Table 1

Comparison of cost estimation methods between the disease specific and all causes approaches

Cost estimation methods
Types of costsDisease specific approachAll causes approach
PAR(I)i, PAR(O)i, and PAR(D)i refer to population attributable risk of hospitalisation, outpatient visits, and death, respectively, with the disease “i” caused by smoking.
PYLL refers to potential years of life lost.
i=1 . . .nth disease.
Direct costs
• Medical costs of hospitalisationTotal number of inpatientsi * PAR(I)i * average per capita annual hospitalisation costsiDifference in average per capita annual hospitalisation costs b/t smokers and non-smokers * total number of smokers
• Medical costs of physician visitsTotal number of outpatientsi * PAR(O)i * average per capita annual outpatient visit costsiDifference in average per capita annual outpatient visit costs b/t smokers and non-smokers * total number of smokers
• Transportation costs to health providersTotal number of outpatientsi * PAR(O)i * average per capita annual outpatient visitsi * average transportation costs per visitDifference in average per capita annual outpatient visits b/t smokers and non-smokers * total number of smokers * average transportation costs per visit
• Caregiver’s costsTotal number of inpatientsi * PAR(I)i * average per capita annual inpatient daysi * daily rates of nursing careDifference in average per capita annual inpatient days b/t smokers and non-smokers * total number smokers * daily rates of professional caregiver’s service
Indirect costs
• Costs of absence from work(Total number of inpatientsi * PAR(I)i * average per capita annual inpatient daysi * average daily wage rate) + (total number of outpatientsi * PAR(O)i * average per capita annual outpatient visitsi * average visit time * average hourly wage rate)(Difference in average per capita annual inpatient days b/t smokers and non-smokers * total number of smokers * average daily wage) + (difference in average per capita annual outpatient visits b/t smokers and non-smokers * total number of smokers * average visit time * average hourly wage)
• Premature death costsTotal number of deaths * PAR(D)i * total expected future earnings during PYLLDifference in probability of death b/t smokers and non-smokers * total number of smokers * total expected future earnings during PYLL