Małopłytkowość – Rokowania, prognozy i postęp choroby

Małopłytkowość
Rokowania, prognozy i postęp choroby

Małopłytkowość (trombocytopenia) jest istotnym markerem prognostycznym w opiece nad pacjentami na oddziałach intensywnej terapii (OIT), niezależnie od innych wskaźników ciężkości choroby. Zarówno niski nadir liczby płytek, jak i ich znaczący spadek korelują z gorszymi wynikami leczenia, w tym wyższą śmiertelnością (HR=1,45; 95% CI: 1,36-1,56). Szczególnie w sepsie małopłytkowość wiąże się z wyższym ryzykiem ostrego uszkodzenia nerek (44,1% vs. 29,5%, P<0,01), dłuższym czasem stosowania leków wazopresyjnych (mediana 37 vs. 23 h, P<0,01) oraz wydłużonym pobytem na OIT (mediana 3,1 vs. 2,1 dni, P<0,01). Brak ustąpienia małopłytkowości jest silnie związany ze zwiększoną 28-dniową śmiertelnością (48,2% vs. 38,5%, P<0,01), co pozostaje istotne po korekcie o wiek, wynik APACHE III i protokół resuscytacji. U pacjentów geriatrycznych po operacjach małopłytkowość występuje u 18,2% i koreluje z ciężkością choroby, wyższą śmiertelnością oraz dłuższym pobytem w szpitalu. W pediatrii, trombokryt i liczba płytek są czułymi wskaźnikami prognostycznymi, przewyższającymi nawet skalę PRISM i CRP.

Prognozy małopłytkowości – przewidywanie wyników leczenia

Małopłytkowość jako wskaźnik ryzyka śmiertelności
Małopłytkowość w sepsie i na OIT
Małopłytkowość u pacjentów geriatrycznych
Małopłytkowość w populacji pediatrycznej
Czynniki ryzyka związane ze śmiertelnością w małopłytkowości
Immunologiczna małopłytkowość u dorosłych

Modele predykcyjne w małopłytkowości

Nomogramy prognostyczne
Zastosowanie sztucznej inteligencji w prognozowaniu małopłytkowości
Ograniczenia i kierunki przyszłych badań
Kolejne rozdziały

Prognozy małopłytkowości – przewidywanie wyników leczenia

Małopłytkowość (trombocytopenia) to stan medyczny charakteryzujący się obniżoną liczbą płytek krwi. Ocena prognostyczna małopłytkowości ma kluczowe znaczenie dla personelu medycznego w planowaniu skutecznych strategii terapeutycznych oraz zarządzaniu opieką nad pacjentem. Odpowiednia stratyfikacja ryzyka oraz wczesna ocena rokowania są niezbędne dla optymalizacji leczenia i poprawy wyników klinicznych.¹²

Małopłytkowość jako wskaźnik ryzyka śmiertelności

Małopłytkowość stanowi prosty i łatwo dostępny marker ryzyka śmiertelności na oddziałach intensywnej terapii (OIT), działający niezależnie od i komplementarnie do ustalonych wskaźników ciężkości choroby. Zarówno niski nadir liczby płytek krwi, jak i znaczący spadek liczby płytek prognozują gorsze wyniki leczenia u dorosłych pacjentów OIT.³ Klinicznie, obecność małopłytkowości może być bardzo niebezpieczna i wiąże się z niekorzystnym rokowaniem u pacjentów z powodu nadmiernego krwawienia, jeśli nie zostanie szybko zdiagnozowana i leczona.⁴⁵

Badania wykazały, że zarówno małopłytkowość, jak i trombocytoza są niezależnie związane z krótszym całkowitym przeżyciem u osób starszych. Po uwzględnieniu wszystkich zmiennych towarzyszących, stwierdzono istotny związek ze zwiększoną śmiertelnością zarówno dla małopłytkowości (HR=1,45; 95% CI: 1,36-1,56), jak i trombocytozy (HR=1,75; 95% CI: 1,56-1,97) w porównaniu z normalnymi wartościami płytek krwi. Efekt ten jest modyfikowany przez pochodzenie etniczne.⁶

Małopłytkowość w sepsie i na OIT

Sepsa jest głównym czynnikiem ryzyka rozwoju małopłytkowości. Pacjenci z małopłytkowością w przebiegu sepsy mają wyższą częstość występowania ostrego uszkodzenia nerek (44,1% vs. 29,5%, P<0,01), wymagają dłuższego wsparcia lekami wazopresyjnymi (mediana (IQR): 37 (17-76) vs. 23 (13-46) h, P<0,01) oraz dłuższego pobytu na OIT (mediana (IQR): 3,1 (1,6-7,8) vs. 2,1 (1,2-4,4) dni, P<0,01).⁷

Co istotne, nie sama małopłytkowość, ale brak jej ustąpienia był związany ze zwiększoną 28-dniową śmiertelnością. Związek między nieustępującą małopłytkowością a śmiertelnością pozostał istotny nawet po skorygowaniu o wiek, wynik APACHE III i zgodność z protokołem resuscytacji w sepsie (P<0,01). Pacjenci z utrzymującą się małopłytkowością mieli wyższą śmiertelność na OIT, w szpitalu oraz 28-dniową w porównaniu do pacjentów, u których liczba płytek krwi znormalizowała się podczas hospitalizacji.⁸

Wyniki badań pokazują, że 28-dniowa śmiertelność u pacjentów z sepsą i małopłytkowością była znacząco wyższa niż u pacjentów bez małopłytkowości na początku hospitalizacji (48,2% vs. 38,5%).⁹

Małopłytkowość u pacjentów geriatrycznych

Małopłytkowość wystąpiła u 18,2% pooperacyjnych pacjentów geriatrycznych przebywających na OIT, co wiązało się z wyższym wskaźnikiem śmiertelności.¹⁰ Badania wykazały, że stopień spadku liczby płytek krwi był bezpośrednio skorelowany z ciężkością choroby – im poważniejsza redukcja płytek, tym wyższy wskaźnik śmiertelności pacjentów, dłuższy pobyt w szpitalu i wyższe koszty hospitalizacji.¹¹ Jest to zgodne z wcześniejszymi ustaleniami badawczymi, co sugeruje, że nasilenie spadku liczby płytek może pomóc we wczesnej identyfikacji rokowania pacjenta.¹²

Małopłytkowość w populacji pediatrycznej

U dzieci z ciężką sepsą, liczba płytek krwi i PCT (trombokryt) były znacząco niższe (p<0,001), a MPV (średnia objętość płytek) znacząco wyższa u pacjentów, którzy nie przeżyli w porównaniu do tych, którzy przeżyli (p=0,004).¹³ Trombokryt był najbardziej czułym parametrem prognozującym zgon, z małopłytkowością na drugim miejscu, mającą taką samą czułość jak skala PRISM. MPV w tym badaniu był najmniej czułym parametrem płytkowym, ale nadal bardziej czułym niż CRP, który jest najczęściej używanym markerem zapalnym ocenianym u dzieci z sepsą.¹⁴

