Table 2

Studies on the risk of cerebrovascular events associated with antipsychotic medicines
Study Population Outcome Propensity for bias / Study quality Procedures to minimize bias Follow-up Outcome Rate in reference group Result
LEVEL I Evidence: Meta-Analyses
Studies that compared atypical antipsychotic (ATYP) treatment to placebo (PLA)
Ballard [1] 1954 dementia patients (age> 60) CV Events Low/High Yes 10-13 weeks 1.0% RR (RISP v PLA); 3.64 95% CI; 1.72-7.69
Schneider [6] 5,110 Dementia patients (mean age 81.2) CV Events Low/High Yes 6-26 weeks 0.9% 1.0% OR (ATYP v PLA); 2.13 95% CI; 1.20-3.75 OR(RISP v PLA); 3.43 95%CI; 1.60-7.32
DeDeyn [4] 1,155 Institutionalised patients (age>=55) CV Events Low/High Yes 12 weeks 1.6% Rate in Risperidone group 3.9% RR not reported
Hermann [32] 1,721 Dementia patients (age>=55) CV Events Low/High Yes 12 weeks 1.1% RR (RISP v PLA); 3.2 95% CI; 1.4-7.2
Katz [5] 895 Institutionalised dementia patients (age>=55) CV Events Low/High Yes 12 weeks 0.8% Rate in Risperidone group 1.6% RR not reported
DeDeyn [4] 1,155 Institutionalised patients (age>=55) Serious CV Event requiring hosp. Low/High Yes 12 weeks 0.7% Rate in Risperidone group 1.6% RR not reported
Hermann [32] 1,721 Dementia patients (age>=55) Serious CV Event requiring hosp. Low/High Yes 12 weeks 0.6% RR (RISP v PLA); 2.3 95% CI; 0.5-10.7
LEVEL III Evidence: Observational Studies
Studies that compared atypical antipsychotic (ATYP) treatment to non-use (NU)
COHORT STUDIES
Sacchett [33] 74,162 All patients, General Practitioner Health Search database (Italy) (age>=65) Diagnosis of stroke (GPs’ medical records) ICD9: 434.9, 438.0, 342 Moderate/Moderate Yes (covariate adjustment, subgroup analysis) 3.5 months 12 per 1000py RR (ATYP v UNEX); 2.46 95% CI; 1.07-5.65
Barnett [34] 14,029 Dementia patients, Veterans Affairs Clients (US) (age>=65) Hospital Admission for CV event ICD9 435, 437, 430, 432, 433, 434 Moderate/Moderate Yes (covariate adjustment, sensitivity analysis) 18 months 3.2% HR (ATYP v NU); 1.20 95% CI; 0.83-1.74
CASE-CONTROL STUDIES
Kolanowski [35] 959 Dementia patients, health care insured on Southeast US (age>70) Diagnosis of Stroke Moderate/ Low Yes (unmatchedcovariate adjustment) 45 days NA OR (ATYP v NU); 0.98 95% CI; 0.64-1.52
Liperoti [36] 1130 cases, 3658 controls, institutionalised dementia patients in six states in the US (Ohio, Maine, Illinois, Mississippi, South Dakota, New York) (age>85) Hospital Admission for CV event ICD9 433.0-434.9 (Ischaemic stroke) 435-435.9 (TIA) Moderate/ Low Yes (matching on admitted to same facility for septicemia, UTI) NR NA OR (RISP v NU); 0.87 95% CI; 0.67-1.12 OR (OLA v NU); 1.32 95% CI; 0.83-2.11 OR (Other ATYP v NU); 1.57 95% CI; 0.65-3.82
SELF-CONTROLLED CASE-SERIES STUDIES
Douglas [37] 6790 All patients with incident diagnosis of stroke General Practice Research Database (UK) (no age restriction) Diagnosis of stroke (GPs’ medical records, excluding TIA) Moderate/Moderate Yes (within patient design) NR NA IRR (ATYP v NU); 2.32 95% CI; 1.73-3.10
Pratt [38] 10638 Australian Department of Veterans Affairs Veterans/spouses with hospitalization for stroke (Australia) (age>=65) Hospitalis-ation for stroke ICD-10: I60-I64 Moderate/Moderate Yes (within patient design) 4 years NA IRR (ATYP v NU); 0.9 95% CI; 0.7-1.0
Studies that compared conventional antipsychotic (CONV) treatment to non-use (NU)
COHORT STUDIES
Sacchetti [33] 74,162 All patients, General Practitioner Health Search database (Italy) (age>=65) Diagnosis of stroke (GPs’ medical records) ICD9: 434.9, 438.0, 342 Moderate/Moderate Yes (covariate adjustment, subgroup analysis) 3.5 months 12 per 1000py RR (BUTY v UNEX); 3.55 95% CI; 1.56-8.07 RR (PHENO v UNEX); 5.79 95% CI; 3.07-10.9
Barnett [34] 14,029 Dementia patients, Veterans Affairs Clients (US) (age>=65) Hospital Admission for CV event ICD9 435, 437, 430, 432, 433, 434 Moderate/Moderate Yes (covariate adjustment, sensitivity analysis) 18 months 3.2% HR (CONV v NU); 1.20 95% CI; 0.48-3.47
CASE-CONTROL STUDIES
Kolanowski [35] 959 Dementia patients, health care insured on Southeast US (age>70) Diagnosis of Stroke Moderate/ Low Yes (Unmatched, covariate adjustment) 45 days NA OR (CONV v NU); 1.18 95% CI; 0.63-2.24
