[R] ICD9 codes

Sat May 29 13:47:45 CEST 2010

I have received the "Matched a Filter Rule" message a number of times.  
I have some control over what the headers of my messages look like, so 
if I was told what the filter rule is, I could perhaps modify the 
messages to avoid it.  You probably don't want to post that information 
to the list, but it might be worthwhile to send it privately to authors 
of false positive messages.

Duncan Murdoch

On 29/05/2010 6:43 AM, (Ted Harding) wrote:
> It is perhaps time to bring to people's general attention that
> there is an issue with postings from gmail.com addresses being
> held for moderation because "The headers matched a filter rule".
> It is not only gmail.com, but there is a marked excess of these.
> Vishwanath (@gmail.com) is one of many victims.
>
> There is a group of about a dozen people who act as moderators,
> approving or discarding messages which have been held. Reasons
> for holding are almost always either "Posting from Non-Member"
> or "Matched a Filter Rule". We do this somewhat non-systematically,
> looking at the list of held messages on a "convenience" basis.
> However, it would be unlikely that any valid message would be
> held for more than an hour or so; usually it would be less.
>
> The situation has been discussed recently amongst the moderators,
> including the list manager Martin Maechler. We are undecided
> about the best action (if any) -- while the excessive false
> trapping of gmail.com postings is at least an inconvenience
> for the trapped, and sometimes leads to visible distress or
> embarrassment, the presence of the filter rules does prevent
> spam from reaching the list.
>
> Vishwanath's reaction in repeatedly re-posting is perhaps
> understandable on the part of someone who is not familiar with
> how the moderation system operates. In fact he posted the same
> message 4 times (morning & evening on 27 May, and again on 28 May)
> as can be seen from the archives. He is subscribed to the list,
> so would normally receive copies of his own postings via the list
> (and therefore know that it had got through) unless he has opted
> to not receive his own postings. In that case, having received
> the "held for moderation" message, not receiving his posting via
> the list, and not thinking of looking in the archives, he could
> form the impression that it/they had vanished into a black hole,
> and try again.
>
> Since we began to discuss this about a fortnight ago, I have
> been keeping a count of those held messages which I have seen
> myself. Not counting true spam, only genuine messages, the results
> are as follows. "NM" denote a posting from a non-member, "FR"
> denotes a posting that matched a filter rule, "GM" denotes a
> posting from a gmail.com address, "NGM" one not from gmail.com.
>
>       FR     NM
> ----+-----+-----
> GM  | 45  |   4
> ----+-----+-----
> NGM | 19  |  24
>
> Fisher test: P=7.3e-07, OR=13.7, CI=(4.0, 62.0)
>
> About 30% of R-help subscribers have gmail.com addresses, so they
> are clearly over-represented in the "FR" group (70%)!
>
> It is suspected that the gmail.com "filter rule" may perhaps be
> triggered by HTML, though the reason is not yet clear.
>
> Meanwhile, a reminder to people who receive notification that
> their posting has been held for moderation: Please check the
> archives to see whether your message has reached the list within
> a reasonable time (say 2 hours -- there can be a delay before a
> message is placed in the archives) beforetrying to do anything
> about it: https://stat.ethz.ch/pipermail/r-help
>
> Hoping this helps to clarify an issue which can lead to unwanted
> consequences.
>
> Ted.
>
> On 29-May-10 05:42:27, Vishwanath Sindagi wrote:
>   
>> Dear Prof Frank Harrel:
>>
>> I am extremely sorry for having reposted the same question numerous
>> times. Earlier when I had posted I got  replies stating that my post
>> had matched a filter rule and hence was being held by the moderator.
>> So I assumed that the question was never posted and I reposted with
>> different subject lines just to make sure that it gets posted.
>>
>> I sincerely apologise for the inconvenience caused.
>>
>> Regards,
>> Vishwanath
>>
>> On Sat, May 29, 2010 at 4:10 AM, Frank E Harrell Jr
>> <f.harrell at vanderbilt.edu> wrote:
>>     
>>> Your notes are bordering on harassment. _Do you expect that everyone
>>> who
>>> reads this list will reply "I do not have anything that will help you"
>>> if
>>> they don't? _By my count this is your 4th note asking for this help.
>>>
>>> That being said I hope that you do find help somewhere or implement it
>>> yourself and share the result, as your question is an important one.
>>>
>>> Also, please be sure to state your affiliation on your notes.
>>>
>>> Frank
>>>
>>> On 05/28/2010 02:19 PM, Vishwanath Sindagi wrote:
>>>       
>>>> Hello:
>>>>
>>>> I am working on getting some statistics related to clinical trials
>>>> and
>>>> stuff. I have to work with ICD9 codes.
>>>>
>>>> Is anyone aware of any R method that deals with ICD9 codes
>>>> verification and manipulation.
>>>>
>>>> Thanks
>>>> Vishwanath
>>>>
>>>>         
>>> --
>>> Frank E Harrell Jr _ Professor and Chairman _ _ _ _School of Medicine
>>> _ _ _ _ _ _ _ _ _ _ Department of Biostatistics _ Vanderbilt
>>> University
>>>
>>>       
>> ______________________________________________
>> R-help at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide
>> http://www.R-project.org/posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
>>     
>
> --------------------------------------------------------------------
> E-Mail: (Ted Harding) <Ted.Harding at manchester.ac.uk>
> Fax-to-email: +44 (0)870 094 0861
> Date: 29-May-10                                       Time: 11:43:23
> ------------------------------ XFMail ------------------------------
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>