Czynniki ryzyka związane ze śmiertelnością w małopłytkowości

Identyfikacja czynników ryzyka śmiertelności u pacjentów z ciężką małopłytkowością może pomóc klinicystom w opracowaniu zindywidualizowanych strategii leczenia i ulepszonego zarządzania opieką, aby zmniejszyć ryzyko zdarzeń niepożądanych i poprawić wyniki leczenia pacjentów.¹⁵ Na podstawie badań zidentyfikowano następujące niezależne czynniki ryzyka śmiertelności wewnątrzszpitalnej u pacjentów z ciężką małopłytkowością:

Wiek¹⁶
Obecność chorób naczyniowo-mózgowych¹⁷
Nowotwór złośliwy¹⁸
Saturacja tlenem¹⁹
Częstość akcji serca²⁰
Częstość oddechów²¹
Średnie ciśnienie tętnicze²²
Wentylacja mechaniczna²³
Stosowanie leków wazopresyjnych²⁴
CRRT (ciągła terapia nerkozastępcza)²⁵
PT (czas protrombinowy)²⁶
PTT (czas częściowej tromboplastyny)²⁷
BUN (azot mocznikowy we krwi)²⁸

Badania wykazały, że BUN jest niezależnie związany z rokowaniem pacjentów z ciężką małopłytkowością, prawdopodobnie dlatego, że pacjenci z niewydolnością nerek mają również złożone zaburzenia hemostazy.²⁹

Immunologiczna małopłytkowość u dorosłych

Immunologiczna małopłytkowość (ITP) u dorosłych wykazuje tendencję do progresji do formy przewlekłej u większości pacjentów. Niższa liczba płytek krwi może wskazywać na bardziej korzystne rokowanie.³⁰ Po 12 miesiącach tylko 37% pacjentów, którzy nie byli poddawani interwencjom modyfikującym przebieg choroby, osiągnęło wyleczenie.³¹

Wyższa liczba płytek krwi lub brak ciężkiego krwawienia na początku choroby były związane ze zwiększonym ryzykiem przewlekłości po 12 miesiącach.³² Badanie potwierdza, że ITP występuje częściej u kobiet niż u mężczyzn i często wiąże się z krwawieniem skórnym, chociaż krwawienia zagrażające życiu są rzadkie.³³

Obecność wywiadu rodzinnego chorób autoimmunologicznych stwierdzono u 9 (8,3%) ze 109 badanych pacjentów z ITP. Ogólnie, posiadanie krewnego pierwszego stopnia z historią zaburzeń autoimmunologicznych nie było związane ze zwiększonym ryzykiem rozwoju ITP.³⁴³⁵

Modele predykcyjne w małopłytkowości

Nomogramy prognostyczne

Nomogram jest powszechną wizualną prezentacją modeli predykcyjnych. W porównaniu z tradycyjnymi systemami oceny pacjentów krytycznie chorych (SOFA i SAPS II), stratyfikacja ryzyka oparta na modelach nomogramowych wykazała wyższą korzyść kliniczną.³⁶ Opracowany model nomogramu może pomóc w ilościowej ocenie czynników ryzyka prognostycznego u pacjentów z ciężką małopłytkowością. Model ten ma doskonałą wydajność predykcyjną i może stanowić dobre odniesienie do oceny wskaźnika śmiertelności wewnątrzszpitalnej u pacjentów z ciężką małopłytkowością.³⁷

Zastosowanie sztucznej inteligencji w prognozowaniu małopłytkowości

Ponieważ sztuczna inteligencja (AI) jest w stanie łączyć i oceniać wiele zmiennych liniowych i nieliniowych jednocześnie, wykazała duży potencjał w zastosowaniu we wczesnej diagnostyce, ocenie rokowania i przewidywaniu rozkładu pacjentów z małopłytkowością.³⁸ Dostępne dowody sugerują, że AI może być skuteczna w przewidywaniu i ocenie rokowania pacjentów z małopłytkowością.³⁹⁴⁰

Opracowano interpretowalny algorytm uczenia maszynowego, który prospektywnie przewiduje ryzyko małopłytkowości u starszych pacjentów w stanie krytycznym podczas ich pobytu na oddziale intensywnej terapii (OIT), ostatecznie wspierając podejmowanie decyzji klinicznych i poprawiając opiekę nad pacjentem.⁴¹ Model C5.0 wykazał doskonałą zdolność przewidywania małopłytkowości u pooperacyjnych pacjentów geriatrycznych OIT, a jego AUC, dokładność, precyzja, swoistość i czułość były zadowalające, zapewniając wiarygodne odniesienie dla klinicystów.⁴²

Model predykcyjny do oceny ciężkości małopłytkowości wykazał dokładność 0,84 z przedziałem ufności 0,81-0,86 dla walidacji wewnętrznej. W przypadku walidacji zewnętrznej z wykorzystaniem bazy danych MIMIC, model wykazał dokładność 0,80 z przedziałem ufności 0,79-0,81.⁴³

Model losowego lasu (Random Forest) miał AUC 0,901 po zewnętrznej walidacji, co świadczy o dobrej zdolności predykcyjnej ryzyka krytycznych krwawień.⁴⁴ Wyniki sugerują, że algorytmy uczenia maszynowego mogą okazać się korzystne w przewidywaniu małopłytkowości związanej z sepsą (SAT) i ciężkiej małopłytkowości, a tym samym pomogłyby we wczesnym zarządzaniu pacjentami z grupy ryzyka.⁴⁵⁴⁶

Ograniczenia i kierunki przyszłych badań

Mimo obiecujących wyników, proponowane modele wymagają dalszej walidacji w wieloośrodkowych badaniach prospektywnych.⁴⁷ Model zaproponowany przez Linga i współpracowników może okazać się bardziej użyteczny, jeśli uwzględni więcej zmiennych, które zapewniłyby wyższą zdolność predykcyjną dla 28-dniowej śmiertelności.⁴⁸⁴⁹

Czynniki prognostyczne chroniczności były badane prospektywnie po raz pierwszy u dorosłych, sugerując, że wyższe liczby płytek krwi przy rozpoznaniu są negatywnie skorelowane z ryzykiem chroniczności. Jednakże potrzebne są dalsze badania, aby potwierdzić to odkrycie i potencjalnie zidentyfikować inne czynniki predykcyjne dla chroniczności w momencie diagnozy.⁵⁰

Pomimo ograniczeń, mocne strony badań obejmują wykorzystanie różnych algorytmów uczenia maszynowego oraz szczegółową ocenę i walidację modeli, aby zapewnić ich odporność i wiarygodność.⁵¹

