The present study results suggest that sweating could predict poor prognosis of rhabdomyolysis that did not require renal replacement therapies on admission. No previous reports have focused on signs and symptoms as predictive factors of serious consequences in such patients. As sweating is a physical finding that can be identified easily by a physician without any clinical instruments or laboratory tests, it is extremely valuable in the daily clinical setting. Dehydration has been reported to be one of the aggravating factors for renal failure in patients with rhabdomyolysis [10
]. Theoretically, sweating exacerbates dehydration, which can subsequently make the prognosis of rhabdomyolysis poorer. However, in this study, the relationship between dehydration and aggravation of rhabdomyolysis, and between sweating and dehydration, were not statistically significant by univariate analysis. Autonomic dysfunction is considered to be one of the causes of sweating in rhabdomyolysis, especially in NMS, [12
] which indicates the possibility that autonomic dysfunction is a more plausible cause of the association between sweating and aggravation of rhabdomyolysis than dehydration.
In this study, there was no significant difference between Groups A and B for age, sex, and serum creatinine and CK on admission and peak CK during hospitalization. Serum albumin <3.3 mg/dL, acidosis, prothrombin time <82%, and maximum CK level >12,750 IU/L were reported to be predictors of acute kidney injury in a study in patients with CK levels >5000 IU/L [6
]. Another study in patients with rhabdomyolysis caused by illicit drug and alcohol use showed a correlation between maximum CK and serum creatinine level during hospitalization without showing a correlation between the introduction of hemodialysis and death [3
]. In addition, in a study on the predictive factors of the prognosis of acute kidney injury in rhabdomyolysis, Gearoid et al. reported that age, sex, causes of rhabdomyolysis, serum creatinine, CK, phosphate, calcium and HCO3−
levels on admission were independent predictive factors for death or the need for renal replacement therapies [13
]. However, in the current study, there were no statistical differences between age, sex, serum creatinine, CK, albumin on admission, and maximum CK during hospitalization, between Groups A and B. Although serum creatinine and CK levels on admission had been expected to be associated with the prognosis of rhabdomyolysis considering the pathophysiology, serious consequences might have been prevented in our study because of early and sufficient rehydration on admission, resulting in no significant differences for these parameters between Groups A and B.
The mortality rate of rhabdomyolysis was 7.5% in this study, but ranged from 1.7 to 59% in previous reports in patients with different backgrounds [1
]. A study in patients with rhabdomyolysis caused by illicit drug and alcohol use showed mortality was 3.4% [3
] Other studies in patients with rhabdomyolysis treated in the intensive care unit showed the mortality rate was 59% with renal failure and 22% without renal failure [15
]. Because half the cases in our study were caused by prescribed drugs, the mortality rate was different from that for rhabdomyolysis caused by illicit drugs and alcohol. In addition, patients who needed renal replacement therapies on admission (and were thus expected to have a higher mortality rate), were not included in our study; this might explain the lower mortality rate in our study compared with that reported in the study in patients treated in intensive care units. In our study, 3.7% of patients required the introduction of renal replacement therapies, which was lower than that in previous studies (8.0-9.5%) [6
]. The lower frequency in our study might be because we excluded patients who required renal replacement therapies on admission.
In previous studies, the predominant causes of rhabdomyolysis were illicit drugs, alcohol, prescribed drugs, myopathies, trauma, NMS and seizure disorders, and prolonged immobility [3
]. The main causes in our study were prescribed drugs, which accounted for 55% of cases, followed by myopathies, metabolic diseases, and heat stroke/dehydration. Our department mainly treats patients with diseases of general internal medicine, with a low frequency of illicit drug use and trauma, which resulted in prescribed drugs being the main cause. Our department belongs to the university hospital, which facilitates our ability to treat patients referred to us by psychiatrists. This fact leads psychiatrists in other local hospitals to refer their patients when they experience serious diseases of internal medicine. As a result, we have many opportunities to treat patients with rhabdomyolysis caused by psychiatric medicines. The causes of rhabdomyolysis vary according to patient background and so does the probability of serious consequences. This is the reason it is imperative to take patient background into consideration when we predict the prognosis of rhabdomyolysis.
NMS can be a fatal disease which occurs in 0.07-3.23% of patients using antipsychotic drugs, with a considerably high mortality rate of about 10% [7
]. There were 11 cases (20.8%) fulfilling Levenson’s criteria in this study [8
]. Of those patients, only three (27%) had a serious prognosis. Because sweating is one of the important symptomatic factors of Levenson’s criteria, [8
] it may have some relationship with the aggravation of rhabdomyolysis with NMS. Our study, however, did not show a significant difference between the rate of patients fulfilling Levenson’s criteria between Groups A and B.
The present study has several limitations. This study was a retrospective chart review. Although we presumed that physicians in our department were in agreement about treatments including hydration and the introduction of renal replacement, these treatments were provided in an arbitrary manner. In addition, the sample size was too small to conduct multivariate analysis, which made it impossible to adjust for various confounding factors such as age, fever, and dehydration. With regard to laboratory data, we didn’t analyze pH, prothrombin time, or calcium, because they had rarely been investigated on admission. Although urinary myoglobin was reported to be a useful marker of predicting poor prognosis, [1
] we didn’t analyze this because it usually takes several days to obtain the result in our institute, which negates its use in predicting prognosis on admission.