The Cro-Magnon 1 skeleton corresponds to a 28?000 BCE Homo sapiens male individual that was discovered in 1868 in a rock shelter in Les Eyzies, France.1 Since its discovery, various diagnoses have been proposed with regards to a round polycyclic osteolytic lesion on the right frontal bone, measuring 37 mm?x?27 mm (appendix): post-mortem alteration due to the soil,2 rickets,3 actinomycosis,4 and Langerhans cell histiocytosis.5

We read with interest the Lancet Editorial on artificial intelligence (AI) in health care (Dec 23, 2017, p 2739).1 Deep learning as a form of AI risks being overhyped. Deep neural networks contain multiple layers of nodes connected by adjustable weights. Learning occurs by adjusting these weights until the desired input-to-output function is achieved.2 With many millions of weights, huge amounts of data are required for learning, a process facilitated by recent increases in computational power. However, the learning algorithm, known as the error back-propagation algorithm, was invented in the 1980s and has been used to train neural networks ever since.

Our research into the UN Relief and Works Agency (UNRWA)'s delivery of health services to Palestinian refugees during the Syria crisis1 puts us in a unique position to anticipate the challenges of the organisation's current funding crisis.2 We have conducted over 90 interviews with health workers and managers, a series of systems modelling sessions, and rigorous analysis of UNRWA health data from 2007–16, and conclude the following.

We read with great interest the Seminar (Feb 9, 2017, p 2239)1 on type 2 diabetes by Sudesna Chatterjee and colleagues. However, we were surprised by the articles selected and believe that detailed selection criteria with the level of evidence of reported studies would have been useful to the reader. According to the research method described, we would expect other papers to be cited, including meta-analyses of randomised controlled trials that could have balanced the authors' outlook.2–6 For example, intensive glycaemic control probably has some beneficial effect on diabetic complications, such as non-fatal myocardial infarctions3–5 or retinopathy assessed with the Early Treatment Diabetic Retinopathy Study scale.

In the summary of their Seminar,1 Sudesna Chatterjee and colleagues state that the incidence of type 2 diabetes “continues to rise globally”. There is no evidence to support this claim and most recent literature suggests that, in developed countries, incidence peaked sometime in the last decade and then levelled off or slightly decreased.2–5

We thank A Rosemary Tate for her insightful comments on our Seminar.1 Tate indicates that there is no evidence to support the rise in global incidence of type 2 diabetes mentioned in the summary by citing a number of references published between 2014 and 2017 in the USA and UK.

The global need for new antimalarial drugs and new combinations is enormous and urgent,1,2 but their successful delivery needs resilience to overcome the barriers imposed by expensive and lengthy clinical development plans. Attention is often directed to areas such as southeast Asia, where some antimalarial combinations are failing but transmission intensities are much lower than in sub-Saharan African countries. Children in Africa have frequent and life-threatening malaria infections as they grow up, and these need to be treated safely.

Pyronaridine–artesunate and dihydroartemisinin–piperaquine treatment and retreatment of malaria were well tolerated with efficacy that was non-inferior to first-line ACTs. Greater access to these efficacious treatments in west Africa is justified.

Sudden infant death syndrome (SIDS) remains a leading cause of infant mortality, despite a steadily decreasing incidence since the 1990s.1 The reasons for this decline are debated, but it could be due to methodological reasons (eg, changes in reporting or advances in diagnosis of specific diseases) or a reduction of risks, such as an increase in supine sleeping position for infants, as advocated by the Back to Sleep campaign.2 A better understanding of the causes of SIDS is needed to identify infants at high risk and to develop interventions and guidelines that will prevent SIDS for all infants.

Rare SCN4A variants that directly alter NaV1.4 function occur in infants who had died from SIDS. These variants are predicted to significantly alter muscle membrane excitability and compromise respiratory and laryngeal function. These findings indicate that dysfunction of muscle sodium channels is a potentially modifiable risk factor in a subset of infant sudden deaths.

