Table 2.

Overview on pros and cons of different outcome indicators in economic evaluations of health promotion for older people

Outcome Indicators	Cons	Pros
Cost effectiveness analysis (CEA)
Natural indicators	- Effects are reduced to a single parameter - Not all effects are covered (e.g. intersectoral effects) - Relevant health promotion outcomes are often difficult to operationalize - Often proxy outcomes are used (causality to patient-relevant endpoints has to be proven) - Comparability of different interventions is very limited	- Depending on the operationalization mostly easy and clearly measurable - Disease specific comparisons are very easily possible - Even small health gains can be documented - Simple and straightforward comparability of the one given indicator
Cost utility analysis (CUA)
Aggregated indicators in general	- Aggregation of different aspects will always represent a limited perspective	- Multidimensionality of effects can be covered - Different types of interventions can be compared
e.g. QALY	- Only health related - Social benefits are not covered - Intersectoral benefits are not covered - The linear conjunction of time and quality is set externally (not preference based) - Limited comparability for HPA - Small health gains are measured poorly - Does not reflect preferences of older people appropriately - May discriminate against older people	- Widely accepted reference standard and well established instrument that is used in many economic evaluations
e.g. ICE-CAP-O	- So far not widely validated - No comparison across age groups possible - Limited in capturing health dimensions	- Developed according to preference weights of older people
Cost benefit analysis (CBA)
Measurement of health costs avoided only	- Health gains or social benefits are attributed no value	- Subjective elements are excluded - Allows the comparison of measures in different fields of policy
Monetary valuation of outcomes e.g. by willingness to pay	- Political reservations against monetary valuation of health benefits - HPA have features of public goods, resulting in low willingness-to-pay - Results are not independent of who is asked: thus an age bias is possible - Willingness to pay is dependent on ability to pay	- Allows the comparison of measures in different fields of policy - Non-health benefits can be included
Cost consequence analysis (CCA)
No standalone method CEA or CUA with multiple endpoints	- Limited comparability of different interventions - Complex results may be difficult to interpret	- Very transparent - Intersectoral costs and benefits can be covered - Contentious cost categories can be included - Broadly spread effects are at least described qualitatively - Description of equity effects can be included - Enables decision makers to decide based on a differentiated assessment of the various effects or benefits - Allows the setting of specific priorities