February 2022 - Burn Trauma

February 2022
EAST Monthly Literature Review

"Keeping You Up-to-Date with Current Literature"
Brought to you by the EAST Manuscript and Literature Review Committee

This issue was prepared by EAST Burn Surgery Committee Members Arpana Jain, MD and Shawn Tejiram, MD.

Thank you to Haemonetics for supporting the EAST Monthly Literature Review.

In This Issue: Burn Trauma

Scroll down to see summaries of these articles

Article 1 reviewed by Arpana Jain, MD
Impact of Opioid-Minimizing Pain Protocols After Burn Injury. Donthula D, Conner CR, Truong VTT, Green C, Jiang C, Wandling MW, Komak S, Huzar TF, Adams SD, Freet DJ, Wainwright DJ, Wade CE, Kao LS, Harvin JA. J Burn Care Res. 2021 Nov 24;42(6):1146-1151. 

Article 2 reviewed by Arpana Jain, MD
Application of Critical Care Scores in Severely Burned Patients. Dastagir K, Krezdorn N, Bingoel AS, Mett TR, Radtke C, Jokuszies A, Vogt PM. J Burn Care Res. 2021 Nov 24;42(6):1176-1180.

Article 3 reviewed by Shawn Tejiram, MD
Impact of Multimodal Analgesia in Critically Ill Burn Patients. Young SR, Duby JJ, Louie EL, Palmieri TL.
 Burn Care Res. 2021 Sep 30;42(5):986-990.

Article 4 reviewed by Shawn Tejiram, MD
The utility of arterial blood gas parameters and chest radiography in predicting appropriate intubations in burn patients with suspected inhalation injury - A retrospective cohort study. Chotalia M, Pirrone C, Ali M, Mullhi R, Torlinska B, Mangham T, England K, Torlinski T. Burns. 2021 Dec;47(8):1793-1801.

Article 1
Impact of Opioid-Minimizing Pain Protocols After Burn Injury. Donthula D, Conner CR, Truong VTT, Green C, Jiang C, Wandling MW, Komak S, Huzar TF, Adams SD, Freet DJ, Wainwright DJ, Wade CE, Kao LS, Harvin JA. J Burn Care Res. 2021 Nov 24;42(6):1146-1151.

Takeaway: Use of protocolized pain management leads to decrease in the overall use of opioids during hospitalization and at the time of discharge without negative impact on acute pain assessment scores.
Summary: Pain management after burn injury remains a challenging task for burn care providers and an elusive state for burn injured patients. Recently, this field has seen several advances with use of non-opioid based medication and non-pharmacologic adjuncts. This paper examines the deployment of a protocolized pain management at a busy urban burn center. This is a retrospective cohort study, where the comparison groups are before and after the implementation of pain management protocols. The study site already used non-opioid agents before implementing the protocol, however the usage was not standardized. The protocol was developed in 2019 and they compared opioid usage in Pre-group (n=495) and Post- group (n=174). The study reports total morphine milligram equivalents (MME) during hospital stay as the primary outcome. They also report MME/day and pain scores as other outcomes. The paper includes two supplements to give detailed insights into their protocol. It spells out guidelines for both baseline and procedural pain. Authors use Bayesian linear models to estimate treatment effects. In the Post-group, there was less exposure to MME 551 MMEs [449, 687] vs Pre-group725 MMEs [642, 823], 98% posterior probability. Also, less use of MME per day, Post-group 53 MMEs/day [47, 61] vs Pre-group 67 MMEs/day [62, 72], 99% posterior probability. They did not find difference in pain scores, Post-group 0.33 [0.30, 0.36] vs Pre-group 0.34 [0.32, 0.36], 62% posterior probability. Incidentally, the study finds shorter hospital length of stay, Post-group (RR 0.61, 95% CrI 0.51–0.74, posterior probability 100%), but not in intensive care unit or ventilator days. The Post-group also had a lower risk of being prescribed opioids at discharge (RR 0.77, 95% CrI 0.68–0.86, posterior probability 100%).

