Joint analysis of duration of ventilation, length of intensive care, and mortality of COVID-19 patients: a multistate approach

Abstract Background The clinical progress of patients hospitalized due to COVID-19 is often associated with severe pneumonia which may require intensive care, invasive ventilation, or extracorporeal membrane oxygenation (ECMO). The length of intensive care and the duration of these supportive therapies are clinically relevant outcomes. From the statistical perspective, these quantities are challenging to estimate due to episodes being time-dependent and potentially multiple, as well as being determined by the competing, terminal events of discharge alive and death. Methods We used multistate models to study COVID-19 patients’ time-dependent progress and provide a statistical framework to estimate hazard rates and transition probabilities. These estimates can then be used to quantify average sojourn times of clinically important states such as intensive care and invasive ventilation. We have made two real data sets of COVID-19 patients (n = 24* and n = 53**) and the corresponding statistical code publically available. Results The expected lengths of intensive care unit (ICU) stay at day 28 for the two cohorts were 15.05* and 19.62** days, while expected durations of mechanical ventilation were 7.97* and 9.85** days. Predicted mortality stood at 51%* and 15%**. Patients mechanically ventilated at the start of the example studies had a longer expected duration of ventilation (12.25*, 14.57** days) compared to patients non-ventilated (4.34*, 1.41** days) after 28 days. Furthermore, initially ventilated patients had a higher risk of death (54%* and 20%** vs. 48%* and 6%**) after 4 weeks. These results are further illustrated in stacked probability plots for the two groups from time zero, as well as for the entire cohort which depicts the predicted proportions of the patients in each state over follow-up. Conclusions The multistate approach gives important insights into the progress of COVID-19 patients in terms of ventilation duration, length of ICU stay, and mortality. In addition to avoiding frequent pitfalls in survival analysis, the methodology enables active cases to be analyzed by allowing for censoring. The stacked probability plots provide extensive information in a concise manner that can be easily conveyed to decision makers regarding healthcare capacities. Furthermore, clear comparisons can be made among different baseline characteristics.

Download Full-text

Joint analysis of duration of ventilation, length of intensive care, and mortality of COVID-19 patients: a multistate approach

10.21203/rs.3.rs-32236/v1 ◽

2020 ◽

Cited By ~ 1

Author(s):

Derek Hazard ◽

Klaus Kaier ◽

Maja von Cube ◽

Marlon Grodd ◽

Lars Bugiera ◽

...

Keyword(s):

Intensive Care ◽

Transition Probabilities ◽

Time Dependent ◽

Joint Analysis ◽

Invasive Ventilation ◽

Entire Cohort ◽

Risk Of Death ◽

Icu Stay ◽

Probability Plots ◽

Duration Of Ventilation

Abstract Background The clinical progress of patients hospitalized due to COVID-19 is often associated with severe pneumonia which may require intensive care, invasive ventilation, or extracorporeal membrane oxygenation (ECMO). The length of intensive care and the duration of these supportive therapies are clinically relevant outcomes. From the statistical perspective, these quantities are challenging to estimate due to episodes being time-dependent and potentially multiple, as well as being determined by the competing, terminal events of discharge alive and death. Methods We used multistate models to study COVID-19 patients’ time-dependent progress and provide a statistical framework to estimate hazard rates and transition probabilities. These estimates can then be used to quantify average sojourn times of clinically important states such as intensive care and invasive ventilation. We have made two real data sets of COVID-19 patients (n = 24* and n = 53**) and the corresponding statistical code publically available. Results The expected lengths of intensive care unit (ICU) stay at day 28 for the two cohorts were 15.05* and 19.62** days, while expected durations of mechanical ventilation were 7.97* and 9.85** days. Predicted mortality stood at 51%* and 15%**. Patients mechanically ventilated at the start of the example studies had a longer expected duration of ventilation (12.25*, 14.57** days) compared to patients non-ventilated (4.34*, 1.41** days) after 28 days. Furthermore, initially ventilated patients had a higher risk of death (54%* and 20%** vs. 48%* and 6 %**) after 4 weeks. These results are further illustrated in stacked probability plots for the two groups from time zero, as well as for the entire cohort which depicts the predicted proportions of the patients in each state over follow-up. Conclusions The multistate approach gives important insights into the progress of COVID-19 patients in terms of ventilation duration, length of ICU stay, and mortality. In addition to avoiding frequent pitfalls in survival analysis, the methodology enables active cases to be analyzed by allowing for censoring. The stacked probability plots provide extensive information in a concise manner that can be easily conveyed to decision makers regarding healthcare capacities. Furthermore, clear comparisons can be among different baseline characteristics.

