Delirium – Rokowania, prognozy i postęp choroby

Delirium
Rokowania, prognozy i postęp choroby

Delirium, definiowane jako ostra zmiana uwagi, świadomości i funkcji poznawczych, jest istotnym czynnikiem ryzyka zwiększającym śmiertelność (30-dniowa śmiertelność 9% vs. 1% bez delirium), wydłużenie hospitalizacji oraz konieczność umieszczenia w placówkach opiekuńczych, szczególnie u osób starszych i pacjentów w stanie krytycznym. W kontekście COVID-19 delirium wiąże się z gorszymi wynikami klinicznymi, w tym dłuższą terapią i pogorszeniem funkcji poznawczych. Wczesna identyfikacja pacjentów zagrożonych delirium oraz wdrożenie interwencji wielodomenowych (leczenie przyczyn, przegląd leków, zarządzanie stresem i środowiskiem) może zapobiec 30-40% przypadków. Modele predykcyjne oparte na uczeniu maszynowym, takie jak PRE-DELIRIC, DELIKT czy modele wykorzystujące dane fizjologiczne (np. Random Forest z AUROC 0,82), umożliwiają identyfikację ryzyka delirium przed wystąpieniem objawów, co pozwala na ukierunkowane działania zapobiegawcze.

Rokowanie delirium: przewidywanie wyników leczenia

Delirium, stan charakteryzujący się ostrą zmianą w obszarze uwagi, świadomości i funkcji poznawczych, spowodowany stanem medycznym, którego nie można lepiej wyjaśnić wcześniej istniejącym zaburzeniem neurokognitywnym, wiąże się z poważnymi konsekwencjami dla pacjentów ¹. Wczesna identyfikacja pacjentów zagrożonych rozwojem delirium i wdrożenie środków zapobiegawczych mogłoby zapobiec nawet 40% przypadków delirium ². Jest to szczególnie istotne, ponieważ delirium można zapobiec w około 30-40% przypadków ³.

Konsekwencje delirium dla rokowania pacjentów

Delirium wiąże się ze znaczącym zwiększeniem niekorzystnych wyników leczenia u pacjentów. Metaanalizy wykazały, że u osób starszych delirium jest związane z gorszymi wynikami (śmiertelność, instytucjonalizacja i demencja), niezależnie od istotnych czynników zakłócających ⁴. Badania potwierdzają, że delirium jest powiązane z:

Zwiększoną chorobowością i śmiertelnością ⁵
Wydłużonym czasem hospitalizacji ⁶
Wyższym wskaźnikiem umieszczania w placówkach opieki ⁷
Trzydziestodniowa śmiertelność jest znacząco wyższa (9%) u pacjentów z delirium w porównaniu do pacjentów bez delirium (1%) ⁸

Delirium jest również istotnym markerem śmiertelności wśród osób starszych odwiedzających oddział ratunkowy, przy czym ryzyko śmiertelności jest najbardziej wyraźne w pierwszych 3 miesiącach po wizycie na SOR ⁹. W kontekście COVID-19, delirium związane z tą chorobą zostało powiązane z gorszymi wynikami u pacjentów, w tym dłuższą terapią, wyższymi wskaźnikami powikłań, zwiększoną śmiertelnością oraz, u osób, które przeżyły COVID-19, z następczym pogorszeniem funkcji poznawczych ¹⁰.

Delirium jako czynnik prognostyczny długoterminowych zaburzeń poznawczych

Coraz więcej dowodów wskazuje, że delirium może być predyktorem długoterminowych zaburzeń poznawczych (LTCI – Long-Term Cognitive Impairment). Badania prospektywne wykazały, że pacjenci w stanie krytycznym są narażeni na ryzyko LTCI po ciężkiej chorobie, a to nowe LTCI może utrzymywać się przez 3 i 12 miesięcy po wyjściu ze szpitala i jest związane z czasem trwania delirium ¹¹. Delirium jest również uznawane za predyktor długoterminowych zaburzeń poznawczych u osób, które przeżyły stan krytyczny ¹².

Ponadto, delirium może wpływać na długoterminową jakość życia związaną ze zdrowiem i funkcjonowanie poznawcze u pacjentów w stanie krytycznym ¹³. Badania wykazały, że ostre delirium jest niezależnym czynnikiem ryzyka śmiertelności u starszych pacjentów na oddziałach intensywnej terapii (OIT) ¹⁴.

Modele predykcyjne delirium

W ostatnich latach obserwuje się szybko rosnącą liczbę badań dotyczących wykorzystania uczenia maszynowego (ML) do przewidywania ryzyka delirium w środowiskach szpitalnych ¹⁵. Modele predykcyjne opracowane przy użyciu metod ML mogą potencjalnie identyfikować osoby zagrożone rozwojem delirium przed wystąpieniem objawów, do których można kierować strategie zapobiegawcze, co z kolei może zmniejszyć częstość występowania delirium i poprawić wyniki leczenia pacjentów ¹⁶.

Istniejące modele predykcyjne i ich skuteczność

Istnieje kilka modeli predykcyjnych delirium, które różnią się pod względem metodologii, populacji docelowej i skuteczności:

PRE-DELIRIC i E-PRE-DELIRIC: Oba modele wykazują umiarkowaną do dobrej skuteczność. Model PRE-DELIRIC lepiej przewiduje delirium, ale lekarze OIT ocenili wygodę użytkowania E-PRE-DELIRIC jako lepszą niż PRE-DELIRIC ¹⁷ ¹⁸.
DELIKT: Model oparty na rutynowo zbieranych danych w ciągu pierwszych 24 godzin od przyjęcia, które mogą być zintegrowane z narzędziem prewencyjnym do automatycznego przewidywania delirium u hospitalizowanych starszych pacjentów ¹⁹.
Modele ML wykorzystujące dane fizjologiczne: Nowe podejście wykorzystujące ciągłe dane fizjologiczne z monitorowania pacjentów na OIT. Model oparty na algorytmie Random Forest (RF) wykazał solidną skuteczność w walidacji wewnętrznej (AUROC: 0,82; AUPRC: 0,62) i zachował swoją dokładność w walidacji czasowej (AUROC: 0,73; AUPRC: 0,85) ²⁰.
Model dla pacjentów z zaawansowanym rakiem: Wieloośrodkowe prospektywne badanie kohortowe opracowało i zwalidowało model do przewidywania delirium u pacjentów z zaawansowanym rakiem podczas ich pobytu w oddziałach opieki paliatywnej, który wykazał doskonałą dyskryminację, kalibrację i praktyczność kliniczną ²¹.

Przegląd systematyczny modeli predykcyjnych majaczenia u osób starszych wykazał jednak, że modele te mają zmienne i zazwyczaj nieadekwatne zdolności predykcyjne. Wskaźnik AUROC w zewnętrznie zwalidowanych modelach predykcyjnych delirium wahał się od 0,52 do 0,94. Spośród tych modeli najlepiej działający model (AUROC 0,94, 95% CI 0,91 do 0,97) został opracowany i zwalidowany w populacji chirurgicznej ²².

Nowe podejścia do przewidywania delirium

Prowadzone są badania nad nowymi metodami przewidywania delirium, w tym:

Ilościowa elektroencefalografia (qEEG): Badania wykazały, że qEEG może wiarygodnie przewidywać delirium po operacji kardiochirurgicznej. Przy użyciu zaledwie 4 elektrod i 20-minutowego zapisu w pierwszej godzinie po przyjęciu na OIT, qEEG może z wysoką czułością i specyficznością przewidywać pooperacyjne delirium (POD) ²³ ²⁴.
Modele dynamiczne ML: W przeciwieństwie do modeli statycznych, które wykorzystują dane z pierwszych 24 godzin po przyjęciu na OIT, modele dynamiczne mogą przewidywać delirium z wyprzedzeniem do 12 godzin z rozsądną dyskryminacją i kalibracją ²⁵ ²⁶.
Modele wykorzystujące dane z elektronicznych kartotek medycznych (EHR): Modele ML trenowane przy użyciu rutynowo zbieranych danych z EHR dokładnie przewidują delirium na OIT, wspierając dynamiczne prognozowanie wrażliwe na czas ²⁷.

Integracja przewidywań ML z klinicznymi przepływami pracy daje obietnicę zrewolucjonizowania opieki zdrowotnej ²⁸. Ostatecznym celem modelu predykcyjnego delirium jest bezproblemowa integracja z klinicznymi przepływami pracy za pośrednictwem systemów elektronicznych kart zdrowia i urządzeń do monitorowania przyłóżkowego ²⁹.

