remains high.
Methadone	Methadone has a very long half-life (8—59 h), with an average of 24 h. Methadone has no ceiling effect.	■ Methadone is best avoided in dementia.
Morphine	Active metabolites penetrate the blood–brain barrier and are renally excreted, so they may accumulate in renal impairment, making older and frail patients more susceptible to adverse effects.	■ Very effective analgesic but likely to cause cognitive problems and other adverse effects in older patients. ■ Due to the blood–brain barrier penetration, slow-release morphine increases the risk of delirium in dementia with Lewy bodies.
Oxycodone	Has a short half-life (t1/2 = 3 h). Therefore, oxycodone has fewer drug–drug interactions and a more predictable dose–response relationship compared to other opiates. Oxycodone is associated with a lower risk of delirium.	■ Due to its more predictable dose–response relationships, oxycodone is a good candidate for oral analgesia in dementia patients.
Pentazocine	Risk of neurotoxicity.	■ Best avoided.
Pethidine	Pethidine was associated with a higher risk of causing delirium than oxycodone, as anticholinergic metabolites accumulate rapidly in renal impairment.	■ Pethidine carries a particularly high risk of cognitive impairment in patients with dementia.
Tramadol	High risk of drug interactions, considerable variation in response, and adverse effects may induce seizures. Tramadol IM should only be used after careful consideration in older adults.	■ May be poorly tolerated in dementia. ■ Avoid in patients with epilepsy or on other epileptogenic drugs, for example, psychotropics and memantine.

• Note: Adapted from [S11].

3.2.2 Antipsychotics

Typical and atypical antipsychotics are a significant risk factor for delirium, particularly those with anticholinergic effects [S46] such as clozapine [S84], olanzapine [S94], and levomepromazine [S46, S57]. While some reports suggest that haloperidol carries an increased risk of inducing a persistent delirium (OR 2.88; 95% CI, 1.38–6.02), particularly when administered intravenously [S2, S79, S82]. However, a meta-analysis and GRADE-level evidence ratings could not show an increased delirium risk for haloperidol (OR 0.96; 95% CI, 0.72–1.28) [S82]. Quetiapine is known to cause increased agitation, temporal disorientation, and visual hallucinations [S6]. Quetiapine is often used as a sedation adjunct for ventilated patients in the ICU setting with a t_1/2 of 6 h [S84, S85]. Despite reports about their significant potential in the causation of delirium, the majority of reported information relates to their off-label use in the prevention and management of delirium. The effectiveness of antipsychotics in the prevention and management of delirium and its associated symptoms is, however, questionable as a meta-analysis found that antipsychotic use was not associated with changes in the duration or severity of delirium [S14, S16]. In addition, QTc prolongation is a common side effect for many antipsychotics, with IV haloperidol significantly increasing the risk of arrhythmias and other cardiovascular concerns [S84]. This is reported to be less of a problem with atypical antipsychotics [S84].

3.2.3 Anxiolytics/Hypnotics

Long-acting benzodiazepines are associated with a higher risk of delirium compared to short-acting ones [S4, S13, S17, S19, S28, S33, S43, S60]. Cumulative daily doses of benzodiazepines above 2 mg (HR, 2.04; 95% CI, 1.05–3.97) and 5 mg (OR 3.5; 95% CI, 1.4–8.8) have been reported to carry a delirium risk [S11, S34]. Benzodiazepine withdrawal has also been reported to be a risk factor [S105]. While the retrieved literature does not describe the mechanism by which benzodiazepines cause delirium, Z-substances (zolpidem, zopiclone, zaleplon, and eszopiclone) risk seems to describe a disruption in the dopaminergic and GABAergic systems, with single doses of 5 mg zolpidem or 7.5 mg zopiclone having been linked to symptoms such as visual hallucinations, disorientation, and agitation [S47].

3.2.4 Antidepressants

Tricyclic antidepressants (TCAs) and selective serotonin reuptake inhibitors (SSRIs) in particular are reported to be associated with an increased risk of developing delirium and akathisia [S46]. Paroxetine is the SSRI reported to carry the highest risk [S17, S4], while amitriptyline is the most frequently mentioned TCA in this SR [S5, S15, S18, S46, S76, S77].

3.2.5 Corticosteroids

While all steroids are considered to carry a risk of delirium [S4, S13, S17, S19, S33, S34, S36, S46, S51, S54, S58, S72, S77], patients exposed to cumulative daily doses of corticosteroids above 15 mg (HR, 2.67; 95% CI, 1.18–6.03) presented an increased risk of developing delirium within 28 days of hospitalization according to a study of 261 hospitalized cancer patients [S34].

3.2.6 Cardiovascular

Especially cardiovascular (CV) drugs with central nervous system (CNS) bioavailability carry the highest risk of acute confusion and delirium, which may even result in more chronic changes in cognition (disopyramide, quinidine, digoxin, clonidine, methyldopa, propranolol, reserpine) [S4, S13, S22, S49, S67]. Digoxin, due to its neurotoxic effects and its risk of causing hypokalemia and hypomagnesemia, especially in combination with thiazide and loop diuretics, is most frequently mentioned in association with delirium [S4, S11, S17, S43, S67]. However, no specific mechanisms, dosing instructions, or treatment alternatives are reported. While calcium channel blockers have been reported to be associated with a higher rate of delirium than renin–angiotensin system agents (∼40% higher) but less common than in beta blockers (20% lower), no specific information about dosing, and so forth is provided beyond the generic statement [S38].