No abstract available

No abstract available

Objectives: The optimum duration of antimicrobial treatment for patients with bacteremia is unknown. Our objectives were to determine duration of antimicrobial treatment provided to patients who have bacteremia in ICUs, to assess pathogen/patient factors related to treatment duration, and to assess the relationship between treatment duration and survival. Design: Retrospective cohort study. Settings: Fourteen ICUs across Canada. Patients: Patients with bacteremia and were present in the ICU at the time culture reported positive. Interventions: Duration of antimicrobial treatment for patients who had bacteremia in ICU. Measurements and Main Results: Among 1,202 ICU patients with bacteremia, the median duration of treatment was 14 days, but with wide variability (interquartile range, 9–17.5). Most patient characteristics were not associated with treatment duration. Coagulase-negative staphylococci were the only pathogens associated with shorter treatment (odds ratio, 2.82; 95% CI, 1.51–5.26). The urinary tract was the only source of infection associated with a trend toward lower likelihood of shorter treatment (odds ratio, 0.67; 95% CI, 0.42–1.08); an unknown source of infection was associated with a greater likelihood of shorter treatment (odds ratio, 2.14; 95% CI, 1.17–3.91). The association of treatment duration and survival was unstable when analyzed based on timing of death. Conclusions: Critically ill patients who have bacteremia typically receive long courses of antimicrobials. Most patient/pathogen characteristics are not associated with treatment duration; survivor bias precludes a valid assessment of the association between treatment duration and survival. A definitive randomized controlled trial is needed to compare shorter versus longer antimicrobial treatment in patients who have bacteremia.

Objective: Despite telemedicine’s potential to improve patients’ health outcomes and reduce costs in the ICU, hospitals have been slow to introduce telemedicine in the ICU due to high up-front costs and mixed evidence on effectiveness. This study’s first aim was to conduct a cost-effectiveness analysis to estimate the incremental cost-effectiveness ratio of telemedicine in the ICU, compared with ICU without telemedicine, from the healthcare system perspective. The second aim was to examine potential cost saving of telemedicine in the ICU through probabilistic analyses and break-even analyses. Design: Simulation analyses performed by standard decision models. Setting: Hypothetical ICU defined by the U.S. literature. Patients: Hypothetical adult patients in ICU defined by the U.S. literature. Interventions: The intervention was the introduction of telemedicine in the ICU, which was assumed to affect per-patient per-hospital-stay ICU cost and hospital mortality. Telemedicine in the ICU operation costs included the telemedicine equipment-installation (start-up) costs with 5-year depreciation, maintenance costs, and clinician staffing costs. Telemedicine in the ICU effectiveness was measured by cumulative quality-adjusted life years for 5 years after ICU discharge. Measurements and Main Results: The base case cost-effectiveness analysis estimated telemedicine in the ICU to extend 0.011 quality-adjusted life years with an incremental cost of $516 per patient compared with ICU without telemedicine, resulting in an incremental cost-effectiveness ratio of $45,320 per additional quality-adjusted life year (= $516/0.011). The probabilistic cost-effectiveness analysis estimated an incremental cost-effectiveness ratio of $50,265 with a wide 95% CI from a negative value (suggesting cost savings) to $375,870. These probabilistic analyses projected that cost saving is achieved 37% of 1,000 iterations. Cost saving is also feasible if the per-patient per-hospital-stay operational cost and physician cost were less than $422 and less than $155, respectively, based on break-even analyses. Conclusions: Our analyses suggest that telemedicine in the ICU is cost-effective in most cases and cost saving in some cases. The thresholds of cost and effectiveness, estimated by break-even analyses, help hospitals determine the impact of telemedicine in the ICU and potential cost saving.

