The effect of proactive versus reactive treatment of hypotension on postoperative disability and outcome in surgical patients under anaesthesia (PRETREAT): clinical trial protocol and considerations.

Background: Intraoperative hypotension has been extensively studied for its association with adverse outcomes. However, small sample sizes and methodological issues limit the causal inference that can be drawn. Methods: In this multicentre, adaptive, randomised controlled trial, we will include 5000 adult inpatients scheduled for elective non-cardiac surgery under general or central neuraxial anaesthesia. Patients will be either randomly allocated to the intervention or care-as-usual group using computer-generated blocks of four, six, or eight, with an allocation ratio of 1:1. In the intervention arm patients will be divided into low-, intermediate-, and high-risk groups based on their likelihood to experience intraoperative hypotension, with resulting mean blood pressure targets of 70, 80, and 90 mm Hg, respectively. Anaesthesia teams will be provided with a clinical guideline on how to keep patients at their target blood pressure. During the first 6 months of the trial the intervention strategy will be evaluated and further revised in adaptation cycles of 3 weeks if necessary, to improve successful impact on the clinical process. The primary outcome is postoperative disability after 6 months measured with the World Health Organization Disability Assessment Score (WHODAS) 2.0 questionnaire. Ethics and dissemination: This study protocol has been approved by the Medical Ethics Committee of the University Medical Centre Utrecht (20-749) and all protocol amendments will be communicated to the Medical Ethics Committee. The study protocol is in adherence with the Declaration of Helsinki and the guideline of Good Clinical Practice. Dissemination plans include publication in a peer-reviewed journal. Clinical trial registration: The Dutch Trial Register, NL9391. Registered on 22 March 2021.

Prognostic factors for multi-organ dysfunction in pediatric oncology patients admitted to the pediatric intensive care unit.

Background: Pediatric oncology patients who require admission to the pediatric intensive care unit (PICU) have worse outcomes compared to their non-cancer peers. Although multi-organ dysfunction (MOD) plays a pivotal role in PICU mortality and morbidity, risk factors for MOD have not yet been identified. We aimed to identify risk factors at PICU admission for new or progressive MOD (NPMOD) during the first week of PICU stay. Methods: This retrospective cohort study included all pediatric oncology patients aged 0 to 18 years admitted to the PICU between June 2018 and June 2021. We used the recently published PODIUM criteria for defining multi-organ dysfunction and estimated the association between covariates at PICU baseline and the outcome NPMOD using a multivariable logistic regression model, with PICU admission as unit of study. To study the predictive performance, the model was internally validated by using bootstrap. Results: A total of 761 PICU admissions of 571 patients were included. NPMOD was present in 154 PICU admissions (20%). Patients with NPMOD had a high mortality compared to patients without NPMOD, 14% and 1.0% respectively. Hemato-oncological diagnosis, number of failing organs and unplanned admission were independent risk factors for NPMOD. The prognostic model had an overall good discrimination and calibration. Conclusion: The risk factors at PICU admission for NPMOD may help to identify patients who may benefit from closer monitoring and early interventions. When applying the PODIUM criteria, we found some opportunities for fine-tuning these criteria for pediatric oncology patients, that need to be validated in future studies.

Continuous Data-Driven Monitoring in Critical Congenital Heart Disease: Clinical Deterioration Model Development.

