Introduction
Background
When appropriately applied to massive extremity hemorrhages, tourniquets have become an irrefutable life-saving intervention. Reference Butler1–Reference Kragh, Dubick and Aden4 For this reason, tourniquets have become the standard of care for civilian and military prehospital treatment of massive extremity hemorrhages. 5–10 A properly applied tourniquet stops all blood flow distal of the device via circumferential pressure. Multiple retrospective studies have been conducted examining the military’s use of tourniquets in the last 17 years, demonstrating minimal complications related to tourniquet usage. Rare complications of tourniquet use in military populations include neuropathy, decreased limb function, amputation, and impaired kidney function. Reference Butler1–Reference Kotwal, Montgomery and Kotwal3,Reference Kragh, O’Neill and Walters11–Reference Lakstein, Blumenfeld and Sokolov17 Studies within military populations may not translate well to civilian populations due to differences in age, fitness, and baseline health. This disparity may apply to the most foundational of military research.
There have been civilian case reports of acute kidney injury (AKI) after tourniquet usage during surgical cases and experimental studies in which mice experience renal pathology. Reference Orsi, Brunelli and Zorzi18–Reference Vold and Weiss23 Additionally, more recent studies have demonstrated increased rates of post-traumatic AKIs in rats with obesity and increased oxidative stress, both of which can be found in the civilian population. Reference Mittwede, Xiang, Lu, Clemmer and Hester24–Reference Xiang, Thompson, Clemmer, Mittwede, Khan and Hester28 Important health-related characteristics of the military population found within the previously mentioned studies that demonstrated minimal to no complications related to tourniquet usage may not reflect the population in which prehospital tourniquets are applied within the civilian sectors. Reference Hoerster, Lehavot, Simpson, McFall, Reiber and Nelson29 When prehospital tourniquets are applied to a civilian population with increased underlying risk factors such as obesity and medical illness, the complication rate could exceed studies in military populations. Specifically, the rates of associated AKI could be increased due to pre-existing renal stressors such as hypertension and diabetes mellitus in a population already prone to AKI due to injury-related hypovolemia and the need for necessary interventions that may adversely affect the kidneys. Reference Hsu, Ordonez, Chertow, Fan, McCulloch and Go30–Reference Bargman and Skorecki32 If pre-existing conditions increased the risk of AKI after the use of a prehospital tourniquet, this would be of importance to clinical decision making as AKIs increase the likelihood of death up to four-times the baseline in hospitalized and critically injured patients. Reference Bihorac, Delano and Schold33,Reference Ulger, Pehlivanlar Kucuk and Kucuk34
Two recently published retrospective chart reviews by Inaba, et al and Scerbo, et al documented incidental AKIs in trauma patients who received civilian prehospital tourniquets at a frequency of less than 3.2%. Reference Inaba, Siboni and Resnick35,Reference Scerbo, Mumm and Gates36 After aggregating the incidence of AKI found by Inaba, et al (two out of 87; 2.3%) and Scerbo, et al (three out of 105; 2.9%), there was an aggregated incidence of five AKIs out of 192 (2.6%) civilian trauma patients. Reference Inaba, Siboni and Resnick35,Reference Scerbo, Mumm and Gates36 Inaba, et al and Scerbo, et al did not examine patient characteristics that may increase tourniquet-associated complications, nor did they imply causation between tourniquet usage and AKI. Reference Inaba, Siboni and Resnick35,Reference Scerbo, Mumm and Gates36 Yet, they provide an associated incidence of AKI with civilian tourniquet usage that surpassed previous results in the military population published by Kragh, et al (seven out of 427; 1.6%). Reference Kragh, O’Neill and Walters11
Harrois, et al provided significant insight into the multifactorial pathophysiology of AKIs in trauma patients. Reference Harrois, Libert and Duranteau37 Iodinated contrast, inflammation, hemorrhagic shock, rhabdomyolysis, abdominal compartment syndrome, and fluid excess all increased the chances of an AKI. Reference Harrois, Libert and Duranteau37 Tourniquet usage could increase rhabdomyolysis and inflammation, which Harrois, et al further delineated as occurring at significantly higher levels during ischemia-reperfusion events. Reference Harrois, Libert and Duranteau37 In theory, tourniquet-related inflammation and muscular apoptosis with subsequent myoglobinemia could tip the scales into an AKI, even if neither event would cause an AKI when experienced in isolation. This additive and deleterious effect could be further increased in patients with renally-injurious comorbidities such as diabetes mellitus or hypertension.
