Setting the Standard: Using the ABA Burn Registry to Benchmark Risk Adjusted Mortality.

Reports of single center experience and studies of larger databases have identified several predictors of burn center mortality, including age, burn size, and inhalation injury. None of these analyses has been broad enough to allow benchmarking across burn centers. The purpose of this study was to derive a reliable, risk-adjusted, statistical model of mortality based on real-life experience at many burn centers in the U.S. We used the American Burn Association 2020 Full Burn Research Dataset, from the Burn Center Quality Platform (BCQP) to identify 130,729 subjects from July 2015 through June 2020 across 103 unique burn centers. We selected 22 predictor variables, from over 50 recorded in the dataset, based on completeness (at least 75% complete required) and clinical significance. We used gradient-boosted regression, a form of machine learning, to predict mortality and compared this to traditional logistic regression. Model performance was evaluated with AUC and PR curves. The CatBoost model achieved a test AUC of 0.980 with an average precision of 0.800. The logistic regression produced an AUC of 0.951 with an average precision of 0.664. While AUC, the measure most reported in the literature, is high for both models, the CatBoost model is markedly more sensitive, leading to a substantial improvement in precision. Using BCQP data, we can predict burn mortality allowing comparison across burn centers participating in BCQP.

Using a National Burn Registry to Develop a Model for Risk-Adjusted Length of Stay Benchmarking.

Length of stay (LOS) is a frequently reported outcome after a burn injury. LOS benchmarking will benefit individual burn centers as a way to measure their performance and set expectations for patients. We sought to create a nationwide, risk-adjusted model to allow for LOS benchmarking based on the data from a national burn registry. Using data from the American Burn Association's Burn Care Quality Platform, we queried admissions from 7/2015 to 6/2020 and identified 130,729 records reported by 103 centers. Using 22 predictor variables, comparisons of unpenalized linear regression and Gradient boosted (CatBoost) regressor models were performed by measuring the R2 and concordance correlation coefficient on the application of the model to the test dataset. The CatBoost model applied to the bootstrapped versions of the entire dataset was used to calculate O/E ratios for individual burn centers. Analyses were run on 3 cohorts: all patients, 10-20% TBSA, >20% TBSA. The CatBoost model outperformed the linear regression model with a test R2 of 0.67 and CCC of 0.81 compared with the linear model with R2=0.50, CCC=0.68. The CatBoost was also less biased for higher and lower LOS durations. Gradient-boosted regression models provided greater model performance than traditional regression analysis. Using national burn data, we can predict LOS across contributing burn centers while accounting for patient and center characteristics, producing more meaningful O/E ratios. These models provide a risk-adjusted LOS benchmarking using a robust data source, the first of its kind, for burn centers.

Specific Patterns of Postoperative Temperature Elevations Predict Blood Infection in Pediatric Burn Patients.

Using readily available temperature data, we seek to propose a scoring criteria that can facilitate accurate and immediate prediction of blood infection. The standard in diagnosing blood infection is a positive blood culture result that may take up to 3 days to process, requiring providers to make a prediction about which febrile patient is actually bacteremic. This prediction is difficult in burned children as systemic inflammation can cause fever in the absence of infection. An ability to make this prediction more accurate using readily available information would be useful. A retrospective chart review was performed for 28 pediatric patients, with a burn size 20% or greater, admitted to the burn unit between 2010 and 2014. All children had blood cultures drawn. They were divided into either infection (positive blood cultures) or control (negative blood cultures) groups. Median temperature and mean number of temperature elevations were compared between the two groups. We evaluated the predictive accuracy of using temperature elevation, pattern, and timing to predict blood infection. A significant difference was seen in the mean number of temperature elevations above 39°C. This was significant for each time stage, especially in the 0- to 24-hour post-surgery period. We found the most predictive accuracy in the 0- to 12-, 12- to 38-, and 12- to 48-hour time periods. We found a strong association between mean number of fever spikes above 39°C and blood infection, especially 12 to 24 hours after surgery. This readily available data can be useful to clinicians as they access children with burns.