Czynniki wpływające na rokowanie w delirium

Czynniki ryzyka rozwoju delirium

Identyfikacja czynników ryzyka rozwoju delirium jest kluczowa dla wczesnej interwencji i poprawy rokowania. Do najważniejszych czynników ryzyka zalicza się:

Delirium w wywiadzie medycznym, zaawansowany wiek i wynik ASA (American Society of Anesthesiologists) ³⁰
Ciężkość choroby mierzona skalą APACHE IV ³¹
Sepsa i wentylacja mechaniczna ³²
Temperatura ciała ³³
Skala Glasgow (GCS) ³⁴
Skumulowane obciążenie antycholinergiczne mierzone Kliniczną Skalą Antycholinergiczną (CrAS) ³⁵

W kontekście COVID-19, ryzyko delirium można stratyfikować na podstawie ciężkości choroby COVID-19 oraz podobnych czynników jak w przypadku delirium związanego z innymi infekcjami układu oddechowego: zaawansowany wiek, historia chorób neurodegeneracyjnych oraz obecność podwyższonych parametrów infekcji i retencji nerkowej ³⁶.

Strategie zapobiegania i leczenia wpływające na rokowanie

Skuteczne strategie zapobiegania i leczenia delirium mogą znacząco poprawić rokowanie pacjentów. Obecnie najlepsze strategie zarządzania to interwencje wielodomenowe, które skupiają się na:

Leczeniu stanów wywołujących delirium ³⁷
Przeglądzie leków ³⁸
Zarządzaniu stresem ³⁹
Łagodzeniu powikłań ⁴⁰
Utrzymaniu zaangażowania w kwestie środowiskowe ⁴¹

Zdolność do przewidywania wystąpienia delirium u osób wysokiego ryzyka może pozwolić na wdrożenie strategii zapobiegawczych lub leczniczych w bardziej ukierunkowany lub nawet spersonalizowany sposób ⁴².

Warto zauważyć, że przegląd systematyczny stwierdził, że istnieje niewystarczająca liczba dowodów potwierdzających rutynowe stosowanie leków przeciwpsychotycznych w leczeniu delirium ⁴³. Ponadto, w wieloośrodkowym, randomizowanym badaniu z podwójnie ślepą próbą i kontrolą placebo (REDUCE) porównującym profilaktyczne stosowanie haloperidolu z placebo w zapobieganiu delirium u dorosłych w stanie krytycznym, haloperidol nie poprawił przeżycia w ciągu 28 dni; dlatego profilaktyczne stosowanie haloperidolu nie jest zalecane w celu zmniejszenia śmiertelności u dorosłych w stanie krytycznym ⁴⁴.

Wyzwania w badaniu i poprawie rokowania delirium

Mimo znaczących postępów w zrozumieniu delirium i rozwoju modeli predykcyjnych, istnieje wiele wyzwań w badaniu i poprawie rokowania delirium:

Relatywnie niewiele modeli predykcyjnych zostało poddanych zewnętrznej walidacji, a gdy tak się stało, wykazywały one pogorszenie skuteczności modelu ⁴⁵
Niewiele badań przeszło prospektywną ocenę w rzeczywistych warunkach klinicznych ⁴⁶
Zmienne włączenie i zastosowane definicje w modelach predykcyjnych delirium są heterogeniczne, co utrudnia porównania ⁴⁷
Skuteczne wdrożenie strategii wykrywania, leczenia i zapobiegania delirium pozostaje głównym wyzwaniem dla organizacji opieki zdrowotnej na całym świecie ⁴⁸

Ponadto, nadal nie jest jasne, czy delirium jest objawem innych powikłań pooperacyjnych, czy też delirium zwiększa ryzyko niekorzystnych zdarzeń pooperacyjnych ⁴⁹. W kontekście COVID-19, niektóre badania nie zaobserwowały istotnych związków między wystąpieniem delirium a długością pobytu w szpitalu lub śmiertelnością, co może być częściowo wyjaśnione przez specjalistyczną wiedzę ośrodków w zarządzaniu ARDS (zespół ostrej niewydolności oddechowej) i faktem, że do momentu rozpoczęcia rekrutacji, istotne doświadczenie w zarządzaniu COVID-19 zostało już zdobyte podczas pierwszej fali pandemii ⁵⁰.

Wnioski kliniczne dotyczące rokowania w delirium

Na podstawie dostępnych dowodów, można wyciągnąć następujące wnioski kliniczne dotyczące rokowania w delirium:

Delirium jest związane ze znaczącym zwiększeniem zachorowalności, śmiertelności, długości pobytu w szpitalu i umieszczania w placówkach opieki ⁵¹ ⁵²
Wczesna identyfikacja pacjentów zagrożonych rozwojem delirium jest ważna, ponieważ odpowiednie, dobrze zaplanowane interwencje mogłyby zapobiec wystąpieniu delirium i związanym z nim niekorzystnym wynikom ⁵³
Modele predykcyjne oparte na uczeniu maszynowym mają potencjał do identyfikacji osób zagrożonych rozwojem delirium przed wystąpieniem objawów, co może pomóc w ukierunkowaniu strategii zapobiegawczych ⁵⁴
Dla pacjentów w podeszłym wieku zagrożonych delirium, EVAR (Endovascular Aneurysm Repair) może być preferowany w stosunku do operacji otwartej ⁵⁵
Biorąc pod uwagę ograniczenia obecnych modeli predykcyjnych delirium, nie powinny one być postrzegane jako substytuty dla opinii ekspertów klinicystów ⁵⁶

Skuteczne zapobieganie delirium mogłoby skorzystać z automatycznej stratyfikacji ryzyka u starszych pacjentów szpitalnych przy użyciu rutynowo zbieranych danych klinicznych ⁵⁷. Zastosowanie modelu predykcyjnego pomoże personelowi opieki paliatywnej zidentyfikować pacjentów z wysokim ryzykiem delirium i ułatwi ukierunkowane inicjowanie środków zapobiegawczych ⁵⁸.

Przyszłe kierunki badań nad rokowaniem delirium

Aby poprawić modele predykcyjne delirium, przyszłe modele powinny rozważyć użycie standardowych zmiennych i definicji, aby pracować w kierunku narzędzia predykcyjnego, które jest możliwe do zastosowania w kilku populacjach w ramach zrozumienia związku z wydarzeniem wywołującym ⁵⁹. Ponadto, potrzebne są dalsze badania w następujących obszarach:

Rozwój solidnych modeli do przewidywania delirium u starszych pacjentów hospitalizowanych ⁶⁰
Lepsze zrozumienie wieloczynnikowego pochodzenia delirium związanego z COVID-19 ⁶¹
Badanie związku między predykcją delirium a długoterminowymi zaburzeniami poznawczymi ⁶²
Rozwój i walidacja modeli predykcyjnych w specyficznych populacjach, takich jak pacjenci z zaawansowanym rakiem w opiece paliatywnej ⁶³
Ocena potencjału modeli ML w przewidywaniu delirium, z konsekwentną skutecznością w różnych scenariuszach walidacji ⁶⁴

Ciągły charakter pomiarów fizjologicznych pozwala modelom ML zapewnić ocenę ryzyka w czasie rzeczywistym, potencjalnie wychwytując subtelne zmiany fizjologiczne, które mogą poprzedzać wystąpienie delirium ⁶⁵. To podejście, wraz z integracją modeli predykcyjnych z systemami elektronicznych kart zdrowia, może zrewolucjonizować zarządzanie delirium i poprawić wyniki leczenia pacjentów.

Kolejne rozdziały

Zapraszamy do dalszego czytania naszego leksykonu.

Wybierz kolejny rozdział z menu poniżej, aby otworzyć nową podstronę kompedium wiedzy i uzyskać szczegółowe informację o leku, substancji lub chorobie.