This study presents a good argument in favor of adopting protocolized care for pain management, however there may be a limit on generalizing this result to other centers. The study design lends itself to unaccounted confounders that may distort the results before and after the protocol came into place. The nature of intervention (protocol-based care) is not very amenable for blinded observations. Lastly, the authors themselves note that use of pain scale and averaging the scores may not give a granular assessment of actual success or failure of protocol. In future it will be interesting to see the effect of pain protocols of patient and staff satisfaction in the burn unit.

Article 2
Application of Critical Care Scores in Severely Burned Patients. Dastagir K, Krezdorn N, Bingoel AS, Mett TR, Radtke C, Jokuszies A, Vogt PM. J Burn Care Res. 2021 Nov 24;42(6):1176-1180.

Takeaway: Diagnosis of sepsis in severely burned patient should be nuanced. For instance, in patient with burn injury less than 15% TBSA, a combination of SIRS, SOFA and procalcitonin demonstrates correlation with sepsis, however in patients with injury TBSA over 55%, SIRS and SOFA scores lose their predictive power.

Summary: Sepsis is a major cause of morbidity and mortality after burn injury. The diagnosis of sepsis depends on clinical suspicion and positive blood cultures. Sepsis-3 definition developed by SCCM and European society of Intensive care medicine employs SOFA score to predict outcome in suspected sepsis. However, due to intense inflammatory response in patient with severe burn injury, the usual SIRS and SOFA score may not be applicable. This study tries to tie the reliability of SIRS, SOFA and procalcitonin in the diagnosis of sepsis with progressively severe burn injury defined as high TBSA. This study provides retrospective analysis of 651 patients from a single burn center in Germany from year 2014 to 2018. Authors stratified the study population into six cohorts based on TBSA, with ceiling being over 55%TBSA and excluding less the 10%TBSA. They use ROC curve and AUC to assess diagnostic accuracy of SIRS, SOFA and procalcitonin. They defined sepsis as presence of clinical suspicion and positive blood cultures. Both SIRS and SOFA show a drop in specificity with increase in TBSA, especially for patients over 55%TBSA (AUC of SIRS 0.527, AUC of SOFA 0.522). Similar trend was noted for procalcitonin, however, performed better than SIRS or SOFA for patients with injury over 55%TBSA (AUC 0.746). The study also found that, SIRS, SOFA and procalcitonin enjoyed higher AUCs, 0.727, 0.754 and 0.876 respectively. This led the authors to conclude that perhaps procalcitonin has a value in predicting sepsis in severely injured burn patients. At the same time, clinicians must use caution when interpreting SIRS criteria or SOFA scores in presence of severe burn injury. 

Article 3
Impact of Multimodal Analgesia in Critically Ill Burn Patients. Young SR, Duby JJ, Louie EL, Palmieri TL. 
J Burn Care Res. 2021 Sep 30;42(5):986-990.

Burn injuries cause significant pain and remain a challenge due to the background pain of injury itself, breakthrough pain, and periprocedural pain related to operative intervention and dressing changes. As a result, opioids remain a mainstay of treatment for burn pain. While there are potential advantages to opioid use in this setting, there are risks as well, including respiratory depression, oversedation, constipation, itching, and dependence. Multimodal analgesia has been introduced as an alternative method of pain control that utilizes both opioid and nonopioid medications with different mechanisms to reduce opioid needs. This study examined the impact of multimodal therapy on postoperative pain control in a burn intensive care unit.

