Update on the Epimed Monitor Adult ICU Database: 15 years of its use in national registries, quality improvement initiatives and clinical research.

In recent decades, several databases of critically ill patients have become available in both low-, middle-, and high-income countries from all continents. These databases are also rich sources of data for the surveillance of emerging diseases, intensive care unit performance evaluation and benchmarking, quality improvement projects and clinical research. The Epimed Monitor database is turning 15 years old in 2024 and has become one of the largest of these databases. In recent years, there has been rapid geographical expansion, an increase in the number of participating intensive care units and hospitals, and the addition of several new variables and scores, allowing a more complete characterization of patients to facilitate multicenter clinical studies. As of December 2023, the database was being used regularly for 23,852 beds in 1,723 intensive care units and 763 hospitals from ten countries, totaling more than 5.6 million admissions. In addition, critical care societies have adopted the system and its database to establish national registries and international collaborations. In the present review, we provide an updated description of the database; report experiences of its use in critical care for quality improvement initiatives, national registries and clinical research; and explore other potential future perspectives and developments.

Post COVID-19 mucormycosis in critical care settings: A prospective cohort study in a tertiary care center in Egypt.

BACKGROUND: The emergence of mucormycosis as a life-threatening fungal infection after the coronavirus disease of 2019 (COVID-19) is a major concern and challenge, but there is limited information on the risk factors for mortality in patients. METHODS: We conducted a prospective cohort study from May 2021 to April 2022 to determine the in-hospital outcomes of post-COVID-19 mucormycosis during the intensive care unit (ICU) stay. The sample of the study was collected as consecutive sampling using all accessible patients in the study period. The Statistical Package for Social Sciences (SPSS), version 25 (IBM, Chicago, Illinois, USA) was used for statistical analysis. RESULTS: Among 150 patients with post-COVID-19 mucormycosis, the majority had a primary sinus infection (86.0 %), while 11.3 % had both sinus and ocular infections, and 2.7 % had sinus and cutaneous infections. Around 21 % (n = 31) of patients deceased after staying in the ICU for a median (range) of 45.0 (10.0-145.0) days. The majority of the patients who deceased had pneumonia patches on computed tomography (CT) (90.3 %) while none of the patients who were discharged had pneumonia patches (p < 0.001). The deceased group had higher rates of pulmonary embolism (93.5 %) compared to the surviving groups (21.8 %). In a multivariate Cox regression analysis, the risk of death was higher in older patients above 60 years old (hazard ratio (95 %CI): 6.7 (1.73-15.81)), increase among patient with history of steroid administration (hazard ratio (95 %CI): 5.70 (1.23-10.91)), who had facial cutaneous infection with mucormycosis (hazard ratio (95 %CI): 8.76 (1.78-25.18)), patients with uncontrolled diabetes (hazard ratio (95 %CI): 10.76 (1.78, 65.18)), and total leukocytic count (TLC>10 ×103 mcL) (hazard ratio (95 %CI): 10.03 (3.29-30.61)). CONCLUSIONS: Identifying high-risk patients especially old diabetic patients with corticosteroid administration and detecting their deterioration quickly is crucial in reducing post-COVID-19 mucormycosis mortality rates, and these factors must be considered when developing treatment and quarantine strategies.

Epidemiological analysis of a 10-year retrospective study of pediatric trauma in intensive care.

Pediatric trauma plays a crucial role in pediatric mortality, with traffic injuries and falls frequently cited as leading causes of significant injuries among children. A comprehensive investigation, including geographical factors, is essential for developing effective strategies to prevent injuries and alleviate the burden of pediatric trauma. This study involved a retrospective analysis of clinical data from pediatric patients admitted to our hospital's intensive care unit (ICU) due to trauma over a 10-year period. Comprehensive analyses were conducted to elucidate trends, demographics, injury patterns, and risk factors associated with these admissions. This retrospective study included 951 pediatric patients (mean age: 4.79 ± 3.24 years; mean weight: 18.45 ± 9.02 kg; median time to ICU admission post-injury: 10.86 ± 14.95 h). Among these patients, 422 (44.4%) underwent emergency surgery, and 466 (49%) required mechanical ventilation support, with a mean duration of 70.19 ± 146.62 h. The mean duration of ICU stay was 6.24 ± 8.01 days, and the overall mean hospitalization duration was 16.08 ± 15.56 days. The predominant cause of unintentional injury was traffic accidents (47.9%), followed by falls (42.5%) and burns/scalds (5.3%). Most incidents involved children aged 0-6 years (70.7%), with males comprising 60.0% of patients. Injury incidents predominantly occurred between 12 and 6 PM (44.5%) and on non-workdays (37.6%). The most common locations where injuries occurred were roadsides (49%) and rural areas (64.35%). Single-site injuries (58.78%) were more prevalent than multiple-site injuries (41.22%), and head injuries were the most common among single-site injuries (81.57%). At ICU admission, the mean injury severity score was 18.49 ± 8.86. Following active intervention, 871 patients (91.59%) showed improvement, while 80 (8.41%) succumbed to their injuries. Traffic injuries remain the primary cause of pediatric trauma leading to ICU admission, underscoring the importance of using appropriate child restraint systems and protective gear as fundamental preventive measures. The increased incidence of injuries among children aged < 6 years and those residing in rural areas highlights the need for targeted preventive strategies, necessitating tailored interventions and public policy formulations that address these high-risk populations.

