The American Journal of Surgery (2017) 213, 73-79

Clinical Science

Bicarbonate and mannitol treatment for

traumatic rhabdomyolysis revisited
Jamison S. Nielsen, D.O., M.B.A., F.A.C.S.a,
Mitchell Sally, M.D., F.A.C.S.a,b, Richard J. Mullins, M.D., F.A.C.S.a,
Matthew Slater, M.D., F.A.C.S.a, Tahnee Groat, B.A., M.P.H.b,
Xiang Gao, B.S.a, J. Salvador de la Cruz, M.D.a,
Margaret K. M. Ellis, M.D.a, Martin Schreiber, M.D., F.A.C.S.a,
Darren J. Malinoski, M.D., F.A.C.S.a,b,*

a
Department of Surgery, Division of Trauma, Surgical Critical Care and Acute Care Surgery, Oregon
Health & Science University, Portland, OR, USA; bSurgical Critical Care Section, Veterans
Administration Portland Healthcare System, Portland, OR, USA

KEYWORDS: Abstract
Rhabdomyolysis; BACKGROUND: A rhabdomyolysis protocol (RP) with mannitol and bicarbonate to prevent acute
Mannitol; renal dysfunction (ARD, creatinine .2.0 mg/dL) remains controversial.
Creatine kinase; METHODS: Patients with creatine kinase (CK) greater than 2,000 U/L over a 10-year period were
Alkaline diuresis; identified. Shock, Injury Severity Score, massive transfusion, intravenous contrast exposure, and RP
Acute kidney injury use were evaluated. RP was initiated for a CK greater than 10,000 U/L (first half of the study) or greater
than 20,000 U/L (second half). Multivariable analyses were used to identify predictors of ARD and the
independent effect of the RP.
RESULTS: Seventy-seven patients were identified, 24 (31%) developed ARD, and 4 (5%) required
hemodialysis. After controlling for other risk factors, peak CK greater than 10,000 U/L (odds ratio 8.6,
P 5 .016) and failure to implement RP (odds ratio 5.7, P 5 .030) were independent predictors of ARD.
Among patients with CK greater than 10,000, ARD developed in 26% of patients with the RP versus
70% without it (P 5 .008).
CONCLUSION: Reduced ARD was noted with RP. A prospective controlled study is still warranted.
Published by Elsevier Inc.

Controversy remains regarding the use of a forced

The authors declare no conflicts of interest.
The authors are solely responsible for the contents of the article, and
the opinions do not necessarily represent the views of the Centers for
Disease Control and Prevention. The views expressed herein are those of
the authors and do not reflect the official policy of the Department of
the Army, Department of Defense, or the U.S. Government.
* Corresponding author. Tel.: 11-503-220-8262; fax: 11-503-494-
6519.
E-mail address: darren.malinoski@va.gov
Manuscript received December 22, 2015; revised manuscript March 9,
2016
alkaline diuresis with mannitol and bicarbonate for the
treatment of trauma-induced rhabdomyolysis.1–3 Recent
reviews and international consensus statements have not
advocated for such a protocol.3,4 While the results of
numerous preclinical studies led to a widespread practice
of inducing a forced alkaline diuresis to prevent myoglobi-
nuric renal failure, clinical observational studies fail to
consistently demonstrate a protective effect. The objective
of this study is to examine the impact of using a mannitol

0002-9610/$ - see front matter Published by Elsevier Inc.