The authors performed a six-year retrospective, observational, cohort study of critically ill burn patients at a tertiary medical center. Patients with burns greater than or equal to 10% TBSA and an intensive care unit length of stay greater than 7 days were eligible for inclusion. Patients were excluded if they were less than 18 years of age, had a hospital or ICU length of stay less than 7 days, did not undergo surgery, or received an opioid infusion greater than 48 hours. The 48-hour limit was chosen due to the difficulty in determining if the indication was for pain control or sedation. Multimodal non-opioid analgesia was ordered at the discretion of the managing burn team and included acetaminophen, nonsteroidal anti-inflammatory drugs (celecoxib, ibuprofen, ketorolac, etc.), and neuropathic agents (gabapentin). Opioid use was recorded as oral morphine equivalents (OME) for comparison purposes. Primary endpoints examined cumulative OME from post-operative days 1 through 5. Secondary endpoints included OME doses stratified by TBSA, average pain score on post-operative day 5, ICU discharge, ICU LOS, hospital LOS, ventilatory days, and naloxone and methylnaltrexone use. There were 98 patients in the nonmultimodal group and 97 in the multimodal group. No differences were noted in demographics, prior to admission opioid use, TBSA, number of surgeries, type of excision and grafting, surgery performed, size of area grafted, or size of donor site between groups. Mean cumulative opioid dose was significantly lower in the multimodal group (1028 ± 651 mg vs 1438 ± 1085 mg, P = .0031). The multimodal group also used significantly lower mean daily OME doses on all post-operative days compared to the nonmultimodal group. When stratified by burn size, patients with burns between 10% and 20% TBSA in the multimodal group had reduced mean cumulative OME doses on postoperative days 1 through 5, but this was not statistically significant (940 mg vs 1282 mg, P = .058). Patients with greater than 20% burns had a larger reduction in mean opioid equivalents in the multimodal group (1106 vs 1594 mg, P = .009). All patients in the multimodal group received at least one multimodal analgesic medication, 45% received two agents, and 17% received more than two agents. Acetaminophen was the most commonly prescribed agent (81%), followed by gabapentin (71%) and NSAIDs (21%). There was no difference in mean pain scores on postoperative day 5 (6.2 ± 2.2 vs 5.5 ± 2.3, P = .07) or at intensive care unit discharge (4.7 ± 2.4 vs 4.7 ± 2.8, P = .99). No differences were noted in ICU LOS, hospital LOS, mean ventilator days, or naloxone/methynaltrexone use between both groups.

In summary, the use of multimodal analgesia significantly reduced cumulative opioid equivalent dose without compromising pain control and was sustained throughout the 5-day period. The effect of mitigating opioid use on other clinical endpoints, such as length of stay and mechanical ventilation, is minimal. Limitations of this study include those inherent to retrospective study design and may not be adequately powered. Additionally, administration of opioids was also performed at the discretion of the bedside nurse and may have affected the amount of OME received. The amount of opioid exposure traditionally in the burn population highlights the likelihood of opioid tolerance and dependence compared to nonburned surgical patients. Inadequate pain control in the burn population occurs for a number of reasons, including variability in burn pain intensity and altered pharmacokinetics and pharmacodynamics in response to hemodynamic changes to postburn injury, making individualized pain control strategies paramount. Previous studies examining multimodal analgesia focused on the impact of single multimodal agents rather than a bundled approach, making this a novel study in this context.

Article 4
The utility of arterial blood gas parameters and chest radiography in predicting appropriate intubations in burn patients with suspected inhalation injury - A retrospective cohort study. Chotalia M, Pirrone C, Ali M, Mullhi R, Torlinska B, Mangham T, England K, Torlinski T. Burns. 2021 Dec;47(8):1793-1801.

Inhalation injury is detected in 20% of patients that undergo hospitalization in the UK due to burn injuries and is associated with increased morbidity and mortality. Indications for endotracheal intubation include upper and lower airway thermal injury and carbon monoxide poisoning. Traditional practice has been a low threshold for early intubation, but new evidence suggests this may lead to unnecessary intubations that is associated with its own morbidity and mortality. Additional methods to improve prediction of appropriate intubations in this population is currently being sought. Intubation criteria, such as that by the American Burn Association, has been demonstrated sensitive and specific for detecting appropriate intubations and inhalation injury presence. However, there is a paucity of literature examining the impact of additional adjuncts, such as chest imaging and arterial blood gases, on this prediction. This study evaluates the utility of arterial blood gas (ABG) parameters and chest radiography in predicting intubation need in patients with burn injuries with suspected inhalation injury.