The 2022-2023 Pediatric Respiratory Illness Surge: Survey of Acute and Critical Care Resource Use.

OBJECTIVE: A surge of pediatric respiratory illnesses beset the United States in late 2022 and early 2023. This study evaluated within-surge hospital acute and critical care resource availability and utilization. The study aimed to determine pediatric hospital acute and critical care resource use during a respiratory illness surge. METHODS: Between January and February 2023, an online survey was sent to the sections of hospital medicine and critical care of the American Academy of Pediatrics, community discussion forums of the Children's Hospital Association, and PedSCCM-a pediatric critical care website. Data were summarized with median values and interquartile range. RESULTS: Across 35 hospitals with pediatric intensive care units (PICU), increase in critical care resource use was significant. In the month preceding the survey, 26 (74%) hospitals diverted patients away from their emergency department (ED) to other hospitals, with 46% diverting 1-5 patients, 23% diverting 6-10 patients, and 31% diverting more than 10 patients. One in 5 hospitals reported moving patients on mechanical ventilation from the PICU to other settings, including the ED (n = 2), intermediate care unit (n = 2), cardiac ICU (n = 1), ward converted to an ICU (n = 1), and a ward (n = 1). Utilization of human critical care resources was high, with PICU faculty, nurses, and respiratory therapists working at 100% capacity. CONCLUSIONS: The respiratory illness surge triggered significant hospital resource use and diversion of patients away from hospitals. Pediatric public health emergency-preparedness should innovate around resource capacity.

[The potential role of the Swedish Intensive Care Registry data for decision-making in intensive care]. / Intensivvårdsregistrets roll vid riskbedömning och prioritering.

The Swedish Intensive Care Registry collects and analyses data regarding intensive care in Sweden. Based on this data an overview can be created regarding available hospital beds and occupancy rates in Swedish intensive care, as well as demographic and clinical characteristics of the patients, and their survival. Through this, identification of patient groups with poorer prognosis is possible, facilitating reflection of the appropriateness of intensive care and invasive procedures that may cause discomfort to the patient.

Limitations of life-sustaining treatments in intensive care units in Croatia: a multicenter retrospective study.

AIM: In order to gain insight into the current prevailing practices regarding the limitation of life-sustaining treatment in intensive care units (ICUs) in Croatia, we assessed the frequency of limitation and provision of certain treatment modalities, as well as the associated patient and ICU-related factors. METHODS: A multicenter retrospective cross-sectional study was conducted in 17 ICUs in Croatia. We reviewed the medical records of patients deceased in 2017 and extracted data on demographic, clinical, and health care variables. A logistic regression analysis was conducted to determine the associations between these variables and treatment modalities. RESULTS: The study enrolled 1095 patients (55% male; mean age 69.9±13.7). Analgesia and sedation were discontinued before the patient's death in 23% and 34% of the cases, respectively. Patients older than 71 years were less often mechanically ventilated (P<0.001), and less frequently received inotropes and vasoactive therapy (P=0.002) than younger patients. Patients hospitalized in the ICU for less than 7 days less frequently had discontinuation of mechanical ventilation and inotropes and vasoactive therapy than patients hospitalized for 8 days and longer (P<0.001). Logistic regression analysis showed that ICU type was a crucial determinant, with multidisciplinary and surgical ICUs being associated with higher odds of intubation, mechanical ventilation, vasoactive and inotropic therapy, analgesia, and sedation. CONCLUSION: Older patients and those diagnosed with stroke and intracranial hemorrhage received fewer therapeutic modalities. All the observed treatment modalities were more frequently discontinued in patients who were hospitalized in the ICU for a prolonged time.