The safety and efficacy of parenteral nutrition among pediatric patients with burn injuries.

OBJECTIVE: Although enteral nutrition is the ideal mode of nutritional support following burn injury, it is often interrupted during episodes of severe sepsis and hemodynamic instability, leading to significant energy and protein deficits. Parenteral nutrition is not commonly used in burn centers due to concerns that it will lead to hyperglycemia, infection, and increased mortality. However, parenteral nutrition is often utilized in our burn unit when goal rate enteral nutrition is not feasible.To determine the safety and efficacy of a standardized protein-sparing parenteral nutrition protocol in which glucose infusion is limited to 5-7 mg/kg/hour. DESIGN: Retrospective observational study. SETTING: Pediatric burn hospital. PATIENTS: A retrospective medical record review of all children admitted to our hospital with burns ≥ 30% total body surface area was conducted. Only patients admitted within one week of injury and who survived > 24 hours after admission were included in this study. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Of the 105 patients who met the inclusion criteria, 96 (91%) received parenteral nutrition or a combination of parenteral nutrition and enteral nutrition at some point during their care. Nine patients received only enteral nutrition. Demographic data were similar between groups. Protein intake was significantly higher in the parenteral nutrition group. Incidence of catheter-related blood infections did not differ between groups. Use of parenteral nutrition was not associated with blood or respiratory infections. Overall mortality rate was low (4%), as most patients (96%) achieved wound closure and were discharged home. CONCLUSIONS: Judicious use of parenteral nutrition is a safe and effective means of nutritional support when goal enteral nutrition cannot be achieved. A hypocaloric, high-nitrogen parenteral nutrition solution can reduce energy and protein deficits while minimizing complications commonly associated with parenteral nutrition usage.

Selection of appropriate empiric gram-negative coverage in a multinational pediatric burn hospital.

The choice of appropriate empiric antimicrobial therapy for burn patients with suspected multidrug-resistant organisms remains a challenge. Burn patients transferred from outside the United States seem to be at particularly high risk. Given this perceived risk of multidrug resistance among our international patient population, we set out to determine which empiric antimicrobial therapy should be used at admission. A retrospective analysis was conducted of all burn patients admitted to a pediatric burn specialty hospital between 2006 and 2010. Patients with burns >10% TBSA were included. Demographics, burn data, and routine/nonroutine culture data were collected. Of the 385 total patients, 133 (34.5%) were international. International patients had significantly larger burns (39.73 vs 22.80% TBSA; P < .001) and more inhalational injuries (27.1 vs 16.3%; P < .03) than their U.S. counterparts. International patients presented with a higher incidence of infection in general (66.9 vs 2%; P < .001) as well as a higher prevalence of infection caused by multidrug-resistant bacteria (51.2 vs 1%; P < .001) and pan-multidrug-resistant bacteria (13.5 vs 1.1%; P < .001). Bacterial resistance was not related to the length of time after burn injury or to a delay in transfer. In conclusion, multidrug-resistant and pan-resistant organisms seem to be more prevalent among the international pediatric burn population when compared with the U.S. pediatric burn population. Given the relatively high incidence of pan-resistant gram-negative organisms among international transfers, colistin seems to be a reasonable choice for empiric antimicrobial coverage for presumed infections.

A survey of invasive catheter practices in U.S. burn centers.

