RESEARCH ARTICLE For reprint orders, please contact: [email protected]

Prevention of hemodialysis catheterrelated blood stream infections using a cotrimoxazole-lock technique Azadeh Moghaddas1, Mohammad-Reza Abbasi2, Afshin Gharekhani3, Simin Dashti-Khavidaki*,2, Effat Razeghi2, Atefeh Jafari1 & Hossein Khalili1

ABSTRACT Background & aim: This trial assessed the efficacy of cotrimoxazole lock solution in reducing catheter-related blood stream infections (CRBSIs) among hemodialysis (HD) patients who were dialyzed using tunneled catheters. Method: Patients randomly received either heparin (2500 U/ml) (control group) or a mixture of 10 mg/ml cotrimoxazole (based on trimethoprim) and 2500 U/ml heparin (antibiotic group) as catheters lock solution. Results: Compared with the control group, CRBSIs rates per 1000 catheter-days was significantly lower (0.58 vs 4.4 events; p = 0.002) and cumulative infection-free catheter survival was significantly higher (log rank statistic 5.88; p = 0.015) in the antibiotic group. There were no statistical differences regarding incidences of catheter removal (8.7% in the antibiotic group vs 22% in the control group; p = 0.116) or thrombosis (2.2% in the antibiotic group vs 9.8% in the control group; p = 0.129) between the two groups. Conclusion: cotrimoxazole containing catheter lock solution is effective in reducing CRBSIs incidence and prolonging dialysis catheter survival in HD patients. Despite all efforts to reduce the use of vascular catheters among hemodialysis (HD) patients by implementation of arteriovenous fistula, widespread use of HD vascular catheters exposes these patients to markedly increased risk of catheter-related complications [1] . The risk of death due to sepsis among HD patients has been estimated to be 100 times more than that of the general population [2] and HD catheters are the primary source of bacteremia [3] . The incidence of catheter-related blood stream infections (CRBSIs) among HD patients is reported to be 2.5–5.5 cases per 1000 catheter-days or 0.9–2 episodes per patient-year in patients using double-lumen catheters as their vascular access [4] . About 12% of these bacteremia episodes result in death [5] . Therefore, the prevention of CRBSIs is one of the major challenges in the routine care of HD patients. One promising option to prevent CRBSIs is antibiotic lock technique. In this technique, a solution containing an antibiotic or antiseptic substance (e.g., taurolidine) with an anticoagulant (heparin, citrate or ethylene diamine tetraacetic acid) was inserted into the catheter lumens after each HD session for interdialytic period to prevent catheter thrombosis and infection. The significant reduction of CRBSIs events by antimicrobial catheter lock solutions was declared in five meta-analyses [6–10] . Biofilm formation on the internal and external surfaces of the catheter induces bacteremia. Antibiotic lock solutions can prevent biofilm formation on foreign surfaces [11] . In spite of promising results in CRBSIs reduction among HD patients using antimicrobial catheter lock technique, this technique has not been recommended in the last published clinical practice guidelines by the Infectious Disease Society of America [12] . The recent statement by the European Renal Best Practice on

KEYWORDS 

• antibiotic lock solution • catheter-related bacteremia • cotrimoxazole • tunneled hemodialysis

catheter

Faculty of Pharmacy, Tehran University of Medical Sciences, Tehran, Iran Nephrology Research Center, Tehran University of Medical Sciences, Tehran, Iran 3 Faculty of Pharmacy, Tabriz University of Medical Sciences, Tabriz, Iran *Author for correspondence: Tel.: +98 2166 5815 68; Fax: +98 2166 5815 68; [email protected] 1 2

