Skip to main content

Table 1 Relative sizes of patient populations and outcomes across population types

From: The “RCT augmentation”: a novel simulation method to add patient heterogeneity into phase III trials

Population type % increase in RCT-eligible patient pool (95% CIc) Outcome Average ΔCGI-S (95% CId)
Patients initiating drug D1
 “Real-world population” (SOHO cohort) −0.88 (− 0.90, − 0.85)
 “RCT population” (22.6% of SOHO cohort patients) 0 − 0.78 (− 0.83, − 0.72)
Augmented RCT populations:
  Patients with illness duration between 1 and 3 yearsa 17.9% (15.5, 20.5%) −1.04 (− 1.18, − 0.89)
  Patients with one past suicide attempta 18.1% (15.7, 20.8%) − 0.94 (− 1.08, − 0.81)
  Private practice patientsa 14.7% (12.6, 17.1%) −1.03 (− 1.15, − 0.90)
  Patients with history of alcohol abusea 6.3% (5.0, 7.8%) −0.60 (− 0.79, − 0.42)
  Patients with history of drug abusea 5.6% (4.4, 7.0%) −0.78 (− 0.99, − 0.57)
  Patients with illness duration between 1 and 3 years and/or one past suicide attemptb 37.7% (34.0, 41.9%) −1.00 (− 1.10, − 0.90)
  Patients with illness duration between 1 and 3 years and/or private practice patientsb 35.1% (31.4, 39.0%) −1.05 (− 1.15, − 0.96)
  Patients with illness duration between 1 and 3 years and/or history of alcohol abuseb 25.0% (22.1, 28.3%) − 0.93 (− 1.05, − 0.81)
Patients initiating drug D2
 “Real-world population” (SOHO cohort) −0.71 (− 0.75, − 0.67)
 “RCT population” (25.7% of SOHO cohort patients) 0 −0.64 (− 0.72, − 0.57)
Augmented RCT populations:
  Patients with illness duration of 1–3 yearsa 11.7% (8.9, 15.1%) −0.77 (− 1.03, − 0.51)
  Patients with one past suicide attempta 15.8% (12.6, 19.8%) −0.57 (− 0.71, − 0.42)
  Private practice patientsa 14.9% (11.7, 18.7%) −0.74 (− 0.93, − 0.55)
  Patients with history of alcohol abusea 8.7% (6.4, 11.6%) −0.63 (− 0.97, − 0.30)
  Patients with history of drug abusea 4.3% (2.8, 6.5%) −0.50 (− 0.83, − 0.17)
  Patients with illness duration between 1 and 3 years and/or one past suicide attemptb 29.4% (24.5, 35.0%) −0.62 (− 0.76, − 0.49)
  Patients with illness duration between 1 and 3 years and/or private practice patientsb 29.6% (24.7, 35.2%) −0.79 (− 0.94, − 0.64)
  Patients with illness duration between 1 and 3 years and/or history of alcohol abuseb 20.5% (16.7, 25.1%) −0.73 (− 0.93, − 0.52)
  1. aplus meeting the remaining 5 RCT eligibility criteria. bplus meeting the remaining 4 RCT eligibility criteria; cThe Clopper-Pearson interval was used to calculate the 95% confidence interval; CI confidence interval; dThe confidence interval (CI) was calculated under the assumption that ΔCGI-S had a normal distribution