Hi Experts,
I have one question regarding comparing our hospital’s infection rate with those from NNIS report , suppose my rate is coming from a monthly data with a 300 device-days , is it fairness to compare this rate with a rate coming from 5000 device-days?
I am confused , please help.
Many thanks in advance.
Omar.

Hi Experts,
I have one question regarding comparing our hospital’s infection rate with those from NNIS report , suppose my rate is coming from a monthly data with a 300 device-days , is it fairness to compare this rate with a rate coming from 5000 device-days?

It can be "fair", if you do the comparison on the basis of - are these two rates statistically different, or only vary by expected random noise. Likely you would do a binomial confidence interval in this situation.

It can be "fair", if you do the comparison on the basis of - are these two rates statistically different, or only vary by expected random noise. Likely you would do a binomial confidence interval in this situation.

Dear Mr. Steve,
Actually, we compare our rate with those found in the tables of the NNIS report (the tables contain the pooled mean and the percentiles only) to determine if our rate in high or low, in other words, examiner the percentiles in each of the tables and look for the 50th percentile (or median).
So according to the process above, how do we know if the rates statistically different or no.
Many thanks.

I don't know anything about this subject, but I did just learn a new word. For the other ignorati here (from Wikipedia (http://en.wikipedia.org/wiki/Nosocomial_infection)):
Nosocomial infections are those which are a result of treatment in a hospital or hospital-like setting, but secondary to the patient's original condition. Infections are considered nosocomial if they first appear 48 hours or more after hospital admission. Nosocomial comes from the Greek word nosokomeion meaning hospital (nosos = disease, komeo = to take care of ). The most common nosocomial infections are of the urinary tract, and various pneumonias.

Dear Mr. Steve,
Actually, we compare our rate with those found in the tables of the NNIS report (the tables contain the pooled mean and the percentiles only) to determine if our rate in high or low, in other words, examiner the percentiles in each of the tables and look for the 50th percentile (or median).
So according to the process above, how do we know if the rates statistically different or no.
Many thanks.

It may help if you could share the tables, but what you need to know is how many observations there were for each of the results. For example, let us say I flip a coin 100 times and get 30 heads. Is that significantly different than the expected 50-50 rate?

The standard deviation of the rate of heads at 100 flips is the square root of ( (0.5) times (1 - 0.5) divided by 100 ). If you do the math, you will find that 30% is more than 3 standard deviations from 50%.