Materiały źródłowe

#1 Delirium | Nature Reviews Disease Primers
https://www.nature.com/articles/s41572-020-00223-4
Delirium, a syndrome characterized by an acute change in attention, awareness and cognition, is caused by a medical condition that cannot be better explained by a pre-existing neurocognitive disorder. […] Currently, the best management strategies are multidomain interventions that focus on treating precipitating conditions, medication review, managing distress, mitigating complications and maintaining engagement to environmental issues. […] The effective implementation of delirium detection, treatment and prevention strategies remains a major challenge for health-care organizations globally. […] This prospective longitudinal cohort study demonstrated that critically ill patients are at risk of LTCI after critical illness, that this new LTCI can persist at 3 and 12 months follow-up, and that it is associated with duration of delirium.
#2 Navigating the machine learning pipeline: a scoping review of inpatient delirium prediction models
https://pmc.ncbi.nlm.nih.gov/articles/PMC10335592/
Early identification of inpatients at risk of developing delirium and implementing preventive measures could avoid up to 40% of delirium cases. […] ML models have potential to identify inpatients at risk of developing delirium before symptom onset. However, few models were externally validated and even fewer underwent prospective evaluation in clinical settings. […] This review confirms a rapidly growing body of research into using ML for predicting delirium risk in hospital settings. Our findings offer insights for both developers and clinicians into strengths and limitations of current ML delirium prediction applications aiming to support but not usurp clinician decision-making. […] Prediction models derived using ML methods can potentially identify individuals at risk of developing delirium before symptom onset to whom preventive strategies can be targeted, which may, in turn, reduce incident delirium and improve patient outcomes. This scoping review identified all publications describing ML-based delirium prediction models over the last 5 years, evaluated their stage in the ML evolution pipeline, and assessed their performance and utility. Relatively few were subject to external validation, which, when performed, showed degraded model performance. In addition, while few studies underwent prospective evaluation in real-world clinical settings, performance and user acceptance seemed promising in those that did. However, given the limitations of current delirium prediction models, they should not be seen as substitutes for expert clinician judgement.
#3 Risk Factors and Outcomes for Postoperative Delirium after Major Surgery in Elderly Patients | PLOS One
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0136071
Early identification of patients at risk for delirium is important, since adequate well timed interventions could prevent occurrence of delirium and related detrimental outcomes. […] Postoperative delirium is a frequent complication after major surgery in elderly patients and is related to an increase in adverse events, length of hospital stay, and mortality. […] A delirium in the medical history, advanced age, and ASA-score may assist in defining patients at increased risk for delirium. […] Delirium is related to an increase in morbidity, mortality, length of stay and care home placement. […] Most importantly, delirium could be prevented in approximately 30-40% of the cases. […] Our reported incidence rate of post-operative delirium (15%) is comparable with other studies in recent literature (11-18%).
#4 Delirium | Nature Reviews Disease Primers
https://www.nature.com/articles/s41572-020-00223-4
This meta-analysis provides evidence that, in elderly patients, delirium is associated with poor outcomes (mortality, institutionalization and dementia), independent of important confounders. […] This paper represents the clearest demonstration that progressive cognitive decline is a progressively increasing risk factor for delirium and also demonstrates, in mice, that this decline is correlated with increasing synaptic loss and can precede frank neurodegeneration. […] This systematic review concluded that there is insufficient evidence supporting the routine use of antipsychotic agents for the treatment of delirium. […] In the multisite, randomized, double-blinded, placebo-controlled REDUCE trial comparing prophylactic haloperidol with placebo for delirium prevention in critically ill adults, haloperidol did not improve survival at 28 days; thus, prophylactic haloperidol is not recommended for reducing mortality in critically ill adults.
#5 Risk Factors and Outcomes for Postoperative Delirium after Major Surgery in Elderly Patients | PLOS One
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0136071
Early identification of patients at risk for delirium is important, since adequate well timed interventions could prevent occurrence of delirium and related detrimental outcomes. […] Postoperative delirium is a frequent complication after major surgery in elderly patients and is related to an increase in adverse events, length of hospital stay, and mortality. […] A delirium in the medical history, advanced age, and ASA-score may assist in defining patients at increased risk for delirium. […] Delirium is related to an increase in morbidity, mortality, length of stay and care home placement. […] Most importantly, delirium could be prevented in approximately 30-40% of the cases. […] Our reported incidence rate of post-operative delirium (15%) is comparable with other studies in recent literature (11-18%).
#6 Risk Factors and Outcomes for Postoperative Delirium after Major Surgery in Elderly Patients | PLOS One
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0136071
Delirium was related to multiple adverse events and increase in hospital stay. […] Thirty-day mortality was significantly higher (9%) in the delirious patients compared to the non-delirious patients (1%). […] The question of whether delirium is a symptom of other postoperative complications or whether a delirium increases the risk of postoperative adverse events remains to be answered. […] For elderly patients at risk for delirium, EVAR could be preferable to open surgery.
#7 Risk Factors and Outcomes for Postoperative Delirium after Major Surgery in Elderly Patients | PLOS One
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0136071
Early identification of patients at risk for delirium is important, since adequate well timed interventions could prevent occurrence of delirium and related detrimental outcomes. […] Postoperative delirium is a frequent complication after major surgery in elderly patients and is related to an increase in adverse events, length of hospital stay, and mortality. […] A delirium in the medical history, advanced age, and ASA-score may assist in defining patients at increased risk for delirium. […] Delirium is related to an increase in morbidity, mortality, length of stay and care home placement. […] Most importantly, delirium could be prevented in approximately 30-40% of the cases. […] Our reported incidence rate of post-operative delirium (15%) is comparable with other studies in recent literature (11-18%).
#8 Risk Factors and Outcomes for Postoperative Delirium after Major Surgery in Elderly Patients | PLOS One
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0136071
Delirium was related to multiple adverse events and increase in hospital stay. […] Thirty-day mortality was significantly higher (9%) in the delirious patients compared to the non-delirious patients (1%). […] The question of whether delirium is a symptom of other postoperative complications or whether a delirium increases the risk of postoperative adverse events remains to be answered. […] For elderly patients at risk for delirium, EVAR could be preferable to open surgery.
#9 Delirium as a predictor of mortality in US Medicare beneficiaries discharged from the emergency department: a national claims-level analysis up to 12 months | BMJ Open
https://bmjopen.bmj.com/content/8/5/e021258
Delirium is common among seniors discharged from the emergency department (ED) and associated with increased risk of mortality. […] Our results demonstrate delirium is a significant marker of mortality among seniors in the ED, and mortality risk is most salient in the first 3 months following an ED visit. […] We found that delirium is a significant marker of mortality among seniors visiting the ED, and that mortality risk is most prominent in the first three months following an ED visit. […] Our study of national claims-level data demonstrates that delirium is a significant marker of mortality among seniors visiting the ED, and that mortality risk is most prominent in the first 3 months following an ED visit.
#10 Delirium in hospitalized COVID-19 patients: Predictors and implications for patient outcome | PLOS One
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0278214
Delirium is recognized as a severe complication of coronavirus-disease-2019 (COVID-19). COVID-19-associated delirium has been linked to worse patient outcomes and is considered to be of multifactorial origin. […] A better understanding of COVID-19-associated delirium appears warranted as its occurrence has been linked to worse patient outcomes, including longer treatment duration, higher complication rates, increased mortality and, in COVID-19 survivors, consecutive cognitive decline. […] The risk of delirium in COVID-19 can be stratified based on COVID-19 disease severity and similar to delirium associated with other respiratory infections the factors advanced age, neurodegenerative disease history, and presence of elevated infection and renal-retention parameters. Screening for these risk factors may facilitate early identification of patients at high-risk for COVID-19-associated delirium.
#11 Delirium | Nature Reviews Disease Primers
https://www.nature.com/articles/s41572-020-00223-4
Delirium, a syndrome characterized by an acute change in attention, awareness and cognition, is caused by a medical condition that cannot be better explained by a pre-existing neurocognitive disorder. […] Currently, the best management strategies are multidomain interventions that focus on treating precipitating conditions, medication review, managing distress, mitigating complications and maintaining engagement to environmental issues. […] The effective implementation of delirium detection, treatment and prevention strategies remains a major challenge for health-care organizations globally. […] This prospective longitudinal cohort study demonstrated that critically ill patients are at risk of LTCI after critical illness, that this new LTCI can persist at 3 and 12 months follow-up, and that it is associated with duration of delirium.
#12 Delirium | Nature Reviews Disease Primers
https://www.nature.com/articles/s41572-020-00223-4
Delirium is prevalent in older hospital inpatients and associated with adverse outcomes: results of a prospective multi-centre study on World Delirium Awareness Day. […] Delirium in critically ill patients: impact on long-term health-related quality of life and cognitive functioning. […] Delirium as a predictor of long-term cognitive impairment in survivors of critical illness.