Burn-specific guidelines for optimal catheter rotation, catheter type, insertion methods, and catheter site care do not exist, and practices vary widely from one burn unit to another. The purpose of this study was to define current practices and identify areas of practice variation for future clinical investigation. An online survey was sent to the directors of 123 U.S. burn centers. The survey consisted of 23 questions related to specific practices in placement and maintenance of central venous catheters (CVCs), arterial catheters, and peripherally inserted central catheters (PICCs). The overall response rate was 36%; response rate from verified centers was 52%. Geographic representation was wide. CVC and arterial catheter replacement varied from every 3 days (24% of sites) to only for overt infection (24% of sites); 23% of sites did not use the femoral position for CVC placement. Nearly 60% of units used some kind of antiseptic catheter. Physicians inserted the majority of catheters, and 22% of sites used nonphysicians for at least some insertions. Ultrasound was routinely used by less than 50% of units. A wide variety of post-insertion dressing protocols were followed. PICCs were used in some critically injured patients in 37% of units; the majority of these users did not rotate PICCs. Thus, it can be surmised that wide practice variation exists among burn centers with regard to insertion and maintenance of invasive catheters. Areas with particular variability that would be appropriate targets of clinical investigation are line rotation protocols, catheter site care protocols, and use of PICCs in acute burns.

Incidence of catheter-associated bloodstream infection after introduction of minocycline and rifampin antimicrobial-coated catheters in a pediatric burn population.

The Centers for Disease Control and Prevention guidelines for prevention of intravascular catheter-related infections suggest that antimicrobial-coated catheters can decrease the risk of developing catheter-related bloodstream infection in a variety of adult patient populations. There are limited data on their efficacy in the pediatric population, particularly among children with burn injuries. A study was conducted at Shriners Hospitals for Children®, Boston, to determine whether minocycline/rifampin (MR)-coated catheters could decrease the incidence of catheter-associated bloodstream infection (CABSI) in a pediatric burn population. A historical control group included all patients with double- or triple-lumen catheters inserted in the 18-month period from January 2006 to June 2007. The study group included all patients with MR antimicrobial double- or triple-lumen catheters inserted in the subsequent 18-month period, July 2007 to December 2008. Data collected included name, age, date of burn/injury, date of admission, percent TBSA area burn injury or other diagnosis, catheter site (subclavian, internal jugular, or femoral), method of insertion (new percutaneous stick or guidewire), type of catheter (double or triple lumen), date inserted, duration of catheter placement (days), and positive blood cultures recovered while the central venous catheter was in place. CABSI was defined using the Centers for Disease Control and Prevention definition of laboratory-confirmed bloodstream infection. There were a total of 66 patients with 252 catheters (1780 catheter days) in the control group and 75 patients with 263 catheters (1633 catheter days) in the study group. Age, percent burn injury, catheter site, and method of insertion were not statistically different between the two groups. The percentage of infected catheters and the rate of infection were significantly different for the two groups, with the MR antimicrobial catheters only half as likely to become infected. In a subset of these patients with catheters in place for more than 4 days, the percentage of infected catheters and rate of infection were also significantly different with results similar to those in the entire group. MR antimicrobial-coated catheters significantly reduced the incidence of CABSI in this pediatric burn population compared with noncoated catheters.

The selenium status of pediatric patients with burn injuries.

BACKGROUND: Dietary selenium (Se) requirements during critical illness are not well known. The objective of this study was to assess the longitudinal Se status of pediatric patients with burns. METHODS: Twenty patients admitted to our hospital with burns exceeding 10% of their total body surface area were studied longitudinally during the first 8 weeks of admission or until 95% wound closure was achieved. Dietary Se intake was calculated daily, and plasma and urine samples were collected weekly for analyses of plasma Se, urinary Se, and glutathione peroxidase activity. RESULTS: Patients included in this study were individuals with an average age of 6.5 years ± 5.3 years and with burn injury of a mean total body surface area of 42% ± 21%. Dietary Se intake throughout the study (mean = 60 µg/d ± 39 µg/d) was consistent with established standards for healthy children and did not change throughout the study. Plasma Se (mean = 1.08 µmol/L ± 0.34 µmol/L) and plasma glutathione peroxidase (mean = 3.2 U/g protein ± 1.42 U/g protein) were below reported normal values for healthy American children. Mean urinary Se excretion (65.9 µg/L ± 50 µg/L) exceed dietary Se intake. Plasma Se was inversely related to incidence of total infection (p = 0.04). CONCLUSIONS: Results from this study indicate that Se status is depressed among pediatric patients with burns and that recommended Se intake for healthy children is likely insufficient for this population. Further studies are necessary to elucidate the amount of dietary Se required to maximize Se stores among pediatric patients with burn injuries.

