Enabling a learning healthcare system with automated computer protocols that produce replicable and personalized clinician actions.

Clinical decision-making is based on knowledge, expertise, and authority, with clinicians approving almost every intervention-the starting point for delivery of "All the right care, but only the right care," an unachieved healthcare quality improvement goal. Unaided clinicians suffer from human cognitive limitations and biases when decisions are based only on their training, expertise, and experience. Electronic health records (EHRs) could improve healthcare with robust decision-support tools that reduce unwarranted variation of clinician decisions and actions. Current EHRs, focused on results review, documentation, and accounting, are awkward, time-consuming, and contribute to clinician stress and burnout. Decision-support tools could reduce clinician burden and enable replicable clinician decisions and actions that personalize patient care. Most current clinical decision-support tools or aids lack detail and neither reduce burden nor enable replicable actions. Clinicians must provide subjective interpretation and missing logic, thus introducing personal biases and mindless, unwarranted, variation from evidence-based practice. Replicability occurs when different clinicians, with the same patient information and context, come to the same decision and action. We propose a feasible subset of therapeutic decision-support tools based on credible clinical outcome evidence: computer protocols leading to replicable clinician actions (eActions). eActions enable different clinicians to make consistent decisions and actions when faced with the same patient input data. eActions embrace good everyday decision-making informed by evidence, experience, EHR data, and individual patient status. eActions can reduce unwarranted variation, increase quality of clinical care and research, reduce EHR noise, and could enable a learning healthcare system.

Six subphenotypes in septic shock: Latent class analysis of the PROWESS Shock study.

PURPOSE: Septic shock is a highly heterogeneous condition which is part of the challenge in its diagnosis and treatment. In this study we aim to identify clinically relevant subphenotypes of septic shock using a novel statistic al approach. METHODS: Baseline patient data from a large global clinical trial of septic shock (n = 1696) was analysed using latent class analysis (LCA). This approach allowed investigators to identify subgroups in a heterogeneous population by estimating a categorical latent variable that detects relatively homogeneous subgroups within a complex phenomenon. RESULTS: LCA identified six different, clinically meaningful subphenotypes of septic shock each with a typical profile: (1) "Uncomplicated Septic Shock, (2) "Pneumonia with adult respiratory distress syndrome (ARDS)", (3) "Postoperative Abdominal", (4) "Severe Septic Shock", (5): "Pneumonia with ARDS and multiple organ dysfunction syndrome (MODS)", (6) "Late Septic Shock". The 6-class solution showed high entropy approaching 1 (i.e., 0.92), indicating there was excellent separation between estimated classes. CONCLUSIONS: LCA appears to be an applicable statistical tool in analysing a heterogenous clinical cohort of septic shock. The results may lead to a better understanding of septic shock complexity and form a basis for considering targeted therapies and selecting patients for future clinical trials.

Whole blood microRNA markers are associated with acute respiratory distress syndrome.

BACKGROUND: MicroRNAs (miRNAs) can play important roles in inflammation and infection, which are common manifestations of acute respiratory distress syndrome (ARDS). We assessed if whole blood miRNAs were potential diagnostic biomarkers for human ARDS. METHODS: This nested case-control study (N = 530) examined a cohort of ARDS patients and critically ill at-risk controls. Whole blood miRNA profiles and logistic regression analyses identified miRNAs correlated with ARDS. Stratification analysis also assessed selected miRNA markers for their role in sepsis and pneumonia associated with ARDS. Receiver operating characteristic (ROC) analysis evaluated miRNA diagnostic performance, along with Lung Injury Prediction Score (LIPS). RESULTS: Statistical analyses were performed on 294 miRNAs, selected from 754 miRNAs after quality control screening. Logistic regression identified 22 miRNAs from a 156-patient discovery cohort as potential risk or protective markers of ARDS. Three miRNAs-miR-181a, miR-92a, and miR-424-from the discovery cohort remained significantly associated with ARDS in a 373-patient independent validation cohort (FDR q < 0.05) and meta-analysis (p < 0.001). ROC analyses demonstrated a LIPS baseline area-under-the-curve (AUC) value of ARDS of 0.708 (95% CI 0.651-0.766). Addition of miR-181a, miR-92a, and miR-424 to LIPS increased baseline AUC to 0.723 (95% CI 0.667-0.778), with a relative integrated discrimination improvement of 2.40 (p = 0.005) and a category-free net reclassification index of 27.21% (p = 0.01). CONCLUSIONS: miR-181a and miR-92a are risk biomarkers for ARDS, whereas miR-424 is a protective biomarker. Addition of these miRNAs to LIPS can improve the risk estimate for ARDS.

Intensive care unit-acquired neuromyopathy and corticosteroids in survivors of persistent ARDS.

OBJECTIVES: To determine the incidence and outcomes of intensive care unit-acquired neuromyopathy and to investigate the role of methylprednisolone in survivors of persistent acute lung injury. DESIGN: Secondary analysis of completed randomized placebo-controlled trial. SETTING: Twenty-five hospitals in the NHLBI ARDS Network. PATIENTS AND PARTICIPANTS: Patients enrolled in the ARDS Network study of methylprednisolone versus placebo for persistent ARDS who survived 60 days or to hospital discharge. MEASUREMENTS AND RESULTS: One hundred and twenty-eight study patients survived 60 days. Forty-three (34%) of these patients had evidence by chart review of ICU-acquired neuromyopathy, which was associated with prolonged mechanical ventilation, return to mechanical ventilation, and delayed return to home after critical illness. Treatment with methylprednisolone was not significantly associated with an increase in risk of neuromyopathy (OR 1.5; 95% CI 0.7-3.2). CONCLUSIONS: ICU-acquired-neuromyopathy is common among survivors of persistent ARDS and is associated with poorer clinical outcomes. We did not find a significant association between methylprednisolone treatment and neuromyopathy. Limitations of this study preclude definitive conclusions about the causal relationship between corticosteroids and ICU-acquired neuromuscular dysfunction.

D-dimers in the diagnosis of pulmonary embolism.

The aim of this study was to determine if the absence of circulating D-dimers, as determined by latex agglutination assays, can correctly exclude the presence of pulmonary embolism using pulmonary angiography as the diagnostic endpoint. Blood samples were obtained prospectively at the time of angiography for suspicion of acute pulmonary embolism. Plasma was assayed for D-dimer by five different latex agglutination assays. Angiographic evidence of pulmonary emboli was found in 34% (35/ 103) of patients. The latex agglutination assays had sensitivities of 97 to 100% and specificities of 19 to 29%. The negative predictive value was 94 to 100%. However, a negative D-dimer was rare in patients with recent surgery, malignancy, or total bilirubin > 34 micromol/L (> 2 mg/dl). In 31 patients suspected of pulmonary emboli but without these confounding factors, the five D-dimer assays were negative in 46 to 55% of patients with normal pulmonary angiograms. The negative predictive value in these patients was 100% by all five latex agglutination assays tested. The latex agglutination assays for D-dimer, when the pulmonary angiogram is used as the diagnostic endpoint and in the absence of recent surgery, malignancy, or liver disease, appears to be a clinically useful test in the diagnosis of acute pulmonary embolism.