3.2.7 Cholinesterase Inhibitors

Cholinesterase inhibitors can cause delirium due to their effects on the cholinergic system, especially in individuals with pre-existing cholinergic system disruptions or high anticholinergic burden [S28]. Even CNS-active drugs that are appropriately used in this population can accumulate in amounts that lead to problems such as delirium, sedation, and falls. Symptoms most commonly reported include agitation and hallucinations [S21, S28, S29, S50]. The response to these drugs can vary significantly among individuals, leading to inconsistent outcomes regarding delirium [S1, S4, S17, S69, S76, S83, S85, S106].

3.2.8 Antibiotics

Reports on antibiotics in association with delirium, while frequent, are very disparate in nature. There seems to be no conclusive evidence that any particular antibiotic or class of antibiotics has a consistent or predictable adverse effect on cognition. However, most antibiotics may occasionally be associated with delirious-like episodes [S4, S5, S11, S27, S37, S46, S48, S49, S80, S82, S87, S90, S13]. This is also true for the reported symptoms, which seem to range from general delusions across a wide spectrum to insomnia, restlessness, disorientation, visual hallucination, repetitive behavior, gait disturbance, dysarthria, and delusion [S4, S11]. It is stated that the number of reports of delirium is greatest for quinolone and macrolide antibiotics [S11], results that were corroborated by a recent pharmacovigilance analysis of 4559 reports of delusion [S5].

4 Discussion

Overall, there is a paucity of specific and good-quality medication-related publications associated with the medication risks in relation to delirium, its mechanisms, treatment, and treatment alternatives in adult patients with and without dementia. Neurotransmitter imbalances, pharmacokinetic changes, and physiological processes are frequently reported as medication-related delirium-causing mechanisms. A total of n = 158 individual medications were listed across 20 different drug classes, with the riskiest drug combinations being the multiple use of opioids and dopamine agonists. Despite good-quality evidence from several large real-world data studies, a large disparity in the reporting of risky combinations was identified. No evidence for the use of antipsychotics in the treatment of delirium could be reported. Alternative pharmacological strategies focus on the mindful use of anticholinergic medication and choosing a newer generation of medication. Medication-related disruption of acetylcholine, histamine, and GABA is associated with delirium in patients with an underlying dementia diagnosis.

The inherent lack of understanding of medication risk in relation to either association or causation, as well as treatment of delirium in adult patients with and without dementia, is reflective of the general lack of understanding of the pathophysiology of delirium irrespective of dementia diagnosis (Wilson et al. 2020; Gupta et al. 2008; Iglseder et al. 2010). This fundamental gap in our understanding is particularly important for patients taking multiple medications, as they are often multimorbid and frail, making them especially vulnerable to delirium. While some mechanisms are easily explained such as the strong association of opioids, in combination or alone, with delirium due to the neurodegenerative properties of mitochondrial dysfunction, oxidative stress, and the upregulation of apoptotic proteins, mitochondrial dysfunction itself has been shown to be involved in the pathophysiology of delirium (Guo et al. 2013; Liu et al. 2022; Lu et al. 2020; Gunther et al. 2008). Other mechanisms, such as anticholinergic medication and their association with delirium, are less well understood while the association as a significant risk factor is well documented, the exact mechanisms of neurotransmitter disturbance caused by the different anticholinergic drugs or drug combinations require further investigation (Talabaki et al. 2024; Macha et al. 2023). Interestingly, rivastigmine, an acetylcholinesterase inhibitor, has been identified as a potential treatment for anticholinergic delirium, despite other reports implicating cholinesterase inhibitors in its development (Fratta et al. 2023; Schlake et al. 2024). While the evidence summarized from the papers included in this review did not identify cholinesterase inhibitors as a treatment option, it did seem to implicate this drug group in the development of delirium symptoms, especially in dementia patients [S5, S28, S29, S50, S62, S68–S70, S76, S83, S90, S102]. Specific information on medication risks associated with DSD was sparse. This reflects the only two SRs to date aiming to summarize the published information on prevalence, associated features, outcomes, and management of DSD, despite it having a pooled prevalence estimated to be as high as 48.9% in older hospitalized patients (Fick et al. 2002; Han et al. 2022). The SR by Han et al. (2022) only lists psychotropic medication, antibiotics, and polypharmacy as the medication risk factors for DSD without providing any medication-related detail. Our study was able to provide more detailed medication-related information, identifying cholinesterase inhibitors, antibiotics, analgesics, cardiac drugs, antimanika, and analgesics as carrying an increased delirium risk in patients with dementia (Figure 1), in particular, memantine and donepezil (> 5 mg/day). Among antibiotics, nitrofurantoin, gatifloxacin, clarithromycin, and macrolides were associated with an increased risk of DSD, while detailed opioid analgesic prescribing advice in older patients with and without dementia was summarized (Table 4). This information potentially provides the most comprehensive information on medication-associated delirium risk in dementia patients to date.

