By using db<>fiddle, you agree to license everything you submit by Creative Commons CC0.
CREATE TABLE therapy
(Id int, Hours int, TherapyType int, Value numeric(10,2), Is_Active bit)
;
INSERT INTO therapy (Id, Hours, TherapyType, Value, Is_Active)
VALUES
(9704, 24, 1, 0.5, 'true'),
(9705, 1, 1, 0.5, 'true'),
(9706, 2, 1, 0.5, 'true'),
(9707, 1, 1, 0.5, 'true'),
(9708, 4, 1, 0.6, 'true'),
(9709, 5, 1, 0.75, 'true'),
(97010, 8, 1, 0.75, 'true')
;
7 rows affected
select t.*
from (select t.*,
lag(value) over (order by id) as prev_value
from therapy t
) t
where prev_value is null or prev_value <> value;
| Id | Hours | TherapyType | Value | Is_Active | prev_value |
|---|---|---|---|---|---|
| 9704 | 24 | 1 | 0.50 | True | null |
| 9708 | 4 | 1 | 0.60 | True | 0.50 |
| 9709 | 5 | 1 | 0.75 | True | 0.60 |