Table 1.
Predictora | N | Studies/models | Median/range of predictors in the models | Significantb predictor coefficient median/range | Model D/I/E | Model accuracy measures median/rangec |
---|---|---|---|---|---|---|
Response | ||||||
Non-modifiable Multivariable Candidate Predictors | ||||||
Depressive Disorders (Major Depressive Disorder), adults | ||||||
Patient factors | ||||||
Sex (females) | 3328 | 13/16 | 8/2–15 | OR: 1.53 (1); Beta: 5.28/5.12–5.44 (2) | 16/0/0 | Accuracy: 74.8% (1); R2: 64%/13–65% (3) |
5-HT transporter polymorphisms (yes) | 625 | 4/5 | 5/4–22 | OR: 6.46/2.88–20.11 (8) | 5/0/0 | Accuracy: 74% (1); R2: 13% (1) |
Modifiable Multivariable Candidate Predictors | ||||||
Depressive Disorders (Major Depressive Disorder), adults | ||||||
Patient factors | ||||||
Marital status (married) | 2099 | 5/5 | 7/4–15 | OR: 1.7/1.41–2.89 (3) | 5/0/0 | Accuracy: 70.9%/67–74.8% (2) |
Negative life events (no) | 495 | 3/5 | 7/5–13 | OR: 2.56 (1); Beta: 6.88/6.06-6.92 (4) | 5/0/0 | R2: 64%/23–65% (3) |
Illness factors | ||||||
Depressive symptomatology (lower) | 2639 | 13/16 | 7/4–15 | OR: 1.17/0.90–3.57 (5) | 16/0/0 | Accuracy: 74.8% (1); R2: 65%/64–65% (2) |
Functioning (higher) | 2013 | 3/3 | 10/4–13 | OR: 1.9/1.86–2.3 (3); Wald: 4.902 (1) | 3/0/0 | R2: 64% (1) |
Treatment factors | ||||||
Early treatment response (yes) | 245 | 3/4 | 5/3-9 | OR: 7.76/6.22–9.3 (2); Beta: 1.32 (1) | 2/0/2 | AUC: 91%/62–92% (4); R2: 74/74–74% (2) |
Depressive Disorders (Major Depressive Disorder, Minor Depression), older adults | ||||||
Illness factors | ||||||
Depressive symptomatology (lower) | 883 | 3/12 | 7/4–15 | OR: 3.84 (1); Beta: 0.04 (1) | 12/0/0 | C-statistic: 0.75 (1) |
Remission | ||||||
Non-modifiable Multivariable Candidate Predictors | ||||||
Depressive Disorders (Major Depressive Disorder, Persistent Depressive Disorder), adults | ||||||
Patient Factors | ||||||
Sex (females) | 2810 | 9/17 | 5/4–48 | OR: 1.42/1.21–2.17 (5) | 14/3/0 | Accuracy: 72.2%/65.5–85% (2); NPV: 81%/77–87% (7); PPV: 50%/43–78% (7); Sensitivity: 69%/64–75% (7);Specificity: 70%/62–93% (7) |
Depressive Disorders (Major Depressive Disorder, Minor Depression, Persistent Depressive Disorder), older adults | ||||||
Patient Factors | ||||||
Sex (males) | 2234 | 12/26 | 5/2–16 | OR: 1.49 (1); HR: 1.71/1.58–1.85 (2) | 23/1/2 | Accuracy: 70%/67–89.47% (3); AUC: 71.5%/70–74% (4); Sensitivity: 64%/60–88.89% (5); Specificity: 72%/67–90% (5) |
Schizophrenia-Spectrum and Other Psychotic Disorders (Schizophrenia, Schizoaffective Disorder, Schizophrenia-Spectrum Disorders, FEP), adults | ||||||
Patient Factors | ||||||