#13 Delirium | Nature Reviews Disease Primers
https://www.nature.com/articles/s41572-020-00223-4
Delirium is prevalent in older hospital inpatients and associated with adverse outcomes: results of a prospective multi-centre study on World Delirium Awareness Day. […] Delirium in critically ill patients: impact on long-term health-related quality of life and cognitive functioning. […] Delirium as a predictor of long-term cognitive impairment in survivors of critical illness.
#14 Interpretable machine learning model to predict the acute occurrence of delirium in elderly patients in the intensive care units: a retrospective cohort study | Journal of Big Data | Full Text
https://journalofbigdata.springeropen.com/articles/10.1186/s40537-025-01107-8
Delirium is a severe complication in critical elderly patients. This study aimed to develop interpretable machine-learning (ML) models to predict acute delirium and identify risk factors for medical intervention in elderly patients in the intensive care unit (ICU). […] Acute delirium is an independent risk of mortality in elderly patients in the intensive care unit. APACHEIV, sepsis, mechanical ventilation and temperature were important risk factors of acute delirium, which were potential targets for medication. […] A total of 66,263 elderly patients were selected from eICU-CRD. About 6299 patients (9.5%) developed delirium within 7 days after admission which was rarely reported elsewhere. […] Delirium is prevalent in ICU, especially in elderly patients. Actually, prediction of acute delirium in elderly patients is short of investigation.
#15 Navigating the machine learning pipeline: a scoping review of inpatient delirium prediction models
https://pmc.ncbi.nlm.nih.gov/articles/PMC10335592/
Early identification of inpatients at risk of developing delirium and implementing preventive measures could avoid up to 40% of delirium cases. […] ML models have potential to identify inpatients at risk of developing delirium before symptom onset. However, few models were externally validated and even fewer underwent prospective evaluation in clinical settings. […] This review confirms a rapidly growing body of research into using ML for predicting delirium risk in hospital settings. Our findings offer insights for both developers and clinicians into strengths and limitations of current ML delirium prediction applications aiming to support but not usurp clinician decision-making. […] Prediction models derived using ML methods can potentially identify individuals at risk of developing delirium before symptom onset to whom preventive strategies can be targeted, which may, in turn, reduce incident delirium and improve patient outcomes. This scoping review identified all publications describing ML-based delirium prediction models over the last 5 years, evaluated their stage in the ML evolution pipeline, and assessed their performance and utility. Relatively few were subject to external validation, which, when performed, showed degraded model performance. In addition, while few studies underwent prospective evaluation in real-world clinical settings, performance and user acceptance seemed promising in those that did. However, given the limitations of current delirium prediction models, they should not be seen as substitutes for expert clinician judgement.
#16 Navigating the machine learning pipeline: a scoping review of inpatient delirium prediction models
https://pmc.ncbi.nlm.nih.gov/articles/PMC10335592/
Early identification of inpatients at risk of developing delirium and implementing preventive measures could avoid up to 40% of delirium cases. […] ML models have potential to identify inpatients at risk of developing delirium before symptom onset. However, few models were externally validated and even fewer underwent prospective evaluation in clinical settings. […] This review confirms a rapidly growing body of research into using ML for predicting delirium risk in hospital settings. Our findings offer insights for both developers and clinicians into strengths and limitations of current ML delirium prediction applications aiming to support but not usurp clinician decision-making. […] Prediction models derived using ML methods can potentially identify individuals at risk of developing delirium before symptom onset to whom preventive strategies can be targeted, which may, in turn, reduce incident delirium and improve patient outcomes. This scoping review identified all publications describing ML-based delirium prediction models over the last 5 years, evaluated their stage in the ML evolution pipeline, and assessed their performance and utility. Relatively few were subject to external validation, which, when performed, showed degraded model performance. In addition, while few studies underwent prospective evaluation in real-world clinical settings, performance and user acceptance seemed promising in those that did. However, given the limitations of current delirium prediction models, they should not be seen as substitutes for expert clinician judgement.
#17 Delirium prediction in the intensive care unit: comparison of two delirium prediction models | Critical Care | Full Text
https://ccforum.biomedcentral.com/articles/10.1186/s13054-018-2037-6
Accurate prediction of delirium in the intensive care unit (ICU) may facilitate efficient use of early preventive strategies and stratification of ICU patients by delirium risk in clinical research, but the optimal delirium prediction model to use is unclear. […] While both ICU delirium prediction models have moderate-to-good performance, the PRE-DELIRIC model predicts delirium better. However, ICU physicians rated the user convenience of E-PRE-DELIRIC superior to PRE-DELIRIC. In low-risk patients the delirium prediction further improves after an update with the PRE-DELIRIC model after 24 h. […] It remains unclear which ICU delirium prediction model might be recommended for daily clinical practice, because the comparative predictive performance of the PRE-DELIRIC and the E-PRE-DELIRIC models and clinicians preferences have not been assessed.
#18 Delirium prediction in the intensive care unit: comparison of two delirium prediction models | Critical Care | Full Text
https://ccforum.biomedcentral.com/articles/10.1186/s13054-018-2037-6
The routine use of delirium preventive measures in the ICU is widely endorsed given the high prevalence of delirium, its deleterious effects on patient outcome, and the high costs related to these effects. […] This study shows that statistically both ICU delirium prediction models have moderate-to-good performance. Although the predictive accuracy of the PRE-DELIRIC is greater, the E-PRE-DELIRIC model scores significantly better on user convenience.
#19
https://link.springer.com/article/10.1007/s11096-023-01566-0
Effective delirium prevention could benefit from automatic risk stratification of older inpatients using routinely collected clinical data. […] Primary aim was to develop and validate a delirium prediction model (DELIKT) suitable for implementation in hospitals. […] The DELIKT is a potentially automatic tool with predictors from standard care including the CrAS to identify patients at high risk for delirium. […] Routinely collected data within the first 24 h of admission can be integrated into a prevention tool to automatically predict delirium in hospitalised older patients. […] The cumulative anticholinergic burden measured with the Clinician-rated Anticholinergic Scale is a reversible predictor for incident delirium, thus tailored medication lists with clear alternatives could installed as preventive measures. […] Automatic delirium risk stratification of older inpatients within the first 24 h of hospital admission might be a powerful tool for effective delirium prevention. […] A DELIKT score of more than 20 points was significantly associated with incident delirium.
#20 Journal of Medical Internet Research – Development and Validation of a Machine Learning Model for Early Prediction of Delirium in Intensive Care Units Using Continuous Physiological Data: Retrospective Study
https://www.jmir.org/2025/1/e59520
Background: Delirium in intensive care unit (ICU) patients poses a significant challenge, affecting patient outcomes and health care efficiency. Developing an accurate, real-time prediction model for delirium represents an advancement in critical care, addressing needs for timely intervention and resource optimization in ICUs. […] Our final model showed robust performance in internal validation (area under the receiver operating characteristic curve [AUROC]: 0.82; area under the precision-recall curve [AUPRC]: 0.62) and maintained its accuracy in temporal validation (AUROC: 0.73; AUPRC: 0.85). External validation supported its effectiveness (AUROC: 0.84; AUPRC: 0.77). […] Our study demonstrates the potential of the RF model in predicting delirium, with consistent performance across various validation scenarios. The model uses noninvasive variables, making it applicable to a wide range of ICU patients, with minimal additional risk.
#21 Prediction model for delirium in advanced cancer patients receiving palliative care: development and validation | BMC Palliative Care | Full Text
https://bmcpalliatcare.biomedcentral.com/articles/10.1186/s12904-025-01683-9
Delirium is a common and distressing mental disorder in palliative care. To date, no delirium prediction model is available for the palliative care population. […] The model serves as a reliable tool to predict delirium onset for advanced cancer patients in palliative care units, which will facilitate early targeted preventive measures to reduce the burden of delirium. […] The aims of our study were to develop and validate a model for predicting delirium in advanced cancer patients during their stay in palliative care units. Such a model may help to identify individuals at high risk of delirium and provide an important reference for targeted prevention and control measures, thereby contributing to a decreased incidence of delirium in palliative care units. […] In this multicenter prospective cohort study, we first developed and validated a model for predicting delirium in advanced cancer patients during their stay in palliative care units, which revealed excellent discrimination, calibration, and clinical practicability. […] The application of the prediction model will help palliative care staff to identify patients at high risk of delirium and facilitate targeted initiation of preventive measures.
#22 Systematic review of prediction models for delirium in the older adult inpatient | BMJ Open
https://bmjopen.bmj.com/content/8/4/e019223
Objective To identify existing prognostic delirium prediction models and evaluate their validity and statistical methodology in the older adult (60 years) acute hospital population. […] Conclusions Delirium prediction models for older adults show variable and typically inadequate predictive capabilities. Our review highlights the need for development of robust models to predict delirium in older inpatients. We provide recommendations for the development of such models. […] Reported AUROC in externally validated delirium prediction models ranged from 0.52 to 0.94. Of these models, the highest performing model (AUROC 0.94, 95% CI 0.91 to 0.97) was developed and validated in a surgical population. […] Overall, the variable inclusion and applied definitions in delirium prediction models are heterogeneous, making comparisons difficult. To improve delirium prediction models, future models should consider using standard variables and definitions to work towards a prediction tool that is generalisable to several populations within the remit of understanding the relationship with the precipitating event.