BACKGROUND: Critical congenital heart disease (cCHD)-requiring cardiac intervention in the first year of life for survival-occurs globally in 2-3 of every 1000 live births. In the critical perioperative period, intensive multimodal monitoring at a pediatric intensive care unit (PICU) is warranted, as their organs-especially the brain-may be severely injured due to hemodynamic and respiratory events. These 24/7 clinical data streams yield large quantities of high-frequency data, which are challenging in terms of interpretation due to the varying and dynamic physiology innate to cCHD. Through advanced data science algorithms, these dynamic data can be condensed into comprehensible information, reducing the cognitive load on the medical team and providing data-driven monitoring support through automated detection of clinical deterioration, which may facilitate timely intervention. OBJECTIVE: This study aimed to develop a clinical deterioration detection algorithm for PICU patients with cCHD. METHODS: Retrospectively, synchronous per-second data of cerebral regional oxygen saturation (rSO2) and 4 vital parameters (respiratory rate, heart rate, oxygen saturation, and invasive mean blood pressure) in neonates with cCHD admitted to the University Medical Center Utrecht, the Netherlands, between 2002 and 2018 were extracted. Patients were stratified based on mean oxygen saturation during admission to account for physiological differences between acyanotic and cyanotic cCHD. Each subset was used to train our algorithm in classifying data as either stable, unstable, or sensor dysfunction. The algorithm was designed to detect combinations of parameters abnormal to the stratified subpopulation and significant deviations from the patient's unique baseline, which were further analyzed to distinguish clinical improvement from deterioration. Novel data were used for testing, visualized in detail, and internally validated by pediatric intensivists. RESULTS: A retrospective query yielded 4600 hours and 209 hours of per-second data in 78 and 10 neonates for, respectively, training and testing purposes. During testing, stable episodes occurred 153 times, of which 134 (88%) were correctly detected. Unstable episodes were correctly noted in 46 of 57 (81%) observed episodes. Twelve expert-confirmed unstable episodes were missed in testing. Time-percentual accuracy was 93% and 77% for, respectively, stable and unstable episodes. A total of 138 sensorial dysfunctions were detected, of which 130 (94%) were correct. CONCLUSIONS: In this proof-of-concept study, a clinical deterioration detection algorithm was developed and retrospectively evaluated to classify clinical stability and instability, achieving reasonable performance considering the heterogeneous population of neonates with cCHD. Combined analysis of baseline (ie, patient-specific) deviations and simultaneous parameter-shifting (ie, population-specific) proofs would be promising with respect to enhancing applicability to heterogeneous critically ill pediatric populations. After prospective validation, the current-and comparable-models may, in the future, be used in the automated detection of clinical deterioration and eventually provide data-driven monitoring support to the medical team, allowing for timely intervention.

Automated feedback modestly improves perioperative treatment adherence of postoperative nausea and vomiting.

STUDY OBJECTIVE: Extensive evidence demonstrates that medical record modernization and a vast amount of available data have not overcome the gap between recommended and delivered care. This study aimed to evaluate the use of clinical decision support (CDS) in conjunction with feedback (post-hoc reporting) to improve PONV medication administration compliance and postoperative nausea and vomiting (PONV) outcomes. DESIGN: Single center, prospective observational study between January 1, 2015, and June 30, 2017. SETTING: Perioperative care at a university-affiliated tertiary care center. PATIENTS: 57,401 adult patients who received general anesthesia in a non-emergency setting. INTERVENTION: A multi-phased intervention that consisted of post-hoc reporting for individual providers by email about PONV occurrences in their patients, followed by directive CDS through preoperative daily case emails that provided therapeutic PONV prophylaxis recommendations based on patients' PONV risk scores. MEASUREMENT: Compliance with PONV medication recommendations, as well as hospital rates of PONV were measured. MAIN RESULT: Over the study period, there was a 5.5% (95% CI, 4.2% to 6.4%; p < 0.001) improvement in the compliance of PONV medication administration along with an 8.7% (95% CI, 7.1% to 10.2%, p < 0.001) reduction in PONV rescue medication administration in the PACU. However, there was no statistically or clinically significant reduction in the prevalence of PONV in the PACU. The prevalence of PONV rescue medication administration decreased during the Intervention Rollout Period (odds ratio 0.95 [per month]; 95% CI, 0.91 to 0.99; p = 0.017), and during the Feedback with CDS Recommendation Period (odds ratio, 0.96 [per month]; 95% CI, 0.94 to 0.99; p = 0.013). CONCLUSION: PONV medication administration compliance modestly improves with CDS in conjunction with post-hoc reporting; however, no improvement in PACU rates of PONV occurred.

