The chart review is not useless. Let's start with a review of 100 records---50 from cancer patients and 50 from CHF patients---and see whether the information we collect can help us design a prospective survey.

CORRECT

You may recognize that there are challenges inherent in collecting valid data about fatigue in both retrospective and prospective studies. The chart review data exist and are easily accessible. They provide a snapshot of fatigue and related factors that must be interpreted cautiously, but could indeed inform the design of a prospective study. Suggesting a rather limited chart review when many more records are available is a reasonable use of resources at this time. It acknowledges the weakness of the retrospective data---no control over data collection, potential for biases that do not exist in prospective studies (e.g. missing charts, missing data, etc.), use of an unvalidated measure---but provides an opportunity for some statistical analyses. Of course, limiting the chart review to only 50 per disease may mean that the data collected will be so limited that it could not be reported on its own, but given the problems inherent in the data, this may be acceptable. The chart review would not be done to acquire publishable data, but rather to obtain pilot data useful for writing a grant proposal and helping in the design of a future prospective study. Finally, this approach recognizes that the best design for a prospective study may not be immediately evident. For example, can valid data about fatigue in the home environment be acquired from cancer patients and CHF patients after admission for some type of acute problem? The chart review will not answer this, but may provide some background that could later help clarify this issue.