sql- Number of actual records added - sql

Question

I am trying to generate a report based on counselling and actual to see how many counseling were converted to actual.
I have table which has Patient, Booking Category, Doctor, Booking.
SQL Fiddle
Following are the business rules
Fetch all the bookings made by patients if they made booking after counseling
The user can take counseling booking first and book one or more bookings that falls under the same counselling group. Example: They can do a first booking Counseling Dental then can book anything mapped to that group. i. e Dentail surgery and Teeth repair. These information maintained in CounselingXActualBookingCategoryMap table
If patient book more than one actual booking consider only one, which is not cancelled
If counselling is cancelled dont consider.
If actual is cancelled dont consider.
Get a total number of counseling bookings(it need not to be a actual booking doctor) and actual bookings for each doctor. Any one can be a counseling doctor. But report we consider only the doctor who did actual
Here is the SQL i tried. But its taking all records instead of one actual for that group.
SELECT P.Name As Patient, SurgeryB.BookingId, SurgeryB.BookingDate, C.CategoryName,
d.name as doctor
FROM BOOKING CounsB
INNER JOIN BOOKING SurgeryB
ON CounsB.PatientId = SurgeryB.PatientId AND ISNULL(SurgeryB.IsCancelled,0)=0
INNER JOIN CounselingXActualBookingCategoryMap MAP
ON MAP.CounselingCategoryId = CounsB.CategoryId
AND MAP.ActualBookingCategoryId = SurgeryB.CategoryId
AND SurgeryB.BookingDate > CounsB.BookingDate
INNER JOIN PATIENT P ON P.PATIENTID = SurgeryB.PATIENTID
inner join category c on c.categoryid= SurgeryB.CategoryId
inner join doctor d on SurgeryB.doctorid = d.doctorid
WHERE ISNULL(CounsB.IsCancelled,0)=0
order by p.patientid
Issue:
I do not want to see a record that Alex made booking for Teeth repair, since he already did actual for same group Dental and category Dental Surgery
I want to get counseling and actual made bookings. Example 100 counseling made and only 70 actual booked like that for each counseling category.

Answer 1

I have added to your code as below:
SELECT P.Name As Patient, SurgeryB.BookingId, SurgeryB.BookingDate, C.CategoryName, map.GroupId, map.CounselingCategoryId, map.ActualBookingCategoryId,
d.name as doctor
into #temp
FROM BOOKING CounsB
INNER JOIN BOOKING SurgeryB
ON CounsB.PatientId = SurgeryB.PatientId AND ISNULL(SurgeryB.IsCancelled,0)=0
INNER JOIN CounselingXActualBookingCategoryMap MAP
ON MAP.CounselingCategoryId = CounsB.CategoryId
AND MAP.ActualBookingCategoryId = SurgeryB.CategoryId
AND SurgeryB.BookingDate > CounsB.BookingDate
INNER JOIN PATIENT P ON P.PATIENTID = SurgeryB.PATIENTID
inner join category c on c.categoryid= SurgeryB.CategoryId
inner join doctor d on SurgeryB.doctorid = d.doctorid
WHERE ISNULL(CounsB.IsCancelled,0)=0
order by p.patientid
;with cte as
(
select Patient, BookingId, BookingDate, CategoryName, Doctor, ROW_NUMBER() over (partition by Patient, GroupId order by BookingDate) row_num
from #temp t
)
select Patient, BookingId, BookingDate, CategoryName, Doctor from cte where row_num = 1
drop table #temp
HTH!