Among the lack of insight into complex medication-related risk factors, the search for obvious patterns and unifying mechanisms continues. Considering the drug classes associated with delirium, their varying effect on the vagus nerve seems to offer one such possible explanation. The vagus nerve is key in the transmission of signals between the gastrointestinal tract and the central nervous system, the gut–brain axis (GBA) (Bonaz et al. 2018). Both the pathophysiology of cognitive impairment and dementia have been linked to the vagus nerve and GBA (Ticinesi et al. 2018; Zhang et al. 2022). Medications, such as antibiotics, corticosteroids, and proton pump inhibitors, have been reported to significantly alter the gut microbiome, impacting both its composition and function and leading to a profound effect on health, including drug efficacy and toxicity (Barrio et al. 2022; Zhao et al. 2023; Simpson et al. 2021). Especially, medication such as antibiotics, corticosteroids, and proton pump inhibitors significantly impact the gut microbiome (Fossmark and Olaisen 2024; Couch et al. 2023). A recent study investigating whether microbiome-targeted therapies (MTTs) improve cognition and prevent postoperative cognitive dysfunctions such as delirium reported that MTTs decrease inflammation and have a positive effect on cognition. While this line of research is still in its infancy, the SR concludes that MTTs could be a potential new preventative strategy for cognitive impairment after surgery (Sugita et al. 2023; Garcez et al. 2023). A study investigating the association between gut microbiota and delirium occurrence in acutely ill older adults identified that gut microbiome diversity and composition were significantly different in acutely ill hospitalized older adults who experienced delirium (Garcez et al. 2023).

As long as “delirium” is used as a blanket term for a plethora of symptoms and etiological heterogeneity without its categorization into distinct etiological subgroups, it will be very difficult to design high-quality studies that help to develop clear prescribing advice and delirium interventions (Ormseth et al. 2023; Oh et al. 2020). However, it is evident that no single pharmacological intervention will demonstrate effectiveness, such a comprehensive guideline can help to support multicomponent approaches such as the Hospital Elder Life Program (HELP), which have demonstrated to reduce the odds of delirium by > 50% in the intervention group compared to controls (OR 0.47; 95% CI, 0.37–0.59; I² = 28%; n = 3605 patients) (Hshieh et al. 2018). In the meantime, the increased use of artificial intelligence in healthcare poses additional opportunities for delirium care (Gong et al. 2023; Xie et al. 2022). Very little information however is available on subsequent treatment and pharmacological intervention strategies much less how AI can help with identifying underlying medication mechanisms and treatment strategies (Weidmann and Watson 2024).

5 Implications for Practice

This is the only review to provide a comprehensive summary of medication-related information in relation to delirium and dementia, which can be used to support prescribing decision-making in adult patients and can help to support multicomponent approaches to delirium care. In addition to the heterogeneity of reported underlying mechanisms of action of common medications associated with delirium, it discusses a possible association between the effect of medication on the gut microbiome and delirium. Results presented in this SR support the call for categorizing delirium into distinct etiological subgroups to facilitate the design of high-quality studies that help to distinguish contributing medication factors and to develop clear prescribing advice and delirium interventions.

Papers specifically looking at delirium superimposed on dementia (DSD) were not included in the search strategy resulting in incomplete information presented. The research team registerred andconducted a separate Systematic Review focused specifically on delirium superimposed on dementia (DSD) [PROSPERO: CRD42024546118] [Aslam et al. 2024].

6 Future Research

Given the plethora and heterogeneity of identified contributing factors that contribute to the development of delirium according to Ormseth et al. (2023), it seems unlikely to adequately address the paucity of good-quality studies that address all these factors. In relation to the role of medication in delirium, future studies should focus on the development of a comprehensive guideline and electronic prescribing resource summarizing detailed pharmacological information on the causation, treatment, and prevention of delirium in different patient groups to support existing and future multicomponent treatment approaches and support prescribing practice and patient care. Further, the use of artificial intelligence to support clinical pharmacy and clinical pharmacology research in the identification and development of new and novel therapeutic approaches to delirium in different patient groups should be explored. Finally, the association between the impact of medications on the GBA and gut microbiome along with their associated risk in the causation, prevention, and treatment of delirium requires further investigation.

7 Conclusion

There is still an inherent lack of understanding of medication risk associated with the causation and treatment of delirium in adult patients with and without dementia, which is reflective of the general lack of our understanding of the pathophysiology of delirium. While it is not clear if medication is a contributing or precipitating factor for delirium, this review has identified many more drug classes and individual drugs than are typically listed in standard practice guidance. The reported effect of medication on the gut microbiome and vagus nerve may pose a unifying mechanism for medication involvement in delirium, especially in patients whose health is already extremely challenged. Clinical pharmacists, as experts in medicine, should always be included in multicomponent approaches to the treatment of delirium.

Author Contributions