Objective: To determine the efficacy of anakinra (recombinant interleukin-1 receptor antagonist) in improving 28-day survival in sepsis patients with features of macrophage activation syndrome. Despite equivocal results in sepsis trials, anakinra is effective in treating macrophage activation syndrome, a similar entity with fever, disseminated intravascular coagulation, hepatobiliary dysfunction, cytopenias, and hyperferritinemia. Hence, sepsis patients with macrophage activation syndrome features may benefit from interleukin-1 receptor blockade. Design: Reanalysis of deidentified data from the phase III randomized interleukin-1 receptor antagonist trial in severe sepsis. Setting: Multicenter study recruiting through 91 centers from 11 countries in Europe and North America. Patients: Sepsis patients with multiorgan dysfunction syndrome and/or shock (original study) were regrouped based on the presence or the absence of concurrent hepatobiliary dysfunction and disseminated intravascular coagulation as features of macrophage activation syndrome. The non–hepatobiliary dysfunction/disseminated intravascular coagulation group included patients with only hepatobiliary dysfunction, only disseminated intravascular coagulation, or neither. Intervention: Treatment with anakinra or placebo. Measurements and Main Results: Main outcome was 28–day mortality. Descriptive and comparative statistics were performed. Data were available for 763 adults from the original study cohort, randomized to receive either anakinra or placebo. Concurrent hepatobiliary dysfunction/disseminated intravascular coagulation was noted in 43 patients (5.6% of total; 18–75 years old; 47% women). The 28-day survival was similar in both anakinra and placebo-treated non–hepatobiliary dysfunction/disseminated intravascular coagulation patients (71.4% vs 70.8%; p = 0.88). Treatment with anakinra was associated with significant improvement in the 28-day survival rate in hepatobiliary dysfunction/disseminated intravascular coagulation patients (65.4% anakinra vs 35.3% placebo), with hazard ratio for death 0.28 (0.11–0.71; p = 0.0071) for the treatment group in Cox regression. Conclusions: In this subgroup analysis, interleukin-1 receptor blockade was associated with significant improvement in survival of patients with sepsis and concurrent hepatobiliary dysfunction/disseminated intravascular coagulation. A prospective randomized trial using features of macrophage activation syndrome for mortality risk stratification should be undertaken to confirm the role of interleukin-1 blockage.

Objectives: A low or moderate expired tidal volume can be difficult to achieve during noninvasive ventilation for de novo acute hypoxemic respiratory failure (i.e., not due to exacerbation of chronic lung disease or cardiac failure). We assessed expired tidal volume and its association with noninvasive ventilation outcome. Design: Prospective observational study. Setting: Twenty-four bed university medical ICU. Patients: Consecutive patients receiving noninvasive ventilation for acute hypoxemic respiratory failure between August 2010 and February 2013. Interventions: Noninvasive ventilation was uniformly delivered using a simple algorithm targeting the expired tidal volume between 6 and 8 mL/kg of predicted body weight. Measurements: Expired tidal volume was averaged and respiratory and hemodynamic variables were systematically recorded at each noninvasive ventilation session. Main Results: Sixty-two patients were enrolled, including 47 meeting criteria for acute respiratory distress syndrome, and 32 failed noninvasive ventilation (51%). Pneumonia (n = 51, 82%) was the main etiology of acute hypoxemic respiratory failure. The median (interquartile range) expired tidal volume averaged over all noninvasive ventilation sessions (mean expired tidal volume) was 9.8 mL/kg predicted body weight (8.1–11.1 mL/kg predicted body weight). The mean expired tidal volume was significantly higher in patients who failed noninvasive ventilation as compared with those who succeeded (10.6 mL/kg predicted body weight [9.6–12.0] vs 8.5 mL/kg predicted body weight [7.6–10.2]; p = 0.001), and expired tidal volume was independently associated with noninvasive ventilation failure in multivariate analysis. This effect was mainly driven by patients with PaO2/FIO2 up to 200 mm Hg. In these patients, the expired tidal volume above 9.5 mL/kg predicted body weight predicted noninvasive ventilation failure with a sensitivity of 82% and a specificity of 87%. Conclusions: A low expired tidal volume is almost impossible to achieve in the majority of patients receiving noninvasive ventilation for de novo acute hypoxemic respiratory failure, and a high expired tidal volume is independently associated with noninvasive ventilation failure. In patients with moderate-to-severe hypoxemia, the expired tidal volume above 9.5 mL/kg predicted body weight accurately predicts noninvasive ventilation failure.