10.2217/FMB.14.116 © 2015 Future Medicine Ltd

Future Microbiol. (2015) 10(2), 169–178

part of

ISSN 1746-0913

169

Research article  Moghaddas, Abbasi, Gharekhani et al. HD catheters is the first one that recommends preventive use of an antimicrobial catheter lock to reduce the rate of CRBSIs [13] . The rationale for such a conservative approach is the concern of promoting the emergence of resistant bacteria. There are some reports on the emergence of gentamicin-resistant bacteremia among HD patients treated with g­ entamicin-based a­ ntibiotic locks solutions [14,15] . Trimethoprim-sulfamethoxazole, also known as cotrimoxazole, is a combination of two bacteriostatic antimicrobial agents that act synergistically against a wide variety of aerobic Gram-positive and Gram-negative bacteria [16] . Except for nocardiosis and of pneumocystis pneumonia treatment, cotrimoxazole is not usually used as first choice agent for infectious treatment among HD patients especially due to the available precautions on its systemic use among ESRD patients [17–19] . Therefore, there is less fear regarding the development of its resistance among HD patients upon administration as antibiotic lock solution. To our knowledge, there is no published evidence around the use of this second-line antibiotic as a lock’s solution. All mentioned considerations prompted us to compare cotrimoxazole-heparin with heparin alone as an interdialytic catheter lock solution. Method ●●Study deign & patients

This multicenter, randomized, clinical trial compared cotrimoxazole-heparin lock with heparin as an interdialytic catheter lock. This study was conducted at HD wards of three hospitals including Imam Khomeini Hospital Complex, Sina and Ghiasi Hospitals affiliated to Tehran University of Medical Sciences between the start of February 2013 and the end of March 2014. Adult patients who were dialyzed by tunneled, cuffed HD catheter using polysulfone, low-flux dialyzer and bicarbonate buffer solution three times a week for 4 h in each dialysis sessions were included. Patients with the history of infection within the week before study’s entrance or who were treating with an antibiotic, and whom had known sulfa antibiotics hypersensitivity or glucose-6-phosphate dehydrogenase enzyme deficiency were excluded. Patients were randomly assigned to either antibiotic or control groups (Figure 1) . The antibiotic group was prescribed cotrimoxazole– heparin lock solution as a catheter lock solution. This solution was a mixture of 10 mg/ml

170

Future Microbiol. (2015) 10(2)

cotrimoxazole (based on trimethoprim) and 2500 U/ml unfractionated heparin. Patients in the control group were prescribed with a solution containing only heparin (2500 U/ml), as is common in dialysis wards of these three hospitals. After each HD session, the lumen of dialysis catheter was flushed by 10 ml of normal saline solution and then was locked for whole interdialytic period with the either antibiotic-lock or heparin lock solution in a volume equivalent to the internal volume of catheter’s lumen. At each HD session, catheter data including blood flow, exit site and tunnel status were recorded. Catheter exit-site wrapping was changed after each dialysis session. Randomization was made using cluster randomization among three dialysis units. According to our limitation, the study investigator who assessed outcomes and the staff who were involved in the preparation of catheter lock solution were not blinded. The major end point of the study was the incidences of CRBSIs between the two comparative groups. Based on the classification by the Center for Disease Control (CDC) of the United States, CRBSI was defined in three distinctive conditions including (i) definite bloodstream infection which was the isolation of the same organism from a peripheral blood sample and catheter-driven blood sample of a symptomatic patient (fever, chills, hypotension or mental confusion) while there was no other probable source of infection; (ii) probable bloodstream infection in which infectious symptoms disappeared after catheter removal while either blood cultures or culture of the catheter tip, but not both, confirmed infection in a symptomatic patient and there was no other suspicious source of infection; and (iii) possible bloodstream infection that was defined as disappearance of infectious symptoms in a symptomatic patients with negative blood and catheter tip cultures after catheter removal while there was no other probable source of infection [20] . After CRBSIs diagnosis, treatment of infection was the responsibility of related individual physicians and we had no special responsibility for patient’s management according to the study protocol. Catheter-related blood stream infectionfree survival was defined as the number of days from the start of the study to diagnosis of CRBSIs as defined above. Catheter dysfunction was defined as the requirement for catheter removal or the need for thrombolytic drugs

future science group

Prevention of hemodialysis catheter-related blood stream infections 

Research article

96 patients were assessed for eligibility

9 were excluded: – 4 did not meet inclusion criteria or met exclusion criteria – 1 declined to participate – 4 often used taurolidine lock

87 underwent randomization

46 were assigned to receive cotrimoxazole-heparin catheter-lock

41 were assigned to receive heparin catheter-lock alone

All patients completed the 6 months follow-up

All patients completed the 6-month follow-up

16 patients could not complete the 1-year follow-up because: – 5 patients died during follow-up – 3 patients switched to arteriovenous fistula – 3 patients were moved to another dialysis center – 5 were newly included patients at the time of statistical analyses