http://dx.doi.org/10.1016/j.amjsurg.2016.03.017
74
and bicarbonate protocol to deter the development of acute
renal dysfunction (ARD) after traumatic rhabdomyolysis.
The hypothesis in this study was that the use of mannitol
and bicarbonate is associated with preserved renal function.
Methods
Treatment protocol
The trauma service at Oregon Health & Science
University developed and implemented a practice guideline
for the identification and treatment of trauma patients with
severe rhabdomyolysis in 1992. Based on a review of the
literature, a serum creatine kinase (CK) level of 10,000 U/L
was initially chosen as the indicator of severe rhabdomyol-
ysis and the threshold for initiating a forced alkaline
diuresis via the rhabdomyolysis protocol (RP; Fig. 1).
Serum CK was chosen as the marker of rhabdomyolysis
degree due to its speed of measurement, relatively low
cost compared with other biochemical markers, and
frequent use in previous studies.2,5–9 The RP goals were
to induce a brisk flow of urine within 1 hour and to raise
the urine pH to greater than 6.0. The objective of choosing
a specific treatment threshold was to include all patients at
significant risk of developing ARD (defined as a serum
creatinine [Cr] . 2.0 mg/dL), while excluding patients at
such low risk that use of an extensive protocol was unnec-
essary. Among patients who did not meet the threshold CK
level for institution of the RP standard, nonprotocolized
fluid management was provided, typically a urine output
goal of .5 mL/kg/hour.
An interim analysis in 1997 revealed that no patients
with a peak CK level less than 20,000 U/L required dialysis
and only 13% of patients with a peak CK between 10,000
and 20,000 U/L developed ARD. This led to an increase in
the CK threshold for initiating the protocol from 10,000 to
20,000 U/L. Consequently, there were patients with similar
CK levels who did and did not receive RP to include in our
analysis.
Data collection
The Oregon Health & Science University trauma
registry was queried to identify patients with rhabdomyol-
ysis treated at our Level 1 trauma center over a 10-year
period (1993 to 2002). This time span included the 5 years
before and after the increase in the CK treatment threshold
from 10,000 to 20,000 U/L. Patients were selected for
detailed chart review if they had an International Classifi-
cation for Disease-9th Revision code for acute renal failure
(584.9, 985.5), crush injury (925 to 929), rhabdomyolysis
(728.88), myoglobinuria (791.3), or a procedure code for
fasciotomy. Patient medical records were reviewed and
individuals with a CK greater than 2,000 U/L (approxi-
mately 5 times the upper limit of normal, 4 to 397 U/L)
were included in the study. Patients were excluded from the
The American Journal of Surgery, Vol 213, No 1, January 2017
data analysis if they had pre-existing chronic renal
insufficiency or a clearly established cause for developing
renal failure other than rhabdomyolysis.
Data retrieved from the trauma registry for this analysis
include age, gender, mechanism of injury, Injury Severity
Score (ISS), shock on admission (systolic blood pressure ,
90 mm Hg), massive transfusion (.10 units of blood during
the first 24 hours), interhospital transfer status, time of
injury, time of arrival at a medical facility, time to treatment
(time from injury to admission), and mortality. Information
recovered from the medical record included exposure to
intravenous contrast during the first 48 hours of admission,
laboratory values, the use of dialysis, and whether the RP
was employed.
Data analysis
The primary outcome measure was ARD, defined as
having a serum Cr greater than 2.0 mg/dL. The secondary
outcome measure was receiving renal replacement therapy.
Univariate analyses were used to identify variables associ-
ated with ARD. Data with normal distributions were
analyzed with the Pearson chi-square, Fisher’s exact, and
independent-samples t test, when appropriate. The Mann–
Whitney U test was utilized for nonparametric data.
Variables with a P value less than .2 on univariate analysis
were included in a multivariable analysis using binary
logistic regression to determine independent predictors of
ARD with a P value less than .05. Variables that were
inherently related were analyzed using separate models.
Statistical analysis was performed using SPSS 22.0
software (SPSS, Inc., Chicago, IL) and STATA 13.1 soft-
ware (StataCorp, College Station, TX). Our Institutional
Review Board approved this project and a waiver of
consent was granted.
Results
Using the inclusion criteria described, the records of
77 patients were analyzed. A peak CK less than 10,000 U/L
was noted in 21 of these patients, 22 had a peak CK
between 10,000 to 20,000 U/L, and 34 had a peak CK
greater than 20,000 U/L. Most patients were seriously
injured (mean ISS 5 23 6 14), including 35% who
presented in shock, 29% who required a massive trans-
fusion, and 27% who received a fasciotomy for extremity
compartment syndrome (Table 1). Two thirds (66.3%) of
patients had exposure to radiographic contrast agents
(computed tomography scans and/or angiography).
The mean time from admission to peak CK level was
21.7 6 20.7 hours. Overall, 24 patients (31%) developed
ARD, 4 (5%) received dialysis, and 9 (12%) died.
Fig. 2 illustrates the distribution of patients based on
peak CK levels, use of RP, and renal function. Of the 21
patients with a CK less than or equal to 10,000 U/L,
none were treated with the protocol. Forty-six of 56 patients
J.S. Nielsen et al. Bicarbonate/mannitol for rhabdomyolysis
Figure 1 Rhabdomyolysis protocol.
with a CK greater than 10,000 U/L received RP. Among
patients who had a CK greater than 10,000 U/L, a greater
percentage progressed to ARD in the no protocol group
(70% vs 26%, P 5 .008).
When all patients were included in univariate analyses,
age, shock, and peak CK were the only identified risk
factors for the development of ARD with a cutoff P value
less than .2 for further evaluation on multivariate analysis
75
(Table 2). The median age of patients with ARD was
40 years vs 33 years for those patients without ARD
(Mann–Whitney U, P 5 .035). ISS, massive transfusion,
and exposure to intravenous contrast did not meet the
criteria for multivariate analysis.
The results of the multivariate analysis are included in
Table 3. After controlling for all variables with a P value
less than .2 on univariate analysis, only the peak CK level
76 The American Journal of Surgery, Vol 213, No 1, January 2017