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 Development and validation of a prediction model for in-hospital mortality of patients with severe thrombocytopenia | Scientific Reports
https://www.nature.com/articles/s41598-022-10438-y
Risk stratification and prognosis evaluation of severe thrombocytopenia are essential for clinical treatment and management. […] Currently, there is currently no reliable predictive model to identify patients at high risk of severe thrombocytopenia. […] The nomogram model we established showed superior predictive performance and can assist in the quantitative assessment of the prognostic risk in patients with severe thrombocytopenia. […] Identifying risk factors for mortality in patients with severe thrombocytopenia can help clinicians develop individualized treatment strategies and enhanced care management to reduce the risk of adverse events and improve patient outcomes. […] We revealed the following as independent risk factors for in-hospital mortality in patients with severe thrombocytopenia: age, presence of cerebrovascular disease, malignant cancer, oxygen saturation, heart rate, respiration rate, mean arterial pressure, mechanical ventilation, vasopressor use, CRRT; PT, PTT, and BUN.
#2 Applications of Artificial Intelligence in Thrombocytopenia
https://www.mdpi.com/2075-4418/13/6/1060
Thrombocytopenia is a medical condition where blood platelet count drops very low. […] Clinically, the presence of thrombocytopenia might be very dangerous and is associated with poor outcomes of patients due to excessive bleeding if not addressed quickly enough. […] Hence, early detection and evaluation of thrombocytopenia is essential for rapid and appropriate intervention for these patients. […] Since artificial intelligence is able to combine and evaluate many linear and nonlinear variables simultaneously, it has shown great potential in its application in the early diagnosis, assessing the prognosis and predicting the distribution of patients with thrombocytopenia. […] The included studies showed that artificial intelligence can potentially enhance the clinical approaches used in the diagnosis, prognosis, and treatment of thrombocytopenia.
#3 A prospective study of thrombocytopenia and prognosis in intensive care
https://pmc.ncbi.nlm.nih.gov/articles/PMC3301937/
To study the incidence and prognosis of thrombocytopenia in an adult critically ill population. […] Thrombocytopenia is a simple and readily available risk marker for ICU mortality, independent of and complementary to established severity of disease indices. Both a low nadir thrombocytosis and a significant fall of platelet count predict a poor vital outcome in adult ICU patients.
#4 Applications of Artificial Intelligence in Thrombocytopenia
https://pmc.ncbi.nlm.nih.gov/articles/PMC10047875/
Thrombocytopenia is a medical condition where blood platelet count drops very low. […] Clinically, the presence of thrombocytopenia might be very dangerous and is associated with poor outcomes of patients due to excessive bleeding if not addressed quickly enough. […] Hence, early detection and evaluation of thrombocytopenia is essential for rapid and appropriate intervention for these patients. […] The included studies showed that artificial intelligence can potentially enhance the clinical approaches used in the diagnosis, prognosis, and treatment of thrombocytopenia. […] The development of thrombocytopenia in sepsis patients is an indication of poor prognosis. […] Hence, the use of AI for early identification and risk prediction in these patients can be of great value. […] As stated previously, thrombocytopenia is associated with poor prognosis in sepsis patients.
#5 Applications of Artificial Intelligence in Thrombocytopenia
https://www.mdpi.com/2075-4418/13/6/1060
Thrombocytopenia is a medical condition where blood platelet count drops very low. […] Clinically, the presence of thrombocytopenia might be very dangerous and is associated with poor outcomes of patients due to excessive bleeding if not addressed quickly enough. […] Hence, early detection and evaluation of thrombocytopenia is essential for rapid and appropriate intervention for these patients. […] Since artificial intelligence is able to combine and evaluate many linear and nonlinear variables simultaneously, it has shown great potential in its application in the early diagnosis, assessing the prognosis and predicting the distribution of patients with thrombocytopenia. […] The included studies showed that artificial intelligence can potentially enhance the clinical approaches used in the diagnosis, prognosis, and treatment of thrombocytopenia.
#6
https://www.haematologica.org/content/99/5/930.long
Even though alterations in platelet counts are presumed to be detrimental, their impact on the survival of patients has not been studied in large cohorts. A significant association with shorter overall survival was found for both thrombocytopenia (HR=1.45; 95% CI: 1.361.56) and thrombocytosis (HR=1.75; 95% CI: 1.561.97) when compared to the survival of patients with normal platelet counts. In conclusion, thrombocytosis and thrombocytopenia are independently associated with shorter overall survival in elderly subjects and this effect is modified by ethnicity. […] The overall mortality rate was 62.6 per 1,000 person-years. After adjusting for all covariates, a significant association with increased mortality was found for both thrombocytopenia (HR=1.45; 95% CI: 1.361.56) and thrombocytosis (HR=1.75; 95% CI: 1.561.97) when compared to normal platelet counts. […] The present study demonstrated that both thrombocytosis and thrombocytopenia independently predicted mortality in a large, outpatient cohort of inner city, elderly subjects and this effect was heterogeneous across different ethnicities.
#7 Thrombocytopenia in adult patients with sepsis: incidence, risk factors, and its association with clinical outcome | Journal of Intensive Care | Full Text
https://jintensivecare.biomedcentral.com/articles/10.1186/2052-0492-1-9
Sepsis is a major risk factor for the development of thrombocytopenia, but few studies have specifically evaluated prognostic importance of thrombocytopenia in patients with sepsis. […] Patients with thrombocytopenia had a higher incidence of acute kidney injury (44.1% vs. 29.5%,P0.01), prolonged vasopressor support (median (IQR): 37 (1776) vs. 23 (1346) h,P0.01), and longer ICU stay (median (IQR): 3.1 (1.67.8) vs. 2.1 (1.24.4) days,P0.01). […] Non-resolution of thrombocytopenia, but not thrombocytopenia itself, was associated with increased 28-day mortality. […] The association between non-resolution of thrombocytopenia and mortality remained significant after adjusting for age, APACHE III score and compliance with a sepsis resuscitation bundle (P 0.01). […] Patients with persistent thrombocytopenia had higher ICU, hospital, and 28-day mortality compared to those patients whose platelet counts normalized during hospitalization. […] The association between non-resolution of thrombocytopenia and 28-day mortality remained significant after adjusting for age, APACHE III score, and compliance with sepsis resuscitation bundle.
#8 Thrombocytopenia in adult patients with sepsis: incidence, risk factors, and its association with clinical outcome | Journal of Intensive Care | Full Text
https://jintensivecare.biomedcentral.com/articles/10.1186/2052-0492-1-9
Sepsis is a major risk factor for the development of thrombocytopenia, but few studies have specifically evaluated prognostic importance of thrombocytopenia in patients with sepsis. […] Patients with thrombocytopenia had a higher incidence of acute kidney injury (44.1% vs. 29.5%,P0.01), prolonged vasopressor support (median (IQR): 37 (1776) vs. 23 (1346) h,P0.01), and longer ICU stay (median (IQR): 3.1 (1.67.8) vs. 2.1 (1.24.4) days,P0.01). […] Non-resolution of thrombocytopenia, but not thrombocytopenia itself, was associated with increased 28-day mortality. […] The association between non-resolution of thrombocytopenia and mortality remained significant after adjusting for age, APACHE III score and compliance with a sepsis resuscitation bundle (P 0.01). […] Patients with persistent thrombocytopenia had higher ICU, hospital, and 28-day mortality compared to those patients whose platelet counts normalized during hospitalization. […] The association between non-resolution of thrombocytopenia and 28-day mortality remained significant after adjusting for age, APACHE III score, and compliance with sepsis resuscitation bundle.
#9 Applications of Artificial Intelligence in Thrombocytopenia
https://www.mdpi.com/2075-4418/13/6/1060