Sex (females) | 2640 | 7/10 | 18/8–26 | OR: 1.29/1.28–1.31 (3); HR: 2.56 (1) | 9/1/0 | NPV: 88%/85–92% (2); PPV: 90%/89–92% (2) |
Substance-Related and Addictive Disorders (Alcohol, Cannabis, Cocaine, Opioid, Tobacco, Substance Use Disorder), adults | ||||||
Patient Factors | ||||||
Sex (females) | 17,912 | 7/14 | 16/9–31 | OR: 2/2–2 (2); HR: 1.38/1.23–1.56 (4) | 14/0/0 | Accuracy: 73%/31–77% (6) |
Modifiable Multivariable Candidate Predictors | ||||||
Anxiety Disorders (Generalized Anxiety Disorder, Panic Disorder, Separation Anxiety Disorder, Social Anxiety Disorder, Specific Phobia), children/adolescents | ||||||
Illness Factors | ||||||
Anxious symptomatology (lower) | 1062 | 5/6 | 10/8–23 | OR: 1.05/1.05–1.07 (3) | 5/1/0 | AUC: 66% (1); R2: 24.8%/20.9–28.7% (2) |
Depressive Disorders (Major Depressive Disorder), children/adolescents | ||||||
Illness Factors | ||||||
Depressive symptomatology (lower) | 1059 | 5/5 | 8/5–15 | OR: 1.07/1.06–1.08 (2); HR: 1.09 (1); RR: 2.63 (1) | 5/0/0 | AUC: 66% (1) |
Depressive Disorders (Major Depressive Disorder, Persistent Depressive Disorder), adults | ||||||
Patient Factors | ||||||
Negative life events (no) | 2960 | 6/25 | 3/3–25 | OR: 2.32/2.22–2.85 (3) | 25/0/0 | AUC: 80% (1); R2: 28% (1); Sensitivity: 77% (1); Specificity: 66% (1) |
General physical health (better) | 1807 | 5/9 | 6/5–48 | OR: 1.2/1.14–3.7 (7) | 9/0/0 | Accuracy: 85% (1); NPV: 87% (1); PPV: 78% (1); Sensitivity: 64% (1); Specificity: 93% (1) |
Education (higher) | 5146 | 4/6 | 22/5–48 | OR: 1.30/1.03–1.47 (5) | 4/1/1 | Accuracy: 64.6%/59.6–85% (3); AUC: 68%/66–70% (2); NPV: 65.3%/65–87% (3); PPV: 65%/56–78% (4); Sensitivity: 62.8%/49.4–64% (3); Specificity: 70.8%/66.2–93% (3) |
Emotional regulation/coping strategies (higher) | 1026 | 4/9 | 5/2–5 | OR: 2.21/2.13–2.3 (2) | 6/3/0 | NPV: 80.5%/77–88% (6); PPV: 49%/43–64% (6); R2: 23% (1); Sensitivity: 70%/67–75% (6); Specificity: 68.5%/62–73% (6) |
Employment status (employed) | 3841 | 4/6 | 15/3–48 | OR: 1.28/1.03–1.75 (5) | 4/1/1 | Accuracy: 64.6%/59.6–85% (3); AUC: 71%/70–73% (2); NPV: 65.3%/65–87% (3); PPV: 64%/56–78% (3); Sensitivity: 62.8%/49.4–64% (3); Specificity: 70.8%/66.2–93% (3) |
Quality of life (higher) | 1003 | 3/4 | 6/2–7 | OR: 1.07/1.04–2.5 (3) | 3/0/0 | R2: 48% (1) |
Illness Factors | ||||||
Anxiety comorbidity (no) | 3693 | 4/5 | 5/4–48 | OR: 3.44/2.27–7.14 (2) | 5/0/0 | Accuracy: 85% (1); NPV: 87% (1); PPV: 78% (1); R2: 6.6%/6.3–48% (3); Sensitivity: 64% (1); Specificity: 93% (1) |
OCD comorbidity (no) | 1270 | 2/6 | 6/6–48 | OR: 1.42/1.40–1.42 (3) | 6/0/0 | Accuracy: 85% (1); NPV: 87% (1); PPV: 78% (1); Sensitivity: 64% (1); Specificity: 93% (1) |
