Table 2.
Moderators included in pooled dataset.
Domain | Sources of Information | Harmonized Variable(s) | % available | N available | Tier/Analysis | Summary statistics and case counts |
---|---|---|---|---|---|---|
Study design features | Study methods | |||||
Study Placebo Type | 100.00 | 720 | Tier 1a | n = 477 psychoactive; n = 243 inert | ||
Crossover or Parallel Arm | 100.00 | 720 | Tier 1a | n = 242 crossover; n = 478 parallel | ||
Psychiatric diagnoses (principal and comorbid) | Patient-level text and ICD/DSM codes; study-level inclusions/exclusions | Dummy-Coded diagnoses: | ||||
Principal Diagnosis | 100.00 | 720 | Tier 1b | n = 579 MDD; n = 66 BD; n = 74 PTSD; n = 1 Borderline Personality Disorder | ||
Major Depressive Disorder (MDD) | 100.00 | 720 | Tier 1b | n = 627 yes; n = 93 no | ||
Bipolar (I or II) Disorder | 99.44 | 716 | Nonec | n = 66 yes; n = 650 no | ||
Post-traumatic Stress Disorder | 72.50 | 522 | Tier 2, M* 2d | n = 118 yes; n = 365 no | ||
Any Anxiety Disorder | 72.36 | 521 | Tier 2, M* 2d | n = 285 yes; n = 236 no | ||
Generalized Anxiety Disorder | 64.58 | 465 | Tier 2 | n = 101 yes; n = 364 no | ||
Social Anxiety Disorder | 58.06 | 418 | Tier 2 | n = 99 yes; n = 319 no | ||
Obsessive-Compulsive Disorder | 57.92 | 417 | Tier 2 | n = 18 yes; n = 399 no | ||
Panic Disorder | 57.78 | 416 | Tier 2 | n = 51 yes; n = 365 no | ||
Agoraphobia | 57.78 | 416 | Tier 2 | n = 36 yes; n = 380 no | ||
Substance Use Disorder (lifetime) | 42.92 | 309 | Tier 2 | n = 26 yes; n = 283 no | ||
Alcohol Use Disorder (lifetime) | 40.28 | 290 | Tier 2 | n = 66 yes; n = 224 no | ||
Eating Disorder | 36.11 | 260 | None | n = 9 yes; n = 251 no | ||
Dysthymia | 30.69 | 221 | None | n = 31 yes; n = 190 no | ||
Any Pain Disorder | 30.28 | 218 | None | n = 48 yes; n = 170 no | ||
Attention-Deficit Disorder | 27.64 | 199 | None | n = 9 yes; n = 190 no | ||
Specific Phobia | 26.94 | 194 | None | n = 21 yes; n = 173 no | ||
Any Personality Disorder | 25.42 | 183 | None | n = 49 yes; n = 134 no | ||
Borderline Personality Disorder | 24.31 | 175 | None | n = 20 yes; n = 155 no | ||
Demographics | Patient-level integers, text, and coded variables | |||||
Age | 99.72 | 718 | Tier 1b | mean = 42.4; SD = 13.3; range = 18–72 | ||
Female or Male | 99.58 | 717 | Tier 1b | n = 386 yes; n = 331 no | ||
Years of Education | 74.44 | 536 | Tier 2, M* 2 g | mean = 14.6; SD = 2.9; range = 4–25 | ||
Marital Status | 50.83 | 366 | Tier 2, M* 2 g | n = 127 married (or equivalent); n = 151 single/never married; n = 57 divorced or separated; n = 6 widowed; n = 25 not married (not further specified) | ||
Income Level/Range | 15.83 | 114 | None | n = 46 $0–25 K; n = 33 $25-50 K; n = 10 $50-70 K; n = 25 > $70 K | ||
Race, Ethnicity, Culture | Patient-level text and coded variables | |||||
Study Performed in the US (y/n) | 100.00 | 720 | Tier 1b | n = 480 yes; n = 240 no | ||
EthnoRacial Category | 89.58 | 645 | Tier 2, M* 2a | n = 469 non-Hispanic white; n = 61 Black; n = 66 Asian; n = 26 Hispanic; n = 14 more than one race; n = 9 other non-White | ||
Hispanic/Latinx (y/n) | 72.22 | 520 | Tier 2 | n = 29 yes; n = 491 no | ||
Depression course and characteristics | Patient-level integers and coded variables; study-level inclusions/exclusions | |||||
Inpatient vs. Outpatient | 100.00 | 720 | Tier 1b | n = 461 outpatient; n = 259 inpatient | ||
