Table 3.
Drug–AE associations and characteristics.
| Relative Risk [(RR) | Source (Type of Study) | Main Methodological Issues | Hazard Function | |||
|---|---|---|---|---|---|---|
| SABA / LABA and AMI | RR > 2 for cardiovascular events vs. placebo [57] ORs 1.7 - 7.3 (new users) for MI vs. non-users [57] | Systematic review (RCT) (Case-control) | Protopathic bias Confounding by indication /severity | Acute onset, transient | ||
| RR = 2.5 for respiratory deaths vs. placebo [58] | Meta-analysis (RCT) | |||||
| salmeterol: all cause mortality Peto OR =1.3 vs. placebo[59] non-fatal AE OR = 1.2 vs. placebo [59] formoterol : non-fatal serious AE OR = 1.6 vs. placebo[60] | Cochrane database systematic review (RCT) | |||||
| OR 1.2 for beta-2 agonists (current users) – 2.5 (IHD patients) [61]RR = 1.6 for SABA (heavy users vs. users of <3 months) [62] RR = 1.1 for LABA (heavy users vs. users of <3 months) [62] | Nested case–control cohort | |||||
| Antimicrobials and ALI | Elevated liver enzymes, cholestasis, and acute liver failure (for betalactam antimicrobials, macrolides, sufonamides, tetracyclines [63] | Case reports/ retrospective cohort | Definition/measurement of the outcome Ascertaining/tracing of exposure (short time window) | Acute/intermediate onset (3-4 weeks) after drug stop | ||
| RRs 2.3 (Amoxicillin without clavulanic acid) – 1299.9 (Isoniazid + rifampicin + pyrazinamide) [64] | Case-population | |||||
| ORs 5.3 (erythromycin) – 94.8 (amoxicillin/clavulanic acid) [37] | Case-control (pop-based) | |||||
| Antidepressants/ BZD and hip fracture | RRs 1.2 - 3.7 for TCA users [65] RRs 1.5 - 8.6 for SSRI users [65] RRs 1.5 - 2.0 for hypnotics including BZD [66] | Case-control/ cohort | Exposure classification (for antidepressants) Selection bias Unmeasured confounding | SSRIs: peak at 6–12 months [67] TCA’s: peak at 1-2 months [67] BZD: acute | ||
| Anticonvulsants and suicide/-attempts | RR = 2 for 11 different groups of the drug (1.5 (psychiatric) 3.5 (epilepsy) risk by indication) [68] | Meta-analysis of RCT | Definition and measurement of outcome | Acute | ||
| RR = 3.1 for current users (lamotrigine, gabapentin, ethosuximide, vigabatrin) [69] OR 2.57 vs. non-users [70] | Nested case-control | |||||
| HRs 1.4 – 2.4 vs. topiramate users [71] | Cohort | |||||
| CCB and cancer | RRs 1.7 (vs. non-users) - 2.6 (breast cancer) [72, 73] RR = 2.1 for verapamil [74] | Cohort | Long latent period Selection bias Unmeasured confounding | Long-term, delayed | ||
SABA = short acting beta-2 agonists
LABA = long acting beta-2 agonists
(A)MI = (acute) myocardial infarction
ALI = acute liver injury
BZD = benzodiazepines
SSRI = selective serotonin reuptake inhibitor
RR = relative risk; OR =odds ratio, HR= hazard ratio
IHD = Ischemic Heart Diseases
AE = adverse event
TCA = tricyclic antidepressants
[Number] = number of reference including these data
CCB = calcium channel blockers