How to Estimate Productivity Costs in Economic Evaluations
Journal: Pharmacoeconomics
Authors and affiliation
Marieke Krol a, b and Werner Brouwer a, b
a Department of Health Policy and Management, Erasmus University,
Rotterdam, The Netherlands
b Institute for Medical Technology Assessment, Erasmus University, Rotterdam,
The Netherlands
Correspondence
Marieke Krol
P.O. Box 1738
3000 DR Rotterdam
The Netherlands
T +31 10 408 8580
F +31 10 408 9081
Electronic Supplementary Material
Summary of key recommendations
1. Recommendations regarding the identification of productivity changes
Productivity costs estimates should include:
- Cost related to absenteeism from paid work
- Cost related to changes in functioning at work (presenteeism)
- Cost related to changes in unpaid productivity
2. Recommendations regarding productivity measurement
Data collection should include:
- Data on absenteeism: days absent, number and duration of absenteeism periods
- Data on presenteeism: days health problems at work, functioning on those days
- Data on unpaid work: days difficulties performing unpaid work, the time a substitute would need to perform the lost unpaid work
Apply an instrument that includes absenteeism, presenteeism and unpaid work, e.g.:
- iMTA Productivity Cost Questionnaire (iPCQ)
- Valuation Of Lost Productivity (VOLP)
Frequency of measurement
- Measurement frequency should be determined by the expected changes in productivity
Recall period
- Recall periods preferably do not exceed two months
3. Recommendations regarding the valuation of productivity changes
For the valuation of paid productivity we recommend applying the following two approaches:
- Human capital approach
- Friction cost approach
For paid work values we recommend applying:
- Average age and gender-specific wage rates
- If these are not representative, patients’ gross wage can be applied.
For the valuation of unpaid work we recommend to apply either:
- The opportunity cost approach
- The proxy good approach
We recommend using the following values for unpaid work:
- Average age- and gender-specific net wage rates (opportunity cost approach)
- Costs of a market substitute (proxy good approach)
4. Calculating costs and reporting outcomes
Cost estimates related to paid work (absenteeism + presenteeism) can be calculated as follows:
- Human capital approach: total time absent x average gross age and sex dependent wage + time decreased functioning at work x ((10 – level of functioning)/10) x average gross age and sex dependent wage
- Friction cost approach: total time absent within friction period(s) x average gross age and sex dependent wage + time decreased functioning at work x ((10 – level of functioning)/10) x average gross age and sex dependent wage. Note that cost calculation per absenteeism period is limited to the friction period.
Valuation approaches unpaid work
- Opportunity cost method: Total time a substitute would need to perform tasks x average net age and sex dependent wage
- Replacement cost method: Total time a substitute would need to perform tasks x costs of a market substitute
Reporting outcomes
- Productivity outcomes should be presented both in quantities and costs
- Method sections should include information about the questionnaire applied (or the exact questions posed), the valuation approach, the values attached to lost productivity, frequency of measurement and the method of extrapolating data beyond the measurement points.
Example calculating productivity costs for patients A, B and C
To further illustrate productivity cost calculations, we here provide an example of how to produce productivity cost estimates for three fictive patients. Assume we have gathered the following information regarding the productivity of patients A, B and C in the preceding four weeks.
Patient A, a male of 26 has no paid profession. He was not able to do all his normal household tasks due to health problems for 10 days during the past four weeks. He estimates that a substitute would have needed one hour a day to do the tasks he was not able to do. Patient B, a male of 53 normally works 50 hours per week spread over five days. Due to ill-health he did not attend work the past 30 weeks. He was not able to do most of his normal unpaid labor tasks the past four weeks. A substitute would have needed 2 hours a day to replace his tasks. Patient C, a female of 34 has a paid job for 40 hours a week spread over five days. She was absent from work for 4 hours the preceding four weeks due to a doctor’s appointment. She suffered from health problems at work for two days. Those two days she was able to do approximately 70% of the work she normally does on a day. She had no problems performing unpaid work.
The total productivity costs for the three patients are summarized in table 1 and calculated as follows. For patient A, there are no productivity costs related to paid labor and 10 hours of lost unpaid labor. According to the opportunity cost method, the value of an hour of unpaid labor at the margin is equal to the value of paid labor. In the Netherlands the value of an hour of lost productivity is estimated to be 24 Euros for a male between 25 and 30 years old (2009 values (46)). Total productivity costs of patient A are therefore 10 *24 = 240 euro. With regard to patient B, the friction period had already ended before our four weeks of interest. With the friction cost approach the productivity costs of paid labor are therefore zero. With the human capital approach the productivity costs of paid labor for patient B would be 50 hours times four weeks times 39 Euros (Dutch 2009 estimate for males between 50 and 55). This would equal 7800 Euros. The costs of unpaid labor would be 28 days times two hours times 39 Euros, equals 2184 Euros. The productivity costs related to absenteeism for patient C would be four hours times 28 Euros (Dutch hourly productivity costs estimate for females between 30 and 35), equals 112 Euros. In this case there are no differences between the human capital approach and the friction cost approach. The presenteeism costs for patient C would be 30% of 16 hours (she was 70% productive as compared to normal) times 28 Euros, is 134 Euros.
As can be seen in table 1, the total four-week productivity cost estimates of these three patients differ substantially between the friction cost approach and the human capital approach due to the long term absenteeism of patient B.
Table 1 Productivity costs of patients A, B and C
Patient A / Patient B / Patient C / TotalAbsenteeism FCA / € 0 / € 0 / 4 * 28 = € 112 / € 112
Absenteeism HCA / € 0 / 50 * 4 * 39 = € 7,800 / 4 * 28 = € 112 / € ,7912
Presenteeism / € 0 / € 0 / 0.3 * 16 * 28 = € 134 / € 134
Unpaid labor / 10 * 24 = € 240 / 28 * 2 * 39 = € 2184 / € 0 / € 2,424
Total FCA / € 240 / € 2,184 / € 246 / € 2,670
Total HCA / € 240 / € 9,984 / € 246 / € 10,470
FCA = friction cost approach, HCA = human capital approach