Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: analysis of cluster of fungal infection in an ICU-unit


From   Austin Nichols <[email protected]>
To   [email protected]
Subject   Re: st: analysis of cluster of fungal infection in an ICU-unit
Date   Wed, 21 Dec 2011 16:58:17 -0500

roland andersson <[email protected]>:
I can't see how this data (consecutive
patients were infected with the same fungal clones) is "suggesting
transmission via the ICU personnel."
The mechanism causing clustering could be any
number of things.

Do you have data on patients who are not positive?

There are a variety of "tests" that will show you what
you have already seen graphically, that the species are clustered.
Is that all you want?
You could just run an -mlogit- of type on last type; see also
-help permute-

On Wed, Dec 21, 2011 at 1:21 PM, roland andersson
<[email protected]> wrote:
> Austin
> For each patient I have date of positive isolate, the species and
> clones of the fungus. By plotting the dates of the positive isolate
> with the species and clone we can see pattern where sometimes 3-4
> consecutive patients have the same infection. We see many such
> clusters with different clones. My intention is to analyse if these
> clusters occurs more often than by chance. I have once done a
> space-time cluster analysis using the method by Knox. It consist of
> analysing the distribution of all possible pairs in a 2x2 chart with
> near distant in time and space. From the margins you can etimate the
> estimated number of close-close pairs and compare with the observed. I
> wonder if this can be done by changing space to fungal clones instead,
> ie analysing the number of pairs of same fungal clones that occur
> close in time?
>
> I do not have any information about the nursing staff. They move
> between the patients all the time and it would probably be too
> complicated.
>
> Any suggestion?
>
> Regards
> Roland Andersson
> Surgeon
>
>
> 2011/12/21 Austin Nichols <[email protected]>:
>> roland andersson <[email protected]>
>> We would need more info on the nature of your problem and your data to
>> answer usefully.
>> Are the same personnel present every moment?  If not, then you might
>> want to model
>> exposure to personnel, or perhaps number of cumulative contacts,
>> directly between
>> patients and staff and indirectly from patient to patient via specific
>> staff members.
>> I.e. staff members 1 and 2 connect patient 3 to patient 4, and
>> staff members 1 and 5 connect patient 6 to patient 7,
>> and patients 3, 4, and 6 have the fungal infection, leading you to put more
>> weight on staff member 2 serving as a transmission mechanism.
>> Are there connections across patients aside from staff?
>> Equipment or locations in the facility?
>> How is your data organized currently?
>>
>> On Wed, Dec 21, 2011 at 7:09 AM, roland andersson
>> <[email protected]> wrote:
>>> We are analysing fungal infections in 792 patients that were admitted
>>> to our ICU unit during 2007 and part of 2008. Some consecutive
>>> patients were infected with the same fungal clones suggesting
>>> transmission via the ICU personnel. So I have data on 78 patients with
>>> positive fungal tests (180 isolates, 6 species and 37 clones) and the
>>> date of the isolate.
>>>
>>> Graphically we can find that the same clone was found in 3-4
>>> consecutive patients on many occasions suggesting true clusters of
>>> infection. The clones varies between these clusters.
>>>
>>> I would be very glad for any suggestion on how to analyse the
>>> probability of this occurrrence of clusters using Stata.
>>>
>>> Best regards
>>> Roland Andersson
>>> Jönköping, Sweden

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index