. 2019 Dec 2;19:310. doi: 10.1186/s12883-019-1547-4

Table 1.

Main data extracted from the included studies

Author, year [reference]	Study design, Country	Aims	Evaluated tools, language	Gold standard/alternative methods(s) considered	Diagnostic practice	Examiner(s)	Patients: inclusion/exclusion criteria	Patients: sample methods and main profile
Infante et al., 2017 [20]	Diagnostic test accuracy study quasi-experimental study^a Italy	1. To assess the effect of DSM-V delirium criteria review and formal training on the ability of neurologists to recognise delirium 2. To evaluate the 4AT for the evaluation of post-stroke delirium	4AT Language: Italian	DSM-V criteria	Delirium was screened with 4AT and assessed with DSM-V criteria at admission and after 7 days of hospitalisation by the same researcher All diagnoses were afterwards reviewed independently by other two expert researchers Period: NR Setting: single tertiary stroke centre	Three neurologists	Inclusion criteria (diagnostic test accuracy study): • > 18 years • diagnosis of acute stroke • GCS > 5 Exclusion criteria: • aphasia • dementia	Consecutive sample n = 100; median age 79 years; gender NR
Kutlubaev et al., 2016 [21]	Diagnostic test accuracy study and observational study^a Russia	1. To identify older patients with high delirium risk 2. To assess the diagnostic value of the 4AT test in this population	4AT Language: Russian	DSM-IV criteria	Patients were examined for delirium within hours after their admission or on the next day; then twice at the interval of 12–24 h during their in-hospital stay Delirium was diagnosed according to the DSM-IV criteria and the 4AT test Period: 2 months (2013–2014) Setting: Neurovascular Department	Neurologist (not specified if the same, or not, who evaluated the delirium presence with both the 4AT and the DSM-IV criteria)	Inclusion criteria: • ≥ 65 years • admitted in the first 3 days of stroke Exclusion criteria: • subarachnoid/subdural haemorrhages without intracerebral haematoma • transient ischaemic attacks • impairment of consciousness as severe as sopor and coma • with significant chronic mental disorders in the past	Consecutive sample n = 73 (over 132 eligible); median age 79 years; male 29%
Lees et al., 2013 [22]	Diagnostic test accuracy study United Kingdom	1. To describe test accuracy properties of various brief screening assessments against an independent clinical diagnosis of cognitive impairment (using MoCA) and delirium 2. To describe the effect of altering the screen-positive cut-point for MoCA using differing predetermined diagnostic thresholds	AMT-10 AMT-4 CDT COG-4 4AT GCS Single Question “Does this patient have cognitive issues?” at the daily multidisciplinary team Language: English	CAM	Patients were assessed during the period of day 1 to day 4 after stroke unit admission Period: 10 weeks (April–June 2012) Setting: Stroke Unit	Two trained medical students: one completed the delirium assessment using the validating tools; one assessed for delirium using the CAM They were blinded	Inclusion criteria: • cerebral ischaemia and haemorrhage • medically stable to allow an attempt at a least part of cognitive assessment Exclusion criteria: NR	Consecutive sample n = 111 (over 138 eligible); median age 74 years; male 50%
Mitasova et al., 2012 [8]	Diagnostic test accuracy study and observational study^a Czech Republic	1. To describe the epidemiology of delirium in a cohort of acute post-stroke patients using the DSM-IV 2. To determine the sensitivity, specificity, and overall accuracy of the CAM-ICU, and 3. To investigate its validity as a routine monitoring instrument for hospitalised patients with stroke by non-psychiatrically trained clinicians	CAM-ICU Language: Czech	DSM-IV criteria	Patients underwent paired daily evaluation with the CAM-ICU The first CAM-ICU evaluation on the first day after stroke onset and admission (day 1) and then daily (6 days/week) on at least 7 consecutive days on which the patient was accessible to testing (RASS ≥ −3). If delirium was present on day 6 or 7, its assessment follow-up continued until at least 2 subsequent days without delirium were recorded In patients with consciousness deterioration the follow-up was stopped The standard DSM evaluation of delirium was performed < 2 h apart daily Period: 18 months (2009–2010) Setting: specialised stroke centre	A trained junior physician assessed patients with the CAM-ICU A panel of specialists, experts on delirium (two neurologists, two neuropsychologists, a psychiatrist and a speech therapist) performed the standard reference DSM evaluation (at least one neurologist and one neuropsychologist)	Inclusion criteria: • cerebral infarction or intracerebral haemorrhage • delirium assessment within 24 h of stroke onset • approval of the patient or his or her relatives Exclusion criteria: • patients who did not speak Czech • duration of stroke symptoms and signs < 24 h • history of severe head trauma or neurosurgery (at any time) • subarachnoid haemorrhage, venous infarction, brain tumour • history of psychosis • patients who were comatose or stuporous on admission and did not improve during the first week post-stroke (RASS ≤ − 4)	Consecutive sample n = 129 (151 initially enrolled, over 331 eligible); mean age 71.3 years; male 55.8%

Author, year [reference]

Study design, Country

Aims

Evaluated tools, language

Gold standard/alternative methods(s) considered

Diagnostic practice

Examiner(s)

Patients: inclusion/exclusion criteria

Patients: sample methods and main profile

Infante et al.,

2017 [20]

