Modifying an Aggregate - sql

Question

I am trying to get my surgeries to calculate at different rates and I am struggling with it. For example, patient 58903 has 4 total surgeries as shown below. However, I would like the first surgery to calculate at 100% of the PPO SURG rate (so $4232), the second one at 50%, and all remaining surgeries at 25% of the main PPO SURG rate. My current code returns $16,929 for patient 5903 which is just $4232*4. My desired output for the SURG Total below is $8,464 (4232+2116+1058+1058).
My Current Code:
SELECT
DISTINCT PATNO,
SUM( PPOSURG) AS 'Surg Total',
SUM( PPONONSURG) AS 'Non Surg Total',
SUM( PPODRUG) AS 'Drug Total',
INSNME,
IIF( SUM( PPOSURG) IS NOT NULL,
SUM( PPOSURG) + SUM(CASE WHEN REV=278 THEN (AMT1)*0.446 END),
ISNULL(SUM( PPODRUG),0)+ISNULL( SUM( PPONONSURG),0)) AS 'Claim Total'
FROM
[OPGRACE$] AS GR --Main Table
LEFT JOIN [BCSURGOP$] AS SRG ON GR.CPTCDA=SRG.[CPTCODESURG] --SURG joined on cpt
LEFT JOIN [BCBSDRUG$] AS DRUG ON GR.CPTCDA=DRUG.[CPT CODE] --DRUG Schedules joined on cpt
LEFT JOIN [BCBSNONSURGOP$] AS NSRG ON GR.CPTCDA=NSRG.[CPT CODE] --Non-SURG joined on cpt
WHERE QTY>0 AND PATNO>0
GROUP BY PATNO,INSNME
ORDER BY PATNO ASC
I tried doing something like this but you can't have an aggregate in a SUM
SUM(CASE WHEN COUNT(CPTCODESURG)=1 THEN PPOSURG ELSE PPOSURG*0.5 END) + SUM(CASE WHEN REV=278 THEN (AMT1)*0.446 END),
Here is my output for just patient 58903
SURG Total
Non Surg Total
Drug Total
CLAIM Total
16929.472
3103
179
22598.84
Here is a blurb of my joined table showing how patient 58903 shows up:
PName
Rev
CPT
PPOSURG
58903
360
29882
4232.368
58903
360
29882
4232.368
58903
360
29882
4232.368
58903
360
29882
4232.368
Would a PARTITION be the way to go here? A subquery? Can I somehow use a case statement? I don't expect anyone to write my code but literally any ideas would be extremely helpful, I have been really stuck on this.

Answer 1

If I understand you correctly you just need a row number partitioned by the patient and then a CASE expression to convert that into a multiplier. I've added an id column to the sample data to allow for an order by (which you need for a row number).
declare #Test table (id int identity(1,1), PName int, Rev int, CPT int, PPOSURG money);
insert into #Test (PName, Rev, CPT, PPOSURG)
values
(58903, 360, 29882, 4232.368),
(58903, 360, 29882, 4232.368),
(58903, 360, 29882, 4232.368),
(58903, 360, 29882, 4232.368);
with cte as (
select *
, row_number() over (partition by PName order by id) rn
from #Test
)
select PName, Rev, CPT
, cast(sum(PPOSURG * case rn when 1 then 1.00 when 2 then 0.50 else 0.25 end) as decimal(9,2)) Total
from cte
group by PName, Rev, CPT;