Table 1 Demographics of patients Comments

Variables* n 5 77
Due to the inconsistent findings in the literature
Age (years) regarding the effectiveness of a forced alkaline diuresis
Mean 6 SD 39 6 17
with mannitol and bicarbonate in preserving renal function
Median 37
among patients with rhabdomyolysis, we conducted a
Sex
Male 61 (79) 10-year retrospective chart review of trauma patients who
Trauma mechanism were treated according to a defined protocol. Our primary
Blunt 72 (93) finding was that RP use was significantly associated with
Penetrating 5 (7) decreased odds for developing ARD among patients with a
ISS peak CK greater than 10,000 U/L. We also found that a
Mean 6 SD 22 6 14 peak CK greater than 10,000 U/L was an independent
Median 22 predictor of developing ARD, after accounting for age,
Shock† 26 (34) shock, and the use of the protocol. Based on these findings
IV contrast (%)‡ 50 (65) we altered the threshold of our treatment protocol and
Massive transfusionx 18 (23)
lowered it back to 10,000 U/L rather than 20,000 U/L over
Fasciotomy 21 (27)
the course of this study period.
ARD 24 (31)
Hemodialysis required 4 (5) Rhabdomyolysis is characterized by the release of
Death within 30 days 9 (12) myoglobin into the circulation, myoglobinuria, and the
risk of progression to renal failure. The mechanisms of
ARD 5 acute renal dysfunction (defined as creatinine .2.0 mg/dL);
CK 5 creatine kinase; ISS 5 Injury Severity Score; IV 5 intravenous;
rhabdomyolysis in trauma populations include extremity
SD 5 standard deviation. crush injury or prolonged muscle ischemia followed by
*Data are expressed as n (%), except as noted otherwise. reperfusion. In either situation, the disruption of skeletal
†
Shock 5 systolic pressure less than 90 mm Hg on admission. muscle cells is followed by the release of intracellular
‡
IV contrast 5 exposure to intravenous contrast agents during the contents into the circulation. Trauma patients with ARD,
first 48 hours of admission.
x
Massive transfusion 5 greater than 10 units of packed red blood
defined as a Cr greater than 2 mg/dL, are at greater risk of
cells during the first 24 hours of admission. death.10 The severity of renal impairment is proportional to
the amount of injured muscle and quantity of released
myoglobin.
was an independent positive predictor for an increased risk Four percent to 33% of patients with rhabdomyolysis
of ARD (odds ratio [OR] 8.639 for CK . 10,000) and the develop ARD due to myoglobinuria.2,5,11 Myoglobin neph-
treatment protocol was found to be protective (OR .175). rotoxicity has been demonstrated through the following
By taking the inverse of OR for the protocol, there was mechanisms: renal vasoconstriction, obstructive cast forma-
nearly a 6-fold increase in the odds of developing ARD tion within the renal tubules, and free radical cytotoxicity
in the absence of a forced alkaline diuresis. leading to lipid peroxidation and cell membrane dissolu-
Being that the treatment threshold was raised from a CK tion. Myoglobin is more likely to precipitate and exhibit
level of 10,000 to 20,000 U/L in 1997 and 4 patients with the aforementioned toxicities in the presence of acidic urine
peak CK levels greater than 20,000 U/L did not receive the and hypovolemia.12,13 Patients with rhabdomyolysis due to
protocol due to either a delay in presentation or diagnosis, it crush or prolonged muscular ischemia are also at risk for
became possible to analyze the effectiveness of our the factors which predict myoglobin toxicitydnamely
protocol in patients with peak CK levels greater than acidosis and hypovolemia.
10,000 U/L. In this subgroup of patients, we found that 12 Early timing of fluid resuscitation is critical to prevent
of 46 (26%) developed ARD if the protocol was used, the development of renal failure. Much of the data
whereas 7 of 10 (70%) developed ARD without the supporting the early use of intravenous fluids are from
protocol (P 5 .008). Of note, the peak CK levels of patients reports of crush injuries from fallen structures after natural
who received the protocol were not significantly different disasters.13–16 In terms of what type of fluid to use, sodium
than those who did not receive the protocol (39,632 vs bicarbonate has been promoted as a means of protecting
49,503 U/L, P 5 .654) (Table 2). against myoglobin cast formation and free-radical tubular
Complications attributed to RP were noted in 4 of the 46 injury by inducing an alkaline environment.1,14,17 Gener-
patients (8.7%). Three (6.5%) patients developed reversible ating a brisk flow of urine with mannitol has also been
intravascular volume overload and were treated with advocated, with the theoretical benefit of flushing
furosemide and decreased fluid administration. One patient myoglobin out of the renal tubules. Mannitol also has
(2.2%) became severely alkalotic (pH 7.78) and was treated free radical scavenging capabilities.18 Being that the RP
with acetazolamide and cessation of bicarbonate adminis- employed in this study utilizes urinary alkalinization as
tration. These complications did not lead to identifiable well as a forced osmotic diuresis, we are unable to deter-
long-term morbidity. mine the relative benefit of each of these components.
J.S. Nielsen et al. Bicarbonate/mannitol for rhabdomyolysis 77

