Potential pediatric intensive care unit demand/capacity mismatch due to novel pH1N1 in Canada.

OBJECTIVE: To investigate the possibility of pediatric intensive care unit shortfalls, using pandemic models for a range of attack rates and durations. The emergence of the swine origin pH1N1 virus has led to concerns about shortfalls in our ability to provide pediatric ventilation and critical care support. DESIGN: Modeling of pediatric intensive care demand based on pH1N1 predictions using simulation techniques. SETTING: Simulation laboratory. PATIENTS: None. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Data collected during the first wave of the pH1N1 in children in Canada were applied to several second wave pandemic models to explore potential pediatric intensive care unit ventilatory demands for Canada and to investigate the impact of vaccination upon these demands. In almost all cases studied, even for relatively low attack rates of 15%, significant pediatric intensive care unit shortages would be expected to occur. Vaccination strategies targeting 50% of the population significantly reduced demand, but shortages may still be expected. Although shortfalls can occur in all provinces, Ontario and British Columbia may experience the greatest supply-demand difference, even at low attack rates. CONCLUSION: Reducing the attack rate among children, whether through vaccination or additional measures, such as social distancing, will be critical to ensure sufficient pediatric intensive care unit capacity for continued pediatric care.

Severe traumatic brain injury in children elevates glial fibrillary acidic protein in cerebrospinal fluid and serum.

OBJECTIVES: 1) To determine the levels of glial fibrillary acidic protein (GFAP) in both cerebrospinal fluid and serum; 2) to determine whether serum GFAP levels correlate with functional outcome; and 3) to determine whether therapeutic hypothermia, as compared with normothermia, alters serum GFAP levels in children with severe traumatic brain injury (TBI). DESIGN: Laboratory-based analyses; postrandomized, controlled trial. SETTING: Four Canadian pediatric intensive care units and a university-affiliated laboratory. PATIENTS: Twenty-seven children, aged 2-17 yrs, with severe TBI (Glasgow Coma Scale score of ≤ 8). INTERVENTIONS: Hypothermia therapy (32.5°C) for 24 hrs with cooling started within 8 hrs of injury and rewarming at a rate of 0.5°C every 2 hrs or normothermia (37.0°C). MEASUREMENTS AND MAIN RESULTS: GFAP was measured in cerebrospinal fluid and serum, using enzyme-linked immunosorbent assay. Levels of GFAP were maximal on day 1 post-TBI, with cerebrospinal fluid GFAP (15.5 ± 6.1 ng/mL) 25-fold higher than serum GFAP (0.6 ± 0.2 ng/mL). Cerebrospinal fluid GFAP normalized by day 7, whereas serum GFAP decreased gradually to reach a steady state by day 10. Serum GFAP measured on day 1 correlated with Pediatric Cerebral Performance Category scores determined at 6 months post-TBI (ρ = 0.527; p = .008) but failed to correlate with the injury scoring on admission, physiologic variables, or indices of injury measured on computerized tomography imaging. The areas under the receiver operating characteristic curves for pediatric intensive care unit day 1 serum GFAP in determining good outcome were 0.80 (pediatric cerebral performance category, 1-2; normal-mild disability) and 0.91 (pediatric cerebral performance category, 1-3; normal-moderate disability). For a serum GFAP cutoff level of 0.6 ng/mL, sensitivity and specificity were 88% to 90% and 43% to 71%, respectively. Serum GFAP levels were similar among children randomized to either therapeutic hypothermia or normothermia. CONCLUSIONS: GFAP was markedly elevated in cerebrospinal fluid and serum in children after severe TBI and serum GFAP measured on pediatric intensive care unit day 1 correlated with functional outcome at 6 months. Hypothermia therapy did not alter serum GFAP levels compared with normothermia after severe TBI in children. Serum GFAP concentration, together with other biomarkers, may have prognostic value after TBI in children.

Characterization of Ebola Virus Risk to Bedside Providers in an Intensive Care Environment.