This cross-sectional study aimed to provide descriptive data concerning AKI incidence and relative risk amongst civilian patients treated with at least one prehospital tourniquet. This study is not meant to affect the usage of tourniquets for massive extremity hemorrhages, nor imply causation, but to investigate a potential increased association demonstrated by Inaba, et al and Scerbo, et al that warrants further investigation. Reference Inaba, Siboni and Resnick35,Reference Scerbo, Mumm and Gates36 The study hypothesized that renal comorbidities, such as diabetes mellitus and hypertension, would increase the incidence of associated AKIs in patients treated with a prehospital tourniquet.
Methods
Expedited study approval was obtained through the University of Texas Health Science Center San Antonio Office of Institutional Research, IRB (San Antonio, Texas USA) HSC20190008H; the University Hospital Research Office (San Antonio, Texas USA), 20190008H; and the US Army Medical Research and Development Command Human Research Protection Office (Fort Detrick, Maryland USA), 13293001, before initiation of the study. This study was conducted in support of the “Remote Trauma Outcomes Research Network (RemTORN): A Civilian Research Model for Translation to Military Pre-Hospital Care.” As this was a retrospective chart review, human research participant consent was waived by the review boards.
Study Design
This was a single-center, observational, retrospective, cross-sectional pilot study involving chart review of patients presenting to University Hospital (San Antonio, Texas USA), a tertiary Level 1 trauma center, after having at least one tourniquet applied by prehospital personnel. This study’s desired outcome was to identify the incidence of an AKI after using at least one prehospital tourniquet and to identify potential predisposing factors that may be associated with an increased incidence and risk of AKI.
Selection of Participants
Patients with prehospital tourniquet usage transported to University Hospital were identified within the multi-Emergency Medical Service (EMS) agency electronic record system maintained by the Southwest Texas Regional Advisory Council’s (STRAC) EMS Electronic Patient Care Record. Patient-related data were then used to match patients within the University Hospital’s electronic medical record. All patient recorded as having a prehospital tourniquet and transported by EMS were captured within this study. Patients that were received with generic “John Doe” identifiers were found within the hospital records using the transport date, time, gender, and injuries as identifiers. Inclusion criteria required the application of at least one tourniquet in the prehospital environment and an age of at least 18 years old. Participants identified as prisoners, pregnant, on hemodialysis, deceased before EMS transport, deceased during the initial resuscitation, and those in which a thoracotomy was performed were excluded from the study.
Measurement and Outcomes
Patient data were extracted from the prehospital and hospital electronic medical records. Inter-rater reliability and precision were achieved with an independent review of 25% of charts. The STRAC prehospital database contains patients transported by EMS in the South Texas region to the hospital used in this study. Using the data gathered at pre-designated time intervals and an approved surrogate value for baseline creatinine when actual baseline creatinine was not available, it was determined whether the patient experienced an AKI after applying a prehospital tourniquet. Consistent with other previously published studies, the surrogate baseline creatinine level was obtained from the Estimated Baseline SCr Based on Modification of Diet in Renal Disease (MDRD) study. Reference Bellomo, Ronco, Kellum, Mehta and Palevsky38–Reference Kellum, Lameire and Aspelin40 For this study, AKI was defined using the validated guidelines provided by the Kidney Disease: Improving Global Outcomes (KDIGO). Reference Ulger, Pehlivanlar Kucuk and Kucuk34,Reference Fujii, Uchino, Takinami and Bellomo41,Reference Khwaja42 Per KDIGO guidelines, an AKI is defined by an increase in serum creatinine by ≥ 0.3mg/dL (≥ 26.5micromol/L) within 48 hours or an increase in serum creatinine to ≥ 1.5-times baseline, which is known or presumed to have occurred within the prior seven days, or urine volume < 0.5mL/kg/hour for six hours. Individual components of the KDIGO guidelines were extracted from the patient records. In addition to the continuous data, a dichotomous AKI indicator (yes or no) was determined using the KDIGO guidelines. Blank or missing information related to the KGIDO guidelines or other data extraction points were left blank and considered null towards qualifying whether an AKI occurred.
Statistical Analysis
Statistical analysis was conducted using IBM SPSS version 26 (IBM; Armonk, New York USA). Analyzing the presence of an AKI involved the use of a two-tailed Pearson Chi-Square test. A z-test comparison of column proportions and Bonferroni adjusted P values was used to prevent Type I error while analyzing the participants’ comorbid conditions, such as diabetes and hypertension. Participant ages amongst those with AKI and those without was compared with independent t-tests to control for the confounder of age. The data were further analyzed using odds ratios and relative risk with 95% confidence intervals (CI).