Association of a Liberal Fasting Policy of Clear Fluids Before Surgery With Fasting Duration and Patient Well-being and Safety.

Importance: Current fasting guidelines for procedures under anesthesia are poorly implemented, leading to negative metabolic sequelae. Recent studies in children showed support of liberal clear fluid intake; adult physiology can support clear fluid intake, but implementation studies are lacking. Objective: To evaluate the successfulness of implementation of a liberal clear fluid policy with regard to fasting duration, well-being, and safety in adults scheduled for anesthesia. Design, Setting, and Participants: This was a quality improvement study conducted from January 2016 to July 2021 at a tertiary referral hospital in the Netherlands. Adults scheduled for nonemergency procedures under anesthesia were included in the study. Patients undergoing obstetrics procedures or those who were intubated preoperatively were excluded. Interventions: Stepwise introduction of a liberal fluid fasting policy, allowing for ingestion of clear fluids until arrival at the operating room. Main Outcomes and Measures: The primary outcome was change in fasting duration. Secondary outcomes were patient well-being, measured as preoperative thirst, amount of fluid ingested, postoperative nausea and vomiting (PONV), and administration of antiemetics. Safety was measured as incidence of regurgitation and aspiration (pneumonia). Results: Of the 76 451 patients (mean [SD] age, 56 [17] years; 39 530 male individuals [52%] 36 921) included in the study, 59 036 (78%) followed the standard policy, and 16 815 (22%) followed the liberal policy. Time series analysis showed an estimated fasting duration decrease of 3:07 hours (IQR, 1:36-7:22; P < .001) after implementation of the liberal policy. Postimplementation median (IQR) fasting duration was 1:20 (0:48-2:24) hours. The incidence of regurgitation changed from 18 (95% CI, 14-21) to 24 (95% CI, 17-32) in 10 000 patients, and the incidence of aspiration changed from 1.7 (95% CI, 0.6-2.7) to 2.4 (95% CI, 0.5-4.7) in 10 000 patients. In the liberal policy, thirst feelings decreased (37% [4982 of 8615] vs 46% [3373 of 7362]; P < .001). PONV incidence decreased from 10.6% (6339 of 59 636) to 9.4% (1587 of 16 815; P < .001) and antiemetic administration decreased from 11.0% (6538 of 59 636) to 9.5% (1592 of 16 815; P < .001). Conclusions and Relevance: Results of this quality improvement study suggest that a liberal fasting policy was associated with a clinically relevant reduction in fasting duration and improved patient well-being with regard to preoperative thirst and PONV. Although a slightly higher incidence of regurgitation could not be ruled out, wider implementation of such a policy may be advocated as results are still within the clinically accepted risks margins. Results suggest that surgical procedures in patients who drink clear fluids within 2 hours before anticipated anesthesia should not be postponed or canceled.

Validation of a modified bedside Pediatric Early Warning System score for detection of clinical deterioration in hospitalized pediatric oncology patients: A prospective cohort study.

BACKGROUND: Hospitalized pediatric oncology patients are at risk of severe clinical deterioration. Yet Pediatric Early Warning System (PEWS) scores have not been prospectively validated in these patients. We aimed to determine the predictive performance of the modified BedsidePEWS score for unplanned pediatric intensive care unit (PICU) admission and cardiopulmonary resuscitation (CPR) in this patient population. METHODS: We performed a prospective cohort study in an 80-bed pediatric oncology hospital in the Netherlands, where care has been nationally centralized. All hospitalized pediatric oncology patients aged 0-18 years were eligible for inclusion. A Cox proportional hazard model was estimated to study the association between BedsidePEWS score and unplanned PICU admissions or CPR. The predictive performance of the model was internally validated by bootstrapping. RESULTS: A total of 1137 patients were included. During the study, 103 patients experienced 127 unplanned PICU admissions and three CPRs. The hazard ratio for unplanned PICU admission or CPR was 1.65 (95% confidence interval [CI]: 1.59-1.72) for each point increase in the modified BedsidePEWS score. The discriminative ability was moderate (D-index close to 0 and a C-index of 0.83 [95% CI: 0.79-0.90]). Positive and negative predictive values of modified BedsidePEWS score at the widely used cutoff of 8, at which escalation of care is required, were 1.4% and 99.9%, respectively. CONCLUSION: The modified BedsidePEWS score is significantly associated with requirement of PICU transfer or CPR. In pediatric oncology patients, this PEWS score may aid in clinical decision-making for timing of PICU transfer.