As stated previously, thrombocytopenia is associated with poor prognosis in sepsis patients. […] Hence, it is important to properly assess the prognosis of patients with SAT to ensure proper care for these patients. […] Several studies have also shown that RDW has significant clinical utility as an independent predictor of poor prognosis in critically ill patients with sepsis and SAT. […] The results of the paper showed that 28-day mortality in sepsis patients with thrombocytopenia was significantly higher than those without thrombocytopenia at the baseline (48.2% vs. 38.5%, respectively). […] The model proposed by Ling and others can prove to be more useful if more variables were included in the model that would provide a higher predictive ability for the 28-day mortality. […] The results of this paper suggest that ML algorithms can prove to be beneficial in prediction of SAT and severe thrombocytopenia and hence would assist in the early management of the patients at risk.
#10 Retrospective analysis of interpretable machine learning in predicting ICU thrombocytopenia in geriatric ICU patients | Scientific Reports
https://www.nature.com/articles/s41598-024-67785-1
We developed an interpretable machine learning algorithm that prospectively predicts the risk of thrombocytopenia in older critically ill patients during their stay in the intensive care unit (ICU), ultimately aiding clinical decision-making and improving patient care. […] Thrombocytopenia occurred in 18.2% of postoperative geriatric patients, with a higher mortality rate. […] Thus, our machine learning-based models have considerable potential in effectively predicting the risk and severity of postoperative thrombocytopenia in geriatric ICU patients for better clinical decision-making and patient care. […] The results showed that the C5.0 model had an excellent ability to predict thrombocytopenia in postoperative geriatric ICU patients, and its AUC, accuracy, precision, specificity, and recall were satisfactory, thus providing reliable reference for clinicians.
#11 Retrospective analysis of interpretable machine learning in predicting ICU thrombocytopenia in geriatric ICU patients | Scientific Reports
https://www.nature.com/articles/s41598-024-67785-1
Our research showed that the extent of platelet drop was directly correlated with the disease severity. In other words, the more severe the reduction in platelets, the higher the patient mortality rate, the longer the hospital stay, and the higher the hospitalization costs. […] This aligns with previous research findings, suggesting that the severity of platelet decrease can assist in the early identification of patient prognosis. […] The predictive model for assessing the severity of thrombocytopenia demonstrated an accuracy of 0.84 with a CI of 0.81-0.86 for internal validation. For external validation using the MIMIC database, the model showed an accuracy of 0.80 with a CI of 0.79-0.81. […] Despite the limitations, the strengths of this study include the use of various machine learning algorithms and detailed model evaluation and validation to ensure the robustness and reliability of the models. […] This study furnishes valid, reliable, and easily interpretable predictive models for assessing the risk and the severity of thrombocytopenia in postoperative geriatric ICU patients.
#12 Retrospective analysis of interpretable machine learning in predicting ICU thrombocytopenia in geriatric ICU patients | Scientific Reports
https://www.nature.com/articles/s41598-024-67785-1
Our research showed that the extent of platelet drop was directly correlated with the disease severity. In other words, the more severe the reduction in platelets, the higher the patient mortality rate, the longer the hospital stay, and the higher the hospitalization costs. […] This aligns with previous research findings, suggesting that the severity of platelet decrease can assist in the early identification of patient prognosis. […] The predictive model for assessing the severity of thrombocytopenia demonstrated an accuracy of 0.84 with a CI of 0.81-0.86 for internal validation. For external validation using the MIMIC database, the model showed an accuracy of 0.80 with a CI of 0.79-0.81. […] Despite the limitations, the strengths of this study include the use of various machine learning algorithms and detailed model evaluation and validation to ensure the robustness and reliability of the models. […] This study furnishes valid, reliable, and easily interpretable predictive models for assessing the risk and the severity of thrombocytopenia in postoperative geriatric ICU patients.
#13 Admission platelet count and indices as predictors of outcome in children with severe Sepsis: a prospective hospital-based study | BMC Pediatrics | Full Text
https://bmcpediatr.biomedcentral.com/articles/10.1186/s12887-020-02278-4
Platelet count and PCT were significantly lower (p0.001) and MPV was significantly higher in non-survivor than survivors (p=0.004). […] Thrombocytopenia, platelet indices and their ratios, especially plateletcrit and MPV/PCT, are readily available, sensitive, prognostic markers, that can identify the severe sepsis patients with poorest outcome. […] The low platelet count in non-survivors may be attributed to the depletion of coagulation factors and platelet consumption during the septic process and the low PCT in the non-survivors may be imputed to that PCT is influenced by number and size of platelets and has a positive relationship with platelet count. […] Plateletcrit was the most sensitive parameter for predicting death, with thrombocytopenia taking second place by having the same sensitivity as the PRISM score. While MPV in this study was the least sensitive platelet parameter, but it was still more sensitive than CRP, which is the most commonly used inflammatory marker to be assessed in children with sepsis. Also, MPV/PCT ratio was the most sensitive ratio to predict mortality in this study.
#14 Admission platelet count and indices as predictors of outcome in children with severe Sepsis: a prospective hospital-based study | BMC Pediatrics | Full Text
https://bmcpediatr.biomedcentral.com/articles/10.1186/s12887-020-02278-4
Platelet count and PCT were significantly lower (p0.001) and MPV was significantly higher in non-survivor than survivors (p=0.004). […] Thrombocytopenia, platelet indices and their ratios, especially plateletcrit and MPV/PCT, are readily available, sensitive, prognostic markers, that can identify the severe sepsis patients with poorest outcome. […] The low platelet count in non-survivors may be attributed to the depletion of coagulation factors and platelet consumption during the septic process and the low PCT in the non-survivors may be imputed to that PCT is influenced by number and size of platelets and has a positive relationship with platelet count. […] Plateletcrit was the most sensitive parameter for predicting death, with thrombocytopenia taking second place by having the same sensitivity as the PRISM score. While MPV in this study was the least sensitive platelet parameter, but it was still more sensitive than CRP, which is the most commonly used inflammatory marker to be assessed in children with sepsis. Also, MPV/PCT ratio was the most sensitive ratio to predict mortality in this study.
#15 Development and validation of a prediction model for in-hospital mortality of patients with severe thrombocytopenia | Scientific Reports
https://www.nature.com/articles/s41598-022-10438-y
Risk stratification and prognosis evaluation of severe thrombocytopenia are essential for clinical treatment and management. […] Currently, there is currently no reliable predictive model to identify patients at high risk of severe thrombocytopenia. […] The nomogram model we established showed superior predictive performance and can assist in the quantitative assessment of the prognostic risk in patients with severe thrombocytopenia. […] Identifying risk factors for mortality in patients with severe thrombocytopenia can help clinicians develop individualized treatment strategies and enhanced care management to reduce the risk of adverse events and improve patient outcomes. […] We revealed the following as independent risk factors for in-hospital mortality in patients with severe thrombocytopenia: age, presence of cerebrovascular disease, malignant cancer, oxygen saturation, heart rate, respiration rate, mean arterial pressure, mechanical ventilation, vasopressor use, CRRT; PT, PTT, and BUN.
#16 Development and validation of a prediction model for in-hospital mortality of patients with severe thrombocytopenia | Scientific Reports
https://www.nature.com/articles/s41598-022-10438-y
Risk stratification and prognosis evaluation of severe thrombocytopenia are essential for clinical treatment and management. […] Currently, there is currently no reliable predictive model to identify patients at high risk of severe thrombocytopenia. […] The nomogram model we established showed superior predictive performance and can assist in the quantitative assessment of the prognostic risk in patients with severe thrombocytopenia. […] Identifying risk factors for mortality in patients with severe thrombocytopenia can help clinicians develop individualized treatment strategies and enhanced care management to reduce the risk of adverse events and improve patient outcomes. […] We revealed the following as independent risk factors for in-hospital mortality in patients with severe thrombocytopenia: age, presence of cerebrovascular disease, malignant cancer, oxygen saturation, heart rate, respiration rate, mean arterial pressure, mechanical ventilation, vasopressor use, CRRT; PT, PTT, and BUN.
