Abstract: FR-PO530
Point-of-Care Ultrasound in Peritoneal Dialysis Catheter-Related Infection: An Observational Study
Session Information
- Peritoneal Dialysis: Modality, Catheter, Infections
November 08, 2019 | Location: Exhibit Hall, Walter E. Washington Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Dialysis
- 703 Dialysis: Peritoneal Dialysis
Authors
- Diniz, Hugo, Centro Hospitalar de São João, Ermesinde, Portugal
- Oliveira, Ana, Centro Hospitalar de São João, Ermesinde, Portugal
- Pereira, Luciano, Centro Hospitalar de São João, Ermesinde, Portugal
Background
Peritoneal dialysis (PD) catheter-related infection (PD-CRI) is the most common complication of this form of renal replacement therapy. The diagnosis of PD-CRI is made with physical examination (PE), but the physical findings lack sensitivity or specificity.
Point-of-care ultrasound (POCUS) is an emerging discipline in the Nephrology community that allows the physician to incorporate real-time information from the ultrasound into his clinical evaluation. POCUS could improve the diagnostic accuracy of PD-CRI and reduce patient exposure to antibiotics.
This single-center observational study aimed to compare the accuracy of POCUS and PE for the diagnosis of PD-CRI.
Methods
POCUS was performed by a Nephrology fellow using a linear transducer when PD-CRI was suspected. POCUS was repeated in every patient visit.
PD-CRI was defined as purulent discharge with or without inflammatory signs, and a positive microbiological culture collected from the exit-site. We considered a positive POCUS as an anechoic collection around the external cuff, and its largest dimension was recorded. PE findings were coded using a validated clinical score.
Results
A total of 25 patients (58 % male) were enrolled. We also recruited nine patients with no signs of PD-CRI as controls. A total of 13 PD-CRIs were diagnosed, from 22 suspected cases. The most common isolated agent was Corynebacterium spp.
In this population, the diagnostic accuracy of PE was low, with an area under the ROC curve (AUROC) of 0,6 (95% CI 0,37-0,84). Purulent drainage alone, although highly specific (100%), showed a low sensitivity (61,9%) for the diagnosis of PD-CRI.
In contrast, POCUS had an AUROC of 0,91 (95% CI: 0,73-1) for PD-CRI diagnosis. All the PD-CRI cases had a positive POCUS evaluation, and we did not find this sonographic sign in controls (p < 0,001). The optimal cut-off point of the collection dimension was equal to or greater than 1,4 mm (Sensitivity 100%; Specificity 89%).
Suspected PD-CRI cases with negative POCUS and not exposed to antibiotic therapy, had similar PD-CRI rates in one-month follow-up as the control group (p = 0,929).
Conclusion
POCUS is superior to PE for the diagnosis of PD-CRI and should be considered by nephrologists with access to an ultrasound machine. POCUS may decrease unnecessary antibiotic exposure in PD patients due to an increase in PD-CRI diagnostic sensitivity.