Dominic Veit wrote:
>
> Hy all,
>
> in a clinical information system I am trying to count for
> DISTINCT patients
> born in a specific period. The data is collected by
> subselects from different
> tables. The query is as follows:
>
> select count( distinct patient_guid ) from t_visite V where
> $V_value AND
> patient_guid in ( select patient_guid from t_visite where
> guid in ( select
> visite_guid from t_demographie where gebdat <
> '".$age[0]."1231' AND ( gebdat
> > '".$age[1]."0101' OR gebdat = '".$age[1]."0101' ) ) AND
> patient_guid =
> V.patient_guid )
>
gebdat < '".$age[0]."1231' --> you do not want to know those people born on the last
day of the year?
gebdat > '".$age[1]."0101' OR gebdat = '".$age[1]."0101 strange, why don't you use
gebdat >= '".$age[1]."0101'
and for the one above gebdat <= '".$age[0]."1231' ?
And V.datum > '20030101' seems to have the same problem >=
Perhaps there is a reason to have t_visite twice in your select.
But will not
select count( distinct patient_guid ) from t_visite V
where
$V_value AND
guid in ( select visite_guid from t_demographie
where gebdat <= '".$age[0]."1231' AND gebdat >= '".$age[1]."0101' )
do what is needed?
> $V_value is: V.datum < '20030601' AND V.datum > '20030101' (a
> specific time
> where the patients to be counted must have visited the hospital).
>
> $age[0] is the lower date of the "age" that shall be requested.
> $age[1] is the upper date of the "age" that shall be requested.
>
> $age[0] and $age[1] are being calculated by subtraction from
> the actual year.
> The input then is:
>
> 0-90
> 0-19
> 20-29
> 30-39
> 40-49
> 50-59
> 60-90
>
> 0-90 would then mean that $age[0] is: 2003-0 = 2003
> and $age[1] = 2003-19 = 1984
>
> In another query I try to get the patients that have visited
> the hospital in
> the requested period (the first part of the given query).
>
> These two queries are both equal in their result when the
> input to the given
> query (with the ages) runs with 0-90 as input. However, if I
> split this query
> up in chunks (0-19 , 20-29 , 30-39 , ... ) and sum up the
> results of the
> chunks, the overall number is always one too high ( 0-90 = 92
> , sum of chunks
> = 93 ).
>
Sorry, but this I do not understand:
0-90 = 91 different values or do you mean that only 92 people were found in that
period?
Are you sure that with every visite exactly one person/gebdat is combined?
Try
SELECT COUNT(gebdat), visite_guid from t_demographie
group by visite_guid
HAVING COUNT(gebdat) > 1
Elke
SAP Labs Berlin
> I am using PERL and the PERL-DB-interface SAP::DBTech::sapdb
>
> Does anyone know what's happening there or does anyone see a
> mistake in my
> procedure? Even the second possibility would be extremely
> helpfull for me.
>
> Regards and thanx in advance,
>
> Dominic
>
> --
> Dominic Veit (Software Developer)
> Medical Center, University of Freiburg
> Dept. of Rheumatology and Clin. Immunology
> Hugstetterstr. 55
> 79106 Freiburg (Germany)
> Tel.: 0761 270 37 84
> Pager: 0168 73 22 60 0
> E-mail: [EMAIL PROTECTED]
>
> _______________________________________________
> sapdb.general mailing list
> [EMAIL PROTECTED]
> http://listserv.sap.com/mailman/listinfo/sapdb.general
>
_______________________________________________
sapdb.general mailing list
[EMAIL PROTECTED]
http://listserv.sap.com/mailman/listinfo/sapdb.general