Associations between intraoperative hypotension, duration of surgery and postoperative myocardial injury after noncardiac surgery: a retrospective single-centre cohort study.

BACKGROUND: Studies of intraoperative hypotension typically specify a blood pressure threshold associated with adverse outcomes. Such thresholds are likely to be study-biased, investigator-biased, or both. We hypothesised that a newly developed modelling method without a threshold, which is biologically more plausible than a threshold-based approach, would reveal a continuous association between exposure to intraoperative hypotension and adverse outcomes. METHODS: Single-centre, retrospective cohort study of subjects ≥60 yr old undergoing noncardiac surgery. We modelled intraoperative hypotension using three different approaches: (1) unweighted, (2) weighted for degree of hypotension (depth), and (3) weighted for duration of hypotension. The primary outcome was myocardial injury, defined as elevated troponin I (>60 ng L-1) measured during the first 3 days after surgery. The associations between the three models, postoperative myocardial injury, and mortality (secondary outcome) were reported as penalised adjusted odds ratios (ORs) scaled between the 75th and 25th percentiles. RESULTS: Myocardial injury occurred in 1812/15 452 (12%) procedures, with 554/15 452 (3.6%) procedures resulting in death before discharge from hospital. The unweighted lower blood pressure measure (OR: 0.26, 95% confidence interval [CI]: 0.12-0.53) and the depth-weighted measure (OR: 4.4, 95% CI: 2.6-7.4) were associated with myocardial injury. The duration-weighted measure was not associated with myocardial injury (OR: 0.89, 95% CI: 0.61-1.3). The unweighted measure (OR 0.08, 95% CI: 0.01-0.40) and the depth-weighted measure (OR: 12, 95% CI, 3.8-35) were associated with in-hospital mortality, but not the duration-weighted measure (OR: 1.3, 95% CI: 0.53-3.0). CONCLUSIONS: Intraoperative hypotension appears to have a graded association with postoperative myocardial injury and mortality, with depth appearing to contribute more than duration.

The predictive performance and impact of pediatric early warning systems in hospitalized pediatric oncology patients-A systematic review.

Pediatric early warning systems (PEWS) arewidely used to identify clinically deteriorating patients. Hospitalized pediatric oncology patients are particularly prone to clinical deterioration. We assessed the PEWS performance to predict early clinical deterioration and the effect of PEWS implementation on patient outcomes in pediatric oncology patients. PubMED, EMBASE, and CINAHL databases were systematically searched from inception up to March 2020. Quality assessment was performed using the Prediction model study Risk-Of-Bias Assessment Tool (PROBAST) and the Cochrane Risk-of-Bias Tool. Nine studies were included. Due to heterogeneity of study designs, outcome measures, and diversity of PEWS, it was not possible to conduct a meta-analysis. Although the studies reported high sensitivity, specificity, and area under the receiver operating characteristics curve (AUROC) of PEWS detecting inpatient deterioration, overall risk of bias of the studies was high. This review highlights limited evidence on the predictive performance of PEWS for clinical deterioration and the effect of PEWS implementation.

Autonomous patient consent for anaesthesia without preoperative consultation: a qualitative feasibility study including low-risk procedures.

