Dataset Viewer
Auto-converted to Parquet Duplicate
trial_id
string
site_id
string
week
int64
enrolled_this_week
int64
screen_fail_rate
float64
protocol_deviation_rate
float64
visit_adherence_rate
float64
dropout_rate
float64
query_backlog_per_patient
float64
outcome_drift_z
float64
dsmb_flags_next_4w
int64
label_trial_fail_risk_next_30d
int64
TRIAL_DF03
S01
4
18
0.22
0.06
0.91
0.05
1.2
0.4
0
0
TRIAL_DF03
S01
6
16
0.25
0.07
0.89
0.06
1.5
0.6
0
0
TRIAL_DF03
S02
8
14
0.29
0.09
0.86
0.08
2.1
0.9
0
0
TRIAL_DF03
S02
10
12
0.33
0.11
0.83
0.1
2.8
1.2
1
1
TRIAL_DF03
S03
12
10
0.36
0.12
0.81
0.12
3.2
1.4
1
1
TRIAL_DF04
S01
5
20
0.2
0.05
0.93
0.04
1
0.3
0
0
TRIAL_DF04
S02
7
17
0.24
0.07
0.9
0.05
1.3
0.5
0
0
TRIAL_DF04
S03
9
13
0.31
0.1
0.84
0.09
2.4
1.1
1
1
TRIAL_DF04
S03
11
11
0.35
0.12
0.8
0.13
3.5
1.6
1
1
TRIAL_DF04
S04
6
15
0.27
0.08
0.88
0.07
1.9
0.8
0
0

Clinical Quad Recruitment Protocol Adherence Outcome Drift v0.1

Each row is a site week snapshot.

Core quad

Recruitment quality
Protocol deviation
Visit adherence
Outcome drift

Target

label_trial_fail_risk_next_30d

Files

data/train.csv
data/tester.csv
scorer.py

Evaluation

Run model on data/tester.csv
Return predictions row aligned
Score with scorer.py

Columns

trial_id
site_id
week
enrolled_this_week
screen_fail_rate
protocol_deviation_rate
visit_adherence_rate
dropout_rate
query_backlog_per_patient
outcome_drift_z
dsmb_flags_next_4w
label_trial_fail_risk_next_30d

License

MIT

Downloads last month
7