Download Full-text

Time-dependent analysis of extra length of stay and mortality due to ventilator-associated pneumonia in intensive-care units of ten limited-resources countries: findings of the International Nosocomial Infection Control Consortium (INICC)

Epidemiology and Infection ◽

10.1017/s0950268811000094 ◽

2011 ◽

Vol 139 (11) ◽

pp. 1757-1763 ◽

Cited By ~ 15

Author(s):

V. D. ROSENTHAL ◽

F. E. UDWADIA ◽

H. J. MUÑOZ ◽

N. ERBEN ◽

F. HIGUERA ◽

...

Keyword(s):

Intensive Care ◽

Length Of Stay ◽

Meta Analysis ◽

Time Dependent ◽

Limited Resources ◽

Ventilator Associated Pneumonia ◽

Risk Of Death ◽

Patient Morbidity ◽

Increased Risk ◽

Time Dependent Analysis

SUMMARYVentilator-associated pneumonias (VAPs) are a worldwide problem that significantly increases patient morbidity, mortality, and length of stay (LoS), and their effects should be estimated to account for the timing of infection. The purpose of the study was to estimate extra LoS and mortality in an intensive-care unit (ICU) due to a VAP in a cohort of 69 248 admissions followed for 283 069 days in ICUs from 10 countries. Data were arranged according to the multi-state format. Extra LoS and increased risk of death were estimated independently in each country, and their results were combined using a random-effects meta-analysis. VAP prolonged LoS by an average of 2·03 days (95% CI 1·52–2·54 days), and increased the risk of death by 14% (95% CI 2–27). The increased risk of death due to VAP was explained by confounding with patient morbidity.

Download Full-text

GASTROINTESTINAL BLEEDING;

The Professional Medical Journal ◽

10.29309/tpmj/2019.26.03.3261 ◽

2019 ◽

Vol 26 (03) ◽

Author(s):

Mujtaba Jaffary ◽

Nida ◽

Saeed Ahmad Khan

Keyword(s):

Intensive Care Unit ◽

Mechanical Ventilation ◽

Intensive Care ◽

Gastrointestinal Bleeding ◽

Mortality And Morbidity ◽

Mechanically Ventilated ◽

Icu Stay ◽

Trust Hospital ◽

Length Of Icu Stay ◽

Duration Of Ventilation

Background: Gastrointestinal bleeding (GIB) among patients with critical illness is one of the leading sources of mortality and morbidity. The prevalence of GIB differs from 15-50 percent during first 24 hours stay in intensive care unit. Mechanical ventilation is a most leading risk factor of GIB among patients admitted in ICU (intensive care unit). Objectives: The objective of the study is to know the prevalence and risk factors associated with gastrointestinal bleeding among mechanically ventilated patients. Study Deign: Retrospectively study. Setting: Ch Rehmat Ali Memorial Trust Hospital, Lahore. Period: 1st October 2017 to 31st March 2018. Materials and Method: A group of 120 patients in intensive care unit who received mechanical ventilation for a period of 48 hours or above were included. Results: Among 56 patients with gastrointestinal bleeding, mean age was 49.2±12.1, mean length of ICU stay was 29.2±16.6 and mean duration of ventilation was 30.2±20.5. Among 64 patients with no gastrointestinal bleeding, mean age was 51.9±15.0, mean length of ICU stay was 12.7±6.8 and mean duration of ventilation was 13.5±7.9. There were 12 (21.1%) mortalities among patients with gastrointestinal bleeding and 10 (15.6%) mortalities among patients with no gastrointestinal bleeding. Conclusion: Study concluded that length of ICU stay, duration of ventilation, renal failure, liver failure and mortalities were more among patients with gastrointestinal bleeding.

Download Full-text

Time-dependent Error in the APACHE II Scoring System

Anaesthesia and Intensive Care ◽

10.1177/0310057x9202000112 ◽

1992 ◽

Vol 20 (1) ◽

pp. 63-65 ◽

Cited By ~ 6

Author(s):

J. W. Sleigh ◽

R. J. Brook ◽

M. Miller

Keyword(s):

Intensive Care Unit ◽

Intensive Care ◽

Scoring System ◽

Predictive Power ◽

Predictive Accuracy ◽

Time Dependent ◽

Apache Ii ◽

Risk Of Death ◽

Predictive Error ◽

Dependent Error

Using the APACHE II scoring system, the risk of death was calculated for 189 patients in the Wanganui Intensive Care Unit and 194 patients in the Harare Intensive Care Unit. Using tables of actual and predicted outcome, the predictive power of the system was compared in patients grouped according to the length of time that they spent in the ICU. The predictive error increased from 15% in those patients staying less than six days, to 38% in those staying six days or more (P < 0.01). The predictive accuracy of the APACHE II system appeared to decrease with the length of time the patient stayed in the Intensive Care Unit.