#17 Development and validation of a prediction model for in-hospital mortality of patients with severe thrombocytopenia | Scientific Reports
https://www.nature.com/articles/s41598-022-10438-y
Risk stratification and prognosis evaluation of severe thrombocytopenia are essential for clinical treatment and management. […] Currently, there is currently no reliable predictive model to identify patients at high risk of severe thrombocytopenia. […] The nomogram model we established showed superior predictive performance and can assist in the quantitative assessment of the prognostic risk in patients with severe thrombocytopenia. […] Identifying risk factors for mortality in patients with severe thrombocytopenia can help clinicians develop individualized treatment strategies and enhanced care management to reduce the risk of adverse events and improve patient outcomes. […] We revealed the following as independent risk factors for in-hospital mortality in patients with severe thrombocytopenia: age, presence of cerebrovascular disease, malignant cancer, oxygen saturation, heart rate, respiration rate, mean arterial pressure, mechanical ventilation, vasopressor use, CRRT; PT, PTT, and BUN.
#18 Development and validation of a prediction model for in-hospital mortality of patients with severe thrombocytopenia | Scientific Reports
https://www.nature.com/articles/s41598-022-10438-y
Risk stratification and prognosis evaluation of severe thrombocytopenia are essential for clinical treatment and management. […] Currently, there is currently no reliable predictive model to identify patients at high risk of severe thrombocytopenia. […] The nomogram model we established showed superior predictive performance and can assist in the quantitative assessment of the prognostic risk in patients with severe thrombocytopenia. […] Identifying risk factors for mortality in patients with severe thrombocytopenia can help clinicians develop individualized treatment strategies and enhanced care management to reduce the risk of adverse events and improve patient outcomes. […] We revealed the following as independent risk factors for in-hospital mortality in patients with severe thrombocytopenia: age, presence of cerebrovascular disease, malignant cancer, oxygen saturation, heart rate, respiration rate, mean arterial pressure, mechanical ventilation, vasopressor use, CRRT; PT, PTT, and BUN.
#19 Development and validation of a prediction model for in-hospital mortality of patients with severe thrombocytopenia | Scientific Reports
https://www.nature.com/articles/s41598-022-10438-y
Risk stratification and prognosis evaluation of severe thrombocytopenia are essential for clinical treatment and management. […] Currently, there is currently no reliable predictive model to identify patients at high risk of severe thrombocytopenia. […] The nomogram model we established showed superior predictive performance and can assist in the quantitative assessment of the prognostic risk in patients with severe thrombocytopenia. […] Identifying risk factors for mortality in patients with severe thrombocytopenia can help clinicians develop individualized treatment strategies and enhanced care management to reduce the risk of adverse events and improve patient outcomes. […] We revealed the following as independent risk factors for in-hospital mortality in patients with severe thrombocytopenia: age, presence of cerebrovascular disease, malignant cancer, oxygen saturation, heart rate, respiration rate, mean arterial pressure, mechanical ventilation, vasopressor use, CRRT; PT, PTT, and BUN.
#20 Development and validation of a prediction model for in-hospital mortality of patients with severe thrombocytopenia | Scientific Reports
https://www.nature.com/articles/s41598-022-10438-y
Risk stratification and prognosis evaluation of severe thrombocytopenia are essential for clinical treatment and management. […] Currently, there is currently no reliable predictive model to identify patients at high risk of severe thrombocytopenia. […] The nomogram model we established showed superior predictive performance and can assist in the quantitative assessment of the prognostic risk in patients with severe thrombocytopenia. […] Identifying risk factors for mortality in patients with severe thrombocytopenia can help clinicians develop individualized treatment strategies and enhanced care management to reduce the risk of adverse events and improve patient outcomes. […] We revealed the following as independent risk factors for in-hospital mortality in patients with severe thrombocytopenia: age, presence of cerebrovascular disease, malignant cancer, oxygen saturation, heart rate, respiration rate, mean arterial pressure, mechanical ventilation, vasopressor use, CRRT; PT, PTT, and BUN.
#21 Development and validation of a prediction model for in-hospital mortality of patients with severe thrombocytopenia | Scientific Reports
https://www.nature.com/articles/s41598-022-10438-y
Risk stratification and prognosis evaluation of severe thrombocytopenia are essential for clinical treatment and management. […] Currently, there is currently no reliable predictive model to identify patients at high risk of severe thrombocytopenia. […] The nomogram model we established showed superior predictive performance and can assist in the quantitative assessment of the prognostic risk in patients with severe thrombocytopenia. […] Identifying risk factors for mortality in patients with severe thrombocytopenia can help clinicians develop individualized treatment strategies and enhanced care management to reduce the risk of adverse events and improve patient outcomes. […] We revealed the following as independent risk factors for in-hospital mortality in patients with severe thrombocytopenia: age, presence of cerebrovascular disease, malignant cancer, oxygen saturation, heart rate, respiration rate, mean arterial pressure, mechanical ventilation, vasopressor use, CRRT; PT, PTT, and BUN.
#22 Development and validation of a prediction model for in-hospital mortality of patients with severe thrombocytopenia | Scientific Reports
https://www.nature.com/articles/s41598-022-10438-y
Risk stratification and prognosis evaluation of severe thrombocytopenia are essential for clinical treatment and management. […] Currently, there is currently no reliable predictive model to identify patients at high risk of severe thrombocytopenia. […] The nomogram model we established showed superior predictive performance and can assist in the quantitative assessment of the prognostic risk in patients with severe thrombocytopenia. […] Identifying risk factors for mortality in patients with severe thrombocytopenia can help clinicians develop individualized treatment strategies and enhanced care management to reduce the risk of adverse events and improve patient outcomes. […] We revealed the following as independent risk factors for in-hospital mortality in patients with severe thrombocytopenia: age, presence of cerebrovascular disease, malignant cancer, oxygen saturation, heart rate, respiration rate, mean arterial pressure, mechanical ventilation, vasopressor use, CRRT; PT, PTT, and BUN.
#23 Development and validation of a prediction model for in-hospital mortality of patients with severe thrombocytopenia | Scientific Reports
https://www.nature.com/articles/s41598-022-10438-y
Risk stratification and prognosis evaluation of severe thrombocytopenia are essential for clinical treatment and management. […] Currently, there is currently no reliable predictive model to identify patients at high risk of severe thrombocytopenia. […] The nomogram model we established showed superior predictive performance and can assist in the quantitative assessment of the prognostic risk in patients with severe thrombocytopenia. […] Identifying risk factors for mortality in patients with severe thrombocytopenia can help clinicians develop individualized treatment strategies and enhanced care management to reduce the risk of adverse events and improve patient outcomes. […] We revealed the following as independent risk factors for in-hospital mortality in patients with severe thrombocytopenia: age, presence of cerebrovascular disease, malignant cancer, oxygen saturation, heart rate, respiration rate, mean arterial pressure, mechanical ventilation, vasopressor use, CRRT; PT, PTT, and BUN.
#24 Development and validation of a prediction model for in-hospital mortality of patients with severe thrombocytopenia | Scientific Reports
https://www.nature.com/articles/s41598-022-10438-y