PTSD comorbidity (no) | 1270 | 2/5 | 6/6–48 | OR: 1.47/1.42–1.58 (4) | 5/0/0 | Accuracy: 85% (1); NPV: 87% (1); PPV: 78% (1); Sensitivity: 64% (1); Specificity: 93% (1) |
Personality disorder comorbidity (no) | 429 | 3/3 | 4/2–5 | OR: 6.25 (1) | 3/0/0 | Accuracy: 80% (1) |
Global severity symptomatology (lower) | 1337 | 3/6 | 6/3–48 | OR: 1.12/1.08–1.13 (4) | 6/0/0 | Accuracy: 85% (1); NPV: 87% (1); PPV: 78% (1); Sensitivity: 64% (1); Specificity: 93% (1) |
Episode/hospitalization duration (lower) | 3053 | 7/7 | 5/4-19 | OR: 1.69/1.04–4.34 (2) | 7/0/0 | Accuracy: 80% (1); AUC: 66% (1); PPV: 67% (1) |
Depressive symptomatology (lower) | 9774 | 22/28 | 8/3-125 | OR: 1.17/1.04–2.7 (7); RR: 1.14 (1) | 19/6/3 | Accuracy: 68.75%/59.6–85% (6); AUC: 73.5%/66–83% (10); NPV: 65.3%/65–87% (3); PPV: 65.5%/56–78% (4); R2: 30%/14.9–48% (9); Sensitivity: 68%/49.4–78% (8); Specificity: 74%/65–93% (8) |
Depressive Disorders (Major Depressive Disorder, Minor Depression, Persistent Depressive Disorder), older adults | ||||||
Patient Factors | ||||||
Social contacts (higher) | 1416 | 6/8 | 7/2–13 | OR: 1.76/1.19-4 (4); Wald: 4.47 (1) | 7/0/1 | AUC: 75%/70–77% (3) |
General physical health (better) | 1989 | 8/15 | 5/4–15 | OR: 1.40/1.14-1.81 (2); HR: 1.44 (1) | 14/0/1 | Accuracy: 70.8%/70–71.6% (2); AUC: 72%/70–74% (2); Sensitivity: 66%/60–72% (2); Specificity: 70.5%/67-74% (2) |
Illness Factors | ||||||
Anxiety comorbidity (no) | 1539 | 4/5 | 5/4–7 | OR: 1.58 (1) | 4/0/1 | Accuracy: 71.6% (1); AUC: 72.5%/70–75% (2) |
Functioning (higher) | 905 | 4/4 | 7/4–13 | OR: 1.03/1.03-1.04 (2); HR: 1.69 (1) | 4/0/0 | AUC: 77% (1) |
Episode/hospitalization duration (lower) | 483 | 3/6 | 5/4–15 | OR: 1.85 (1) | 4/0/2 | Accuracy: 68.5%/67–70% (2); AUC: 71.5%/70–74% (4); Sensitivity: 62.5%/60–72% (4); Specificity: 71.5%/67–74% (4) |
Feeding and Eating Disorders (Anorexia Nervosa, Bulimia Nervosa, Binge-eating Disorder, Eating Disorders not specified), adults | ||||||
Illness Factors | ||||||
Weight (higher) | 1016 | 9/12 | 5/2–14 | OR: 1.72/1.06–2.18 (6); HR: 1.16 (1) | 12/0/0 | Accuracy: 78.2% (1); R2: 38.5%/33–44% (2) |
Eating disorder symptomatology (lower) | 1231 | 7/8 | 5/3–14 | OR: 1.07/0.94–1.08 (6) | 8/0/0 | R2: 40% (1); |
Duration of illness (lower) | 561 | 4/5 | 6/4–10 | OR: 2.56/1.28–2.77 (3) | 5/0/0 | R2: 45% (1) |
Schizophrenia-Spectrum and Other Psychotic Disorders (Schizophrenia, Schizoaffective Disorder, Schizophrenia-Spectrum Disorders, FEP), adults | ||||||
Patient Factors | ||||||
Quality of life (higher) | 891 | 2/2 | 5/5–5 | OR: 1.03 (1) | 2/0/0 | AUC: 86.1 (1); PPV: 86% (1); R2: 63%/54–72% (2); Sensitivity: 89.9% (1); Specificity: 81.2% (1) |