Malnutrition among pediatric burn patients: A consequence of delayed admissions.

Early definitive burn treatment is assumed to improve prognosis, in part because open wounds rapidly stimulate muscle catabolism and systemic inflammation. This study describes the incidence and management of injury associated malnutrition among pediatric burn patients transferred for definitive care 21-166 days following burn injury. Medical records of patients admitted to our hospital between January 2003 and January 2009, at least 3 weeks after burn injury, were retrospectively reviewed. Only children with an initial total body surface area (TBSA) burn of ≥20% were included in this study (n=36). Patients were classified as acutely well nourished or malnourished by the medical team. All patients were admitted with chronic open wounds (31±16% TBSA). Sixty-one percent (n=22) of patients were diagnosed with malnutrition. These patients had a significantly longer delay to transfer (26-166 days) than well nourished patients who transferred at 21-138 days (p<0.05). Average protein (2.8±0.18g/kg), and kilocalorie (1.6±0.1% basal metabolic rate) provision did not differ between groups. Incidence of infection was not different between well nourished and malnourished patients. Malnutrition occurs frequently among pediatric burn patients with delayed admissions. Adequate surgical care, infection control, and nutrition are required for wound healing.

Intravenous colistin for the treatment of multi-drug resistant, gram-negative infection in the pediatric burn population.

The rising incidence of multi-drug resistant (MDR) gram-negative infections in the intensive care unit (ICU) continues to challenge clinicians and has resulted in reemergence of the glycopeptide antibiotic colistin. Over the past 11 years, 14 patients at a tertiary pediatric burn center were treated with colistin for gram-negative infections resistant to all tested antibiotics. This study reviews the safety of such treatment and the outcome for this cohort of patients. All hospitalized patients treated with intravenous colistin between 1990 and 2005 were identified. A retrospective chart review was performed for each patient. Demographic data, along with information regarding the type and severity of injury, were collected. Data with respect to microbiology, renal status, and neurological events were also noted. Over an 11-year period, we identified 14 children infected with pan-resistant gram-negative organisms requiring 16 courses of colistin. Two children (14.3%) developed significant rises in serum creatinine concentration; however, no child required renal replacement therapy or developed neurologic complications attributable to colistin. Favorable response rate was 78.6% (11/14), and overall mortality was 14.3% (2/14); both deaths were attributed to sepsis. In our experience with 14 children treated with intravenous colistin, two developed a significant elevation in serum creatinine concentration during the course of therapy and neurotoxicity was not reported. Colistin should be dispensed with great caution, but it appears to have an acceptable safety profile in children and may be used in select cases of infection with highly resistant gram-negative organisms.

Mechanical and infectious complications of central venous cannulation in children: lessons learned from a 10-year experience placing more than 1000 catheters.