Risk stratification and prognosis evaluation of severe thrombocytopenia are essential for clinical treatment and management. […] Currently, there is currently no reliable predictive model to identify patients at high risk of severe thrombocytopenia. […] The nomogram model we established showed superior predictive performance and can assist in the quantitative assessment of the prognostic risk in patients with severe thrombocytopenia. […] Identifying risk factors for mortality in patients with severe thrombocytopenia can help clinicians develop individualized treatment strategies and enhanced care management to reduce the risk of adverse events and improve patient outcomes. […] We revealed the following as independent risk factors for in-hospital mortality in patients with severe thrombocytopenia: age, presence of cerebrovascular disease, malignant cancer, oxygen saturation, heart rate, respiration rate, mean arterial pressure, mechanical ventilation, vasopressor use, CRRT; PT, PTT, and BUN.
#25 Development and validation of a prediction model for in-hospital mortality of patients with severe thrombocytopenia | Scientific Reports
https://www.nature.com/articles/s41598-022-10438-y
Risk stratification and prognosis evaluation of severe thrombocytopenia are essential for clinical treatment and management. […] Currently, there is currently no reliable predictive model to identify patients at high risk of severe thrombocytopenia. […] The nomogram model we established showed superior predictive performance and can assist in the quantitative assessment of the prognostic risk in patients with severe thrombocytopenia. […] Identifying risk factors for mortality in patients with severe thrombocytopenia can help clinicians develop individualized treatment strategies and enhanced care management to reduce the risk of adverse events and improve patient outcomes. […] We revealed the following as independent risk factors for in-hospital mortality in patients with severe thrombocytopenia: age, presence of cerebrovascular disease, malignant cancer, oxygen saturation, heart rate, respiration rate, mean arterial pressure, mechanical ventilation, vasopressor use, CRRT; PT, PTT, and BUN.
#26 Development and validation of a prediction model for in-hospital mortality of patients with severe thrombocytopenia | Scientific Reports
https://www.nature.com/articles/s41598-022-10438-y
Risk stratification and prognosis evaluation of severe thrombocytopenia are essential for clinical treatment and management. […] Currently, there is currently no reliable predictive model to identify patients at high risk of severe thrombocytopenia. […] The nomogram model we established showed superior predictive performance and can assist in the quantitative assessment of the prognostic risk in patients with severe thrombocytopenia. […] Identifying risk factors for mortality in patients with severe thrombocytopenia can help clinicians develop individualized treatment strategies and enhanced care management to reduce the risk of adverse events and improve patient outcomes. […] We revealed the following as independent risk factors for in-hospital mortality in patients with severe thrombocytopenia: age, presence of cerebrovascular disease, malignant cancer, oxygen saturation, heart rate, respiration rate, mean arterial pressure, mechanical ventilation, vasopressor use, CRRT; PT, PTT, and BUN.
#27 Development and validation of a prediction model for in-hospital mortality of patients with severe thrombocytopenia | Scientific Reports
https://www.nature.com/articles/s41598-022-10438-y
Risk stratification and prognosis evaluation of severe thrombocytopenia are essential for clinical treatment and management. […] Currently, there is currently no reliable predictive model to identify patients at high risk of severe thrombocytopenia. […] The nomogram model we established showed superior predictive performance and can assist in the quantitative assessment of the prognostic risk in patients with severe thrombocytopenia. […] Identifying risk factors for mortality in patients with severe thrombocytopenia can help clinicians develop individualized treatment strategies and enhanced care management to reduce the risk of adverse events and improve patient outcomes. […] We revealed the following as independent risk factors for in-hospital mortality in patients with severe thrombocytopenia: age, presence of cerebrovascular disease, malignant cancer, oxygen saturation, heart rate, respiration rate, mean arterial pressure, mechanical ventilation, vasopressor use, CRRT; PT, PTT, and BUN.
#28 Development and validation of a prediction model for in-hospital mortality of patients with severe thrombocytopenia | Scientific Reports
https://www.nature.com/articles/s41598-022-10438-y
Risk stratification and prognosis evaluation of severe thrombocytopenia are essential for clinical treatment and management. […] Currently, there is currently no reliable predictive model to identify patients at high risk of severe thrombocytopenia. […] The nomogram model we established showed superior predictive performance and can assist in the quantitative assessment of the prognostic risk in patients with severe thrombocytopenia. […] Identifying risk factors for mortality in patients with severe thrombocytopenia can help clinicians develop individualized treatment strategies and enhanced care management to reduce the risk of adverse events and improve patient outcomes. […] We revealed the following as independent risk factors for in-hospital mortality in patients with severe thrombocytopenia: age, presence of cerebrovascular disease, malignant cancer, oxygen saturation, heart rate, respiration rate, mean arterial pressure, mechanical ventilation, vasopressor use, CRRT; PT, PTT, and BUN.
#29 Development and validation of a prediction model for in-hospital mortality of patients with severe thrombocytopenia | Scientific Reports
https://www.nature.com/articles/s41598-022-10438-y
Our study found that BUN is independently related to the prognosis of patients with severe thrombocytopenia, possibly because patients with renal failure also have complex hemostatic disorders. […] A nomogram is a common visual presentation for predictive models. Compared with the traditional critically ill patient scoring systems (SOFA and SAPS II), risk stratification based on our nomogram model showed a higher clinical benefit. […] The nomogram model we developed can help quantitatively assess the prognostic risk factors in patients with severe thrombocytopenia. This model has superior predictive performance and can provide a good reference for evaluating the in-hospital mortality rate in patients with severe thrombocytopenia. However, multi-center prospective studies are required for further verification.
#30
https://haematologica.org/article/view/7816
This prospective observational cohort study aimed to explore the clinical features of incident immune thrombocytopenia in adults and predictors of outcome, while determining if a family history of autoimmune disorder is a risk factor for immune thrombocytopenia. […] After 12 months, only 37% of patients not subject to disease-modifying interventions achieved cure. […] Immune thrombocytopenia in adults has been shown to progress to a chronic form in the majority of patients. A lower platelet count could be indicative of a more favorable outcome. […] The presence of a higher platelet count or the absence of severe bleeding at baseline was found to be associated with increased odds of chronicity at 12 months. […] A familial history of autoimmune disorder was found in 9 (8.3%) out of the 109 case-patients explored.
#31
https://haematologica.org/article/view/7816
This prospective observational cohort study aimed to explore the clinical features of incident immune thrombocytopenia in adults and predictors of outcome, while determining if a family history of autoimmune disorder is a risk factor for immune thrombocytopenia. […] After 12 months, only 37% of patients not subject to disease-modifying interventions achieved cure. […] Immune thrombocytopenia in adults has been shown to progress to a chronic form in the majority of patients. A lower platelet count could be indicative of a more favorable outcome. […] The presence of a higher platelet count or the absence of severe bleeding at baseline was found to be associated with increased odds of chronicity at 12 months. […] A familial history of autoimmune disorder was found in 9 (8.3%) out of the 109 case-patients explored.
#32
https://haematologica.org/article/view/7816
