Table 1.
Overview of meta-analyses on COVID-19 severe outcomes in any mental disorders.
| Pooled odds ratios for any mental disorder compared to people without mental disorders | Pooled odds ratios for severe mental disorders compared to people without severe mental disordersb | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Meta-analysis (search updated to) | No. original studies | Total sample size | Mental disorders sample size | Primary COVID-19 outcomes | Class of evidence a | Crude Odds Ratio (95%CIs, I2) | Adjusted Odds Ratioc (95%CIs, I2) | Crude Odds Ratio (95%CIs, I2) | Adjusted Odd Ratioc (95%CIs, I2) | Additional subgroup analyses | AMSTAR 2 quality |
| Ceban et al. (1 Feb 2021) [8] | 14 | 65,514,469 | NA | Infection | NS | NA | NA | MD: 0·91 (0·58–1·41, 100%) | MD: 1·50 (0·75–2·99, 100%) | Depression, study design | Low |
| 7 | 26,554,397 | NA | Hospitalization | IV | NA | NA | MD: 1·38 (1·17–1·63, 80%) | MD: 1·27 (1·09–1·47, 97%) | |||
| 7 | 83,240 | NA | Severe events | NS | NA | NA | MD: 0·93 (0·85–1·03, 0%) | MD: 0·99 (0·80–1·24) | |||
| 12 | 25,808,660 | NA | Mortality | II | NA | NA | MD: 1·44 (1·23–1·68, 30%) | MD: 1·57 (1·26–1·95, 78%) | |||
| Fond et al. (22 Feb 2021) [10] | 16 | NA | 19,086 | Mortality | III | 1·75 (1·40–2·20, 26%) | 1·38 (1·15–1·65, 0%) | 2·26 (1·18–4·31, 56%) | 1·67 (1·02–2·73, 27%) | Severe mental disorders | Low |
| Liu et al. (7 Jul 2021) [7] | 18 | 72,464,308 | 3,325,988 | Infection | IV | 1·42(1·15–1·76, 95%) | 1·71(1·09–2·69, 100%) | MD: 2·02(1·08–3·76, 100%);d
SZ: 1·72(0·62–4·77, 100%)d |
NA | Type of disorders (mental or neurological) diagnostic category, temporal relationship of exposure and COVID-19 infection (pre-existing vs sequela), sex ratio, mean age, income level of regions | Moderate |
| 19 | 25,767,005 | 3,045,593 | Severity | II | 1·40(1·25–1·57, 80%)d | NA | MD:1·34(1·08–1·67, 66%);d
SZ: 1·22(0·70–2·13, 88%)d |
NA | |||
| 28 | 34,168,377 | 3,211,426 | Mortality | III | 1·47(1·26–1·72, 93%)d | NA | MD: 1·36 (1·15–1·61, 81%);d
SZ: 2·28 (1·40–3·73, 64%)d |
NA | |||
| Toubasi et al. (15 Feb 2021) [9] | 16 | 634,338 | 68,023 | Severity and Mortality | III | 1·76 (1·29–2·41, 94%)d | 1·52 (1·20–1·93, 63%) | NA | NA | Countries | Low |
| Vai et al. (5 Mar 2021) [11] | 23 | 1,469,731 | 43,938 | Mortality | II | 2·00 (1·58–2·54, 93%) | 1·31 (1·13 1·52, 39%) | 2·21 (1·63–2·99, 82%) | 1·55 (1·30–1·85, 29%) | Diagnostic category; Psychopharmacological drug class; severe mental disorders; baseline COVID-19 treatment setting; countries | Moderate |
| ICU admission | IV | 1·77 (1·09–2·89, 93%) | 1·33 (0·87–2·04, 90%) | 1·65 (0·85–3·21, 90%) | 1·08 (0·65–1·79, 69%) | Diagnostic category; severe mental disorders; baseline COVID-19 treatment setting | |||||
| Hospitalization | II | 2·24 (1·70–2·94, 89%) | 1·77 (1·29–2·42, 91%) | 2·48 (1·66–3·69, 55%) | 1·37 (0·79–2·38, 37%) | Diagnostic category; severe mental disorders | |||||
Abbreviations: CI, confidence intervals; ICU, intensive care unit; MD, mood disorders; NA, not available; SZ, schizophrenia.
Meta-analytic evidence are stratified from class I to class IV and to Not significant Evidence (NS); representing respectively convincing evidence, highly suggestive evidence, suggestive evidence and weak evidence.
Severe mental disorders included in Ref. [10] schizophrenia spectrum disorders and/or bipolar disorders, in Ref. [11] psychotic and mood disorders.
Pooled OR adjusted in the original studies for possible confounders (e.g., age, sex, smoking, obesity, socioeconomic status, race or ethnicity, and psychical comorbidities).
Non-fully adjusted model, including Odds Ratios from all the original studies, either crude, or adjusted.