What is the association between deviations from the Parkland Formul for fluid resuscitation and in-hospital mortality in adult burn patients?
- This paper uses data from the German Burn Registry, covering patients from Austria, Germany, and Switzerland.
- Aigner A, Schiefer JL, Reinshagen K, Najem S, Vasileiadis V, Königs I. Too much or too little? Fluid resuscitation in the first 24 h after severe burns: Evaluating the Parkland formula – A retrospective analysis of adult burn patients in Austria, Germany, and Switzerland 2015–2022. Burns. 2025;51(4):107397. doi:10.1016/j.burns.2025.107397
Summary of the Whole Text
- Fluid resuscitation is crucial in acute burn therapy due to the significant fluid shifts following severe burns.
- The Parkland formula is a common method for calculating initial fluid requirements.
- However, its strict adherence varies in practice, and both under- and over-resuscitation carry risks.
- This study aimed to assess how deviations from the Parkland formula, either positive (over-infusion) or negative (under-infusion), affect in-hospital mortality.
- The study found that while deviations are common, over-infusion (positive deviation) is a significant risk factor for increased in-hospital mortality, even after adjusting for various patient characteristics.
- Under-infusion (negative deviation) did not significantly impact mortality.
- The study also explored how patient characteristics like inhalation trauma (IHT), total body surface area (TBSA) affected, and Body Mass Index (BMI) modify this association.
- It concludes that while the Parkland formula remains a useful tool, avoiding over-infusion is critical, and individualized fluid management with close monitoring is essential.
Methods
This was a retrospective analysis using data from the German Burn Registry from January 1, 2015, to December 31, 2022.
- Study Design: Retrospective analysis of registry data.
- Patient Selection: Data from 2235 adult burn patients (age ≥ 16 years) were included.
- Exclusion Criteria: Patients with blistering skin diseases, hospital stays less than 24 hours or missing duration, and those with zero fluid substitution in the first 24 hours.
- Complete-case analysis: Further excluded 491 individuals with missing data for in-hospital mortality, fluid volume, or BMI, resulting in a dataset of 1744 for these specific analyses.
- Data Collection: Anonymized patient data were retrospectively collected from over 40 burn centers in Germany, Austria, and Switzerland.
- Ethical Approval: The study was approved by the ethics committee of the Medical Faculty of the Ruhr University Bochum (No. 21-7184, 18.08.2022).
Population of the Study
The study population consisted of 2235 adult burn patients (age ≥ 16 years) from Germany, Austria, and Switzerland, whose data were recorded in the German Burn Registry between January 1, 2015, and December 31, 2022.
- Overall Cohort Characteristics (n=2229):
- Age: Median 49.0 years (IQR 33.0, 63.0).
- Sex: 24.5% Female, 75.5% Male.
- BMI: Median 26.0 (IQR 23.5, 28.9).
- Inhalation trauma: 29.2%.
- Total Body Surface Area affected: Median 25.0% (IQR 18.0, 37.0).
- In-hospital mortality: 444 patients (19.9%).
- Comparison of patients who died vs. survived (n=444 vs n=1785)
- Age: Patients who died were older (median 65.0 vs 45.0 years).
- Sex: Higher proportion of females among those who died (34.0% vs 22.1%).
- BMI: Similar median BMI (26.3 vs 25.9).
- Sepsis: More frequent in those who died (47.2% vs 17.7%).
- Pneumonia: More frequent in those who died (37.1% vs 20.8%).
- TBSA affected: Nearly twice as high in those who died (40.0% vs 23.0%).
- Inhalation trauma: More frequent in those who died (54.2% vs 23.0%).
Statistical Methods
| Statistical Method | Purpose in Study |
| Descriptive Statistics | To present patient characteristics, including relative and absolute frequencies for categorical variables, and median with interquartile range (IQR) for continuous variables. This was used to summarize the patient cohort and compare characteristics between patients who survived and those who died. |
| Boxplots | To graphically display the relative deviation from the Parkland formula, stratified by in-hospital mortality and the direction of deviation (positive or negative). |
| Mixed Logistic Regression with Random Intercept | To model the association between deviation from Parkland and in-hospital mortality, accounting for potential center effects by including a random intercept per center. This method is appropriate for clustered data (patients within different burn centers) to avoid biased estimates. The model included the logarithms to the base of 2 of the relative extent of the deviation, the direction of the deviation, and their interaction. |
| Logistic Regression (simplified) | For more complex models involving three-way interaction terms, the mixed models did not converge. Given the low intraclass correlation (0.5%) and similar effect measures from the original mixed model, simplified regular logistic regression models were used. This allowed for the evaluation of interactions between Parkland deviation and patient-level characteristics (IHT, TBSA, BMI, age). |
| Multiple Imputation | As a sensitivity analysis to address missing data in outcome variables or covariates. Missing values were imputed 50 times (m=50) to assess the robustness of the results. |
| Odds Ratios and 95% Confidence Intervals | To quantify the strength and precision of the association between the deviation from Parkland and in-hospital mortality. An odds ratio of 2.22 for positive deviation, for example, indicated that the odds of mortality were 2.2 times higher with a twofold increase in volume compared to Parkland. |
Main Results of the Study
- Deviation from Parkland Formula is Common: Deviations from the Parkland formula were very frequent in both positive (over-infusion) and negative (under-infusion) directions.