The authors conducted a retrospective, observational cohort study in patients with moderate to severe burn injuries and suspected inhalation injury to a single center Burn Intensive Care Unit over a three-year period from April 2016 to July 2019. ABG parameters and chest radiograph opacification severity were compared with whether a patient received an appropriate intubation or had inhalation injury. Appropriate intubation was defined as any patient that remained intubated over a 48-hour period due to thermal or chemical injury. Patients extubated within 48 hours were deemed to have an inappropriate intubation. Diagnosis of inhalation injury was determined by direct laryngoscopy and fiberoptic bronchoscopy. Patients who died within 24 hours of admission, underwent endotracheal intubation for reasons other than inhalation injury, or had missing clinical information (i.e. ABG, chest radiography) were excluded. The 2011 ABA intubation criteria were used as opposed to the updated 2018 guidelines (which include >40% TBSA and dysphagia) due to the timing of the authors’ data collection period. Eighty-nine patients were included. The majority (84%; n = 75) were intubated, of which 81% (n = 61) were determined to have received appropriate intubations. Median TBSA was 25% (11 – 51%). No patients died within 48 hours of hospital admission, but an actual 30-day mortality of 21.4% (n = 19) was reported. The majority of patients had a PaO2 >10 kPa (N = 65; 73%) and an acidosis (pH < 7.35; N = 62; 70%) on the initial ABG. A mixed respiratory and metabolic acidosis was most common (N = 40; 65%). Patients that received prolonged intubation had a lower pH, base deficit, and P/F ratio, and higher PaCO2 and hematocrit values than patients that did not receive prolonged intubation. No significant differences were observed in PaO2, COHb and lactate values between patients who did or did not receive prolonged intubation. AUROC for blood gas parameters for appropriate intubation need and inhalation injury showed pH with the highest AUROC at 0.88, followed by P/F ratio (AUROC = 0.81). PaO2 had poor predictive value (AUROC = 0.58). PaCO2 and base deficit had fair ability in predicting both intubation need (AUROC PaCO2 = 0.79; base deficit = 0.79) and inhalation injury (AUROC PaCO2 = 0.76, base deficit = 0.75). Other blood gas parameters (lactate, hematocrit, COHb), were in general poor predictors of appropriate intubation or inhalation injury. Chest radiograph opacification severity score had minimal inter-rater variation, (Cohen’s-Kappa = 0.85), but poor utility in predicting appropriate intubation and inhalation injury (AUROC = 0.69). Addition of pH and P/F ratio to the ABA intubation criteria improved the sensitivity (0.84 to 0.97, p = 0.013) and negative predictive value (0.73 to 0.92, p = 0.002), respectively, in detecting appropriate intubation need and inhalation injury presence.

In summary, in patients suspected inhalation injury, pH and P/F ratio < 40 were good predictors for appropriate intubation. Incorporating these parameters into the ABA criteria improved their sensitivity and negative predictive value. Chest radiograph opacification were poor predictors of intubation need and inhalation injury owing to the acute setting with which smoke inhalation may not be evident on admission chest radiograph. Acidosis was predominantly mixed respiratory and metabolic in nature owing to the amalgamation of PaCO2 and base deficit representing lung and cutaneous burn severity. Most patients had a normal or elevated PaO2 as they were often administrated high FiO2 due to suspected carbon monoxide poisoning. Examining the expression of the physiological response (PaO2) relative to the dose of therapy administered (FiO2) by the P/F ratio improved prediction of necessary intubations and pathology detection.

Have you checked out EAST's Landmark Paper Resource?

 This Literature Review is being brought to you by the EAST Manuscript and Literature Review Committee. Have a suggestion for a review or an additional comment on articles reviewed? Please email litreview@east.orgPrevious issues available on the EAST website.