BACKGROUND: The 2014-2016 Ebola outbreak in West Africa recapitulated that nosocomial spread of Ebola virus could occur and that health care workers were at particular risk including notable cases in Europe and North America. These instances highlighted the need for centers to better prepare for potential Ebola virus cases; including understanding how the virus spreads and which interventions pose the greatest risk. METHODS: We created a fully equipped intensive care unit (ICU), within a Biosafety Level 4 (BSL4) laboratory, and infected multiple sedated non-human primates (NHPs) with Ebola virus. While providing bedside care, we sampled blood, urine, and gastric residuals; as well as buccal, ocular, nasal, rectal, and skin swabs, to assess the risks associated with routine care. We also assessed the physical environment at end-point. RESULTS: Although viral RNA was detectable in blood as early as three days post-infection, it was not detectable in the urine, gastric fluid, or swabs until late-stage disease. While droplet spread and fomite contamination were present on a few of the surfaces that were routinely touched while providing care in the ICU for the infected animal, these may have been abrogated through good routine hygiene practices. CONCLUSIONS: Overall this study has helped further our understanding of which procedures may pose the highest risk to healthcare providers and provides temporal evidence of this over the clinical course of disease.

Critical illness in children with influenza A/pH1N1 2009 infection in Canada.

OBJECTIVE: To describe characteristics, treatment, and outcomes of critically ill children with influenza A/pandemic influenza A virus (pH1N1) infection in Canada. DESIGN: An observational study of critically ill children with influenza A/pH1N1 infection in pediatric intensive care units (PICUs). SETTING: Nine Canadian PICUs. PATIENTS: A total of 57 patients admitted to PICUs between April 16, 2009 and August 15, 2009. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Characteristics of critically ill children with influenza A/pH1N1 infection were recorded. Confirmed intensive care unit cases were compared with a national surveillance database containing all hospitalized pediatric patients with influenza A/pH1N1 infection. Risk factors were assessed with a Cox proportional hazard model. The PICU cohort and national surveillance data were compared, using chi-square tests. Fifty-seven children were admitted to the PICU for community-acquired influenza A/pH1N1 infection. One or more chronic comorbid illnesses were observed in 70.2% of patients, and 24.6% of patients were aboriginal. Mechanical ventilation was used in 68% of children, 20 children (35.1%) had acute lung injury on the first day of admission, and the median duration of ventilation was 6 days (range, 0-67 days). The PICU mortality rate was 7% (4 of 57 patients). When compared with nonintensive care unit hospitalized children, PICU children were more likely to have a chronic medical condition (relative risk, 1.73); aboriginal ethnicity was not a risk factor of intensive care unit admission. CONCLUSIONS: During the first outbreak of influenza A/pH1N1 infection, when the population was naïve to this novel virus, severe illness was common among children with underlying chronic conditions and aboriginal children. Influenza A/pH1N1-related critical illness in children was associated with severe hypoxemic respiratory failure and prolonged mechanical ventilation. However, this higher rate and severity of respiratory illness did not result in an increased mortality when compared with seasonal influenza.

Extracorporeal lung support for patients who had severe respiratory failure secondary to influenza A (H1N1) 2009 infection in Canada.

