Re: Intensivist supervision of resident-placed central venous catheters decreases the incidence of catheter-related blood stream infections Adamson Sinjani Muula, University of Malawi and University of North Carolina at Chapel Hill 2 May 2008 The paper by Papadimos et al  is timely as it comes at a time when the cost of health care for many patients in the United States is unbearable, and any financial savings are welcome amidst a hurting economy. I commend the authors for using Auto-Regressive Integration Moving Average (ARIMA) analysis to assess if the change in outcome was due to the intervention and not "noise". While the choice of method appears to be appropriate, the reader is left to guess if the assumptions necessary for this method were considered. For instance, it is not known whether the residual variations from the data were or assumed to be normally distributed, to what extent were they independent and whether the authors considered whether there was constant variance . These are important assumptions that, if violated significantly, would make the ARIMA method less appropriate.While when ARIMA is used, the 'effect' of seasonality is potentially removed, I believe it would have been informative if the authors had considered presenting a graphic distribution of the rate of infections for each month for all the years followed. This would have provided the reader with a pictorial distribution of any patterns in the rate over the years. This could have been easily presented in a single figure.An interesting observation of Table 1 is that for each year of follow-up after the intervention, in general the rate of infections went down suggesting perhaps that there were cumulative gains or residual effects from the preceding year. And yet the residents were on rotations and new residents were being introduced to the intensive care unit and new patients come to the unit. How is this important? It may suggest that it is the intensivists (the supervisors) who were getting better at the technique! The intensivits were the same over the years, while the residents were new each time a rotation was made. The main conclusion I therefore see in this paper is that: when supervisors are not changed, they get better, and no matter any change in residents, the gains are long lasting and the improvement gains continue. Unless Papadimos et al have other data, I believe my interpretation may be applicable just as Papadimos et al is.Reference1.Papadimos TJ, Hensley SJ, Duggan JM, Hofmann JP, Khuder SA, Borst MJ, Fath JJ.Intensivist supervision of resident-placed central venous catheters decreases the incidence of catheter-related blood stream infections.Patient Saf Surg;2(1):11.2. Marshall S. Epidemiologic analysis of time to event data: Time series analysis. University of North Carolina at Chapel Hill, 2008. Competing interests I have no competing interest to declare.