A lay of the land: a description of academic acute care surgery models in Canada.

BACKGROUND: Patients who require emergency general surgery (EGS) are at a substantially higher risk for perioperative morbidity and mortality than patients undergoing elective general surgery. The acute care surgery (ACS) model has been shown to improve EGS patient outcomes and cost-effectiveness. A recent systematic review has shown extensive heterogeneity in the structure of ACS models worldwide. The objective of this study was to describe the current landscape of ACS models in academic centres across Canada. METHODS: We sent an online questionnaire to the 18 academic centres in Canada. The lead ACS physicians from each institution completed the questionnaire, describing the structure of their ACS models. RESULTS: In total, 16 institutions responded, all of which reported having ACS models, with a total of 29 ACS services described. All services had resident coverage. Of the 29, 18 (62%) had dedicated allied health care staff. The staff surgeon was free from elective duties while covering ACS in 17/29 (59%) services. More than half (15/29; 52%) of the services described protected ACS operating room time, but only 7/15 (47%) had a dedicated ACS room all 5 weekdays. Four of 29 services (14%) had no protected ACS operating room time. Only 1/16 (6%) institutions reported a mandate to conduct ACS research, while 12/16 (75%) found ACS research difficult, owing to lack of resources. CONCLUSION: We saw large variations in the structure of ACS models in academic centres in Canada. The components of ACS models that are most important to patient outcomes remain poorly defined. Future research will focus on defining the necessary cornerstones of ACS models.

Returning to work and health status at 12 months among patients with COVID-19 cared for in intensive care-A prospective, longitudinal study.

OBJECTIVE: Intensive care unit (ICU) stay for a serious illness has a long-term impact on patients' physical and psychological well-being, affecting their ability to return to their everyday life. We aimed to investigate whether there are differences in health status between those who return to work and those who do not, and how demographic characteristics and illness severity impact patients' ability to return to work 12 months after intensive care for COVID-19. RESEARCH METHODOLOGY: This was a prospective longitudinal cohort study. The participants were patients who had been in intensive care for COVID-19 and had worked before contracting COVID-19. Data on return to previous occupational status, demographic data, comorbidities, intensive care characteristics, and health status were collected at a 12-month follow-up visit. SETTING: General ICU at the Uppsala University Hospital in Sweden. RESULTS: Seventy-three participants were included in the study. Twelve months after discharge from the ICU, 77 % (n = 56) had returned to work. The participants who were unable to return to work reported more severe health symptoms. The (odds ratio [OR] for not returning to work was high for critical illness OR, 12.05; 95 % confidence interval [CI], 2.07-70.29, p = 0.006) and length of ICU stay (OR, 1.06; 95 % CI, 1.01-1.11, p = 0.01) CONCLUSION: Two-thirds of the participants were able to return to work within 1 year after discharge from the ICU. The primary factors contributing to the failure to work were duration of the acute disease and presence of severe and persistent long-term symptoms. IMPLICATIONS FOR CLINICAL PRACTICE: Patients' health status must be comprehensively assessed and their ability to return to work should be addressed in the rehabilitation process. Therefore, any complications faced by the patients must be identified and treated early to increase the possibility of their successful return to work.

Linking Dutch Intensive Care Registry Data to Socioeconomic Status Scores.

This study aimed to gain insight into the success rate of linking the NICE registry with SES data from CBS and to examine whether the characteristics of linked and non-linked patients differ. Although clinically relevant differences were found, in total 93,4% of the admissions were successfully linked.

Impact of En Route Critical Care Provider Experience on Lung Protective Ventilation Compliance During Air Transport of Combat Wounded.