Download Full-text

Early Prediction of Seven-Day Mortality in Intensive Care Unit Using a Machine Learning Model: Results from the SPIN-UTI Project

Journal of Clinical Medicine ◽

10.3390/jcm10050992 ◽

2021 ◽

Vol 10 (5) ◽

pp. 992

Author(s):

Martina Barchitta ◽

Andrea Maugeri ◽

Giuliana Favara ◽

Paolo Marco Riela ◽

Giovanni Gallo ◽

...

Keyword(s):

Machine Learning ◽

Intensive Care ◽

Intensive Care Units ◽

Learning Algorithm ◽

Area Under The Curve ◽

Support Vector ◽

Icu Admission ◽

Risk Of Death ◽

Saps Ii ◽

Svm Algorithm

Patients in intensive care units (ICUs) were at higher risk of worsen prognosis and mortality. Here, we aimed to evaluate the ability of the Simplified Acute Physiology Score (SAPS II) to predict the risk of 7-day mortality, and to test a machine learning algorithm which combines the SAPS II with additional patients’ characteristics at ICU admission. We used data from the “Italian Nosocomial Infections Surveillance in Intensive Care Units” network. Support Vector Machines (SVM) algorithm was used to classify 3782 patients according to sex, patient’s origin, type of ICU admission, non-surgical treatment for acute coronary disease, surgical intervention, SAPS II, presence of invasive devices, trauma, impaired immunity, antibiotic therapy and onset of HAI. The accuracy of SAPS II for predicting patients who died from those who did not was 69.3%, with an Area Under the Curve (AUC) of 0.678. Using the SVM algorithm, instead, we achieved an accuracy of 83.5% and AUC of 0.896. Notably, SAPS II was the variable that weighted more on the model and its removal resulted in an AUC of 0.653 and an accuracy of 68.4%. Overall, these findings suggest the present SVM model as a useful tool to early predict patients at higher risk of death at ICU admission.

Download Full-text

Cardiac biomarkers in acute respiratory distress syndrome: a systematic review and meta-analysis

Journal of Intensive Care ◽

10.1186/s40560-021-00548-6 ◽

2021 ◽

Vol 9 (1) ◽

Author(s):

Dilip Jayasimhan ◽

Simon Foster ◽

Catherina L. Chang ◽

Robert J. Hancox

Keyword(s):

Systematic Review ◽

Acute Respiratory Distress Syndrome ◽

Intensive Care ◽

Cardiac Dysfunction ◽

Distress Syndrome ◽

Troponin I ◽

Meta Analysis ◽

Cardiac Injury ◽

Cardiac Biomarkers ◽

Risk Of Death

Abstract Background Acute respiratory distress syndrome (ARDS) is a leading cause of morbidity and mortality in the intensive care unit. Biochemical markers of cardiac dysfunction are associated with high mortality in many respiratory conditions. The aim of this systematic review is to examine the link between elevated biomarkers of cardiac dysfunction in ARDS and mortality. Methods A systematic review of MEDLINE, EMBASE, Web of Science and CENTRAL databases was performed. We included studies of adult intensive care patients with ARDS that reported the risk of death in relation to a measured biomarker of cardiac dysfunction. The primary outcome of interest was mortality up to 60 days. A random-effects model was used for pooled estimates. Funnel-plot inspection was done to evaluate publication bias; Cochrane chi-square tests and I2 tests were used to assess heterogeneity. Results Twenty-two studies were included in the systematic review and 18 in the meta-analysis. Biomarkers of cardiac stretch included NT-ProBNP (nine studies) and BNP (six studies). Biomarkers of cardiac injury included Troponin-T (two studies), Troponin-I (one study) and High-Sensitivity-Troponin-I (three studies). Three studies assessed multiple cardiac biomarkers. High levels of NT-proBNP and BNP were associated with a higher risk of death up to 60 days (unadjusted OR 8.98; CI 4.15-19.43; p<0.00001). This association persisted after adjustment for age and illness severity. Biomarkers of cardiac injury were also associated with higher mortality, but this association was not statistically significant (unadjusted OR 2.21; CI 0.94-5.16; p= 0.07). Conclusion Biomarkers of cardiac stretch are associated with increased mortality in ARDS.

Download Full-text

Post-intensive care syndrome after a critical COVID-19: cohort study from a Belgian follow-up clinic

Annals of Intensive Care ◽

10.1186/s13613-021-00910-9 ◽

2021 ◽

Vol 11 (1) ◽

Author(s):

Anne-Françoise Rousseau ◽

Pauline Minguet ◽

Camille Colson ◽

Isabelle Kellens ◽

Sourour Chaabane ◽

...