Study-level TRD Threshold | 100.00 | 720 | Tier 1b | n = 280 no criterion; n = 71 1 failure; n = 253 2 failures; n = 16 ≥ 3 failures | ||
Suicidality severity at baseline (MADRS item 10) | 70.42 | 507 | Tier 2 | mean = 2.01; SD = 1.5; range = 0–6 | ||
Duration of Current MDE (months)a | 61.11 | 440 | Tier 2, M* 2c | mean = 52.5; SD = 95.7; range = 0.47–768 | ||
Recurrent MDD (y/n) | 55.14 | 397 | Tier 2, M* 2c | n = 301 yes; n = 6 no | ||
Number of Depressive Episodesa | 52.64 | 379 | Tier 2, M* 2c | mean = 8.0; SD = 10.7; range = 1–78 | ||
Age of illness onset | 48.8 | 351 | Tier 2 | mean = 23.7; SD = 12.2; range = 4–60 | ||
Number of failed ADM trials (patient-level) | 42.78 | 308 | Tier 2 | mean = 4.44; SD = 2.3; range = 1–14 | ||
Childhood trauma questionnaire (or any other available measure of childhood trauma) | 20.3 | 146 | None | mean = 56.4; SD = 22.4; range = 0–119 | ||
Concurrent Medications and Substances | Patient-level integers and coded variables; study-level inclusions/exclusions | |||||
Any Concurrent Psych Med (y/n) | 76.25 | 549 | Tier 2, M* 2b | n = 262 yes; n = 287 no | ||
Any Benzodiazepine (y/n) | 65.56 | 472 | Tier 2, M* 2b | n = 42 yes; n = 430 no | ||
Number of Concurrent Psych Meds | 51.53 | 371 | Tier 2 | mean = 0.30; SD = 0.61; range = 0–3 | ||
Smoker (Dummy-Coded) | 53.47 | 385 | Tier 2, M* 2 f | n = 109 yes; n = 76 no | ||
Biological measurements | Patient-level integers | |||||
Ketamine dose (mg) | 82.64 | 595 | Tier 2 | Ketamine arms only: mean = 42.1; SD = 15.4; range = 18.5–100.7b | ||
Dosing weight (kg)a | 60.83 | 438 | Tier 2 | mean = 79.03; SD = 19.8; range = 35–174 | ||
BMIa | 55.56 | 400 | Tier 2, M* 2 f | mean = 27.23; SD = 6.2; range = 17–60.8 | ||
Systolic Blood Pressure | 53.89 | 388 | Tier 2, M* 2e | mean = 121.49; SD = 14.3; range = 86–166 | ||
Diastolic Blood Pressure | 53.89 | 388 | Tier 2, M* 2e | mean = 75.24; SD = 10.2; range = 48–112 | ||
Baseline Pulsea | 52.22 | 376 | Tier 2 | mean = 70.89; SD = 12.7; range = 43–164 | ||
Baseline serum BDNF level (ng/mL) | 16.90 | 122 | None | mean = 921.72; SD = 1629.1; range = 0 9419 | ||
Baseline Pulse-Ox | 17.60 | 127 | None | mean = 97.77; SD = 1.9; range = 90–100 | ||
Baseline Respiration Rate | 10.83 | 78 | None | mean = 15.44; SD = 3.6; range = 5–28 |
aTier 1 variables are those available in >99% of patients in the pooled dataset. These first two Tier 1 variables, which related to study design features rather than to patient-level characteristics, were excluded from all combined moderator analyses, as the purpose was to identify combinations of patient (not study) characteristics that predicted differential improvement following ketamine (relative to placebo).
bTier 1 variables are those available in >99% of patients in the pooled dataset. These 7 Tier 1 variables were related to patient-level characteristics and thus were included in the Tier 1 M*, and also retained in all M* analyses across all Tiers 2a-f, as their inclusion never reduced the number of studies/patients available for any analysis and could only increase predictive power for the data-driven approach.
cBipolar I/II diagnosis was highly non-orthogonal and virtually redundant with the MDD diagnosis variable, as bipolar and unipolar depression diagnoses are mutually exclusive. To eliminate this redundancy in hypothesis tests, this variable was not tested as a moderator in any analysis.