Table 4.

Classification and types of calculation

Method of calculation	Nb of studies (%)	N° (Table 3)
Studies directly calculating the cost of ADE	15 (75)	1, 4, 5, 6, 8, 10, 11, 12, 13, 14, 15, 17, 18, 19, 20
Studies estimating the cost of ADE through external data or studies	5 (25)	2, 3, 7, 9, 16
Number of direct costs considered in the calculation:
1	10 (50)	2, 3, 5, 7, 8, 9, 14, 19, 20
2	1 (5)	11
≥ 3	9 (45)	1, 4, 6, 10, 12, 13, 15, 16, 17, 18
Only direct costs	16 (80)	1, 2, 3, 4, 6, 7, 8, 9, 10, 11, 12, 13, 14, 17, 19, 20
+ indirect costs	3 (15)	5, 15, 18
+ opportunity costs	1 (5)	16
Micro-costing method Real costs of exact resources consumed in the care of each patient	4 (20)	1, 6, 12, 18
ADR vs. non-ADR method (propensity scores matching) Costs difference between ADR patients and matched control (non-ADR patient)	3 (15)	11, 18, 19
Gross-costing method Assign average values from national administrative databases	2 (10)	5, 13
Extended LOS attributable to preventable ADE method Extra medical costs = Hospital admission x number of patients with preventable ADEs × estimated extended LOS	1 (5)	3
Before vs. after ADR method Difference between the total medical care cost of a patient 6 months before the ER visit for an ADR and 6 months after Expenses for a new diagnosis of diseases (other than those diagnosed during the control period) were not included	1 (5)	10
Resource use method 1. ADE identification: manual from medical records, causality assessments between potential ADEs and drug therapies with causality levels (definite/likely/possible) 2. Resource use identification: contribution to healthcare use: assessment of each ADEs contribution to resource use, with contribution levels (dominant/partly/less) 3. Proportion of costs from regional cost per patient register:  ADE contribution dominant = full costs  ADE contributed partly or less = cost for specific resources used for ADEs	1 (5)	17
Proportion of registered costs method 1. ADE identification: manual from medical records, causality assessments between potential ADEs and drug therapies with causality levels (definite/likely/possible) 2. Resource use identification: contribution to healthcare use: assessment of each ADEs contribution to resource use, with contribution levels (dominant/partly/less) 3. Proportion of costs from regional cost per patient register:  ADE contribution dominant = full costs  ADE contributed partly = $1\left./,2\right)$ of costs  ADE contributed less = $1\left./3\right)$ of costs	1 (5)	17
Unit cost method 1. ADE identification: manual from medical records, causality assessments between potential ADEs and drug therapies with causality levels (definite/likely/possible) 2. Resource use identification: contribution to healthcare use: assessment of each ADEs contribution to resource use, with contribution levels (dominant/partly/less) 3. Proportion of costs from national statistics:  ADE contribution dominant = full costs  ADE contributed partly = $1\left./2\right)$ of costs  ADE contributed less = $1\left./3\right)$ of costs	1 (5)	17
Diagnostic code method 1. ADE identification with ICD codes indicating ADEs 2. Resource use identification: all resource use during the healthcare encounter assigned to the ADE 3. Estimating cost with regional or national registers: full costs, $1\left./2\right)$ costs, $1\left./3\right)$ costs	1 (5)	17
Main diagnosis method 1. ADE identification: manual matching with main diagnosis and ICD codes 2. Resource use identification: all resource use during the healthcare encounter assigned to the ADE 3. Estimating cost with regional or national registers: full costs, $1\left./2\right)$ costs, $1\left./3\right)$ costs	1 (5)	17

LOS Long of stay, ICD International code of disease