19 patients could not complete the 1-year follow-up because: – 7 patients died during follow-up – 3 patients switched to arteriovenous fistula – 2 patients were moved to another dialysis center – 2 patients were transplanted by cadaveric donors – 5 were newly included patients at the time of statistical analyses

Figure 1. Patients’ inclusion in the study.

administration via the dialysis catheter because a pump blood flow of more than 250 ml/min that is needed for tunneled, cuffed catheters was not achieved during HD or there was clinical suspicious for thrombosis formation. According to CDC criteria, exit site infection was defined as a symptoms including erythema, tenderness and/or induration within 2 cm of the dialysis catheter exit site with or without purulent exudates or microbiological exit site infection where the exudates lead to microorganism’s growth in the culture [20] . Interventional radiologist had responsibility to implant dialysis catheters for patients. Patients’ follow-up length was 6 months based on the study protocol; however, we followed most, but not all, patients for 1 year (Figure 1) .

future science group

All patients were assessed during each dialysis session for adverse drug reactions and dermatologic complications such as skin rash, pruritus, photosensivity, urticaria and other rare adverse effects such as Stevens-Johnson syndrome or erythema multiform. ●●Ethics

The study protocol was approved by local ethics committee of Tehran University of Medical Sciences and was registered in the Iranian Registry of Clinical Trials (IRCT201305313043N8). All patients signed written informed consent forms. ●●Sample size calculation

According to results of Kim et al. [21] , which were CRB incidence of 1.7% in the antibiotic lock

www.futuremedicine.com

171

Research article  Moghaddas, Abbasi, Gharekhani et al. Table 1. Demographic, clinical and laboratory characteristics of the patients at the initiation of the study. Parameters

Antibiotic group (n = 46) mean ± SD

Control group (n = 41) p-value mean ± SD

Age (years) Sex (male) n (%) BMI (kg/m2) DM (%) Hemoglobin (g/dl) Albumin (g/dl CRP (mg/l)† KT/V Ferritin (μg/l)† Median catheter vintage before study recruitment (days)†

63.63 ± 10.63 20 (43.5%) 25.20 ± 4.23 50 10.49 ± 1.43 4.21 ± 0.48 18 (16–35) 1.14 ± 0.16 422 (282.5–1000) 45 (10–194)

60.68 ± 14.40 23 (56.1%) 23.78 ± 4.24 61 10.82 ± 1.06 4.11 ± 0.46 21 (6–35) 1.48 ± 0.53 453 (276–620) 31 (3.5–151)

0.277 0.240 0.179 0.581 0.223 0.355 0.800 0.308 0.473 0.530

Mann–Whitney U-test analysis due to lack of normality, data shown in median (range). BMI: Body mass index, CRP: C-reactive protein, DM: Diabetes mellitus, KT/V: Dialysis adequacy based on urea clearance. †

z

n1 =

group compared with 11.7% in no-antibiotic group and CRB rate per 1000 catheter-days were 0.44 in the antibiotic group compared with 3.12 in the no-antibiotic group and by considering the α = 0.05 and study power of 80%, early assumption of 6-month follow-up with exclusion of 10% of enrolled patients, sample size was calculated to be 45 patients in each group of the study using the following formula: p2 q2 2

a 1- 2 #

1 pq a1 + k k + z1 - b # D

p1 q1 + k

●●Statistical analyses

All analyses were performed on an intentionto-treat basis using SPSS software (version 19.0; SPSS Inc., IL, USA). Data are represented as means ± standard deviations (SD) or as medians (ranges). Normal distribution of quantitative variables was assessed by Kolmogorov–Smirnov test. The Student’s t-test, Mann–Whitney U-test, chisquare test analyses were used to compare patient characteristics. CRBSIs rates (CRBSIs events for 1000 catheter-days) were compared using the logrank test. Cumulative infection-free catheter survivals and thrombosis-free survivals were determined using the Kaplan–Meier method. p-values

Prevention of hemodialysis catheter-related blood stream infections using a cotrimoxazole-lock technique.

This trial assessed the efficacy of cotrimoxazole lock solution in reducing catheter-related blood stream infections (CRBSIs) among hemodialysis (HD) ...
2MB Sizes 0 Downloads 14 Views