Microbial Burden of Hemodialysis Fluids: A Multi-Center Study Public
Tocci, Nicole Ann (2012)
Published
Abstract
Standards for microbial contamination in hemodialysis fluids aim to protect patients from adverse events such as pyrogenic reactions, bacteremia, and endotoxemia. However, the international community has not achieved consensus on setting upper limits of contamination. At the time of data collection, the Association for the Advancement of Medical Instrumentation (AAMI) proposed new standards for upper limits and culturing methods for microbial contamination of hemodialysis fluids (AAMI RD52:2004), which have since been updated (ANSI/AAMI/ISO 11663:2009). The research performed a methods comparison for measuring microbial contamination in processed water (AAMI method: Trypticase soy agar (TSA) at 37°C for 48 hours vs. standard method: Reasoner's 2A agar (R2A) at 28°C for 7 days). Additionally, the research examined the association between microbial contamination and selected biomarkers provided by retrospective chart review. Nineteen metro-Atlanta hemodialysis facilities participated. Samples of processed water (n=223) and dialysate (n=223) were collected monthly (Jan-Mar, 1997) using aseptic technique, transported to CDC and immediately assayed. Six of the nineteen facilities provided patient data (n=454). Logistic regression was carried out to predict patient biomarkers that fell outside the acceptable target ranges determined by Centers for Medicare and Medicaid Services and National Kidney Foundation Clinical Practice Guidelines. Microbial recovery from processed water was significantly greater when using the standard method (R2A) compared to the AAMI method (TSA) (Wilcoxon's matched-pair's signed-rank test, p<0.0001). A larger percentage of non-compliance with AAMI standards would have been missed using the AAMI method with TSA (19.8%) than using the standard method with R2A (0.9%). Non-compliance with the AAMI standard for endotoxin in dialysate was a consistent predictor of negative patient outcomes as indicated by regression models for erythropoietin dose, serum albumin, and urea reduction ratio. Conversely, non-compliance with the microbial standard for colony counts of bacteria in water, as measured by TSA indicated a protective effect, possibly indicative of the method's lack of sensitivity. The study provides a connection between patient clinical data and hemodialysis fluid contamination. As the dialysis patient population grows, efforts to improve clinical outcomes and eliminate adverse events by minimizing microbial contamination will remain imperative.
Table of Contents
Table of Contents
Background...1
Microbial Contaminants...4
Standards...6
Compliance...7
Biomarkers...7
Microbiologic techniques...9
Methods...12
Study question...12
Selection of facilities...13
Sample collection...13
Microbiologic methods...13
Endotoxin assay...14
Retrospective chart review...14
Variables...14
Statistical analysis...15
Results...17
Microbial recovery...17
Clinical effects...18
Erythropoietin...19
Hematocrit...21
Serum albumin...22
Urea reduction ratio...24
KT/V...27
Discussion...30
Future directions...34
Limitations and strengths...34
Conclusion...35
References...37
Tables...44
Table 1. Colony counts and endotoxin levels in
both water and dialysate at 19 dialysis centers, Atlanta,
GA...44
Table 2. Colony counts and endotoxin levels for processed water at
19 metro-Atlanta hemodialysis centers...45
Table 3. Correlation between colony counts and endotoxin levels of
water and dialysate...46
Table 4. Patient characteristics from the six maintenance
hemodialysis facilities used to evaluate effects of dialysis fluids
on facility outcome data...47
Table 5. Correlation between erythropoietin dose and
predictors...48
Table 6. Logistic regression analysis of 454 patient records for
high erythropoietin dose (≥ 5000 units)...49
Table 7. Logistic regression analysis of 454 patient records for
high erythropoietin dose (≥ 5000 units)...50
Table 8. Correlation between average hematocrit and
predictors...51
Table 9. Logistic regression analysis of 454 patient records for
low hematocrit levels (≤ 37%)...52
Table 10. Logistic regression analysis of 454 patient records for
low hematocrit levels (≤ 37%)...53
Table 11. Correlation between average serum albumin and
predictors...54
Table 12. Logistic regression analysis of 454 patient records for
low serum albumin levels (≤ 3.5 g/dL)...55
Table 13. Logistic regression analysis of 454 patient records for
low serum albumin levels (≤ 3.5 g/dL)...56
Table 14. Correlation between average urea reduction ratio (URR)
and predictors...57
Table 15. Logistic regression analysis of 454 patient records for
low urea reduction ratios (≤ 65%)...58
Table 16. Logistic regression analysis of 454 patient records for
low urea reduction ratios (≤ 65%)...59
Table 17. Correlation between average KT/V and
predictors...60
Table 18. Logistic regression analysis of 454 patient records for
low KT/V (≤ 1.2)...61
Table 19. Logistic regression analysis of 454 patient records for
low KT/V (≤ 1.2)...62
About this Master's Thesis
School | |
---|---|
Department | |
Subfield / Discipline | |
Degree | |
Submission | |
Language |
|
Research Field | |
Mot-clé | |
Committee Chair / Thesis Advisor | |
Committee Members | |
Partnering Agencies |
Primary PDF
Thumbnail | Title | Date Uploaded | Actions |
---|---|---|---|
|
Microbial Burden of Hemodialysis Fluids: A Multi-Center Study () | 2018-08-28 14:14:05 -0400 |
|
Supplemental Files
Thumbnail | Title | Date Uploaded | Actions |
---|