30,60,90 Days Totals in sql query

Question

Hi All,
Here I  have written a query to find the total for 30Days to calculate costs....
These 30Days Amount will be populated into another table based on joins.
I know we can do by case in the final update but lot of logic is going into temptables1 and 2, so please help...
select distinct ip.membernbr,ip.patcom,ip.admitdate,ip.dischargerdate,dateadd(day,1,ip.dishargedate) as afterdishcargedatestart,
dateadd(day,30,ip.dishargedate) as afterdishcargedateend, cd.specificdateofservice as ipdos, cd.specificdateofservicethry
as ipdosthru
into #inpatientdoshospice30
from tbl_inpatient ip
join claimsdetail cd on cd.membernbr=ip.membernbr
join claims c on c.claimsseqnbr=cd.claimnsseqnbr
where c.formnbr='inp'
and cd.specificdateofservice between dateadd(day,1,ip.dishargedate)
 and dateadd(day,30,ip.dishargedate) 
and cd.specificdateofservicethru
between dateadd(day,1,ip.dishargedate)  and dateadd(day,30,ip.dishargedate) 
--find part c hospice claims lines that that are already included in the inpatientcost(c.formnbr='INP')
cost;
--doing this in a separate step in case there are multiple inpatient claims which dos dates that overlap
(e.g; 1/15 to 1/20 and ---1/18 to 1/20) causing us to overstate dollars
select distinct ip.patcom,cd.claimssequencenbr
into #Hospiceinpatientclaimnslines30
from
#inpatientdoshospice30 ip 
join claimsdetail cd on cd.membernbr=ip.membernbr
and cd.specificdateofservice between dateadd(day,1,ip.dishargedate)  and dateadd(day,30,ip.dishargedate) 
and cd.specificdateofservicethru between dateadd(day,1,ip.dishargedate)  and dateadd(day,30,ip.dishargedate) 
join claims c on c.claimsseqnbr=cd.claimnsseqnbr
where c.formnbr='Hospice'
--sum all hospice claims excluding those already included in the inpatient
update pp
set homehospicecost30day=isnull(sub.hospicecost,0)
from inpatient pp
left join(
select ip.patcom,sum(isnull(cd.netamt,0.00)) as hospicecost
from tbl_inpatient ip
join claimsdetail cd on cd.membernbr=ip.membernbr
and cd.specificdateofservice between dateadd(day,1,ip.dishargedate)  and dateadd(day,30,ip.dishargedate) 
and cd.specificdateofservicethru between dateadd(day,1,ip.dishargedate)  and dateadd(day,30,ip.dishargedate) 
join claims c on c.claimsseqnbr=cd.claimnsseqnbr
where (c.formnbr='Hospice' or (c.formnbr='PartB' and cd.placeofservice in ('34')))
and cd.claimsseqnbr not in (select cklaimseqnbr from #hospiceinpatientclaimlines30 sub where sub.patcom=ip.patcom)
group by ip.patcom
)sub
on pp.patcom=sub.patcom
I really appreciate your help.
Thanks,
Kalyan.

Answer 1

The UPDATE can be written this way:
update pp
set    homehospicecost30day = isnull(sub.hospicecost,0)
from   inpatient pp
left   join(select ip.patcom,
                   sum(CASE WHEN cd.specificdateofservice between 
                                 dateadd(day,1,ip.dishargedate)  AND
                                 dateadd(day,30,ip.dishargedate) 
                             AND cd.specificdateofservicethru between
                                 dateadd(day,1,ip.dishargedate)  and 
                                 dateadd(day,30,ip.dishargedate) 
                            THEN isnull(cd.netamt,0.00)
                            ELSE 0
                        END) as hospicecost30
            from   tbl_inpatient ip
            join   claimsdetail cd on cd.membernbr=ip.membernbr
            join   claims c on c.claimsseqnbr=cd.claimnsseqnbr
             where  (c.formnbr='Hospice' or
                    (c.formnbr='PartB' and cd.placeofservice in ('34')))
               and cd.claimsseqnbr not in (select cklaimseqnbr from
                                           #hospiceinpatientclaimlines30 sub
                                           where sub.patcom=ip.patcom)
             group by ip.patcom)sub
on pp.patcom=sub.patcom
The pattern shows how you can add further days.
I was not able to figure out the meaning of the temp tables, and since you said that you typed it from memory without access to the code, I don't trust that you got everything right. Give it a second look on Monday.
Erland Sommarskog, SQL Server MVP, [email protected]