BACKGROUND: From March to July 2009, influenza A (H1N1) 2009 (H1N1-2009) virus emerged as a major cause of respiratory failure that required mechanical ventilation. A small proportion of patients who had this condition developed severe respiratory failure that was unresponsive to conventional therapeutic interventions. In this report, we describe characteristics, treatment, and outcomes of critically ill patients in Canada who had H1N1-2009 infection and were treated with extracorporeal lung support (ECLS). METHODS: We report the findings of a case series of six patients supported with ECLS who were included in a cohort study of critically ill patients with confirmed H1N1-2009 infection. The patients were treated in Canadian adult and pediatric intensive care units (ICUs) from April 16, 2009 to August 12, 2009. We describe the nested sample treated with ECLS and compare it with the larger sample. RESULTS: During the study period, 168 patients in Canada were admitted to ICUs for severe respiratory failure due to confirmed H1N1-2009 infection. Due to profound hypoxemia unresponsive to conventional therapeutic interventions, six (3.6%) of these patients were treated with ECLS in four ICUs. Four patients were treated with veno-venous pump-driven extracorporeal membrane oxygenation (vv-ECMO), and two patients were treated with pumpless lung assist (NovaLung iLA). The mean duration of support was 15 days. Four of the six patients survived (66.6%), one of the surviving patients was supported with iLA and the other three surviving patients were supported with ECMO. The two deaths were due to multiorgan failure, which occurred while the patients were on ECLS. INTERPRETATION: Extracorporeal lung support may be an effective treatment for patients who have H1N1-2009 infection and refractory hypoxemia. Survival of these patients treated with ECLS is similar to that reported for patients who have acute respiratory distress syndrome of other etiologies and are treated with ECMO.

Correction to: Impact of intensive care unit supportive care on the physiology of Ebola virus disease in a universally lethal non-human primate model.

Please note that four authors (Logan Banadyga, Alixandra Albietz, Brad Pickering, and Gary Wong) have been erroneously omitted from the author list in the published original article [1].

Impact of intensive care unit supportive care on the physiology of Ebola virus disease in a universally lethal non-human primate model.

BACKGROUND: There are currently limited data for the use of specific antiviral therapies for the treatment of Ebola virus disease (EVD). While there is anecdotal evidence that supportive care may be effective, there is a paucity of direct experimental data to demonstrate a role for supportive care in EVD. We studied the impact of ICU-level supportive care interventions including fluid resuscitation, vasoactive medications, blood transfusion, hydrocortisone, and ventilator support on the pathophysiology of EVD in rhesus macaques infected with a universally lethal dose of Ebola virus strain Makona C07. METHODS: Four NHPs were infected with a universally lethal dose Ebola virus strain Makona, in accordance with the gold standard lethal Ebola NHP challenge model. Following infection, the following therapeutic interventions were employed: continuous bedside supportive care, ventilator support, judicious fluid resuscitation, vasoactive medications, blood transfusion, and hydrocortisone as needed to treat cardiovascular compromise. A range of physiological parameters were continuously monitored to gage any response to the interventions. RESULTS: All four NHPs developed EVD and demonstrated a similar clinical course. All animals reached a terminal endpoint, which occurred at an average time of 166.5 ± 14.8 h post-infection. Fluid administration may have temporarily blunted a rise in lactate, but the effect was short lived. Vasoactive medications resulted in short-lived improvements in mean arterial pressure. Blood transfusion and hydrocortisone did not appear to have a significant positive impact on the course of the disease. CONCLUSIONS: The model employed for this study is reflective of an intramuscular infection in humans (e.g., needle stick) and is highly lethal to NHPs. Using this model, we found that the animals developed progressive severe organ dysfunction and profound shock preceding death. While the overall impact of supportive care on the observed pathophysiology was limited, we did observe some time-dependent positive responses. Since this model is highly lethal, it does not reflect the full spectrum of human EVD. Our findings support the need for continued development of animal models that replicate the spectrum of human disease as well as ongoing development of anti-Ebola therapies to complement supportive care.

Delivering Prolonged Intensive Care to a Non-human Primate: A High Fidelity Animal Model of Critical Illness.

Critical care needs have been rising in recent decades as populations age and comorbidities increase. Sepsis-related admissions to critical care contribute up to 50% of volume and septic shock carries a 35-54% fatality rate. Improvements in sepsis-related care and mortality would have a significant impact of a resource-intensive area of health care delivery. Unfortunately, research has been hampered by the lack of an animal model that replicates the complex care provided to humans in an intensive care unit (ICU). We developed a protocol to provide full ICU type supportive care to Rhesus macaques. This included mechanical ventilation, continuous sedation, fluid and electrolyte management and vasopressor support in response to Ebolavirus-induced septic shock. The animals accurately recapitulated human responses to a full range of ICU interventions (e.g. fluid resuscitation). This model can overcome current animal model limitations by accurately emulating the complexity of ICU care and thereby provide a platform for testing new interventions in critical care and sepsis without placing patients at risk.