This prospective observational cohort study aimed to explore the clinical features of incident immune thrombocytopenia in adults and predictors of outcome, while determining if a family history of autoimmune disorder is a risk factor for immune thrombocytopenia. […] After 12 months, only 37% of patients not subject to disease-modifying interventions achieved cure. […] Immune thrombocytopenia in adults has been shown to progress to a chronic form in the majority of patients. A lower platelet count could be indicative of a more favorable outcome. […] The presence of a higher platelet count or the absence of severe bleeding at baseline was found to be associated with increased odds of chronicity at 12 months. […] A familial history of autoimmune disorder was found in 9 (8.3%) out of the 109 case-patients explored.
#33
https://haematologica.org/article/view/7816
Overall, having a first-degree relative with a history of autoimmune disorder was not associated with increased odds of developing ITP. […] The study confirms that ITP is more common in women than men, and is frequently associated with cutaneous bleeding, even if life-threatening bleeding is rare. […] The predictors of chronicity were explored prospectively for the first time in adults, suggesting that higher platelet counts at diagnosis are negatively correlated with the risk of chronicity. However, further research is needed to confirm this finding and potentially to identify other predictive factors for chronicity at the time of diagnosis.
#34
https://haematologica.org/article/view/7816
This prospective observational cohort study aimed to explore the clinical features of incident immune thrombocytopenia in adults and predictors of outcome, while determining if a family history of autoimmune disorder is a risk factor for immune thrombocytopenia. […] After 12 months, only 37% of patients not subject to disease-modifying interventions achieved cure. […] Immune thrombocytopenia in adults has been shown to progress to a chronic form in the majority of patients. A lower platelet count could be indicative of a more favorable outcome. […] The presence of a higher platelet count or the absence of severe bleeding at baseline was found to be associated with increased odds of chronicity at 12 months. […] A familial history of autoimmune disorder was found in 9 (8.3%) out of the 109 case-patients explored.
#35
https://haematologica.org/article/view/7816
Overall, having a first-degree relative with a history of autoimmune disorder was not associated with increased odds of developing ITP. […] The study confirms that ITP is more common in women than men, and is frequently associated with cutaneous bleeding, even if life-threatening bleeding is rare. […] The predictors of chronicity were explored prospectively for the first time in adults, suggesting that higher platelet counts at diagnosis are negatively correlated with the risk of chronicity. However, further research is needed to confirm this finding and potentially to identify other predictive factors for chronicity at the time of diagnosis.
#36 Development and validation of a prediction model for in-hospital mortality of patients with severe thrombocytopenia | Scientific Reports
https://www.nature.com/articles/s41598-022-10438-y
Our study found that BUN is independently related to the prognosis of patients with severe thrombocytopenia, possibly because patients with renal failure also have complex hemostatic disorders. […] A nomogram is a common visual presentation for predictive models. Compared with the traditional critically ill patient scoring systems (SOFA and SAPS II), risk stratification based on our nomogram model showed a higher clinical benefit. […] The nomogram model we developed can help quantitatively assess the prognostic risk factors in patients with severe thrombocytopenia. This model has superior predictive performance and can provide a good reference for evaluating the in-hospital mortality rate in patients with severe thrombocytopenia. However, multi-center prospective studies are required for further verification.
#37 Development and validation of a prediction model for in-hospital mortality of patients with severe thrombocytopenia | Scientific Reports
https://www.nature.com/articles/s41598-022-10438-y
Our study found that BUN is independently related to the prognosis of patients with severe thrombocytopenia, possibly because patients with renal failure also have complex hemostatic disorders. […] A nomogram is a common visual presentation for predictive models. Compared with the traditional critically ill patient scoring systems (SOFA and SAPS II), risk stratification based on our nomogram model showed a higher clinical benefit. […] The nomogram model we developed can help quantitatively assess the prognostic risk factors in patients with severe thrombocytopenia. This model has superior predictive performance and can provide a good reference for evaluating the in-hospital mortality rate in patients with severe thrombocytopenia. However, multi-center prospective studies are required for further verification.
#38 Applications of Artificial Intelligence in Thrombocytopenia
https://www.mdpi.com/2075-4418/13/6/1060
Thrombocytopenia is a medical condition where blood platelet count drops very low. […] Clinically, the presence of thrombocytopenia might be very dangerous and is associated with poor outcomes of patients due to excessive bleeding if not addressed quickly enough. […] Hence, early detection and evaluation of thrombocytopenia is essential for rapid and appropriate intervention for these patients. […] Since artificial intelligence is able to combine and evaluate many linear and nonlinear variables simultaneously, it has shown great potential in its application in the early diagnosis, assessing the prognosis and predicting the distribution of patients with thrombocytopenia. […] The included studies showed that artificial intelligence can potentially enhance the clinical approaches used in the diagnosis, prognosis, and treatment of thrombocytopenia.
#39 Applications of Artificial Intelligence in Thrombocytopenia
https://www.mdpi.com/2075-4418/13/6/1060
The aim of this review was to investigate the different applications of ML in thrombocytopenia that are currently available. […] We have shown that AI can also be utilized in the detection of drugs that can cause DITP, and mapping of potential hotspots for SFTS transmission. […] The available evidence in this review paper suggests that AI may be effective in predicting and evaluating the prognosis of patients with thrombocytopenia.
#40 Applications of Artificial Intelligence in Thrombocytopenia
https://pmc.ncbi.nlm.nih.gov/articles/PMC10047875/
The study is the only study included in our paper that has been verified clinically in a prospective cohort. […] The RF model had an AUC of 0.901 when externally validated, which demonstrates good predictive ability for the risk of critical bleeds. […] The available evidence in this review paper suggests that AI may be effective in predicting and evaluating the prognosis of patients with thrombocytopenia.
#41 Retrospective analysis of interpretable machine learning in predicting ICU thrombocytopenia in geriatric ICU patients | Scientific Reports
https://www.nature.com/articles/s41598-024-67785-1
We developed an interpretable machine learning algorithm that prospectively predicts the risk of thrombocytopenia in older critically ill patients during their stay in the intensive care unit (ICU), ultimately aiding clinical decision-making and improving patient care. […] Thrombocytopenia occurred in 18.2% of postoperative geriatric patients, with a higher mortality rate. […] Thus, our machine learning-based models have considerable potential in effectively predicting the risk and severity of postoperative thrombocytopenia in geriatric ICU patients for better clinical decision-making and patient care. […] The results showed that the C5.0 model had an excellent ability to predict thrombocytopenia in postoperative geriatric ICU patients, and its AUC, accuracy, precision, specificity, and recall were satisfactory, thus providing reliable reference for clinicians.
#42 Retrospective analysis of interpretable machine learning in predicting ICU thrombocytopenia in geriatric ICU patients | Scientific Reports
https://www.nature.com/articles/s41598-024-67785-1
We developed an interpretable machine learning algorithm that prospectively predicts the risk of thrombocytopenia in older critically ill patients during their stay in the intensive care unit (ICU), ultimately aiding clinical decision-making and improving patient care. […] Thrombocytopenia occurred in 18.2% of postoperative geriatric patients, with a higher mortality rate. […] Thus, our machine learning-based models have considerable potential in effectively predicting the risk and severity of postoperative thrombocytopenia in geriatric ICU patients for better clinical decision-making and patient care. […] The results showed that the C5.0 model had an excellent ability to predict thrombocytopenia in postoperative geriatric ICU patients, and its AUC, accuracy, precision, specificity, and recall were satisfactory, thus providing reliable reference for clinicians.
#43 Retrospective analysis of interpretable machine learning in predicting ICU thrombocytopenia in geriatric ICU patients | Scientific Reports