We sought to better describe the expected incidence of mechanical and infectious complications associated with central venous cannulation of critically ill children. We undertook a retrospective analysis of a prospective data collection of 1056 consecutive percutaneous central venous catheters inserted under the supervision of an experienced surgeon. There were 245 (23%) subclavian (SC), 118 (11%) internal jugular (IJ), and 693 (66%) femoral (F) catheters placed in 289 children with an average age of 6.4 +/- 5.1 years (range, 4 weeks to 18 years) admitted to a burn intensive care unit. Catheter sepsis occurred in 7.4% of SC, 7.6% of IJ, and 4.9% of F catheters (NS, P = .25), for an overall sepsis rate of 5.8%. The number of catheter lumens did not impact infection rate. Infection rates increased in catheters left in situ more than 10 days, increasing to 37.5% at 14 days. Acute mechanical complications occurred in three insertions (0.3%), including two (0.8%) SC, zero (0%) IJ, and one (0.1%) F catheters (NS, P = .20). All three were arterial cannulations that were recognized and treated successfully without surgery. There were no pneumothoraces, vascular lacerations, acute thromboses, or catheter emboli. There were six (0.6%) cases of deep venous thrombosis that occurred in cannulated sites: one (0.4%) SC, two (1.6.%) IJ, and three (0.4%) F sites (NS, P = .23). Patient age did not influence complication rates. A total of 239 (23%) of the CVCs were placed in infants less than 24 months; 273 (26%) 2 to 5 years, 259 (25%) 6 to 10 years, and 285 (27%) >10 to 18 years. Catheter sepsis occurred in 6.7%, 5.9%, 6.2%, and 4.6%, respectively (NS, P = .75). There was no difference in rates of infection or mechanical complication between younger and older children. When closely supervised by an experienced surgeon, a low rate of infection (5.8%), acute mechanical complication (0.3%), and deep venous thrombosis (0.6%) accompanies central venous cannulation of critically ill children.

Computer equipment used in patient care within a multihospital system: recommendations for cleaning and disinfection.

Computer hardware has been implicated as a potential reservoir for infectious agents. Leaders of a 22-hospital system, which spans North America and serves pediatric patients with orthopedic or severe burns, sought to develop recommendations for the cleaning and disinfection of computer hardware within its myriad patient care venues. A task force comprising representatives from infection control, medical affairs, information services, and outcomes management departments was formed. Following a review of the literature and of procedures within the 22 hospitals, criteria for cleaning and disinfection were established and recommendations made. The recommendations are consistent with general environmental infection control cleaning and disinfection guidelines, yet flexible enough to be applicable to the different locales, different computer and cleaning products available, and different patient populations served within this large hospital system.

Effectiveness of bacteria-controlled nursing units in preventing cross-colonization with resistant bacteria in severely burned children.

Bacteria-controlled nursing units (BCNUs) are laminar air-flow patient isolation units. The rate of cross-colonization with resistant organisms in 66 critically ill pediatric burn patients with massive open wounds and ventilators housed in BCNUs during 5 years was examined and found to be extremely low (3.2 cases per 1,000 patient-days).

Young age is not a predictor of mortality in burns.

OBJECTIVE: Conventional wisdom and recently published reports suggest that children <48 months of age have a higher mortality rate after burns than older children and adolescents with similar injuries and that young age is a predictor of mortality. This study was done to validate or refute this impression. DESIGN: Retrospective review. SETTING: Regional pediatric burn center. PATIENTS: All children (n = 1223) managed over a recent 8-yr interval (1991-1998) for acute thermal burns. INTERVENTIONS: The survival rate of children <48 months of age was compared with the survival rate of children >/=48 months of age. MEASUREMENTS AND MAIN RESULTS: Of the 1112 children with burns covering <30% of the body surface, 721 (65%) were <48 months of age. After exclusion of one child who was brain dead and became a solid organ donor, there were no deaths in this burn size group. There were 111 children admitted with burns covering >/=30% of the body surface: 47 (42%) with an average age of 2.0 yrs (range, 4 wks to 3 yrs and 11 months) were <48 months of age, and 64 (58%) with an average age of 10.9 yrs (range, 4 yrs to 17 yrs) were >/=48 months of age. There were no clinically important differences between the two groups in burn size (51.9% +/- 18.1% [range, 30%-90%] vs. 56.9% +/- 19.4% [range, 30%-97%]; p =.18) or need for mechanical ventilatory support (30/47 [63.8%] vs. 44/64 [68.8%]; p =.59). The mortality rate in the young group was 0% (0/47) and 10.9% (7/64) in the older group (p =.04). All children who died had large burns (average burn size, 82.9% +/- 11.5%) with concurrent inhalation injury. CONCLUSION: Young age is not a predictor of mortality in burns.