#23 Quantitative electroencephalography predicts postoperative delirium in adult cardiac surgical patients from a prospective observational study | Scientific Reports
https://www.nature.com/articles/s41598-024-82422-7
The diagnostic and prognostic value of quantitative electroencephalogram (qEEG) in the the onset of postoperative delirium (POD) remains an area of inquiry. […] The qEEG can reliably predict delirium after heart cardiac surgery. It is helpful for clinicians to early diagnose and manage these patients. […] The present date indicated that aEEG (peak value and valley value of F3-P3/F4-P4 derivation) had significant correlation in relation to the incidence of post-operative delirium. […] Our data showed that with only 4 electrodes and 20 min of in the first one hour after admission to ICU, qEEG recording can reliably predict delirium after heart cardiac surgery. […] The results showed that Q1 and Q4 of peak or valley value of F3-P3/F4-P4 derivation (for example, Q1 of peak value for F3-P3 derivation: OR 12.4, 95% CI 1.7289.76, p=0.012), and age (OR 1.1, 95% CI 1.00-1.14, p=0.039) had higher relationships with the incidence of post-operative delirium.
#24 Quantitative electroencephalography predicts postoperative delirium in adult cardiac surgical patients from a prospective observational study | Scientific Reports
https://www.nature.com/articles/s41598-024-82422-7
In general, the aEEG could achieve high sensitivity, high specificity, and overall good accuracy in predicting POD. For instance, the peak value of F3-P3 derivation as a predictor of POD showed a sensitivity of 90% and specificity of 72% with a cutoff value of 16.4 (p0.001). […] These results demonstrate qEEG is strongly correlate with delirium and could serve as valuable biomarkers for early detection of POD and assist in clinical decision-making.
#25
https://journals.lww.com/10.1097/ALN.0000000000004478
Delirium poses significant risks to patients, but countermeasures can be taken to mitigate negative outcomes. Accurately forecasting delirium in intensive care unit (ICU) patients could guide proactive intervention. Our primary objective was to predict ICU delirium by applying machine learning to clinical and physiologic data routinely collected in electronic health records. […] Machine learning models trained with routinely collected electronic health record data accurately predict ICU delirium, supporting dynamic time-sensitive forecasting. […] The static model using data from the first 24 h after ICU admission to predict delirium at any point during the ICU stay demonstrated higher discrimination compared with a widely cited reference model. […] The dynamic model was able to predict delirium up to 12 h in advance with reasonable discrimination and calibration.
#26
https://journals.lww.com/10.1097/ALN.0000000000004478
The ability to predict delirium onset in high-risk individuals might allow preventive or treatment strategies to be implemented in a more targeted or even personalized fashion. […] We hypothesized that physiologic and clinical variables routinely acquired during intensive care would be associated with the probability of delirium onset. […] The primary outcome variable was delirium, defined as a positive Confusion Assessment Method for the ICU screen, a score of 4 or more on the Intensive Care Delirium Screening Checklist, without any contradictions from diagnostic code information. […] The dynamic models performed overall better than the first 24-h model, with higher performances noted at short lead times. […] We reject the null hypothesis that physiologic and clinical variables routinely acquired during intensive care have no relation to the probability of delirium onset. […] Leveraging machine learning applied to very large datasets, we have developed and externally validated a novel approach for prediction of delirium in the ICU.
#27
https://journals.lww.com/10.1097/ALN.0000000000004478
Delirium poses significant risks to patients, but countermeasures can be taken to mitigate negative outcomes. Accurately forecasting delirium in intensive care unit (ICU) patients could guide proactive intervention. Our primary objective was to predict ICU delirium by applying machine learning to clinical and physiologic data routinely collected in electronic health records. […] Machine learning models trained with routinely collected electronic health record data accurately predict ICU delirium, supporting dynamic time-sensitive forecasting. […] The static model using data from the first 24 h after ICU admission to predict delirium at any point during the ICU stay demonstrated higher discrimination compared with a widely cited reference model. […] The dynamic model was able to predict delirium up to 12 h in advance with reasonable discrimination and calibration.
#28 Interpretable machine learning model to predict the acute occurrence of delirium in elderly patients in the intensive care units: a retrospective cohort study | Journal of Big Data | Full Text
https://journalofbigdata.springeropen.com/articles/10.1186/s40537-025-01107-8
XGB was proved to the best ML model of delirium prediction which was demonstrated in AUC-ROC and calibration curve analysis. […] Among all variables taken in ML model, GCS, APACHEIV and sepsis showed the highest importance. […] The discoveries in the present study provided help for the prevention and prompt treatment of acute delirium. APACHEIV and sepsis were showed to be the most important risk factors of acute delirium which reminded physicians to control systemic infection and improve status of critical patients. […] In summary, integrating ML predictions into clinical workflows holds promise for revolutionizing healthcare.
#29 Journal of Medical Internet Research – Development and Validation of a Machine Learning Model for Early Prediction of Delirium in Intensive Care Units Using Continuous Physiological Data: Retrospective Study
https://www.jmir.org/2025/1/e59520
The continuous nature of these measurements also allows our model to provide real-time, ongoing risk assessment, potentially capturing subtle physiological changes that might precede the onset of delirium. […] The ultimate goal of the delirium prediction model is seamless integration into clinical workflows via electronic health record systems and bedside monitoring devices.
#30 Risk Factors and Outcomes for Postoperative Delirium after Major Surgery in Elderly Patients | PLOS One
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0136071
Early identification of patients at risk for delirium is important, since adequate well timed interventions could prevent occurrence of delirium and related detrimental outcomes. […] Postoperative delirium is a frequent complication after major surgery in elderly patients and is related to an increase in adverse events, length of hospital stay, and mortality. […] A delirium in the medical history, advanced age, and ASA-score may assist in defining patients at increased risk for delirium. […] Delirium is related to an increase in morbidity, mortality, length of stay and care home placement. […] Most importantly, delirium could be prevented in approximately 30-40% of the cases. […] Our reported incidence rate of post-operative delirium (15%) is comparable with other studies in recent literature (11-18%).
#31 Interpretable machine learning model to predict the acute occurrence of delirium in elderly patients in the intensive care units: a retrospective cohort study | Journal of Big Data | Full Text
https://journalofbigdata.springeropen.com/articles/10.1186/s40537-025-01107-8
XGB was proved to the best ML model of delirium prediction which was demonstrated in AUC-ROC and calibration curve analysis. […] Among all variables taken in ML model, GCS, APACHEIV and sepsis showed the highest importance. […] The discoveries in the present study provided help for the prevention and prompt treatment of acute delirium. APACHEIV and sepsis were showed to be the most important risk factors of acute delirium which reminded physicians to control systemic infection and improve status of critical patients. […] In summary, integrating ML predictions into clinical workflows holds promise for revolutionizing healthcare.
#32 Interpretable machine learning model to predict the acute occurrence of delirium in elderly patients in the intensive care units: a retrospective cohort study | Journal of Big Data | Full Text
https://journalofbigdata.springeropen.com/articles/10.1186/s40537-025-01107-8
Delirium is a severe complication in critical elderly patients. This study aimed to develop interpretable machine-learning (ML) models to predict acute delirium and identify risk factors for medical intervention in elderly patients in the intensive care unit (ICU). […] Acute delirium is an independent risk of mortality in elderly patients in the intensive care unit. APACHEIV, sepsis, mechanical ventilation and temperature were important risk factors of acute delirium, which were potential targets for medication. […] A total of 66,263 elderly patients were selected from eICU-CRD. About 6299 patients (9.5%) developed delirium within 7 days after admission which was rarely reported elsewhere. […] Delirium is prevalent in ICU, especially in elderly patients. Actually, prediction of acute delirium in elderly patients is short of investigation.
#33 Interpretable machine learning model to predict the acute occurrence of delirium in elderly patients in the intensive care units: a retrospective cohort study | Journal of Big Data | Full Text
https://journalofbigdata.springeropen.com/articles/10.1186/s40537-025-01107-8
Delirium is a severe complication in critical elderly patients. This study aimed to develop interpretable machine-learning (ML) models to predict acute delirium and identify risk factors for medical intervention in elderly patients in the intensive care unit (ICU). […] Acute delirium is an independent risk of mortality in elderly patients in the intensive care unit. APACHEIV, sepsis, mechanical ventilation and temperature were important risk factors of acute delirium, which were potential targets for medication. […] A total of 66,263 elderly patients were selected from eICU-CRD. About 6299 patients (9.5%) developed delirium within 7 days after admission which was rarely reported elsewhere. […] Delirium is prevalent in ICU, especially in elderly patients. Actually, prediction of acute delirium in elderly patients is short of investigation.
#34 Interpretable machine learning model to predict the acute occurrence of delirium in elderly patients in the intensive care units: a retrospective cohort study | Journal of Big Data | Full Text
https://journalofbigdata.springeropen.com/articles/10.1186/s40537-025-01107-8