Background: Informed consent for anaesthesia is mandatory and requires provision of information and subsequent consent during consultation between anaesthesiologist and patient. Although information can be provided in an electronic format, it is unknown whether this a valid substitute for a consultation. We explored whether provision of digital information is equivalent to oral consultation and whether it enables patients to give electronic informed consent (e-consent) for anaesthesia. Methods: Qualitative feasibility study using semi-structured interviews in 20 low-risk adults scheduled for minor surgery under general anaesthesia or procedural sedation at a university hospital. Data were analysed using a thematic content analysis approach. During the interviews, patients followed an application that provides information and subsequent e-consenting. Results: The mean age was 50 yr and patients had good digital skills. Fifteen patients (75%) had previous experience of anaesthesia. The digital application provided enough information for all patients, but eight (40%) preferred consultation with an anaesthesiologist, mainly for personal contact. Patients had different information needs, with previous experiences leading to lower information needs. Nineteen patients had sufficient information to consent autonomously. Most patients considered separate anaesthesia consent superfluous to the surgical consent. Conclusion: The digital application provided sufficient information and patients valued the information offered and the advantage of processing information at their own pace. This information made patients feel empowered to autonomously consent to anaesthesia without consultation. Remarkably, consent for anaesthesia was considered unimportant, because patients felt they had 'no choice' if they wanted to undergo surgery.

Intraoperative hypotension and delirium among older adults undergoing transcatheter aortic valve replacement.

BACKGROUND: Postoperative delirium (POD) is a frequently observed complication after transcatheter aortic valve replacement (TAVR). The effects of intraoperative hypotension (IOH) on POD occurrence are currently unclear. METHODS: A retrospective observational cohort study of patients who underwent TAVR was conducted. We predefined IOH as area under the threshold (AUT) of five mean arterial blood pressures (MBP), varying from <100 to <60 mmHg. The AUT consisted of the combination of duration and depth under the MBP thresholds, expressed in mmHg*min. All MBP AUTs were computed based on the complete procedure, independent of procedural phase or duration. RESULTS: This cohort included 675 patients who underwent TAVR under general anesthesia (n = 128, 19%) or procedural sedation (n = 547, 81%). Delirium occurred mostly during the first 2 days after TAVR, and was observed in n = 93 (14%) cases. Furthermore, 674, 672, 663, 630, and 518 patients had at least 1 min intraoperative MBP <100, <90, <80, <70, and <60 mmHg, respectively. Patients who developed POD had higher AUT based on all five MBP thresholds during TAVR. The penalized adjusted odds ratio varied between 1.08 (99% confidence interval [CI] 0.74-1.56) for the AUT based on MBP < 100 mmHg and OR 1.06 (99% CI 0.88-1.28) for the AUT based on MBP < 60 mmHg. CONCLUSIONS: Intraoperative hypotension is frequently observed during TAVR, but not independently associated with POD after TAVR. Other potential factors than intraoperative hypotension may explain the occurrence of delirium after TAVR.

Hypotension during endovascular treatment under general anesthesia for acute ischemic stroke.

OBJECTIVE: The effect of anesthetic management (general anesthesia [GA], conscious sedation, or local anesthesia) on functional outcome and the role of blood pressure management during endovascular treatment (EVT) for acute ischemic stroke is under debate. We aimed to determine whether hypotension during EVT under GA is associated with functional outcome at 90 days. METHODS: We retrospectively collected data from patients with a proximal intracranial occlusion of the anterior circulation treated with EVT under GA. The primary outcome was the distribution on the modified Rankin Scale at 90 days. Hypotension was defined using two thresholds: a mean arterial pressure (MAP) of 70 mm Hg and a MAP 30% below baseline MAP. To quantify the extent and duration of hypotension, the area under the threshold (AUT) was calculated using both thresholds. RESULTS: Of the 366 patients included, procedural hypotension was observed in approximately half of them. The occurrence of hypotension was associated with poor functional outcome (MAP <70 mm Hg: adjusted common odds ratio [acOR], 0.57; 95% confidence interval [CI], 0.35-0.94; MAP decrease ≥30%: acOR, 0.76; 95% CI, 0.48-1.21). In addition, an association was found between the number of hypotensive periods and poor functional outcome (MAP <70 mm Hg: acOR, 0.85 per period increase; 95% CI, 0.73-0.99; MAP decrease ≥30%: acOR, 0.90 per period; 95% CI, 0.78-1.04). No association existed between AUT and functional outcome (MAP <70 mm Hg: acOR, 1.000 per 10 mm Hg*min increase; 95% CI, 0.998-1.001; MAP decrease ≥30%: acOR, 1.000 per 10 mm Hg*min; 95% CI, 0.999-1.000). CONCLUSIONS: Occurrence of procedural hypotension and an increase in number of procedural hypotensive periods were associated with poor functional outcome, whereas the extent and duration of hypotension were not. Randomized clinical trials are needed to confirm our hypothesis that hypotension during EVT under GA has detrimental effects.