Diagnostic test accuracy study quasi-experimental study^a

Italy

1. To assess the effect of DSM-V delirium criteria review and formal training on the ability of neurologists to recognise delirium

2. To evaluate the 4AT for the evaluation of post-stroke delirium

4AT Language: Italian

DSM-V criteria

Delirium was screened with 4AT and assessed with DSM-V criteria at admission and after 7 days of hospitalisation by the same researcher

All diagnoses were afterwards reviewed independently by other two expert researchers

Period: NR

Setting: single tertiary stroke centre

Three neurologists

Inclusion criteria (diagnostic test accuracy study):

• > 18 years

• diagnosis of acute stroke

• GCS > 5

Exclusion criteria:

• aphasia

• dementia

Consecutive sample n = 100; median age 79 years; gender NR

Kutlubaev et al.,

2016 [21]

Diagnostic test accuracy study and observational study^a

Russia

1. To identify older patients with high delirium risk

2. To assess the diagnostic value of the 4AT test in this population

4AT Language: Russian

DSM-IV criteria

Patients were examined for delirium within hours after their admission or on the next day; then twice at the interval of 12–24 h during their in-hospital stay

Delirium was diagnosed according to the DSM-IV criteria and the 4AT test

Period: 2 months (2013–2014)

Setting: Neurovascular Department

Neurologist (not specified if the same, or not, who evaluated the delirium presence with both the 4AT and the DSM-IV criteria)

Inclusion criteria:

• ≥ 65 years

• admitted in the first 3 days of stroke

Exclusion criteria:

• subarachnoid/subdural haemorrhages without intracerebral haematoma

• transient ischaemic attacks

• impairment of consciousness as severe as sopor and coma

• with significant chronic mental disorders in the past

Consecutive sample

n = 73 (over 132 eligible); median age 79 years; male 29%

Lees et al.,

2013 [22]

Diagnostic test accuracy study

United Kingdom

1. To describe test accuracy properties of various brief screening assessments against an independent clinical diagnosis of cognitive impairment (using MoCA) and delirium

2. To describe the effect of altering the screen-positive cut-point for MoCA using differing predetermined diagnostic thresholds

AMT-10

AMT-4

CDT

COG-4

4AT

GCS

Single Question “Does this patient have cognitive issues?” at the daily multidisciplinary team

Language: English

CAM

Patients were assessed during the period of day 1 to day 4 after stroke unit admission

Period: 10 weeks (April–June 2012)

Setting: Stroke Unit

Two trained medical students: one completed the delirium assessment using the validating tools; one assessed for delirium using the CAM

They were blinded

Inclusion criteria:

• cerebral ischaemia and haemorrhage

• medically stable to allow an attempt at a least part of cognitive assessment

Exclusion criteria: NR

Consecutive sample

n = 111 (over 138 eligible); median age 74 years; male 50%

Mitasova et al.,

2012 [8]

Diagnostic test accuracy study and observational study^a

Czech Republic

1. To describe the epidemiology of delirium in a cohort of acute post-stroke patients using the DSM-IV

2. To determine the sensitivity, specificity, and overall accuracy of the CAM-ICU, and

3. To investigate its validity as a routine monitoring instrument for hospitalised patients with stroke by non-psychiatrically trained clinicians

CAM-ICU

Language: Czech

DSM-IV criteria

Patients underwent paired daily evaluation with the CAM-ICU

The first CAM-ICU evaluation on the first day after stroke onset and admission (day 1) and then daily (6 days/week) on at least 7 consecutive days on which the patient was accessible to testing (RASS ≥ −3).

If delirium was present on day 6 or 7, its assessment follow-up continued until at least 2 subsequent days without delirium were recorded

In patients with consciousness deterioration the follow-up was stopped

The standard DSM evaluation of delirium was performed < 2 h apart daily

Period: 18 months (2009–2010)

Setting: specialised stroke centre

A trained junior physician assessed patients with the CAM-ICU

A panel of specialists, experts on delirium (two neurologists, two neuropsychologists, a psychiatrist and a speech therapist) performed the standard reference DSM evaluation (at least one neurologist and one neuropsychologist)

Inclusion criteria:

• cerebral infarction or intracerebral haemorrhage

• delirium assessment within 24 h of stroke onset

• approval of the patient or his or her relatives

Exclusion criteria:

• patients who did not speak Czech

• duration of stroke symptoms and signs < 24 h

• history of severe head trauma or neurosurgery (at any time)

• subarachnoid haemorrhage, venous infarction, brain tumour

• history of psychosis

• patients who were comatose or stuporous on admission and did not improve during the first week post-stroke (RASS ≤ − 4)

Consecutive sample

n = 129 (151 initially enrolled, over 331 eligible); mean age 71.3 years; male 55.8%

4AT: 4-Assessment Test for delirium, AMT: Abbreviated Mental Test, CAM: Confusion Assessment Method, CAM-ICU: Confusion Assessment Method for the Intensive Care Unit, CDT: Clock Drawing Test, COG4: Cognitive examination derived from National Institutes of Health Stroke Scale (NIHSS), DSM: Diagnostic and Statistical Manual of mental disorders, GCS: Glasgow Coma Scale, MoCA: Montreal Cognitive Assessment, NR not reported, RASS Richmond Agitation and Sedation Scale.

^aonly data regarding validation phase has been extracted and reported in this Table