XGB was proved to the best ML model of delirium prediction which was demonstrated in AUC-ROC and calibration curve analysis. […] Among all variables taken in ML model, GCS, APACHEIV and sepsis showed the highest importance. […] The discoveries in the present study provided help for the prevention and prompt treatment of acute delirium. APACHEIV and sepsis were showed to be the most important risk factors of acute delirium which reminded physicians to control systemic infection and improve status of critical patients. […] In summary, integrating ML predictions into clinical workflows holds promise for revolutionizing healthcare.
#35
https://link.springer.com/article/10.1007/s11096-023-01566-0
Effective delirium prevention could benefit from automatic risk stratification of older inpatients using routinely collected clinical data. […] Primary aim was to develop and validate a delirium prediction model (DELIKT) suitable for implementation in hospitals. […] The DELIKT is a potentially automatic tool with predictors from standard care including the CrAS to identify patients at high risk for delirium. […] Routinely collected data within the first 24 h of admission can be integrated into a prevention tool to automatically predict delirium in hospitalised older patients. […] The cumulative anticholinergic burden measured with the Clinician-rated Anticholinergic Scale is a reversible predictor for incident delirium, thus tailored medication lists with clear alternatives could installed as preventive measures. […] Automatic delirium risk stratification of older inpatients within the first 24 h of hospital admission might be a powerful tool for effective delirium prevention. […] A DELIKT score of more than 20 points was significantly associated with incident delirium.
#36 Delirium in hospitalized COVID-19 patients: Predictors and implications for patient outcome | PLOS One
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0278214
Delirium is recognized as a severe complication of coronavirus-disease-2019 (COVID-19). COVID-19-associated delirium has been linked to worse patient outcomes and is considered to be of multifactorial origin. […] A better understanding of COVID-19-associated delirium appears warranted as its occurrence has been linked to worse patient outcomes, including longer treatment duration, higher complication rates, increased mortality and, in COVID-19 survivors, consecutive cognitive decline. […] The risk of delirium in COVID-19 can be stratified based on COVID-19 disease severity and similar to delirium associated with other respiratory infections the factors advanced age, neurodegenerative disease history, and presence of elevated infection and renal-retention parameters. Screening for these risk factors may facilitate early identification of patients at high-risk for COVID-19-associated delirium.
#37 Delirium | Nature Reviews Disease Primers
https://www.nature.com/articles/s41572-020-00223-4
Delirium, a syndrome characterized by an acute change in attention, awareness and cognition, is caused by a medical condition that cannot be better explained by a pre-existing neurocognitive disorder. […] Currently, the best management strategies are multidomain interventions that focus on treating precipitating conditions, medication review, managing distress, mitigating complications and maintaining engagement to environmental issues. […] The effective implementation of delirium detection, treatment and prevention strategies remains a major challenge for health-care organizations globally. […] This prospective longitudinal cohort study demonstrated that critically ill patients are at risk of LTCI after critical illness, that this new LTCI can persist at 3 and 12 months follow-up, and that it is associated with duration of delirium.
#38 Delirium | Nature Reviews Disease Primers
https://www.nature.com/articles/s41572-020-00223-4
Delirium, a syndrome characterized by an acute change in attention, awareness and cognition, is caused by a medical condition that cannot be better explained by a pre-existing neurocognitive disorder. […] Currently, the best management strategies are multidomain interventions that focus on treating precipitating conditions, medication review, managing distress, mitigating complications and maintaining engagement to environmental issues. […] The effective implementation of delirium detection, treatment and prevention strategies remains a major challenge for health-care organizations globally. […] This prospective longitudinal cohort study demonstrated that critically ill patients are at risk of LTCI after critical illness, that this new LTCI can persist at 3 and 12 months follow-up, and that it is associated with duration of delirium.
#39 Delirium | Nature Reviews Disease Primers
https://www.nature.com/articles/s41572-020-00223-4
Delirium, a syndrome characterized by an acute change in attention, awareness and cognition, is caused by a medical condition that cannot be better explained by a pre-existing neurocognitive disorder. […] Currently, the best management strategies are multidomain interventions that focus on treating precipitating conditions, medication review, managing distress, mitigating complications and maintaining engagement to environmental issues. […] The effective implementation of delirium detection, treatment and prevention strategies remains a major challenge for health-care organizations globally. […] This prospective longitudinal cohort study demonstrated that critically ill patients are at risk of LTCI after critical illness, that this new LTCI can persist at 3 and 12 months follow-up, and that it is associated with duration of delirium.
#40 Delirium | Nature Reviews Disease Primers
https://www.nature.com/articles/s41572-020-00223-4
Delirium, a syndrome characterized by an acute change in attention, awareness and cognition, is caused by a medical condition that cannot be better explained by a pre-existing neurocognitive disorder. […] Currently, the best management strategies are multidomain interventions that focus on treating precipitating conditions, medication review, managing distress, mitigating complications and maintaining engagement to environmental issues. […] The effective implementation of delirium detection, treatment and prevention strategies remains a major challenge for health-care organizations globally. […] This prospective longitudinal cohort study demonstrated that critically ill patients are at risk of LTCI after critical illness, that this new LTCI can persist at 3 and 12 months follow-up, and that it is associated with duration of delirium.
#41 Delirium | Nature Reviews Disease Primers
https://www.nature.com/articles/s41572-020-00223-4
Delirium, a syndrome characterized by an acute change in attention, awareness and cognition, is caused by a medical condition that cannot be better explained by a pre-existing neurocognitive disorder. […] Currently, the best management strategies are multidomain interventions that focus on treating precipitating conditions, medication review, managing distress, mitigating complications and maintaining engagement to environmental issues. […] The effective implementation of delirium detection, treatment and prevention strategies remains a major challenge for health-care organizations globally. […] This prospective longitudinal cohort study demonstrated that critically ill patients are at risk of LTCI after critical illness, that this new LTCI can persist at 3 and 12 months follow-up, and that it is associated with duration of delirium.
#42
https://journals.lww.com/10.1097/ALN.0000000000004478
The ability to predict delirium onset in high-risk individuals might allow preventive or treatment strategies to be implemented in a more targeted or even personalized fashion. […] We hypothesized that physiologic and clinical variables routinely acquired during intensive care would be associated with the probability of delirium onset. […] The primary outcome variable was delirium, defined as a positive Confusion Assessment Method for the ICU screen, a score of 4 or more on the Intensive Care Delirium Screening Checklist, without any contradictions from diagnostic code information. […] The dynamic models performed overall better than the first 24-h model, with higher performances noted at short lead times. […] We reject the null hypothesis that physiologic and clinical variables routinely acquired during intensive care have no relation to the probability of delirium onset. […] Leveraging machine learning applied to very large datasets, we have developed and externally validated a novel approach for prediction of delirium in the ICU.
#43 Delirium | Nature Reviews Disease Primers
https://www.nature.com/articles/s41572-020-00223-4
This meta-analysis provides evidence that, in elderly patients, delirium is associated with poor outcomes (mortality, institutionalization and dementia), independent of important confounders. […] This paper represents the clearest demonstration that progressive cognitive decline is a progressively increasing risk factor for delirium and also demonstrates, in mice, that this decline is correlated with increasing synaptic loss and can precede frank neurodegeneration. […] This systematic review concluded that there is insufficient evidence supporting the routine use of antipsychotic agents for the treatment of delirium. […] In the multisite, randomized, double-blinded, placebo-controlled REDUCE trial comparing prophylactic haloperidol with placebo for delirium prevention in critically ill adults, haloperidol did not improve survival at 28 days; thus, prophylactic haloperidol is not recommended for reducing mortality in critically ill adults.
#44 Delirium | Nature Reviews Disease Primers
https://www.nature.com/articles/s41572-020-00223-4
This meta-analysis provides evidence that, in elderly patients, delirium is associated with poor outcomes (mortality, institutionalization and dementia), independent of important confounders. […] This paper represents the clearest demonstration that progressive cognitive decline is a progressively increasing risk factor for delirium and also demonstrates, in mice, that this decline is correlated with increasing synaptic loss and can precede frank neurodegeneration. […] This systematic review concluded that there is insufficient evidence supporting the routine use of antipsychotic agents for the treatment of delirium. […] In the multisite, randomized, double-blinded, placebo-controlled REDUCE trial comparing prophylactic haloperidol with placebo for delirium prevention in critically ill adults, haloperidol did not improve survival at 28 days; thus, prophylactic haloperidol is not recommended for reducing mortality in critically ill adults.
#45 Navigating the machine learning pipeline: a scoping review of inpatient delirium prediction models
https://pmc.ncbi.nlm.nih.gov/articles/PMC10335592/