- Negative deviations were more common, but over-infusion was more prevalent among deceased patients.
- Association with Mortality (Unadjusted): Without accounting for patient-level characteristics, both positive and negative deviations from the Parkland formula were associated with a higher probability of in-hospital mortality. The tendency was stronger for positive deviations.
- For example, administering twice the Parkland volume increased mortality probability from ~17.7% to ~32.4%.
- Association with Mortality (Adjusted): After adjusting for sex, age, BMI, inhalation trauma (IHT), third-degree burn, and TBSA, only positive deviations from the Parkland formula remained a significant risk factor for increased in-hospital mortality.
- A two-fold positive deviation increased the probability of mortality from about 4% to 12%. The odds of in-hospital mortality increased three-fold with a two-fold positive deviation. Negative deviation, however, showed a diminished effect and was no longer relevantly associated with mortality in the adjusted model.
- Modification by Patient-Level Characteristics:
- Inhalation Trauma (IHT): The detrimental effect of over-infusion was less pronounced in patients with IHT compared to those without.
- Total Body Surface Area (TBSA): A higher TBSA (e.g., 40% vs 20%) reduced the detrimental effect of deviation from Parkland. Patients with larger injuries tolerated greater positive deviations better.
- Body Mass Index (BMI): In more obese patients (e.g., BMI 35 vs 25), positive deviations were better tolerated. In patients with lower BMI, a greater increase in mortality was observed if the calculated infusion volume was significantly exceeded.
- Age: Patient age did not significantly modify the association between Parkland deviation and in-hospital mortality.
Main Conclusions of the Study and How They Relate to What is Known in the Field
- The study concludes that the Parkland formula remains a useful guide for initial fluid calculations in severe burns, but deviations are common, with over-infusion occurring more frequently, especially among those who die.
- The key finding is that exceeding the Parkland formula’s calculated amount significantly increases mortality, whereas under-infusion does not have a significant impact on mortality.
- This reinforces the idea that “less seems to be better than more” in fluid resuscitation.
- This aligns with a growing awareness in burn care that intravenous fluids should be considered a medication with potential side effects, and over-resuscitation can lead to complications like acute respiratory distress syndrome, compartment syndrome, and multi-organ dysfunction.
- The study’s finding that positive deviations increase mortality is consistent with previous literature highlighting the “fluid creep” phenomenon and its adverse effects. The observed overestimation of TBSA, as a potential reason for over-infusion, is also supported by prior meta-analyses.
- The study also provides novel insights into how patient characteristics such as IHT, TBSA, and BMI modify the impact of fluid deviation.
- The observation that patients with IHT, higher TBSA, and higher BMI may tolerate over-infusion somewhat better suggests that individualized fluid management based on patient-specific factors is crucial.
- This supports the current trend towards individualized early goal-directed therapy using advanced hemodynamic monitoring, rather than strict adherence to formulas.
- The study acknowledges that while urine output is a common monitoring tool, the tendency to increase fluids more readily in response to low urine output than to decrease them with sufficient output can contribute to fluid overload.
The limitations acknowledged by the authors, such as the observational nature and lack of detailed physiological data (e.g., arterial pressure, continuous urine output) and information on specific resuscitation protocols at each center, are important considerations for interpreting the findings and suggest areas for future research.
Questions and Doubts
- Why is the Parkland formula still widely used if deviations are common and over-infusion is detrimental? The study highlights the prevalence of deviations and the negative impact of over-infusion. This raises questions about the formula’s practical utility as a strict guideline versus a starting point, and whether current clinical education adequately emphasizes the risks of over-resuscitation.
- Given that negative deviations (under-infusion) did not significantly affect mortality in the adjusted model, does this imply that a more conservative fluid strategy is generally safer? The conclusion “Less seems to be better than more” suggests this, but the study doesn’t explicitly define an optimal “less” amount. What are the clinical implications for adjusting fluid targets downwards?
- The study mentions that overestimation of TBSA contributes to over-infusion. What strategies or tools could be implemented to improve the accuracy of TBSA estimation in acute burn settings? This is a practical issue that directly impacts fluid calculations.
- How does the lack of information on specific resuscitation protocols and real-time hemodynamic monitoring data (e.g., arterial pressure, CVP) in the registry limit the conclusions about optimal fluid management? These unmeasured variables could be significant confounders or effect modifiers.
- The study suggests that patients with IHT, higher TBSA, and higher BMI tolerate over-infusion “somewhat better.” What are the physiological mechanisms behind this differential tolerance, and how can these insights be translated into personalized fluid management protocols? Further research into these specific patient populations could refine resuscitation guidelines.
- The study is retrospective. Could unmeasured confounding factors, such as the severity of comorbidities or variations in clinical judgment, influence the observed associations between fluid deviation and mortality?
- While “opioid creep” was considered negligible due to esketamine use in participating centers, could other medications or clinical practices not captured in the registry contribute to fluid overload?
- The study’s findings are based on data from specific European countries. How generalizable are these results to other burn care systems globally, especially where different resuscitation protocols or patient demographics might prevail?
Authors
- Annette Aigner
- Jennifer Lynn Schiefer
- Konrad Reinshagen
- Safiullah Najem
- The German Burn Registry
- Vasileios Vasileiadis
- Ingo Königs