Prospective data collection
	A health economic study alongside a clinical study (piggyback). Gathering data during the clinical study programme is an efficient way of collecting patient-level data. Not all such studies are, however, appropriate as a basis for health economic evaluation. It is therefore important to consider if the design of the clinical study can generate valid answers to the research question.
	Observational study. This is a method of collecting data from clinical practice. No intervention occurs in the management of patients, allowing collection of data on effectiveness and resource use in a real-life setting.
	Expert panels. Sometimes it is not feasible to obtain certain data by observing patients or doctors. In such situations, an expert panel is preferable. A Delphi panel is a way of systematically collecting and collating informed judgments from a group of experts in a series or rounds on specific questions or issues. A structured interview study is a simpler approach, where all information is collected in one round.

Retrospective data collection Sweden has a long tradition of keeping registries with information on the health and social conditions of the population. Several of these registries were started many years ago and include data on all Swedish citizens. As personal identification numbers have been applied in a routine manner it is possible to carry out advanced searches, including the linking of registries. As a result, there are unique opportunities in Sweden to study health care resource use and health outcomes, as the data are based on the entire population. We conduct observational, retrospective research based on registry data from national registries, disease-specific registries and electronic patient records.

When all relevant data have been collected, the early-phase model analysis is up-dated, further developed, and a comprehensive health economic analysis carried out. We use dedicated software (TreeAge Pro™ 2005) or the Microsoft Excel program.