https://www.nature.com/articles/s41598-024-67785-1
Our research showed that the extent of platelet drop was directly correlated with the disease severity. In other words, the more severe the reduction in platelets, the higher the patient mortality rate, the longer the hospital stay, and the higher the hospitalization costs. […] This aligns with previous research findings, suggesting that the severity of platelet decrease can assist in the early identification of patient prognosis. […] The predictive model for assessing the severity of thrombocytopenia demonstrated an accuracy of 0.84 with a CI of 0.81-0.86 for internal validation. For external validation using the MIMIC database, the model showed an accuracy of 0.80 with a CI of 0.79-0.81. […] Despite the limitations, the strengths of this study include the use of various machine learning algorithms and detailed model evaluation and validation to ensure the robustness and reliability of the models. […] This study furnishes valid, reliable, and easily interpretable predictive models for assessing the risk and the severity of thrombocytopenia in postoperative geriatric ICU patients.
#44 Applications of Artificial Intelligence in Thrombocytopenia
https://pmc.ncbi.nlm.nih.gov/articles/PMC10047875/
The study is the only study included in our paper that has been verified clinically in a prospective cohort. […] The RF model had an AUC of 0.901 when externally validated, which demonstrates good predictive ability for the risk of critical bleeds. […] The available evidence in this review paper suggests that AI may be effective in predicting and evaluating the prognosis of patients with thrombocytopenia.
#45 Applications of Artificial Intelligence in Thrombocytopenia
https://pmc.ncbi.nlm.nih.gov/articles/PMC10047875/
It is important to properly assess the prognosis of patients with SAT to ensure proper care for these patients. […] The results of this paper suggest that ML algorithms can prove to be beneficial in prediction of SAT and severe thrombocytopenia and hence would assist in the early management of the patients at risk. […] The model proposed by Ling and others can prove to be more useful if more variables were included in the model that would provide a higher predictive ability for the 28-day mortality. […] The early detection of LAT allowed for physicians to maintain patient safety by switching antibiotics when the platelet count drops to a critical level. […] The models presented by Cheng and colleagues show very good performance matrices for the prediction of HAT. […] The models were also generated utilizing single-center retrospective data which limits its generalizability and suffers from missing data and insufficient variables.
#46 Applications of Artificial Intelligence in Thrombocytopenia
https://www.mdpi.com/2075-4418/13/6/1060
As stated previously, thrombocytopenia is associated with poor prognosis in sepsis patients. […] Hence, it is important to properly assess the prognosis of patients with SAT to ensure proper care for these patients. […] Several studies have also shown that RDW has significant clinical utility as an independent predictor of poor prognosis in critically ill patients with sepsis and SAT. […] The results of the paper showed that 28-day mortality in sepsis patients with thrombocytopenia was significantly higher than those without thrombocytopenia at the baseline (48.2% vs. 38.5%, respectively). […] The model proposed by Ling and others can prove to be more useful if more variables were included in the model that would provide a higher predictive ability for the 28-day mortality. […] The results of this paper suggest that ML algorithms can prove to be beneficial in prediction of SAT and severe thrombocytopenia and hence would assist in the early management of the patients at risk.
#47 Development and validation of a prediction model for in-hospital mortality of patients with severe thrombocytopenia | Scientific Reports
https://www.nature.com/articles/s41598-022-10438-y
Our study found that BUN is independently related to the prognosis of patients with severe thrombocytopenia, possibly because patients with renal failure also have complex hemostatic disorders. […] A nomogram is a common visual presentation for predictive models. Compared with the traditional critically ill patient scoring systems (SOFA and SAPS II), risk stratification based on our nomogram model showed a higher clinical benefit. […] The nomogram model we developed can help quantitatively assess the prognostic risk factors in patients with severe thrombocytopenia. This model has superior predictive performance and can provide a good reference for evaluating the in-hospital mortality rate in patients with severe thrombocytopenia. However, multi-center prospective studies are required for further verification.
#48 Applications of Artificial Intelligence in Thrombocytopenia
https://pmc.ncbi.nlm.nih.gov/articles/PMC10047875/
It is important to properly assess the prognosis of patients with SAT to ensure proper care for these patients. […] The results of this paper suggest that ML algorithms can prove to be beneficial in prediction of SAT and severe thrombocytopenia and hence would assist in the early management of the patients at risk. […] The model proposed by Ling and others can prove to be more useful if more variables were included in the model that would provide a higher predictive ability for the 28-day mortality. […] The early detection of LAT allowed for physicians to maintain patient safety by switching antibiotics when the platelet count drops to a critical level. […] The models presented by Cheng and colleagues show very good performance matrices for the prediction of HAT. […] The models were also generated utilizing single-center retrospective data which limits its generalizability and suffers from missing data and insufficient variables.
#49 Applications of Artificial Intelligence in Thrombocytopenia
https://www.mdpi.com/2075-4418/13/6/1060
As stated previously, thrombocytopenia is associated with poor prognosis in sepsis patients. […] Hence, it is important to properly assess the prognosis of patients with SAT to ensure proper care for these patients. […] Several studies have also shown that RDW has significant clinical utility as an independent predictor of poor prognosis in critically ill patients with sepsis and SAT. […] The results of the paper showed that 28-day mortality in sepsis patients with thrombocytopenia was significantly higher than those without thrombocytopenia at the baseline (48.2% vs. 38.5%, respectively). […] The model proposed by Ling and others can prove to be more useful if more variables were included in the model that would provide a higher predictive ability for the 28-day mortality. […] The results of this paper suggest that ML algorithms can prove to be beneficial in prediction of SAT and severe thrombocytopenia and hence would assist in the early management of the patients at risk.
#50
https://haematologica.org/article/view/7816
Overall, having a first-degree relative with a history of autoimmune disorder was not associated with increased odds of developing ITP. […] The study confirms that ITP is more common in women than men, and is frequently associated with cutaneous bleeding, even if life-threatening bleeding is rare. […] The predictors of chronicity were explored prospectively for the first time in adults, suggesting that higher platelet counts at diagnosis are negatively correlated with the risk of chronicity. However, further research is needed to confirm this finding and potentially to identify other predictive factors for chronicity at the time of diagnosis.
#51 Retrospective analysis of interpretable machine learning in predicting ICU thrombocytopenia in geriatric ICU patients | Scientific Reports
https://www.nature.com/articles/s41598-024-67785-1
Our research showed that the extent of platelet drop was directly correlated with the disease severity. In other words, the more severe the reduction in platelets, the higher the patient mortality rate, the longer the hospital stay, and the higher the hospitalization costs. […] This aligns with previous research findings, suggesting that the severity of platelet decrease can assist in the early identification of patient prognosis. […] The predictive model for assessing the severity of thrombocytopenia demonstrated an accuracy of 0.84 with a CI of 0.81-0.86 for internal validation. For external validation using the MIMIC database, the model showed an accuracy of 0.80 with a CI of 0.79-0.81. […] Despite the limitations, the strengths of this study include the use of various machine learning algorithms and detailed model evaluation and validation to ensure the robustness and reliability of the models. […] This study furnishes valid, reliable, and easily interpretable predictive models for assessing the risk and the severity of thrombocytopenia in postoperative geriatric ICU patients.