Keyword(s):

Cohort Study ◽

Intensive Care ◽

Inpatient Rehabilitation ◽

Quadriceps Strength ◽

Biological Parameters ◽

Icu Stay ◽

Icu Discharge ◽

Prolonged Icu Stay ◽

Post Intensive Care Syndrome

Abstract Purpose Many patients with coronavirus disease 2019 (COVID-19) required critical care. Mid-term outcomes of the survivors need to be assessed. The objective of this single-center cohort study was to describe their physical, cognitive, psychological, and biological outcomes at 3 months following intensive care unit (ICU)-discharge (M3). Patients and methods All COVID-19 adults who survived an ICU stay ≥ 7 days and attended the M3 consultation at our multidisciplinary follow-up clinic were involved. They benefited from a standardized assessment, addressing health-related quality of life (EQ-5D-3L), sleep disorders (PSQI), and the three principal components of post-intensive care syndrome (PICS): physical status (Barthel index, handgrip and quadriceps strength), mental health disorders (HADS and IES-R), and cognitive impairment (MoCA). Biological parameters referred to C-reactive protein and creatinine. Results Among the 92 patients admitted to our ICU for COVID-19, 42 survived a prolonged ICU stay and 32 (80%) attended the M3 follow-up visit. Their median age was 62 [49–68] years, 72% were male, and nearly half received inpatient rehabilitation following ICU discharge. At M3, 87.5% (28/32) had not regained their baseline level of daily activities. Only 6.2% (2/32) fully recovered, and had normal scores for the three MoCA, IES-R and Barthel scores. The main observed disorders were PSQI > 5 (75%, 24/32), MoCA < 26 (44%, 14/32), Barthel < 100 (31%, 10/32) and IES-R ≥ 33 (28%, 9/32). Combined disorders were observed in 13/32 (40.6%) of the patients. The EQ-5D-3L visual scale was rated at 71 [61–80]. A quarter of patients (8/32) demonstrated a persistent inflammation based on CRP blood level (9.3 [6.8–17.7] mg/L). Conclusion The burden of severe COVID-19 and prolonged ICU stay was considerable in the present cohort after 3 months, affecting both functional status and biological parameters. These data are an argument on the need for closed follow-up for critically ill COVID-19 survivors.

Download Full-text

Clinical characteristics and outcomes of people living with HIV hospitalized with COVID-19: a nationwide experience

International Journal of STD & AIDS ◽

10.1177/0956462420973106 ◽

2021 ◽

Vol 32 (5) ◽

pp. 435-443

Author(s):

Maria Elena Ceballos ◽

Patricio Ross ◽

Martin Lasso ◽

Isabel Dominguez ◽

Marcela Puente ◽

...

Keyword(s):

Mechanical Ventilation ◽

Intensive Care ◽

General Population ◽

Clinical Characteristics ◽

Cd4 Count ◽

People Living With Hiv ◽

Hiv Diagnosis ◽

Hiv Viral Load ◽

Risk Of Death ◽

Living With Hiv

In this prospective, multicentric, observational study, we describe the clinical characteristics and outcomes of people living with HIV (PLHIV) requiring hospitalization due to COVID-19 in Chile and compare them with Chilean general population admitted with SARS-CoV-2. Consecutive PLHIV admitted with COVID-19 in 23 hospitals, between 16 April and 23 June 2020, were included. Data of a temporally matched-hospitalized general population were used to compare demography, comorbidities, COVID-19 symptoms, and major outcomes. In total, 36 PLHIV subjects were enrolled; 92% were male and mean age was 44 years. Most patients (83%) were on antiretroviral therapy; mean CD4 count was 557 cells/mm3. Suppressed HIV viremia was found in 68% and 56% had, at least, one comorbidity. Severe COVID-19 occurred in 44.4%, intensive care was required in 22.2%, and five patients died (13.9%). No differences were seen between recovered and deceased patients in CD4 count, HIV viral load, or time since HIV diagnosis. Hypertension and cardiovascular disease were associated with a higher risk of death ( p = 0.02 and 0.006, respectively). Compared with general population, the HIV cohort had significantly more men (OR 0.15; IC 95% 0.07–0.31) and younger age (OR 8.68; IC 95% 2.66–28.31). In PLHIV, we found more intensive care unit admission (OR 2.31; IC 95% 1.05–5.07) but no differences in the need for mechanical ventilation or death. In this cohort of PLHIV hospitalized with COVID-19, hypertension and cardiovascular comorbidities, but not current HIV viro-immunologic status, were the most important risk factors for mortality. No differences were found between PLHIV and general population in the need for mechanical ventilation and death.

Download Full-text