Depression, anxiety, bipolar disorder and schizophrenia trials live or die on clinician ratings and self-reports. Capture keeps rater scales, participant questionnaires and suicidality assessment in one casebook, with every score attributed to the person who gave it.
Free sandbox · No credit card · 21 CFR Part 11 aligned
Apparent sadness (0 to 6)
Reported sadness (0 to 6)
Inner tension (0 to 6)
Suicidal thoughts (0 to 6)
High-priority query: follow up per safety plan
1
casebook for clinician scales, self-reports and safety
0
apps for participants to install
100%
of values attributed to a named user
0
setup fees
What matters in psychiatry trials
Endpoints
In most therapeutic areas the primary endpoint is a lab value, an image or an event. In psychiatry it is usually a number assigned by a trained rater after a structured interview. That makes the rating process itself part of the measurement: who rated, how they were trained, whether the same rater followed the participant, and whether they knew anything that could bias the score.
An EDC for psychiatry therefore has to do more than store totals. It should capture each item so scores can be checked and verified, attribute every value to the person who entered it, and make it easy to see when a participant's rater changed. In Capture, each clinician scale is a site form with one field per item and a range check on each, and the field-level audit trail records who entered and changed every value.
Major depressive disorder trials commonly use the MADRS or the 17-item HAM-D as the primary endpoint, with CGI-S and CGI-I, and self-reports such as the PHQ-9. Generalised anxiety disorder trials use the HAM-A and often the GAD-7. Schizophrenia trials rely on the PANSS; bipolar mania trials on the YMRS. Most also include sleep and functioning measures, such as the Insomnia Severity Index.
Side effects can reveal treatment assignment to a rater who also manages safety. Many psychiatry trials therefore separate the efficacy rater from the clinician who reviews adverse events. That separation is set in your procedures and role assignments; in Capture, blinded roles never see treatment allocation, and the audit trail shows which user entered each form.
Participants
96
Same rater
89
Rater changed
7
Rater left in month 3
Safety
Positive answers on suicidality items should reach the investigator the same day. Custom-value edit checks on the relevant C-SSRS questions, PHQ-9 item 9 or a scale's suicidal thoughts item raise a high-priority auto-query in the site action list, and serious events flow into the adverse event workflow with routed safety alerts.
Edit checks / auto-queries
2Type
Range High
Operator
Greater than
Value
180
Priority: High
Type
Range Low
Operator
Less than
Value
80
Priority: Normal
Query raised automatically
Value 192 violates limit (180). Please verify.
Self-report
Self-reported outcomes such as the PHQ-9, GAD-7 and sleep measures run in the phone browser from a secure link, with reminders and completion windows. Participants do not install anything, and sites see completion on the dashboard.
Evening diary
Question 3 of 8
How would you rate your fatigue today?
0 = no fatigue, 10 = worst imaginable
Schedule
Weekly or fortnightly ratings early on, when change is fastest, then less often. Built once, the schedule applies identically at every site, with visit windows flagging late assessments.
| Assessment | Scr | BL | W1 | W2 | W4 | W6 | W8 |
|---|---|---|---|---|---|---|---|
| MADRS | |||||||
| CGI-S / CGI-I | |||||||
| PHQ-9 (ePRO) | |||||||
| C-SSRS | |||||||
| AE review |
Add a clinician scale, a PHQ-9 and a C-SSRS to a sample study in the free sandbox.
Design
High and variable placebo response has sunk many psychiatric drug trials. Much of the answer lies in design (population, endpoints, duration, number of sites), but data practices contribute too. Inflated baseline scores, where raters under enrolment pressure score participants just above the entry threshold, show up as large early "improvement" in both arms. Inconsistent rating between sites adds noise that masks real effects.
Three data practices help. First, record item-level baseline scores so blinded data review can look for clustering just above the inclusion threshold. Second, keep raters consistent and trained, and review rater attribution before interim analyses. Third, collect self-reported outcomes alongside clinician ratings, since a divergence between the two can signal rating problems. None of these needs special software; they need item-level, attributed data in one place.
The primary endpoints are usually clinician rating scales, so rater attribution, item-level capture and consistent administration matter as much as the values. Suicidality assessment at each visit is also common.
Yes, as site forms with one field per item and range checks. Obtain each scale and any licence from its rights holder; licence details can be recorded on the form.
Custom-value edit checks raise high-priority auto-queries to the site, and serious events go through the adverse event workflow with routed alerts and sign-off.
Blinded roles never see treatment allocation, and the audit trail records who entered each form. Keeping raters separate from adverse event review is defined in your procedures and role assignments.
No. Questionnaires open in the phone browser from a secure link.
Yes. The free sandbox has every feature, with no credit card and no time limit.
Keep exploring
Rater scales, self-reports and suicidality in one casebook. Free sandbox.