Rapid Responses to:

PRACTICE: Abraham George, Jerome I Tokars, Elaine J Clutterbuck, Kathleen B Bamford, Charles Pusey, and Alison H Holmes Reducing dialysis associated bacteraemia, and recommendations for surveillance in the United Kingdom: prospective study BMJ 2006; 332: 1435 [Abstract] [Full text]

Rapid Responses: Submit a response to this article

Rapid Responses published:

Seasonal variation in rate of dialysis associated bacteraemia complicates data analysis

Christopher Lawrence, Margaret Farrell, Michael K. Almond   (23 June 2006)

Seasonal variation in rate of dialysis associated bacteraemia complicates data analysis 23 June 2006
Christopher Lawrence, Specialist registrar in nephrology Renal Unit, Southend University Hospital NHS Foundation Trust, Essex, SS0 0RY, Margaret Farrell, Michael K. Almond Send response to journal: Re: Seasonal variation in rate of dialysis associated bacteraemia complicates data analysis	Editor – In support of George et al(1) we too believe that dialysis associated bacteraemia is a significant contributor to morbidity, mortality and antibiotic usage. We have undertaken formal line infection surveillance by recording episodes of line infection and bacteraemia for the last three years and regularly perform clinical audit and review of practice. George et al present a decline in the rate of dialysis associated bacteraemia using two quarters (June to September 2002 and September to December 2004). The data bears closer scrutiny. Our own data shows a seasonal variation in dialysis related bacteraemia with bacteraemia rates for any given year being consistently higher in the June to September quarter and significantly different to the September to December quarter of the same year. In George et al’s study the crude rate for access related bacteraemia would appear to have almost halved however it would be interesting to know whether the results are less impressive if shown comparing June to September 2002 with the same period in 2004. In our unit the dialysis associated bacteraemia rate was 6.75 per 100 patient months in June to September 2003 and 2.41 per 100 patient months in the subsequent quarter (Chi square test p<0.01). In June to September 2005 the bacteraemia rate was also 2.41 per 100 patient months and the rate in the subsequent quarter was 1.71. Seasonal variation in dialysis related bacteraemia could be presented as showing a large reduction between June to September 2003 and September to December 2005 (6.75 per 100 patient months to 1.71 per 100 patient months) or a small reduction between June to September 2003 and June to September 2005 (2.41 per 100 patient months to 1.71 per 100 patient months), without any overt change in clinical practice except an increased use of thrombolytic agents to preserve line patency (Alteplase). Seasonal variations in bacteraemia rate could be expected to translate in to seasonal variations in the rates of admission to hospital and rates of antibiotic usage. A longer period of observation with a sustained reduction in bacteraemia rates is required across several quarters to confirm the results of George’s intervention. The authors raise two important points. Firstly that there are likely to be significantly higher rates of bacteraemia related mortality, morbidity and cost in units employing a high proportion of semi permanent dialysis catheters. Secondly that there is no standardized method of surveillance in the United Kingdom and to this end it is unfortunate that there is not, as yet, a Renal Association Standard(2) for haemodialysis related bacteraemia as there is for peritoneal dialysis related infection. References: 1. George A et al Reducing dialysis associated bacteraemia, and recommendations for surveillance in the United Kingdom: prospective study. BMJ 332:1435 2. The Renal Association: Treatment of adults and children with renal failure. Standards and audit measures. 3rd Edition (2002). Royal College of Physicians. Competing interests: None declared