Early identification of inpatients at risk of developing delirium and implementing preventive measures could avoid up to 40% of delirium cases. […] ML models have potential to identify inpatients at risk of developing delirium before symptom onset. However, few models were externally validated and even fewer underwent prospective evaluation in clinical settings. […] This review confirms a rapidly growing body of research into using ML for predicting delirium risk in hospital settings. Our findings offer insights for both developers and clinicians into strengths and limitations of current ML delirium prediction applications aiming to support but not usurp clinician decision-making. […] Prediction models derived using ML methods can potentially identify individuals at risk of developing delirium before symptom onset to whom preventive strategies can be targeted, which may, in turn, reduce incident delirium and improve patient outcomes. This scoping review identified all publications describing ML-based delirium prediction models over the last 5 years, evaluated their stage in the ML evolution pipeline, and assessed their performance and utility. Relatively few were subject to external validation, which, when performed, showed degraded model performance. In addition, while few studies underwent prospective evaluation in real-world clinical settings, performance and user acceptance seemed promising in those that did. However, given the limitations of current delirium prediction models, they should not be seen as substitutes for expert clinician judgement.
#46 Navigating the machine learning pipeline: a scoping review of inpatient delirium prediction models
https://pmc.ncbi.nlm.nih.gov/articles/PMC10335592/
Early identification of inpatients at risk of developing delirium and implementing preventive measures could avoid up to 40% of delirium cases. […] ML models have potential to identify inpatients at risk of developing delirium before symptom onset. However, few models were externally validated and even fewer underwent prospective evaluation in clinical settings. […] This review confirms a rapidly growing body of research into using ML for predicting delirium risk in hospital settings. Our findings offer insights for both developers and clinicians into strengths and limitations of current ML delirium prediction applications aiming to support but not usurp clinician decision-making. […] Prediction models derived using ML methods can potentially identify individuals at risk of developing delirium before symptom onset to whom preventive strategies can be targeted, which may, in turn, reduce incident delirium and improve patient outcomes. This scoping review identified all publications describing ML-based delirium prediction models over the last 5 years, evaluated their stage in the ML evolution pipeline, and assessed their performance and utility. Relatively few were subject to external validation, which, when performed, showed degraded model performance. In addition, while few studies underwent prospective evaluation in real-world clinical settings, performance and user acceptance seemed promising in those that did. However, given the limitations of current delirium prediction models, they should not be seen as substitutes for expert clinician judgement.
#47 Systematic review of prediction models for delirium in the older adult inpatient | BMJ Open
https://bmjopen.bmj.com/content/8/4/e019223
Objective To identify existing prognostic delirium prediction models and evaluate their validity and statistical methodology in the older adult (60 years) acute hospital population. […] Conclusions Delirium prediction models for older adults show variable and typically inadequate predictive capabilities. Our review highlights the need for development of robust models to predict delirium in older inpatients. We provide recommendations for the development of such models. […] Reported AUROC in externally validated delirium prediction models ranged from 0.52 to 0.94. Of these models, the highest performing model (AUROC 0.94, 95% CI 0.91 to 0.97) was developed and validated in a surgical population. […] Overall, the variable inclusion and applied definitions in delirium prediction models are heterogeneous, making comparisons difficult. To improve delirium prediction models, future models should consider using standard variables and definitions to work towards a prediction tool that is generalisable to several populations within the remit of understanding the relationship with the precipitating event.
#48 Delirium | Nature Reviews Disease Primers
https://www.nature.com/articles/s41572-020-00223-4
Delirium, a syndrome characterized by an acute change in attention, awareness and cognition, is caused by a medical condition that cannot be better explained by a pre-existing neurocognitive disorder. […] Currently, the best management strategies are multidomain interventions that focus on treating precipitating conditions, medication review, managing distress, mitigating complications and maintaining engagement to environmental issues. […] The effective implementation of delirium detection, treatment and prevention strategies remains a major challenge for health-care organizations globally. […] This prospective longitudinal cohort study demonstrated that critically ill patients are at risk of LTCI after critical illness, that this new LTCI can persist at 3 and 12 months follow-up, and that it is associated with duration of delirium.
#49 Risk Factors and Outcomes for Postoperative Delirium after Major Surgery in Elderly Patients | PLOS One
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0136071
Delirium was related to multiple adverse events and increase in hospital stay. […] Thirty-day mortality was significantly higher (9%) in the delirious patients compared to the non-delirious patients (1%). […] The question of whether delirium is a symptom of other postoperative complications or whether a delirium increases the risk of postoperative adverse events remains to be answered. […] For elderly patients at risk for delirium, EVAR could be preferable to open surgery.
#50 Delirium in hospitalized COVID-19 patients: Predictors and implications for patient outcome | PLOS One
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0278214
Our study observed a negative effect of the occurrence of delirium on patients discharge modality, hereby confirming previous studies. Yet, we did not observe significant associations with the length of hospital stay or mortality, as reported in previous research. This finding might be partially explained by the specialized expertise of our center in the management of ARDS and the fact that, up to the start of our recruitment period, relevant experience in the management of COVID-19 had already been acquired during the first wave of the pandemic.
#51 Risk Factors and Outcomes for Postoperative Delirium after Major Surgery in Elderly Patients | PLOS One
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0136071
Early identification of patients at risk for delirium is important, since adequate well timed interventions could prevent occurrence of delirium and related detrimental outcomes. […] Postoperative delirium is a frequent complication after major surgery in elderly patients and is related to an increase in adverse events, length of hospital stay, and mortality. […] A delirium in the medical history, advanced age, and ASA-score may assist in defining patients at increased risk for delirium. […] Delirium is related to an increase in morbidity, mortality, length of stay and care home placement. […] Most importantly, delirium could be prevented in approximately 30-40% of the cases. […] Our reported incidence rate of post-operative delirium (15%) is comparable with other studies in recent literature (11-18%).
#52 Risk Factors and Outcomes for Postoperative Delirium after Major Surgery in Elderly Patients | PLOS One
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0136071
Delirium was related to multiple adverse events and increase in hospital stay. […] Thirty-day mortality was significantly higher (9%) in the delirious patients compared to the non-delirious patients (1%). […] The question of whether delirium is a symptom of other postoperative complications or whether a delirium increases the risk of postoperative adverse events remains to be answered. […] For elderly patients at risk for delirium, EVAR could be preferable to open surgery.
#53 Risk Factors and Outcomes for Postoperative Delirium after Major Surgery in Elderly Patients | PLOS One
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0136071
Early identification of patients at risk for delirium is important, since adequate well timed interventions could prevent occurrence of delirium and related detrimental outcomes. […] Postoperative delirium is a frequent complication after major surgery in elderly patients and is related to an increase in adverse events, length of hospital stay, and mortality. […] A delirium in the medical history, advanced age, and ASA-score may assist in defining patients at increased risk for delirium. […] Delirium is related to an increase in morbidity, mortality, length of stay and care home placement. […] Most importantly, delirium could be prevented in approximately 30-40% of the cases. […] Our reported incidence rate of post-operative delirium (15%) is comparable with other studies in recent literature (11-18%).
#54 Navigating the machine learning pipeline: a scoping review of inpatient delirium prediction models
https://pmc.ncbi.nlm.nih.gov/articles/PMC10335592/
Early identification of inpatients at risk of developing delirium and implementing preventive measures could avoid up to 40% of delirium cases. […] ML models have potential to identify inpatients at risk of developing delirium before symptom onset. However, few models were externally validated and even fewer underwent prospective evaluation in clinical settings. […] This review confirms a rapidly growing body of research into using ML for predicting delirium risk in hospital settings. Our findings offer insights for both developers and clinicians into strengths and limitations of current ML delirium prediction applications aiming to support but not usurp clinician decision-making. […] Prediction models derived using ML methods can potentially identify individuals at risk of developing delirium before symptom onset to whom preventive strategies can be targeted, which may, in turn, reduce incident delirium and improve patient outcomes. This scoping review identified all publications describing ML-based delirium prediction models over the last 5 years, evaluated their stage in the ML evolution pipeline, and assessed their performance and utility. Relatively few were subject to external validation, which, when performed, showed degraded model performance. In addition, while few studies underwent prospective evaluation in real-world clinical settings, performance and user acceptance seemed promising in those that did. However, given the limitations of current delirium prediction models, they should not be seen as substitutes for expert clinician judgement.
#55 Risk Factors and Outcomes for Postoperative Delirium after Major Surgery in Elderly Patients | PLOS One
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0136071
Delirium was related to multiple adverse events and increase in hospital stay. […] Thirty-day mortality was significantly higher (9%) in the delirious patients compared to the non-delirious patients (1%). […] The question of whether delirium is a symptom of other postoperative complications or whether a delirium increases the risk of postoperative adverse events remains to be answered. […] For elderly patients at risk for delirium, EVAR could be preferable to open surgery.
#56 Navigating the machine learning pipeline: a scoping review of inpatient delirium prediction models
https://pmc.ncbi.nlm.nih.gov/articles/PMC10335592/