Objective: Knowledge on characteristics and outcome of ICU patients with AIDS is highly limited. We aimed to determine the main reasons for admission and outcome in ICU patients with AIDS and trends over time therein. Design: A retrospective study within the Dutch National Intensive Care Evaluation registry. Setting: Dutch ICUs. Patients: We used data collected between 1997 and 2014. Characteristics of patients with AIDS were compared with ICU patients without AIDS, matched for age, sex, admission type, and admission year. Joinpoint regression analysis was applied to study trends over time. Interventions: None. Measurements and Main Results: We included 1,127 patients with AIDS and 4,479 matched controls. The main admission diagnoses of patients with AIDS were respiratory infection (28.6%) and sepsis (16.9%), which were less common in controls (7.7% and 7.5%, respectively; both p < 0.0001). Patients with AIDS had increased severity of illness and in-hospital mortality (28.2% vs 17.8%; p < 0.0001) compared with controls, which was associated with a higher rate of infections at admission in patients with AIDS (58.4% vs 25.5%). Over time, the proportion of patients with AIDS admitted with an infection decreased (75% in 1999 to 56% in 2013). Mortality declined in patients with AIDS (39% in 1999 to 16% in 2013), both in patients with or without an infection. Mortality also declined in matched controls without AIDS, but to a lesser extent. Conclusion: Infections are still the main reason for ICU admission in patients with AIDS, but their prevalence is declining. Outcome of patients with AIDS continued to improve during a time of widespread availability of combination antiretroviral therapy, and mortality is reaching levels similar to ICU patients without AIDS.

Objective: The approach to applying positive end-expiratory pressure in morbidly obese patients is not well defined. These patients frequently require prolonged mechanical ventilation, increasing the risk for failed liberation from ventilatory support. We hypothesized that lung recruitment maneuvers and titration of positive end-expiratory pressure were both necessary to improve lung volumes and the elastic properties of the lungs, leading to improved gas exchange. Design: Prospective, crossover, nonrandomized interventional study. Setting: Medical and surgical ICUs at Massachusetts General Hospital. Patients: Critically ill, mechanically ventilated morbidly obese (body mass index > 35 kg/m2) patients (n = 14). Interventions: This study evaluated two methods of titrating positive end-expiratory pressure; both trials were done utilizing positive end-expiratory pressure titration and recruitment maneuvers while measuring hemodynamics and respiratory mechanics. Measurements were obtained at the baseline positive end-expiratory pressure set by the clinicians, at zero positive end-expiratory pressure, at best positive end-expiratory pressure identified through esophageal pressure measurement before and after a recruitment maneuver, and at best positive end-expiratory pressure identified through a best decremental positive end-expiratory pressure trial. Measurements and Main Results: The average body mass index was 50.7 ± 16.0 kg/m2. The two methods of evaluating positive end-expiratory pressure identified similar optimal positive end-expiratory pressure levels (20.7 ± 4.0 vs 21.3 ± 3.8 cm H2O; p = 0.40). End-expiratory pressure titration increased end-expiratory lung volumes (?11 ± 7 mL/kg; p < 0.01) and oxygenation (?86 ± 50 torr; p < 0.01) and decreased lung elastance (?5 ± 5 cm H2O/L; p < 0.01). Recruitment maneuvers followed by titrated positive end-expiratory pressure were effective at increasing end-expiratory lung volumes while decreasing end-inspiratory transpulmonary pressure, suggesting an improved distribution of lung aeration and reduction of overdistension. The positive end-expiratory pressure levels set by the clinicians (11.6 ± 2.9 cm H2O) were associated with lower lung volumes, worse elastic properties of the lung, and lower oxygenation. Conclusions: Commonly used positive end-expiratory pressure by clinicians is inadequate for optimal mechanical ventilation of morbidly obese patients. A recruitment maneuver followed by end-expiratory pressure titration was found to significantly improve lung volumes, respiratory system elastance, and oxygenation.