INTRODUCTION: The primary objective of this study was to evaluate the association between the U.S. Air Force Critical Care Air Transport (CCAT) provider operational experience with compliance for lung protective ventilation (LPV) volumes recommended by Acute Respiratory Distress Syndrome Clinical Network guidelines. MATERIALS AND METHODS: We performed a retrospective cohort study of CCAT providers transporting combat casualties requiring mechanical ventilation from the Middle East to Germany from 2007 to 2012. We reviewed CCAT medical records from 2007 to 2012 for the total number of patient transports by CCAT physicians and respiratory care practitioners (RCPs). Center for Sustainment of Trauma and Readiness Skills Cincinnati process improvement questionnaire data described provider demographics and clinical backgrounds. We linked these data to patient demographics and in-flight ventilation management from a prior CCAT cohort study. Patient inclusion criteria included transport by CCAT from the Middle East to Germany for traumatic injury requiring mechanical ventilation between 2007 and 2012. We excluded patients with no documented height or tidal volume. LPV compliance was defined as tidal volumes ≤8 mL/kg of predicted body weight during en route critical care transport. We performed a logistic regression analysis. This study was reviewed and approved by the 59th Medical Wing institutional review board (IRB). RESULTS: We analyzed 491 patient transports conducted by 71 (RCPs and 84 physicians. Patients had a median age of 25 years (IQR 22-30), 98% were male, median injury severity score was 24 (IQR 17-34), and median preflight PaO2/FiO2 was 285 (IQR 220-365). Median experience was 26 missions (IQR 13-40) for RCPs and 23 missions (IQR 12-38) for physicians. All in-flight tidal volumes were LPV compliant in 58.3% of missions. Unadjusted analysis showed higher LPV compliance for RCPs with in-garrison critical care experience. Multivariate models did not find an association between missions flown and LPV compliance but did demonstrate a positive association with physician specialty of medical intensivist (OR 3.0, 95% CI 1.6-5.7) and a negative association with flights in 2008 (OR 0.4, 95% CI 0.2-0.7) for LPV compliance. CONCLUSION: No association was found between number of missions flown by CCAT providers and lung protective tidal volume compliance. Linkage of multiple data sources enabled investigation of clinical and operational currency associations with a care quality metric compliance in the combat en route care environment. Future studies should evaluate the impact of ongoing CCAT training and quality improvement interventions on LPV compliance.

Incidence of post-extubation dysphagia among critical care patients undergoing orotracheal intubation: a systematic review and meta-analysis.

BACKGROUND: Post-extubation dysphagia (PED) emerges as a frequent complication following endotracheal intubation within the intensive care unit (ICU). PED has been strongly linked to adverse outcomes, including aspiration, pneumonia, malnutrition, heightened mortality rates, and prolonged hospitalization, resulting in escalated healthcare expenditures. Nevertheless, the reported incidence of PED varies substantially across the existing body of literature. Therefore, the principal objective of this review was to provide a comprehensive estimate of PED incidence in ICU patients undergoing orotracheal intubation. METHODS: We searched Embase, PubMed, Web of Science, Cochrane Library, China National Knowledge Infrastructure (CNKI), Wanfang Database, China Science, Technology Journal Database (VIP), and SinoMed databases from inception to August 2023. Two reviewers independently screened studies and extracted data. Subsequently, a random-effects model was employed for meta-statistical analysis utilizing the "meta prop" command within Stata SE version 15.0 to ascertain the incidence of PED. In addition, we performed subgroup analyses and meta-regression to elucidate potential sources of heterogeneity among the included studies. RESULTS: Of 4144 studies, 30 studies were included in this review. The overall pooled incidence of PED was 36% (95% confidence interval [CI] 29-44%). Subgroup analyses unveiled that the pooled incidence of PED, stratified by assessment time (≤ 3 h, 4-6 h, ≤ 24 h, and ≤ 48 h), was as follows: 31.0% (95% CI 8.0-59.0%), 28% (95% CI 22.0-35.0%), 41% (95% CI 33.0-49.0%), and 49.0% (95% CI 34.0-63.0%), respectively. When sample size was 100 < N ≤ 300, the PED incidence was more close to the overall PED incidence. Meta-regression analysis highlighted that sample size, assessment time and mean intubation time constituted the source of heterogeneity among the included studies. CONCLUSION: The incidence of PED was high among ICU patients who underwent orotracheal intubation. ICU professionals should raise awareness about PED. In the meantime, it is important to develop guidelines or consensus on the most appropriate PED assessment time and assessment tools to accurately assess the incidence of PED.

The intensive care population profile in Denmark before and during the first wave of the SARS-CoV-2 pandemic; a national register-based study.