Liperoti [36] 1130 cases, 3658 controls, institutionalised dementia patients in six states in the US (Ohio, Maine, Illinois, Mississippi, South Dakota, New York) (age>85) Hospital Admission for CV event ICD9 433.0-434.9 (Ischaemic stroke) 435-435.9 (TIA) Moderate /Low Yes (matched; on admitted to same facility for septicemia, UTI) NR NA OR (CONV v NU); 1.24 95% CI; 0.95-1.63
SELF-CONTROLLED CASE-SERIES STUDIES
Douglas [37] 6790 All patients with incident diagnosis of stroke General Practice Research Database (UK) (no age restriction) Diagnosis of stroke (GPs’ medical records, excluding TIA) Moderate/Moderate Yes (within patient design) NR NA IRR (CONV v NU); 1.60 95% CI; 1.55-1.84
Pratt [38] 10638 Australian Department of Veterans Affairs Veterans/spouses with hospitalization for stroke (Australia) (age>=65) Hospitalis-ation for stroke ICD-10: I60-I64 Moderate/Moderate Yes (within patient design) 4 years NA IRR (CONV v NU); 1.0 95% CI; 0.8-1.2
Studies that compared all antipsychotics (ANTIP) with no treatment (NU)
COHORT STUDIES
Percudani [39] 1645978 All patients in Lombardy Italy with CV related Outcome in 2002 (age>=65) Hospital Admission for CV related outcome ICD9 430--438 Moderate/Low Yes (unmatched covariate adjustment) 2 years 2.15 % OR (ANTIP v NU); 1.24 95% CI; 1.16-1.32
Sacchetti [40] 134488 All patients, General Practitioner Health Search database (Italy) (age>50) Diagnosis of stroke (GPs’ medical records) ICD9: 434.9, 438.0, 342 Moderate/Moderate Yes (covariate adjustment) 6 months 3.6 per 1000py 1 month: RR (ANTIP v NU); 12.4 95% CI; 8.4-18.1
Kleijer [41] 2448 Patients in community pharmacy practice, PHARMO Database (Netherlands) (age>50) Hospital Admission for stroke (inc TIA) ICD9 430-436 Moderate/Low Yes (matched: age/sex) 1 year NA Current use: OR (ANTIP v NU); 1.6 95% CI 1.3-2.0 0-7 days: OR (ANTIP v NU); 9.9 (5.7-17.2) 8-14 days :OR (ANTIP v NU);2.6 (1.3-5.3) 15-30 days: OR (ANTIP v NU); 2.1 (1.0-4.5) 31-90 days: OR (ANTIP v NU); 1.5 (1.0-2.2) >90 days: OR (ANTIP v NU); 1.0 (0.7-1.3)
Studies that compared conventional antipsychotic (C) and atypical antipsychotic (A) treatment
COHORT STUDIES
Hermann [42] 11400 All patients, population based cohort (Canada) (age>65) Hospital Admission for stroke ICD9 430-436 Moderate/Moderate Yes (covariate adjustment) 5 years 5.7 per 1000py RR (RISP v CONV); 1.4 95% CI; 0.7-2.8 RR (OLA v CONV); 1.1 95% CI; 0.5-2.3
Gill [43] 32710 Dementia patients, administrative health care database (Canada) (age>=65) Hospital Admission for ischaemic stroke ICD9 431, 434,436 Moderate/Moderate Yes (covariate adjustment, subgroup analysis) 5 years 6.3 % HR(ATYP v CONV); 1.01 95% CI; 0.81-1.26
Finkel [44] 18477 Dementia patients, Medicaid database (US) (age>60) Hospital Admission for stroke ICD( 430-432, 434-436, 437.1, 437.9 Moderate/Moderate Yes (covariate adjustment) 3 months 0.87 % OR (OLA v RISP); 1.1 95% CI; 0.6-1.7 OR (QUE v RISP); 0.78 95% CI; 0.2-1.9 OR (HAL v RISP); 1.9 95% CI; 1.0-3.6
Percudani [39] 1645978 All patients with CV related outcome in 2002 (Italy) (age>=65) Hospital Admission for CV related outcome ICD9 430--438 Moderate/Low Yes (unmatched covariate adjustment) 2 years 2.37 % OR (ATYP v CONV); 1.42 95% CI; 1.24-1.64
Wang [45] 22890 Patients in Pharmacy Assistance Contract for Elderly Program (US) (age>=65) Hospital Admission for stroke, cerebral hemorrhagic and ischemic events Moderate/Moderate Yes (Propensity score adjustment, instrumental variable analysis) 180 days Not Reported 30 days: HR (CONV v ATYP); 1.08 95% CI; 0.99-1.18 60 days: HR (CONV v ATYP) 1.10 95% CI; 1.02-1.19 180 days: HR (CONV v ATYP) 1.09 95% CI; 1.02-1.16 IV analyses not reported
Sacchetti [33] 74,162 All patients, General Practitioner Health Search database (Italy) (age>=65) Diagnosis of stroke (GPs’ medical records) ICD9: 434.9, 438.0, 342 Moderate/Moderate Yes (covariate adjustment, subgroup analysis) 3.5 months 47.4 per 1000py RR (BUTY v ATYP); 1.44 95% CI; 0.55-3.76 RR (PHENO v ATYP); 2.34 95% CI; 1.01-5.41

Pratt et al.

Pratt et al. BMC Medical Research Methodology 2012 12:72   doi:10.1186/1471-2288-12-72

Open Data