Objectives: Bradycardia is common during targeted temperature management, likely being a physiologic response to lower body temperature, and has recently been associated with favorable outcome following out-of-hospital cardiac arrest in smaller observational studies. The present study sought to confirm this finding in a large multicenter cohort of patients treated with targeted temperature management at 33°C and explore the response to targeted temperature management targeting 36°C. Design: Post hoc analysis of a prospective randomized study. Setting: Thirty-six ICUs in 10 countries. Patients: We studied 447 (targeted temperature management = 33°C) and 430 (targeted temperature management = 36°C) comatose out-of-hospital cardiac arrest patients with available heart rate data, randomly assigned in the targeted temperature management trial from 2010 to 2013. Interventions: Targeted temperature management at 33°C and 36°C. Measurements and Main Results: Endpoints were 180-day mortality and unfavorable neurologic function (cerebral performance category 3–5). Patients were stratified by target temperature and minimum heart rate during targeted temperature management (< 50, 50–59, and ? 60 beats/min [reference]) at 12, 20, and 28 hours after randomization. Heart rates less than 50 beats/min and 50–59 beats/min were recorded in 132 (30%) and 131 (29%) of the 33°C group, respectively. Crude 180-day mortality increased with increasing minimum heart rate (< 50 beats/min = 32%, 50–59 beats/min = 43%, and ? 60 beats/min = 60%; plog-rank < 0.0001). Bradycardia less than 50 beats/min was independently associated with lower 180-day mortality (hazard ratioadjusted = 0.50 [0.34–0.74; p < 0.001]) and lower odds of unfavorable neurologic outcome (odds ratioadjusted = 0.38 [ 0.21–0.68; p < 0.01]) in models adjusting for potential confounders including age, initial rhythm, time to return of spontaneous circulation, and lactate at admission. Similar, albeit less strong, independent associations of lower heart rates and favorable outcome were found in patients treated with targeted temperature management at 36°C. Conclusions: This study confirms an independent association of bradycardia and lower mortality and favorable neurologic outcome in a large cohort of comatose out-of-hospital cardiac arrest patients treated by targeted temperature management at 33°C. Bradycardia during targeted temperature management at 33°C may thus be a novel, early marker of favorable outcome.

Objective: Administrative data are used for research, quality improvement, and health policy in severe sepsis. However, there is not a sepsis-specific tool applicable to administrative data with which to adjust for illness severity. Our objective was to develop, internally validate, and externally validate a severe sepsis mortality prediction model and associated mortality prediction score. Design: Retrospective cohort study using 2012 administrative data from five U.S. states. Three cohorts of patients with severe sepsis were created: 1) International Classification of Diseases, 9th Revision, Clinical Modification codes for severe sepsis/septic shock, 2) Martin approach, and 3) Angus approach. The model was developed and internally validated in International Classification of Diseases, 9th Revision, Clinical Modification, cohort and externally validated in other cohorts. Integer point values for each predictor variable were generated to create a sepsis severity score. Setting: Acute care, nonfederal hospitals in New York, Maryland, Florida, Michigan, and Washington. Subjects: Patients in one of three severe sepsis cohorts: 1) explicitly coded (n = 108,448), 2) Martin cohort (n = 139,094), and 3) Angus cohort (n = 523,637) Interventions: None. Measurements and Main Results: Maximum likelihood estimation logistic regression to develop a predictive model for in-hospital mortality. Model calibration and discrimination assessed via Hosmer-Lemeshow goodness-of-fit and C-statistics, respectively. Primary cohort subset into risk deciles and observed versus predicted mortality plotted. Goodness-of-fit demonstrated p value of more than 0.05 for each cohort demonstrating sound calibration. C-statistic ranged from low of 0.709 (sepsis severity score) to high of 0.838 (Angus cohort), suggesting good to excellent model discrimination. Comparison of observed versus expected mortality was robust although accuracy decreased in highest risk decile. Conclusions: Our sepsis severity model and score is a tool that provides reliable risk adjustment for administrative data.