Risk factors for colistin-associated nephrotoxicity and mortality in critically ill patients

Turkish Journal of Medical Sciences Turk J Med Sci (2017) 47: 1165-1172 © TÜBİTAK doi:10.3906/sag-1604-60 http://journals.tubitak.gov.tr/medical/ Research Article Risk factors for colistin-associated nephrotoxicity and mortality in critically ill patients 1 1, 1 2 Hüseyin ÖZKARAKAŞ , Işıl KÖSE *, Çiler ZİNCİRCİOĞLU , Sibel ERSAN , 3 1 3 1 Gürsel ERSAN , Nimet ŞENOĞLU , Şükran KÖSE , Rıza Hakan ERBAY 1 Department of Anesthesiology, İzmir Tepecik Training and Research Hospital, İzmir, Turkey 2 Department of Nephrology, İzmir Tepecik Training and Research Hospital, İzmir, Turkey 3 Department of Infectious Disease and Clinic Microbiology, İzmir Tepecik Training and Research Hospital, İzmir, Turkey Received: 13.04.2016 Accepted/Published Online: 31.03.2017 Final Version: 23.08.2017 Background/aim: Colistin is gaining popularity against multidrug-resistant bacteria. The primary concern with colistin is its nephrotoxicity (NT). The aim of this study was to evaluate the incidence and risk factors for NT and to evaluate the risk factors for mortality in the toxicity group. Materials and methods: NT was defined according to the RIFLE criteria. Data of patients who did or did not develop NT were compared. Positive and negative predictive values, risk ratio, and correlation coefficients were calculated. Results: NT was seen in 39 patients (70%). Hypoalbuminemia, old age, and the use of vasopressors (VPs) were associated with NT. The use of VPs had the highest positive predictive value, while age had the highest negative predictive value and risk ratio. The only variable that was associated with mortality in the toxicity group was VP use. Conclusion: Aging, hypoalbuminemia, and the use of VPs were shown to be risk factors for NT, while the last of these was the only significant risk factor for mortality in the toxicity group. Key words: Colistin, risk factors, nephrotoxicity, mortality 1. Introduction Over the last few decades, there has been a marked increase in the incidence of nosocomial infections caused by multidrug-resistant gram-negative bacteria, particularly in critically ill patients (1). The shortage of new antibiotics that specifically target these pathogens has recently rekindled interest in colistin, a historical antimicrobial agent, discovered in 1949. The use of colistin declined from the early 1970s to the early 2000s, due to its frequent adverse effects, including a high incidence of nephrotoxicity (NT) (2,3). It has been reported in many studies that NT was associated with a poor outcome and high mortality rates in critical patients (4–6). The reported incidence of NT varies from 0% to 60% (7–11). The wide range in colistinassociated acute kidney injury (AKI) may be attributed to inconsistency in how AKI was defined. Most of the earlier studies did not clearly define the NT. In several recent studies, NT was defined with specific criteria; however, varying definitions have complicated study comparisons (12–17). In 2004, the Acute Dialysis Quality Initiative Group published their consensus definition * Correspondence: for AKI; the Risk, Injury, Failure, Loss, End stage renal disease (RIFLE) classification is based on the change in the glomerular filtration rate (GFR), serum creatinine levels, or urinary output. The RIFLE approach can detect AKI with high sensitivity and specificity (18). Studies assessing colistin-associated NT according to RIFLE are scarce in the literature (19). In this study, we aimed to determine the risk of renal damage according to the RIFLE criteria in critically ill patients in the intensive care unit and secondarily aimed to determine the predictive factors that facilitate the development of renal damage. 2. Materials and methods 2.1. Study design, setting, and patient population After approval from the hospital ethics committee, a retrospective, cohort study was conducted in the 27-bed intensive care unit (ICU) of a tertiary care hospital (İzmir Tepecik Training and Research Hospital) in İzmir. Patients who received intravenous colistin between 1 January 2013 and 1 April 2014 were included in the study. Patients aged over 18 years and who received intravenous colistin for at least 72 h were included. 1165 ÖZKARAKAŞ et al. / Turk J Med Sci Patients were excluded if they had been previously diagnosed with chronic renal disease, regardless of the need for hemodialysis. Patients with heart failure, liver disease, and malignancy were also excluded. If a patient met the inclusion criteria on multiple occasions, only the first qualifying episode was evaluated. The product used in this study was colimycin (colistimethate sodium produced by Koçak Farma, Turkey) available as 150 mg of colistin base activity per vial. Colistimethate sodium dosage is regulated as 2.5–5 mg/kg (maximum 300 mg) in patients who do not have renal insufficiency. We therefore used 5 mg/kg colistin loading dose followed by two doses of 5 mg/kg (maximum 300 mg) in accordance with the infectious diseases physician’s recommendations; subsequent dosages were adjusted according to the serum creatinine (Scr) levels. The decision to use colistin was based on the recommendation of an infectious disease specialist and the results of susceptibility testing given resistance to all antibiotics apart from colistin. 2.2. Study variables, definitions, and measurements The electronic medical records for the study patients were retrospectively reviewed using a standardized data collection form. The following baseline characteristics of the patients were collected from the medical records: age, sex, and the simplified acute physiology score (SAPS II). Causative microorganisms, coexistence of severe sepsis, and septic shock were recorded. Severe sepsis and septic shock were defined according to the Surviving Sepsis Guidelines (20). Concomitant use of other nephrotoxins (nephrotoxic antimicrobials, loop diuretics, intravenous dye, nonsteroidal antiinflammatory drugs (NSAID), angiotensin-converting-enzyme (ACE) inhibitors, and calcineurin inhibitors), use of vasopressors (VPs), serum albumin levels, and duration of colistin therapy were recorded. NT was determined by the RIFLE criteria. RIFLE has three severity classes, which were assessed by relative changes in either Scr or the GFR from a baseline value, or by a decrease in urine output at a predetermined time interval. The two outcome criteria were defined by the duration of the loss of kidney function, which was assessed at the 4th week (loss) and 3rd month (end-stage kidney disease) (Table 1). If data were not available for assessing the GFR at the 4th week and 3rd month, due to patient death or discharge, only the acute degree of NT was considered. The need for renal replacement therapy was also recorded. 2.3. Statistical analysis IBM SPSS version 22.0 (IBM Corp, Somers, NY, USA) was used for statistical analysis. Normality was assessed using the Kolmogorov–Smirnov test. Continuous normally distributed data, expressed a (...truncated)