Often, the first step in costing is to determine the intervention costs. For interventions that already receive reimbursement or are sold on the market, this is easy to obtain (i.e., reimbursement amount or out-of-pocket cost). However, since behavioral interventions are often not reimbursed and/or are not available on the market, defining the cost of the intervention can be more challenging.
For example, Get Busy Get Better is a novel nonpharmacological home-based intervention designed to reduce depression in older African Americans (Gitlin et al., 2012). Get Busy Get Better is not currently reimbursed by health care payers and is not available on the market. To determine the cost of implementing Get Busy Get Better, the components of the intervention were decomposed into its various parts detailed in Table 18.4. The cost-effectiveness analysis of Get Busy Get Better was planned during the original study design to enable a prospective approach, and therefore many of the necessary cost components were included in the data collection instruments so that costs could be captured prospectively. However, in many cases, cost-effectiveness is evaluated post hoc or after data collection has been completed. In such cases, it is necessary to estimate the cost of delivering the intervention retrospectively.
TABLE 18.4 Cost Framework for the Get Busy Get Better Intervention Trial
Screener's time spent screening participants
Time spent screening potential participants multiplied by screener's wage rate + fringe benefit costs divided by sample size
Interventionists' time spent with subjects
Time conducting intervention multiplied by wage rate of interventionist + fringe benefit costs divided by sample size
Participant contact outside of intervention delivery
Interventionists' time spent in communication with subjects outside of designated intervention visits
Time spent in contact outside of intervention multiplied by wage rate + fringe benefit costs of interventionist divided by sample size
Interventionists' time spent traveling to and from appointments with subjects
Wage rate + fringe benefit costs of interventionist multiplied by time spent in travel divided by sample size
Reimbursement for interventionists' auto expenses
Miles traveled multiplied by mileage reimbursement rate divided by sample size
Supervisors' time spent supervising interventionists as well as reviewing recordings of intervention visits
Wage rate + fringe benefit costs of supervisor and interventionist multiplied by time spent supervising employee then divided by sample size
Cost of training the interventionists
Wage rate + fringe benefit costs of screeners why screeners? and trainers, multiplied by time spent in training divided by sample size
Materials needed for the study included GPS devices, electronic recorders, educational pamphlets, and paper encounter forms
Cost of materials used for screening and during the intervention divided by sample size
Time spent by MD responders to alerts
Supervisor wage rate + fringe benefit costs, multiplied by time spent dealing with adverse events divided by sample size
Note: Originally published in BMC Geriatrics (Gitlin et al., 2012).