Identifying the critically ill paediatric oncology patient: a study protocol for a prospective observational cohort study for validation of a modified Bedside Paediatric Early Warning System score in hospitalised paediatric oncology patients.

INTRODUCTION: Hospitalised paediatric oncology patients are at risk to develop acute complications. Early identification of clinical deterioration enabling adequate escalation of care remains challenging. Various Paediatric Early Warning Systems (PEWSs) have been evaluated, also in paediatric oncology patients but mostly in retrospective or case-control study designs. This study protocol encompasses the first prospective cohort with the aim of evaluating the predictive performance of a modified Bedside PEWS score for non-elective paediatric intensive care unit (PICU) admission or cardiopulmonary resuscitation in hospitalised paediatric oncology patients. METHODS AND ANALYSIS: A prospective cohort study will be conducted at the 80-bed Dutch paediatric oncology hospital, where all national paediatric oncology care has been centralised, directly connected to a shared 22-bed PICU. All patients between 1 February 2019 and 1 February 2021 admitted to the inpatient nursing wards, aged 0-18 years, with an International Classification of Diseases for Oncology (ICD-O) diagnosis of paediatric malignancy will be eligible. A Cox proportional hazard regression model will be used to estimate the association between the modified Bedside PEWS and time to non-elective PICU transfer or cardiopulmonary arrest. Predictive performance (discrimination and calibration) will be assessed internally using resampling validation. To account for multiple occurrences of the event of interest within each patient, the unit of study is a single uninterrupted ward admission (a clinical episode). ETHICS AND DISSEMINATION: The study protocol has been approved by the institutional ethical review board of our hospital (MEC protocol number 16-572/C). We adapted our enrolment procedure to General Data Protection Regulation compliance. Results will be disseminated at scientific conferences, regional educational sessions and publication in peer-reviewed journals. TRIAL REGISTRATION NUMBER: Netherlands Trial Registry (NL8957).

Reducing medical device alarms by an order of magnitude: A human factors approach.

The intensive care unit (ICU) is one of the most technically advanced environments in healthcare, using a multitude of medical devices for drug administration, mechanical ventilation and patient monitoring. However, these technologies currently come with disadvantages, namely noise pollution, information overload and alarm fatigue-all caused by too many alarms. Individual medical devices currently generate alarms independently, without any coordination or prioritisation with other devices, leading to a cacophony where important alarms can be lost amongst trivial ones, occasionally with serious or even fatal consequences for patients. We have called this approach to the design of medical devices the single-device paradigm, and believe it is obsolete in modern hospitals where patients are typically connected to several devices simultaneously. Alarm rates of one alarm every four minutes for only the physiological monitors (as recorded in the ICUs of two hospitals contributing to this paper) degrades the quality of the patient's healing environment and threatens patient safety by constantly distracting healthcare professionals. We outline a new approach to medical device design involving the application of human factors principles which have been successful in eliminating alarm fatigue in commercial aviation. Our approach comprises the networked-device paradigm, comprehensive alarms and humaniform information displays. Instead of each medical device alarming separately at the patient's bedside, our proposed approach will integrate, prioritise and optimise alarms across all devices attached to each patient, display information more intuitively and hence increase alarm quality while reducing the number of alarms by an order of magnitude below current levels.

