2091

HAEMODIALYSIS ACCESS ASSOCIATED INFECTIONS – THE NEED FOR A UNIFORM DEFINITION

Makanjuola D1, Steele M1, Clark J2

Renal Unit1, Department of Microbiology2, St. Helier Hospital, Carshalton, Surrey

INTRODUCTION: The UK Renal Association Clinical Practice Guidelines recommend that any individual who commences haemodialysis (HD) should do so with an arteriovenous fistula (AVF) as first choice, an arteriovenous graft (AVG) as second choice, a tunnelled venous catheter as third choice and a non-tunnelled catheter as an option of necessity.It also states thatAVGs have the second highest rate of infection in dialysis patients; infection of an AVG has a worse prognosis than infection of an AVF; and that bacteraemia is common and may occur during cannulation and responds well to antibiotics. We present a review of the rates of AVF/AVG infections at our unit.

METHODS: We did not find a universally accepted definition of what constitutes an AVF/AVG associated infection. We defined AVF/AVG infections as either; the presence of pus at the wound/needle site, or other features of possible infection such as clinical signs of inflammation (erythema, swelling, tenderness) in conjunction with raised inflammatory markers.In order to ensure that we did not miss any cases, the presence or absence of a positive blood culture or swab was noted but was not a pre-requisite for the definition of an infection, as some of these patients had been commenced on antibiotics on clinical grounds by their general practitioners prior to samples being sent for microbiological testing.

All patients who had AVF/AVG surgery between January 2011 and August 2013 were included. We collected data on: time from surgery to infection, type of access involved, creation vs rescue, needled vs never needled, Surgeon/Theatre for access procedure, time to review by vascular access nurse post-surgery, type of organism, previous/current use of dialysis line, impact of co-morbidities and HD in a satellite unit vs main centre.

RESULTS: 954 patients had vascular access surgery in the study period (829 AVFs, 125 AVGs). 23 (22.5%) of the patients who had AVG surgery and 48 (6.1%) of those who had AVF surgery developed infections. Of these 71 patients (M:F = 1.4:1), 14 (20%) had positive blood cultures. 66% were diabetic. 46% never had a line for dialysis. 71% had gram positive organisms of which 45% were MSSA, and 4% were MRSA. The rate of all AVF/AVG infections was 0.3 per 100 patient months (bacteraemias 0.06 per 100 patient months). The median time to infection was 15 days for infections occurring within 30 days of the operation (Table 1). Needling, surgeon, creation or rescue surgery and main centre or satellite dialysis, were not associated with an increased risk of infection.

Table 1: Days from surgery to infection in patients presenting < or > 30days post-surgery

< 30 days post surgery / 30days post surgery
Range / 2-29 / 31-2460
Median / 15 / 361

CONCLUSION: Our results showed an increased incidence of infections in AVGs as compared to AVFs. With our very broad definition, our rate was 0.3 per 100 patient months. The rate for bacteraemia was much less. There is no guideline for an ‘acceptable’ rate of infection episodes and no clear definition of what constitutes an AVF/AVG infection. With the current emphasis on avoiding lines for dialysis access, we believe that it would be important for the renal community to come up with an agreed definition of AVF/AVG infections and a guideline about what constitutes an ‘acceptable infection rate order to aid auditing and interpreting results from different centres.