# Registry codebook: which ClinicalTrials.gov registrations are trials this review would include

Each record is a registration on ClinicalTrials.gov found by searching interventions for ketamine, esketamine, or arketamine and any field for depression or mood. You see its titles, conditions, design, arms, interventions, outcome measures, and status, but no results. Decide, for each, whether the trial it registers is one the review would include if it reported.

## What counts

A registration is **eligible** when the trial it describes is:

1. **Randomized**, with
2. **an arm that receives ketamine or esketamine** (racemic, S-, or arketamine, any route), and
3. **an inert control arm** (saline, placebo, or no study drug, with otherwise the same care),
4. **around a procedure or childbirth**: the drug is given before, during, or after surgery, another procedure under general anesthesia, regional anesthesia, or sedation, or labor or delivery (electroconvulsive therapy doesn't count),
5. **with a depression measure after it**: a depression scale or the presence of depression among its primary or secondary outcomes.

Otherwise give the first reason that applies: `not_randomized`, `no_ketamine_arm`, `no_inert_control`, `not_perioperative` (including psychiatric trials outside any procedure), `no_depression_outcome`, `ect`, `other` (explain in `note`).

For eligible registrations, also give `conscious_dosing` as the design codebook defines it (`none` if every dose is given under general anesthesia and ends by the end of surgery, `some` if any is given while the participant is conscious, `unclear`), from the arm and intervention descriptions.

## Output

One JSON object per record, in the order given, as a JSON array written to the file named in your task:

```json
{"nct": "NCT01234567", "eligible": true, "exclusion_reason": null, "conscious_dosing": "none", "note": ""}
```

Decide from the record alone. Don't look up publications, results, or anyone else's coding.