Answer 2

I have an answer to the query you gave as an example, which is described in business rule 1 and to the one in business rule 6.
I made some slight changes to your tables, since I had to re-create them to test the query anyway.
create table patient
(
id int,
name varchar(8)
);
create table doctor
(
id int,
name varchar(8)
);
create table category
(
id int,
name varchar(20),
counseling_or_surgery varchar(10)
);
create table map
(
counseling_category_id int,
surgery_category_id int,
group_id int
);
create table booking
(
id int,
patient_id int,
category_id int,
doctor_id int,
booking_date date,
cancelled bit
);
Here are the inserts to set up the data:
insert into patient (id, name) values (1, 'Alex');
insert into patient (id, name) values (2, 'John');
insert into patient (id, name) values (3, 'Mark');
insert into patient (id, name) values (4, 'Stephen');
insert into doctor (id, name) values (1, 'Doctor 1');
insert into doctor (id, name) values (2, 'Doctor 2');
insert into doctor (id, name) values (3, 'Doctor 3');
insert into category (id, name, counseling_or_surgery) values (1, 'Counseling Dental', 'Counseling');
insert into category (id, name, counseling_or_surgery) values (2, 'Counseling Nose', 'Counseling');
insert into category (id, name, counseling_or_surgery) values (3, 'Dental surgery', 'Surgery');
insert into category (id, name, counseling_or_surgery) values (4, 'Teeth repair', 'Surgery');
insert into category (id, name, counseling_or_surgery) values (5, 'Nose Surgery', 'Surgery');
insert into category (id, name, counseling_or_surgery) values (6, 'Nose adjustment', 'Surgery');
insert into map (counseling_category_id, surgery_category_id, group_id) values (1, 3, 1);
insert into map (counseling_category_id, surgery_category_id, group_id) values (1, 4, 1);
insert into map (counseling_category_id, surgery_category_id, group_id) values (2, 5, 2);
insert into map (counseling_category_id, surgery_category_id, group_id) values (2, 6, 2);
insert into booking (id, patient_id, category_id, doctor_id, booking_date, cancelled) values ( 1, 1, 1, 1, '2017-11-20', 'false');
insert into booking (id, patient_id, category_id, doctor_id, booking_date, cancelled) values ( 2, 1, 3, 3, '2017-11-25', 'false');
insert into booking (id, patient_id, category_id, doctor_id, booking_date, cancelled) values ( 3, 1, 4, 2, '2017-11-25', 'false');
insert into booking (id, patient_id, category_id, doctor_id, booking_date, cancelled) values ( 4, 1, 2, 1, '2017-11-21', 'true');
insert into booking (id, patient_id, category_id, doctor_id, booking_date, cancelled) values ( 5, 1, 2, 2, '2017-11-26', 'false');
insert into booking (id, patient_id, category_id, doctor_id, booking_date, cancelled) values ( 6, 1, 5, 3, '2017-11-28', 'false');
insert into booking (id, patient_id, category_id, doctor_id, booking_date, cancelled) values ( 7, 3, 2, 2, '2017-11-26', 'false');
insert into booking (id, patient_id, category_id, doctor_id, booking_date, cancelled) values ( 8, 3, 6, 2, '2017-11-26', 'true');
insert into booking (id, patient_id, category_id, doctor_id, booking_date, cancelled) values ( 9, 4, 2, 2, '2017-11-29', 'false');
insert into booking (id, patient_id, category_id, doctor_id, booking_date, cancelled) values (10, 4, 5, 1, '2017-11-29', 'false');
Given these tables, here is the query for business rule 1:
select
c.name as patient,
b.id as booking_id,
b.booking_date as first_surgery_date,
d.name as first_surgery,
e.name as doctor
from
(
select
r.patient_id,
min(r.id) as booking_id,
t.group_id
from
booking r
join
category s
on r.category_id = s.id
join
map t
on t.surgery_category_id = r.category_id
where
r.cancelled = 0 and
s.counseling_or_surgery = 'Surgery'
group by
r.patient_id,
t.group_id
) a
join
booking b
on a.booking_id = b.id
join
patient c
on b.patient_id = c.id
join
category d
on b.category_id = d.id
join
doctor e
on b.doctor_id = e.id;
and the results are:
patient booking_id first_surgery_date first_surgery doctor
---------------- ----------- ------------------ ---------------- ----------------
Stephen 10 2017-11-29 Nose Surgery Doctor 1
Alex 2 2017-11-25 Dental surgery Doctor 3
Alex 6 2017-11-28 Nose Surgery Doctor 3
(3 rows affected)
Here's the query for business rule 6. It counts cancellations too; easy to change if that's not what you want.
select
z.name as doctor,
y.total_counseling_bookings,
x.total_surgery_bookings
from
(
select
doctor_id,
count(*) as total_surgery_bookings
from
booking a
join
category b
on a.category_id = b.id
where
b.counseling_or_surgery = 'Surgery'
group by
doctor_id
) x
full outer join
(
select
doctor_id,
count(*) as total_counseling_bookings
from
booking c
join
category d
on c.category_id = d.id
where
d.counseling_or_surgery = 'Counseling'
group by
doctor_id
) y
on x.doctor_id = y.doctor_id
join
doctor z
on x.doctor_id = z.id
and here are the results for this one:
doctor total_counseling_bookings total_surgery_bookings
-------- ------------------------- ----------------------
Doctor 1 2 1
Doctor 2 3 2
Doctor 3 NULL 2
(3 rows affected)