80 Patients with
traumatic

3 Patients
excluded
77 Patients included in
analysis

CK level < 10,000 CK level >10,000

n=21 (27%) n=56 (73%)

No Protocol No Protocol Protocol

n=21 n=10 n=46

16 (76%) normal renal funcƟon 3 (30%) normal renal funcƟon 34 (74%) normal renal funcƟon
5 (24%) ARD 7 (70%) ARD 12 (26%) ARD
0 Required hemodialysis 4 required hemodialysis 0 required hemodialysis

Figure 2 Rhabdomyolysis patients stratified by peak CK level, treatment, and outcome. ARD 5 acute renal dysfunction, creatinine .2.0
mg/dL; CK 5 creatinine phosphokinase; normal renal function 5 creatinine %2.0 mg/dL; Hemodialysis 5 subset of patients with ARD
requiring hemodialysis.

Future studies aimed at delineating the impact of achieving
the urine pH and volume goals of the protocol are needed
so that an optimal protocol can be developed for use in a
formal randomized controlled trial.
Among the papers reviewed over the past 10 years, there
is a general acknowledgement that the effect of forced
alkaline diuresis is largely unknown. Recent reviews and
consensus statements have supported the use of crystalloid
only regimens over the use of bicarbonate and mannitol for
rhabdomyolysis patients. The paucity of data supporting
bicarbonate and mannitol and the conclusions of the Brown
et al study of 2004 in favor of crystalloid therapy are
commonly cited as rationale.2–4 Thus, more groups are
effectively advocating against bicarbonate and mannitol
based on lack of information rather than negative studies.
In contrast, our data support the use of a forced alkaline
diuresis and suggest that the discussion should be reopened.
Central to our study and the management of rhabdo-
myolysis is the use of serum CK as an indicator of severity.
Other published investigators have reported that CK is a
reliable, quantitative indicator of risk of renal damage from
myoglobinuria. Veenstra et al reviewed 93 patients with