OBJECTIVE: To describe the change in admission rate and demographic profile of patients admitted to ICUs throughout Denmark before and during first wave of the SARS-CoV-2 pandemic. METHODS: A register-based national observational study of all patients admitted to ICU from December 2019 until April 2020, comparing ICU admission before and after lockdown. RESULTS: The number of admissions declined, especially in the age groups below 18 and above 70. The sex distribution and the comorbidity-level remained unchanged. The length of hospital stay prior to ICU admission increased. Overall fewer patients were admitted electively. CONCLUSION: Fewer patients were admitted to ICU and waited longer for admission during the first wave of the COVID-19 pandemic.

Potentially inappropriate medication as a predictor of poor prognosis of COVID-19 in older adults: a South Korean nationwide cohort study.

OBJECTIVES: To investigate the association between exposure to potentially inappropriate medication (PIM) and poor prognosis of COVID-19 in older adults, controlling for comorbidity and sociodemographic factors. DESIGN AND SETTING: Nationwide retrospective cohort study based on the national registry of COVID-19 patients, established through the linkage of South Korea's national insurance claims database with the Korea Disease Control and Prevention Agency registry of patients with COVID-19, up to 31 July 2020. PARTICIPANTS: A total of 2217 COVID-19 patients over 60 years of age who tested positive between 20 January 2022 and 4 June 2020. Exposure to PIM was defined based on any prescription record of PIM during the 30 days prior to the date of testing positive for COVID-19. PRIMARY OUTCOME MEASURES: Mortality and utilisation of critical care from the date of testing positive until the end of isolation. RESULTS: Among the 2217 COVID-19 patients over 60 years of age, 604 were exposed to PIM prior to infection. In the matched cohort of 583 pairs, PIM-exposed individuals exhibited higher rates of mortality (19.7% vs 9.8%, p<0.0001) and critical care utilisation (13.4% vs 8.9%, p=0.0156) compared with non-exposed individuals. The temporal association of PIM exposure with mortality was significant across all age groups (RR=1.68, 95% CI: 1.23~2.24), and a similar trend was observed for critical care utilisation (RR: 1.75, 95% CI: 1.26~2.39). The risk of mortality and critical care utilisation increased with exposure to a higher number of PIMs in terms of active pharmaceutical ingredients and drug categories. CONCLUSION: Exposure to PIM exacerbates the poor outcomes of older patients with COVID-19 who are already at high risk. Effective interventions are urgently needed to address PIM exposure and improve health outcomes in this vulnerable population.

A Bayesian non-stationary heteroskedastic time series model for multivariate critical care data.

We propose a multivariate GARCH model for non-stationary health time series by modifying the observation-level variance of the standard state space model. The proposed model provides an intuitive and novel way of dealing with heteroskedastic data using the conditional nature of state-space models. We follow the Bayesian paradigm to perform the inference procedure. In particular, we use Markov chain Monte Carlo methods to obtain samples from the resultant posterior distribution. We use the forward filtering backward sampling algorithm to efficiently obtain samples from the posterior distribution of the latent state. The proposed model also handles missing data in a fully Bayesian fashion. We validate our model on synthetic data and analyze a data set obtained from an intensive care unit in a Montreal hospital and the MIMIC dataset. We further show that our proposed models offer better performance, in terms of WAIC than standard state space models. The proposed model provides a new way to model multivariate heteroskedastic non-stationary time series data. Model comparison can then be easily performed using the WAIC.

Measuring the impact of maternal critical care admission on short- and longer-term maternal and birth outcomes.

PURPOSE: Factors increasing the risk of maternal critical illness are rising in prevalence in maternity populations. Studies of general critical care populations highlight that severe illness is associated with longer-term physical and psychological morbidity. We aimed to compare short- and longer-term outcomes between women who required critical care admission during pregnancy/puerperium and those who did not. METHODS: This is a cohort study including all women delivering in Scottish hospitals between 01/01/2005 and 31/12/2018, using national healthcare databases. The primary exposure was intensive care unit (ICU) admission, while secondary exposures included high dependency unit admission. Outcomes included hospital readmission (1-year post-hospital discharge, 1-year mortality, psychiatric hospital admission, stillbirth, and neonatal critical care admission). Multivariable Cox and logistic regression were used to report hazard ratios (HR) and odds ratios (OR) of association between ICU admission and outcomes. RESULTS: Of 762,918 deliveries, 1449 (0.18%) women were admitted to ICU, most commonly due to post-partum hemorrhage (225, 15.5%) followed by eclampsia/pre-eclampsia (133, 9.2%). Over-half (53.8%) required mechanical ventilation. One-year hospital readmission was more frequent in women admitted to ICU compared with non-ICU populations [24.5% (n = 299) vs 8.9% (n = 68,029)]. This association persisted after confounder adjustment (HR 1.93, 95% confidence interval [CI] 1.33, 2.81, p < 0.001). Furthermore, maternal ICU admission was associated with increased 1-year mortality (HR 40.06, 95% CI 24.04, 66.76, p < 0.001), stillbirth (OR 12.31, 95% CI 7.95,19.08, p < 0.001) and neonatal critical care admission (OR 6.99, 95% CI 5.64,8.67, p < 0.001) after confounder adjustment. CONCLUSION: Critical care admission increases the risk of adverse short-term and long-term maternal, pregnancy and neonatal outcomes. Optimizing long-term post-partum care may benefit maternal critical illness survivors.