Results
From March 2009 through June 2019, 313 records were identified within the prehospital STRAC database that included the use of at least one tourniquet. After applying the inclusion and exclusion criteria, 255 participants were included in this study, with characteristics shown in Table 1.
Abbreviations: AKI, acute kidney injury; TQ, tourniquet.
Among the 29 participants with known diabetes mellitus, five (17.2%) experienced an AKI. Participants with a history of diabetes mellitus had a significantly higher incidence of AKI as compared to those without diabetes (12 out of 226; 5.3%; X 2[1, N=255, 5.881]; adjusted P = .031; Table 2). The analysis demonstrated no difference in the incidence of AKI without diabetes mellitus (12 out of 226) as compared to the previously published and aggregated civilian data published by Inaba, et al and Scerbo, et al (five out of 192; 2.6%; X 2[1, N=418, 1.316]; P = .251). Reference Inaba, Siboni and Resnick35,Reference Scerbo, Mumm and Gates36 The characteristics of the participants who experienced an AKI in association with the use of a prehospital tourniquet are shown in Table 3. Those who experienced an AKI had a higher incidence of open fractures (four out of 17) compared to the overall group (14 out of 255; X 2[1, N=272, 8.545]; P = .003). No other differences were found within the categorical characteristics of those with and without an AKI. The mean ages of those with (44.41; SD = 19.11) and without diabetes mellitus (64.60; SD = 15.40) that had an associated AKI were not statistically different (t [15] = -2.084; P = .055) with equal variances due to non-significant Levene’s Test of homogeneity. Forty-five participants had a documented history of hypertension and twenty participants were documented to have both diabetes mellitus and hypertension. There was no difference in the incidence of AKI in those with hypertension or diabetes mellitus and hypertension (X 2[1, N=255, 3.903]; P = .090 and X 2[1, N=255, 2.422]; P = .138).
Abbreviation: AKI, acute kidney injury.
Abbreviations: AKI, acute kidney injury; TQ, tourniquet.
Odds ratio analysis revealed an increased odds of AKI with a known history of diabetes mellitus in association with the use of a prehospital tourniquet (OR = 3.715; 95% CI, 1.206 - 11.448). Participants with known diabetes had an increased relative risk of AKI associated with the use of a prehospital tourniquet (RR = 3.247; 95% CI, 1.232 - 8.557; Figure 1).
Analysis revealed that 17 of the 255 (6.67%) included participants experienced an AKI associated with using a prehospital tourniquet. An analysis using the aggregated civilian incidence published by Inaba, et al and Scerbo, et al (five out of 192; 2.6%) demonstrated no statistically higher incidence of AKI amongst this study’s participants in association with the use of a prehospital tourniquet (X 2[1, N=447, 3.044]; P = .081). Reference Inaba, Siboni and Resnick35,Reference Scerbo, Mumm and Gates36 Yet, the incidence of AKI amongst the included participants was statistically higher than what was previously reported in the military population by Kragh, et al (seven out of 427; 1.6%; X 2[1, N=682, 10.450]; P = .001) in association with the use of a prehospital tourniquet. Reference Kragh, O’Neill and Walters11
Discussion
Within the included participants, those with a known history of diabetes mellitus experienced a higher incidence of AKI, in association with the use of a prehospital tourniquet, compared to those without the disorder. Participants with a known history of diabetes mellitus also had an increased incidence of AKI, in association with using a prehospital tourniquet, compared to previously published incidence data. Reference Inaba, Siboni and Resnick35,Reference Scerbo, Mumm and Gates36 Additionally, participants with a known history of diabetes mellitus had an increased odds ratio and relative risk of AKI associated with the use of a prehospital tourniquet. There is no causality to be inferred from this descriptive, retrospective study. However, the increased incidence, odds, and risk of AKI found in this study’s participants with a known history of diabetes mellitus in association with the use of a prehospital tourniquet does raise the question as to whether the well-documented cumulative and deleterious effects of diabetes mellitus on the renal system can increase the incidence of AKI after a traumatic injury requiring a prehospital tourniquet. Reference Hsu, Ordonez, Chertow, Fan, McCulloch and Go30,Reference Bargman and Skorecki32 The baseline renal glomerular changes associated with diabetes mellitus could increase AKI incidence when a person’s kidneys are faced with potentially deleterious trauma-related factors. Reference Hsu, Ordonez, Chertow, Fan, McCulloch and Go30–Reference Bargman and Skorecki32,Reference Harrois, Libert and Duranteau37