Table 2 Univariate analysis of risk factors for the development of ARD

Variable Normal renal function (n 5 53) ARD (n 5 24) P value
Age (year)* 33 40 .029
Male sex† 77.4% 79.2% .859
Injury Severity Score‡,x 21.5 6 13.6 23.8 6 16.4 .517
Shock on admission (SBP , 90 mm Hg)† 28.3% 45.8% .1319
Time to treatment (hours from injury to admission)* 1 1.14 .243
Massive transfusion (.10 units in 24 hours)† 20.8% 29.2% .419
Intravenous contrast exposure† 66.0% 62.5% .763
Peak creatinine phosphokinase (U/L)* 16,400 24,767 .054
Bold value meets statistical significance of P , .05.
ARD 5 acute renal dysfunction (defined as serum creatinine .2.0 mg/dL); SBP 5 systolic blood pressure.
*Nonparametric data expressed as median values, evaluated with Mann–Whitney U test.
†
Categorical data expressed as percentage and analyzed with Pearson chi-square.
‡
Parametric data expressed as mean 6 standard deviation and analyzed with independent samples t test.
x
The Injury Severity Score ranges from 1 to 75, with higher scores indicating more severe injury.
78
Table 3 Independent risk factors for the development of ARD by multivariate logistic regression
Variable
Age*
Shock on admission (SBP , 90 mm Hg)
Treatment protocol
Peak CK 2,000 to 10,000 U/L†
Peak CK . 10,000†
Bold value meets statistical significance of P , .05.
ARD 5 acute renal dysfunction (serum Cr . 2.0 mg/dL); CK 5 creatine kinase; SBP 5 systolic blood pressure; Treatment protocol 5 patient received
forced alkaline diuresis according to our rhabdomyolysis protocol.
*Age as a continuous variable and OR reflects the increase associated with each year of life.
†
Peak CK ranges were compared with the reference category of 2,000 to 10,000 U/L.
rhabdomyolysis from several etiologies, and concluded that
patients whose peak CK level exceeded 15,000 U/L had a
significantly higher rate of ARD. In Veenstra’s study,
patients with transient renal dysfunction had a higher
mortality.8 In another study of 157 patients, Ward5 reported
that patients with a peak CK level greater than 16,000 U/L
had the highest risk of developing ARD. Ward also found
that a majority of patients had a maximum CK elevation
within 24 hours of admission. Similarly, our study found
that, of patients with times recorded for both initial and
peak CK, 64% peaked within 24 hours and 86% peaked
within 48 hours. This is consistent with the practice of
measuring serial CK levels in high-risk patients (crushed
extremities, revascularized extremity vessels following
more than 4 hours of ischemia, dark urine) during the first
48 hours following injury. Thus, patients with significant
muscle disruption and associated myoglobinuria can be
identified and treated early.
Additional patient risk stratification among patients with
elevated serum CK can be identified. Eneas et al1 found that
only patients with a peak CK greater than 20,000 U/L failed
to respond to a mannitol–bicarbonate diuresis and went on
to require dialysis. This observation supports our position
that the protocol should be started before the CK exceeds
20,000 U/L. This study is also consistent with other manu-
scripts that indicate injured patients with rhabdomyolysis
whose CK remains under 10,000 U/L do not require
therapy beyond standard intravenous infusion of isotonic
crystalloid resuscitation.
Brown et al2 concluded from their retrospective review
of 382 trauma patients that an elevation in CK above
5,000 U/L puts patients at risk for renal failure (defined
as a serum Cr . 2.0 mg/dL). Furthermore, these authors
concluded that the infusion of mannitol and bicarbonate
is ineffective at preserving renal function. However, they
did comment that there was a trend for patients whose
CK level exceeded 30,000 U/L to benefit from an induced
alkaline osmotic diuresis; fewer patients needed dialysis
and there was a trend toward a lower mortality rate
(P 5 .09). A limitation of their conclusion regarding the
general ineffectiveness of a forced alkaline diuresis is that
the patients who received mannitol and bicarbonate had
The American Journal of Surgery, Vol 213, No 1, January 2017
Odds ratio P value
1.023 .162
2.587 .087
.175 .030
1 (reference)
8.639 .016
significantly higher peak CK levels than those who did
