SQL Server 2017

By using db<>fiddle, you agree to license everything you submit by Creative Commons CC0.

create table t1(id int, dateval datetime,title varchar(10),userid int)
insert into t1 values
(1,'2019-04-29 14:15:55.567','A',1222),
(2,'2019-04-29 14:25:13.530','A',1222),
(3,'2019-04-29 15:17:07.210','A',1222),
(4,'2019-04-29 16:05:49.067','B',1048),
(5,'2019-04-29 16:36:37.330','C',681),
(6,'2019-04-29 16:37:16.250','C',681),
(7,'2019-04-29 16:37:49.160','D',681),
(8,'2019-04-29 16:38:48.803','C',681)
8 rows affected

select id,dateval, title, userid from
(
select *, lag(title) over(order by dateval desc) as prevtitle
from t1
)A where title<>prevtitle or prevtitle is null
order by dateval


id
dateval
title
userid


3
2019-04-29 15:17:07.210
A
1222

4
2019-04-29 16:05:49.067
B
1048

6
2019-04-29 16:37:16.250
C
681

7
2019-04-29 16:37:49.160
D
681

8
2019-04-29 16:38:48.803
C
681

id	dateval	title	userid
3	2019-04-29 15:17:07.210	A	1222
4	2019-04-29 16:05:49.067	B	1048
6	2019-04-29 16:37:16.250	C	681
7	2019-04-29 16:37:49.160	D	681
8	2019-04-29 16:38:48.803	C	681