The overall incidence of AKI amongst the included participants was statistically higher than the previously published data by Kragh, et al. Reference Kragh, O’Neill and Walters11 This significant difference could be due to the previously described deleterious effects of diabetes on baseline renal health and the increased incidence of diabetes amongst this study’s participants compared to the military. Reference Hsu, Ordonez, Chertow, Fan, McCulloch and Go30–Reference Bargman and Skorecki32 The incidence of diabetes mellitus within this study’s participants was 29 out of 255 (11.4%). The incidence of diabetes mellitus in the deployed military population, in which the previously published data were collected by Kragh, et al, could be considered non-existent due to its presence preventing deployment. 43 It could therefore be hypothesized that the increased presence of diabetes mellitus could account for the increased incidence of AKIs found in civilians as compared to military patients, in association with the use of a prehospital tourniquet. There was no difference in the incidence of AKI between this study’s participants and the incidence reported by Inaba, et al and Scerbo, at al. Reference Inaba, Siboni and Resnick35,Reference Scerbo, Mumm and Gates36 The incidence of diabetes mellitus amongst the participants resulted in Inaba’s, et al and Scerbo’s, et al studies were not reported. Therefore, similar comparisons concerning the incidence of diabetes mellitus could not be calculated, nor could its association with their participants developing an AKI be evaluated to refute or support this study’s findings. Reference Inaba, Siboni and Resnick35,Reference Scerbo, Mumm and Gates36
This pilot study provides novel results into prehospital tourniquet use and associated incidence of AKIs from those previously published using military populations. It cannot be stressed enough that this study is not meant to affect prehospital tourniquet usage, as it is a life-saving intervention. In this vain, the new data found within this pilot study warrant further investigation. Additional multi-center studies would support or refute this single-center pilot study and increase the availability of clinically relevant information to receiving emergency staff to recalibrate their suspicion that a patient with diabetes and a prehospital tourniquet could develop an AKI and increase their mortality.
Limitations
There are several limitations centered around the retrospective nature of records, such as the accuracy and completeness of the databases and the ability to control for potential confounders. The STRAC prehospital database contains all patients transported by EMS to the hospital studied. The prehospital patient care record requests that responding EMS professionals document either the presence of a tourniquet prior to their arrival or if EMS placed a tourniquet during treatment and transport to the receiving University Hospital. It is possible that under the circumstances of severe trauma, a patient’s entire medical history was not solicited and, therefore, it was not complete. It is also possible that patients with a prehospital tourniquet were transported by personal vehicle to the receiving hospital, which would have prevented this study from capturing their medical data.
This study sought to find whether a patient experienced an AKI in association with using a prehospital tourniquet within 48 hours of the injury. It is possible that the medical record did not reflect a transient AKI in the participants’ records because the appropriate laboratory study was not conducted during the period of renal insult. Some of the included records were also found to be missing accurate measurements of urine output. While one would assume that oliguria or anuria would have been reflected in the participants’ records, this cannot be guaranteed. If these limitations did affect the resultant data, it would have biased the data in favor of fewer incidence of AKIs and the acceptance of the null hypothesis, which was not the case in this study. Likewise, single-center data provides well-known limitations. This study did not factor in hemorrhagic shock’s effect on renal function and the subsequent increased risk of AKI. Yet, the effect of exsanguination and hypovolemic shock would have been theoretically distributed across all the included participants.
Within this study, participants with an open fracture were over-represented amongst those who experienced an AKI compared to the overall participant pool. This difference could indicate that the pathology involved in an open fracture increased participants’ chances of experiencing an AKI, confounding this analysis. While the other characteristics were found to be similar (Tables 2 and 3), there was limited ability to control for other potential confounders.
Conclusion
This study revealed new information regarding AKIs associated with prehospital tourniquets in patients with diabetes mellitus. This area of research warrants further investigation with a large multi-center retrospective study or one of a prospective nature spanning geographic regions to further investigate the potential impact of diabetes mellitus and prehospital tourniquet use on renal function. The proposed study may also be beneficial to investigate whether renally impacting environmental and nutritional stressors can lead to increased incidence and prevalence of AKIs amongst military personnel who receive a prehospital tourniquet.
Conflicts of interest/funding
The authors have no conflicts of interest to declare.
Acknowledgement
The authors would like to acknowledge Lawrence Roakes and his essential data extraction from the STRAC record system, without which this study would not have been possible.