not (23,492 vs 9,819 U/L, P , .0001). By treating patients
with a higher peak CK, the authors may have effectively
neutralized their risk of renal failure and created a
perceived lack of benefit to a forced alkaline diuresis.
Brown et al also reported that 45% of the 217 rhabdomyol-
ysis patients in their series whose Cr exceeded 2.0 mg/dL
required dialysis. In contrast, only 17% of the patients in
our series with a similar elevation in Cr required dialysis.
We agree with the authors of the Brown study that the
utility of a forced alkaline diuresis should be evaluated in
a randomized control trial.
Conclusion
Our data support using a protocol including fluids,
mannitol, and bicarbonate for patients with traumatic
rhabdomyolysis and a CK over 10,000 U/L. It remains to
be answered if there is a ‘‘point of no return’’ in which
delay of therapy, extreme elevation of CK, or prolonged
oliguria will always progress to renal failure. We advocate
for a multicenter trial comparing the use of crystalloid
titrated to urine output versus a protocolized forced alkaline
diuresis in patients with a CK greater than 10,000 U/L.
Acknowledgments
This study was supported by grant R49/CCR-006283
from the U.S. Public Health Service, Centers for Disease
Control and Prevention, National Center for Injury Preven-
tion and Control, Atlanta, GA.
References
1. Eneas JF, Schoenfeld PY, Humphreys MH. The effect of infusion of
mannitol-sodium bicarbonate on the clinical course of myoglobinuria.
Arch Intern Med 1979;139:801–5.
2. Brown CV, Rhee P, Chan L, et al. Preventing renal failure in patients
with rhabdomyolysis: do bicarbonate and mannitol make a difference?
J Trauma 2004;56:1191–6.
J.S. Nielsen et al. Bicarbonate/mannitol for rhabdomyolysis
3. Zimmerman JL, Shen MC. Rhabdomyolysis. Chest 2013;144:
1058–65.
4. Brochard L, Abroug F, Brenner M, et al. An official ATS/ERS/ESICM/
SCCM/SRLF statement: prevention and management of acute renal
failure in the ICU patient: an international consensus conference in
intensive care medicine. Am J Respir Crit Care Med 2010;181:
1128–55.
5. Ward MM. Factors predictive of acute renal failure in rhabdomyolysis.
Arch Intern Med 1988;148:1553–7.
6. Oda J, Tanaka H, Yoshioka T, et al. Analysis of 372 patients with crush
syndrome caused by the Hanshin-Awaji earthquake. J Trauma 1997;
42:470–6.
7. Zager RA. Studies of mechanisms and protective maneuvers in myo-
globinuric acute renal injury. Lab Invest 1989;60:619–29.
8. Veenstra J, Smit WM, Krediet RT, et al. Relationship between elevated
creatine phosphokinase and the clinical spectrum of rhabdomyolysis.
Nephrol Dial Transplant 1994;9:637–41.
9. Knottenbelt J. Traumatic rhabdomyolysis from severe beatingdexper-
ience of volume diuresis in 200 patients. J Trauma 1994;37:214–9.
10. Harbrecht BG, Rosengart MR, Zenati MS, et al. Defining the contribu-
tion of renal dysfunction to outcome after traumatic injury. Am Surg
2007;73:836–40.
79
11. Slater MS, Mullins RJ. Rhabdomyolysis and myoglobinuric renal fail-
ure in trauma and surgical patients: a review. J Am Coll Surg 1998;
186:693–716.
12. Huerta-Alardı́n AL, Varon J, Marik PE. Bench-to-bedside review:
rhabdomyolysisdan overview for clinicians. Crit Care 2005;9:
158–69.
13. Malinoski DJ, Slater MS, Mullins RJ. Crush injury and rhabdomyoly-
sis. Crit Care Clin 2004;20:171–92.
14. Desforges JF, Better OS, Stein JH. Early management of shock and
prophylaxis of acute renal failure in traumatic rhabdomyolysis.
N Engl J Med 1990;322:825–9.
15. Ron D, Taitelman U, Michaelson M, et al. Prevention of acute renal fail-
ure in traumatic rhabdomyolysis. Arch Intern Med 1984;144:277–80.
16. Epstein FH, Odeh M. The role of reperfusion-induced injury in the
pathogenesis of the crush syndrome. N Engl J Med 1991;324:1417–22.
17. Moore KP, Holt SG, Patel RP, et al. A causative role for redox cycling
of myoglobin and its inhibition by alkalinization in the pathogenesis
and treatment of rhabdomyolysis-induced renal failure. J Biol Chem
1998;273:31731–7.
18. Zager RA. Combined mannitol and deferoxamine therapy for myohe-
moglobinuric renal injury and oxidant tubular stress. Mechanistic and
therapeutic implications. J Clin Invest 1992;90:711–9.