Education (higher) | 2529 | 6/11 | 18/8–30 | OR: 1.62/1.49–1.74 (2) | 7/1/3 | Accuracy: 64.75%/62.5-67 (2); AUC: 63%/62–73% (5); NPV: 75.9%/60.6–86% (4); PPV: 78.4%/64.5–92% (4); Sensitivity: 59.5%/58.4–60.6 % (2); Specificity: 69.8%/64.5–75.1% (2) |
Illness Factors | ||||||
Duration of untreated illness (lower) | 3784 | 10/15 | 13/3–30 | OR: 1.20/1.01–2.32 (6) | 12/1/2 | Accuracy: 67% (1); AUC: 68%/62–70.3% (3); NPV: 85%/66.7–86% (3); PPV: 89%/67.9–92% (3); Sensitivity: 58.4% (1); Specificity: 75.1% (1) |
Negative psychotic symptomatology (lower) | 3208 | 9/13 | 18/9-30 | OR: 1.07/1.03–1.29 (6) | 9/1/3 | Accuracy: 64.7%/62.5–67% (2); AUC: 65.5%/62–86.1% (6); NPV: 75.8%/60.6–86% (4); PPV: 86%/64.5–92% (5); R2: 63%/54–72% (2); Sensitivity: 60.6%/58.4–89.9% (3); Specificity: 75.1%/64.5–81.2% (3) |
Positive psychotic symptomatology (lower) | 3208 | 9/15 | 18/5–26 | OR: 1.21/1.12–1.56 (3) | 11/1/3 | Accuracy: 64.7%/62.5–67% (2); AUC: 65.5%/62–86.1% (6); NPV: 75.8%/60.6–86% (4); PPV: 86%/64.5–92% (5); R2: 63%/54–72% (2); Sensitivity: 60.6%/58.4–89.9% (3); Specificity: 75.1%/64.5–81.2% (3) |
Depressive symptomatology (higher) | 166 | 3/5 | 18/13–26 | HR: 1.18 (1) | 3/1/1 | Accuracy: 62.5% (1); AUC: 63%/63–63% (2); NPV: 85%/60.6–86% (3); PPV: 89%/64.5–92% (3); Sensitivity: 60.6% (1); Specificity: 64.5% (1) |
Functioning (higher) | 2765 | 4/7 | 18/9–30 | OR: 4.35/1.02-7.69 (2) | 5/0/2 | Accuracy: 67% (1); AUC: 68%/62–70.3% (3); NPV: 66.7% (1); PPV: 67.9% (1); Sensitivity: 58.4% (1); Specificity: 75.1% (1) |
Substance-Related and Addictive Disorders (Alcohol, Cannabis, Cocaine, Opioid, Tobacco, Substance Use Disorder), adults | ||||||
Patient Factors | ||||||
Marital status (married) | 16,247 | 6/12 | 31/6–31 | OR: 2.86/2.27–3.45 (2); HR: 1.29 (1) | 12/0/0 | Accuracy: 73%/31–77% (6) |
Education (higher) | 11,404 | 5/10 | 14/9–31 | OR: 2.28/1.5–3.03 (2); HR: 1.1 (1) | 10/0/0 | Accuracy: 73%/31–77% (6) |
Illness Factors | ||||||
SUD lifetime comorbidity, other (no) | 5320 | 5/7 | 11/6–16 | OR: 4/3.12–5.55 (3); HR: 2.56/1.56–9.09 (4) | 7/0/0 | Accuracy: 64% (1) |
Depressive comorbidity (yes) | 1403 | 4/4 | 6/5–9 | OR: 2.15/1.82–2.49 (2) | 4/0/0 | Accuracy: 77.2% (1); R2: 6.4% (1) |
AUC area under the curve, D development sample, E external validation sample, HR hazard ratio, I internal validation sample, k number of studies, N number of subjects, NA not available, NPV negative predictive value, OR odds ratio, PPV positive predictive value, RR relative risk, SUD substance use disorder.
aIn brackets value of predictor associated with remission.
bCoefficient of predictors significantly associated with outcome in multivariable models.
cStudies might have reported more than one accuracy measure, so the number of studies can exceed k.