Application of the Sepsis-3 criteria to describe sepsis epidemiology in the Amsterdam UMCdb intensive care dataset.

INTRODUCTION: Sepsis is a major cause of morbidity and mortality worldwide. In the updated, 2016 Sepsis-3 criteria, sepsis is defined as life-threatening organ dysfunction caused by a dysregulated host response to infection, where organ dysfunction can be represented by an increase in the Sequential Organ Failure Assessment (SOFA) score of 2 points or more. We sought to apply the Sepsis-3 criteria to characterise the septic cohort in the Amsterdam University Medical Centres database (Amsterdam UMCdb). METHODS: We examined adult intensive care unit (ICU) admissions in the Amsterdam UMCdb, which contains de-identified data for patients admitted to a mixed surgical-medical ICU at a tertiary academic medical centre in the Netherlands. We operationalised the Sepsis-3 criteria, defining organ dysfunction as an increase in the SOFA score of 2 points or more, while infection was defined as a new course of antibiotics or an escalation in antibiotic therapy, with at least one antibiotic given intravenously. Patients with sepsis were determined to be in septic shock if they additionally required the use of vasopressors and had a lactate level >2 mmol/L. RESULTS: We identified 18,221 ICU admissions from 16,408 patients in our cohort. There were 6,312 unique sepsis episodes, of which 30.2% met the criteria for septic shock. A total of 4,911/6,312 sepsis (77.8%) episodes occurred on ICU admission. Forty-seven percent of emergency medical admissions and 36.7% of emergency surgical admissions were for sepsis. Overall, there was a 12.5% ICU mortality rate; patients with septic shock had a higher ICU mortality rate (38.4%) than those without shock (11.4%). CONCLUSIONS: We successfully operationalised the Sepsis-3 criteria to the Amsterdam UMCdb, allowing the characterization and comparison of sepsis epidemiology across different centres.

Pneumocystis pneumonia in intensive care: clinical spectrum, prophylaxis patterns, antibiotic treatment delay impact, and role of corticosteroids. A French multicentre prospective cohort study.

PURPOSE: Severe Pneumocystis jirovecii pneumonia (PJP) requiring intensive care has been the subject of few prospective studies. It is unclear whether delayed curative antibiotic therapy may impact survival in these severe forms of PJP. The impact of corticosteroid therapy combined with antibiotics is also unclear. METHODS: This multicentre, prospective observational study involving 49 adult intensive care units (ICUs) in France was designed to evaluate the severity, the clinical spectrum, and outcomes of patients with severe PJP, and to assess the association between delayed curative antibiotic treatment and adjunctive corticosteroid therapy with mortality. RESULTS: We included 158 patients with PJP from September 2020 to August 2022. Their main reason for admission was acute respiratory failure (n = 150, 94.9%). 12% of them received antibiotic prophylaxis for PJP before ICU admission. The ICU, hospital, and 6-month mortality were 31.6%, 35.4%, and 40.5%, respectively. Using time-to-event analysis with a propensity score-based inverse probability of treatment weighting, the initiation of curative antibiotic treatment after 96 h of ICU admission was associated with faster occurrence of death [time ratio: 6.75; 95% confidence interval (95% CI): 1.48-30.82; P = 0.014]. The use of corticosteroids for PJP was associated with faster occurrence of death (time ratio: 2.48; 95% CI 1.01-6.08; P = 0.048). CONCLUSION: This study showed that few patients with PJP admitted to intensive care received prophylactic antibiotic therapy, that delay in curative antibiotic treatment was common and that both delay in curative antibiotic treatment and adjunctive corticosteroids for PJP were associated with accelerated mortality.