Early identification of inpatients at risk of developing delirium and implementing preventive measures could avoid up to 40% of delirium cases. […] ML models have potential to identify inpatients at risk of developing delirium before symptom onset. However, few models were externally validated and even fewer underwent prospective evaluation in clinical settings. […] This review confirms a rapidly growing body of research into using ML for predicting delirium risk in hospital settings. Our findings offer insights for both developers and clinicians into strengths and limitations of current ML delirium prediction applications aiming to support but not usurp clinician decision-making. […] Prediction models derived using ML methods can potentially identify individuals at risk of developing delirium before symptom onset to whom preventive strategies can be targeted, which may, in turn, reduce incident delirium and improve patient outcomes. This scoping review identified all publications describing ML-based delirium prediction models over the last 5 years, evaluated their stage in the ML evolution pipeline, and assessed their performance and utility. Relatively few were subject to external validation, which, when performed, showed degraded model performance. In addition, while few studies underwent prospective evaluation in real-world clinical settings, performance and user acceptance seemed promising in those that did. However, given the limitations of current delirium prediction models, they should not be seen as substitutes for expert clinician judgement.
#57
https://link.springer.com/article/10.1007/s11096-023-01566-0
Effective delirium prevention could benefit from automatic risk stratification of older inpatients using routinely collected clinical data. […] Primary aim was to develop and validate a delirium prediction model (DELIKT) suitable for implementation in hospitals. […] The DELIKT is a potentially automatic tool with predictors from standard care including the CrAS to identify patients at high risk for delirium. […] Routinely collected data within the first 24 h of admission can be integrated into a prevention tool to automatically predict delirium in hospitalised older patients. […] The cumulative anticholinergic burden measured with the Clinician-rated Anticholinergic Scale is a reversible predictor for incident delirium, thus tailored medication lists with clear alternatives could installed as preventive measures. […] Automatic delirium risk stratification of older inpatients within the first 24 h of hospital admission might be a powerful tool for effective delirium prevention. […] A DELIKT score of more than 20 points was significantly associated with incident delirium.
#58 Prediction model for delirium in advanced cancer patients receiving palliative care: development and validation | BMC Palliative Care | Full Text
https://bmcpalliatcare.biomedcentral.com/articles/10.1186/s12904-025-01683-9
Delirium is a common and distressing mental disorder in palliative care. To date, no delirium prediction model is available for the palliative care population. […] The model serves as a reliable tool to predict delirium onset for advanced cancer patients in palliative care units, which will facilitate early targeted preventive measures to reduce the burden of delirium. […] The aims of our study were to develop and validate a model for predicting delirium in advanced cancer patients during their stay in palliative care units. Such a model may help to identify individuals at high risk of delirium and provide an important reference for targeted prevention and control measures, thereby contributing to a decreased incidence of delirium in palliative care units. […] In this multicenter prospective cohort study, we first developed and validated a model for predicting delirium in advanced cancer patients during their stay in palliative care units, which revealed excellent discrimination, calibration, and clinical practicability. […] The application of the prediction model will help palliative care staff to identify patients at high risk of delirium and facilitate targeted initiation of preventive measures.
#59 Systematic review of prediction models for delirium in the older adult inpatient | BMJ Open
https://bmjopen.bmj.com/content/8/4/e019223
Objective To identify existing prognostic delirium prediction models and evaluate their validity and statistical methodology in the older adult (60 years) acute hospital population. […] Conclusions Delirium prediction models for older adults show variable and typically inadequate predictive capabilities. Our review highlights the need for development of robust models to predict delirium in older inpatients. We provide recommendations for the development of such models. […] Reported AUROC in externally validated delirium prediction models ranged from 0.52 to 0.94. Of these models, the highest performing model (AUROC 0.94, 95% CI 0.91 to 0.97) was developed and validated in a surgical population. […] Overall, the variable inclusion and applied definitions in delirium prediction models are heterogeneous, making comparisons difficult. To improve delirium prediction models, future models should consider using standard variables and definitions to work towards a prediction tool that is generalisable to several populations within the remit of understanding the relationship with the precipitating event.
#60 Systematic review of prediction models for delirium in the older adult inpatient | BMJ Open
https://bmjopen.bmj.com/content/8/4/e019223
Objective To identify existing prognostic delirium prediction models and evaluate their validity and statistical methodology in the older adult (60 years) acute hospital population. […] Conclusions Delirium prediction models for older adults show variable and typically inadequate predictive capabilities. Our review highlights the need for development of robust models to predict delirium in older inpatients. We provide recommendations for the development of such models. […] Reported AUROC in externally validated delirium prediction models ranged from 0.52 to 0.94. Of these models, the highest performing model (AUROC 0.94, 95% CI 0.91 to 0.97) was developed and validated in a surgical population. […] Overall, the variable inclusion and applied definitions in delirium prediction models are heterogeneous, making comparisons difficult. To improve delirium prediction models, future models should consider using standard variables and definitions to work towards a prediction tool that is generalisable to several populations within the remit of understanding the relationship with the precipitating event.
#61 Delirium in hospitalized COVID-19 patients: Predictors and implications for patient outcome | PLOS One
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0278214
Delirium is recognized as a severe complication of coronavirus-disease-2019 (COVID-19). COVID-19-associated delirium has been linked to worse patient outcomes and is considered to be of multifactorial origin. […] A better understanding of COVID-19-associated delirium appears warranted as its occurrence has been linked to worse patient outcomes, including longer treatment duration, higher complication rates, increased mortality and, in COVID-19 survivors, consecutive cognitive decline. […] The risk of delirium in COVID-19 can be stratified based on COVID-19 disease severity and similar to delirium associated with other respiratory infections the factors advanced age, neurodegenerative disease history, and presence of elevated infection and renal-retention parameters. Screening for these risk factors may facilitate early identification of patients at high-risk for COVID-19-associated delirium.
#62 Delirium | Nature Reviews Disease Primers
https://www.nature.com/articles/s41572-020-00223-4
Delirium is prevalent in older hospital inpatients and associated with adverse outcomes: results of a prospective multi-centre study on World Delirium Awareness Day. […] Delirium in critically ill patients: impact on long-term health-related quality of life and cognitive functioning. […] Delirium as a predictor of long-term cognitive impairment in survivors of critical illness.
#63 Prediction model for delirium in advanced cancer patients receiving palliative care: development and validation | BMC Palliative Care | Full Text
https://bmcpalliatcare.biomedcentral.com/articles/10.1186/s12904-025-01683-9
Delirium is a common and distressing mental disorder in palliative care. To date, no delirium prediction model is available for the palliative care population. […] The model serves as a reliable tool to predict delirium onset for advanced cancer patients in palliative care units, which will facilitate early targeted preventive measures to reduce the burden of delirium. […] The aims of our study were to develop and validate a model for predicting delirium in advanced cancer patients during their stay in palliative care units. Such a model may help to identify individuals at high risk of delirium and provide an important reference for targeted prevention and control measures, thereby contributing to a decreased incidence of delirium in palliative care units. […] In this multicenter prospective cohort study, we first developed and validated a model for predicting delirium in advanced cancer patients during their stay in palliative care units, which revealed excellent discrimination, calibration, and clinical practicability. […] The application of the prediction model will help palliative care staff to identify patients at high risk of delirium and facilitate targeted initiation of preventive measures.
#64 Journal of Medical Internet Research – Development and Validation of a Machine Learning Model for Early Prediction of Delirium in Intensive Care Units Using Continuous Physiological Data: Retrospective Study
https://www.jmir.org/2025/1/e59520
Background: Delirium in intensive care unit (ICU) patients poses a significant challenge, affecting patient outcomes and health care efficiency. Developing an accurate, real-time prediction model for delirium represents an advancement in critical care, addressing needs for timely intervention and resource optimization in ICUs. […] Our final model showed robust performance in internal validation (area under the receiver operating characteristic curve [AUROC]: 0.82; area under the precision-recall curve [AUPRC]: 0.62) and maintained its accuracy in temporal validation (AUROC: 0.73; AUPRC: 0.85). External validation supported its effectiveness (AUROC: 0.84; AUPRC: 0.77). […] Our study demonstrates the potential of the RF model in predicting delirium, with consistent performance across various validation scenarios. The model uses noninvasive variables, making it applicable to a wide range of ICU patients, with minimal additional risk.
#65 Journal of Medical Internet Research – Development and Validation of a Machine Learning Model for Early Prediction of Delirium in Intensive Care Units Using Continuous Physiological Data: Retrospective Study
https://www.jmir.org/2025/1/e59520
The continuous nature of these measurements also allows our model to provide real-time, ongoing risk assessment, potentially capturing subtle physiological changes that might precede the onset of delirium. […] The ultimate goal of the delirium prediction model is seamless integration into clinical workflows via electronic health record systems and bedside monitoring devices.