A Human(e) Factor in Clinical Decision Support Systems.

The overwhelming amount, production speed, multidimensionality, and potential value of data currently available-often simplified and referred to as big data -exceed the limits of understanding of the human brain. At the same time, developments in data analytics and computational power provide the opportunity to obtain new insights and transfer data-provided added value to clinical practice in real time. What is the role of the health care professional in collaboration with the data scientist in the changing landscape of modern care? We discuss how health care professionals should provide expert knowledge in each of the stages of clinical decision support design: data level, algorithm level, and decision support level. Including various ethical considerations, we advocate for health care professionals to responsibly initiate and guide interprofessional teams, including patients, and embrace novel analytic technologies to translate big data into patient benefit driven by human(e) values.

Myocardial Injury in Critically Ill Patients with Community-acquired Pneumonia. A Cohort Study.

Rationale: Myocardial injury, as reflected by elevated cardiac troponin levels in plasma, is common in patients with community-acquired pneumonia (CAP), but its temporal dynamics and etiology remain unknown. Objectives: Our aim was to determine the incidence of troponin release in patients with CAP and identify risk factors that may point to underlying etiologic mechanisms. Methods: We included consecutive patients admitted with severe CAP to two intensive care units in the Netherlands between 2011 and 2015. High-sensitivity cardiac troponin I was measured daily during the first week. We used multivariable linear regression to identify variables associated with troponin release on admission, and we used mixed-effects regression to model the daily rise and fall of troponin levels over time. Results: Of 200 eligible patients, 179 were included, yielding 792 observation days. A total of 152 (85%) patients developed raised troponin levels greater than 26 ng/L. Baseline factors independently associated with troponin release included coronary artery disease (176% increase; 95% confidence interval [CI], 11-589), smoking (248% increase; 95% CI, 33-809), and higher Acute Physiology and Chronic Health Evaluation IV score (2% increase; 95% CI, 0.8-3.3), whereas Staphylococcus aureus as a causative pathogen was protective (70% reduction; 95% CI, 18-89). Time-dependent risk factors independently associated with daily increase in troponin concentrations included reduced platelet count (2.3% increase; 95% CI, 0.6-4), tachycardia (1.5% increase; 95% CI, 0.1-2.9), hypotension (6.2% increase; 95% CI, 2.1-10.6), dobutamine use (44% increase; 95% CI, 12-85), prothrombin time (8.2% increase; 95% CI, 0.2-16.9), white cell count (1.7% increase; 95% CI, 0-3.5), and fever (22.7% increase; 95% CI, 0.1-49.6). Conclusions: Cardiac injury develops in a majority of patients with severe CAP. Myocardial oxygen supply-demand mismatch and activated inflammation/coagulation are associated with this injury. Clinical trial registered with www.clinicaltrials.gov (NCT01905033).

Evaluating the impact of prediction models: lessons learned, challenges, and recommendations.

An important aim of clinical prediction models is to positively impact clinical decision making and subsequent patient outcomes. The impact on clinical decision making and patient outcome can be quantified in prospective comparative-ideally cluster-randomized-studies, known as 'impact studies'. However, such impact studies often require a lot of time and resources, especially when they are (cluster-)randomized studies. Before envisioning such large-scale randomized impact study, it is important to ensure a reasonable chance that the use of the prediction model by the targeted healthcare professionals and patients will indeed have a positive effect on both decision making and subsequent outcomes. We recently performed two differently designed, prospective impact studies on a clinical prediction model to be used in surgical patients. Both studies taught us new valuable lessons on several aspects of prediction model impact studies, and which considerations may guide researchers in their decision to conduct a prospective comparative impact study. We provide considerations on how to prepare a prediction model for implementation in practice, how to present the model predictions, and how to choose the proper design for a prediction model impact study.