Apneic Oxygenation during Rapid Sequence Intubation in Critically Ill Children.

This prospective case series documented hypoxemia and potential complications associated with apneic oxygenation in critically ill pediatric patients during rapid sequence intubation. Forty-four patients received apneic oxygenation via nasal cannula at rates of 5, 10, and 15 L/min for ages <4, 4 to 12, and 12 to 18 years, respectively. Pre- and postintubation attempt mean Spo 2 were 98.9 ± 2.95 and 90.7 ± 1.95%, respectively. Postintubation Spo 2 < 80% were significantly less with one intubation attempt, compared with multiple attempts (p < 0.001). No serious complications were noted. Apneic oxygenation was well tolerated in critically ill children.

Building a novel pediatric intensive care iPad handover aid for residents.

There is no consensus on how effective patient handover in the pediatric intensive care unit should occur. Complex patients with rapidly moving clinical trajectories are difficult to summarize and comprehend. We aimed to redesign our current handover instrument to encourage higher-level cognitive interactions, questioning and understanding for pediatric residents. Through an iterative process with five pediatric intensive care unit intensivists, the current hardcopy handover tool was reviewed and compared to other formats published in medical and non-medical domains. Several handover formats on the iPad were created and proposed, but continual feedback from intensive care unit physicians led to a selection, and further revisions were made. A mock handover with completed data fields allowed pediatric residents to provide the final feedback in both a semi-structured group review and a survey with responses on a 1-5 Likert scale (1 = strongly disagree; 2 = disagree; 3 = neither agree or disagree; 4 = agree; 5 = strongly agree). A novel two-axis data grid combining the 'systems' and 'problem list' approach was developed, with read back prompts, and off the shelf applications such as "burn notices". Residents found ease with the organizational format (4.0 ± 0.67 Likert scale response) and the transition to the iPad device (4.3 ± 0.67). Improving physician handovers may be achieved efficiently and economically through physician led iterative processes. Pediatric residents were at ease when the novel handover was combined with newer iPad technologies and applications.

Air transported pediatric rescue extracorporeal membrane oxygenation: a single institutional review.

BACKGROUND: Pediatric extracorporeal membrane oxygenation (ECMO) programs are sophisticated endeavors usually found only in high-volume cardiac surgical programs. Worldwide, many cardiology programs do not have on-site pediatric cardiac surgery expertise. Our single-center experience shows that an organized multidisciplinary rescue-ECMO program, in collaboration with an accepting facility, can achieve survival rates comparable to modern era on-site ECMO. METHODS: A retrospective review was conducted of all patients initiated on rescue-ECMO from 2004 to 2009 in a single academic pediatric hospital without a pediatric cardiac surgery program. All aspects of ECMO were formalized using Failure Mode Effects Analysis. RESULTS: Eight patients were initially cannulated for ECMO at our institution. Six were subsequently transported by air to the receiving facility 1,305 km away. Extracorporeal membrane oxygenation was initiated in 0.2% of our Pediatric Intensive Care Unit admissions and in 0.52% of all our pediatric cardiac patients. Mean age was 4.0 years (7 weeks to 15 years). Indications for ECMO initiations were cardiogenic shock (n = 5) and acute respiratory distress syndrome (n = 3). Six had veno-arterial- and two had veno-veno ECMO. Two patients were not transported (one death and one weaned locally). Six patients were successfully transported within 2 to 24 hours, with a survival to hospital discharge rate of 67% (four of six). Median total time on ECMO was 5.5 days. Complication rate was 50% (4/8). CONCLUSIONS: Our rescue-ECMO survival results were comparable to that of current published results from established pediatric ECMO programs. Air transport of ECMO patients can be performed safely using an organized multidisciplinary team approach.