We appreciate the information provided by Drs Oldridge, Lim, and Guyatt regarding their tool, the MacNew Heart Disease Health-related Quality of Life questionnaire, formerly known as the Quality of Life after Myocardial Infarction questionnaire. The intent of our review article was to provide thorough and practical information for cardiovascular researchers utilizing quality of life instruments. The response of Dr Oldridge and colleagues highlights two problems that frequently arise when publishing review articles; the first is the lag time from the literature review to publication and the second is the limitations associated with computerized literature searches. Our original literature search was completed in early 2000 and was updated in 2001. Consequently, 2 of the articles mentioned in their response were not available to us (8,9).
We intentionally restricted the review to published articles identified by the search. We made this decision out of concern for fairness, since the Ferrans and Powers Quality of Life Index (QLI) was one of the instruments reviewed. This meant that articles for the QLI that were newly published or in review were excluded from consideration.
Our review was performed systematically, applying the same rules for all four instruments. We searched the CINAHL and MEDLINE databases (English only) using the keyword quality of life, as well as the names of the four instruments that we reviewed. We also performed an ancestry search in which we reviewed citations from the studies that were included in the review. Finally, we looked specifically for studies that examined the psychometric properties of the instruments. We evaluated every study identified by the search. Perhaps the greatest limitation to a comprehensive literature search that has both high recall (sensitivity) and high precision (specificity) are problems related to indexing. 1 Varying coding systems and improperly classified studies make the retrieval of relevant literature difficult. 2 An example of this is the fact that we did not locate any studies coded with the new name for the Quality of Life after Myocardial Infarction questionnaire.
We also would like to clarify three points. First, Drs Oldridge, Lim, and Guyatt raise the question of how strong a correlation needs to be to provide evidence of validity. The answer depends on whether the two measures are thought to assess similar constructs or different ones. For similar constructs, in our review we cited three authoritative sources, all of which agreed that the correlations should be "substantial" and "high" for evidence of validity. The values of 0.70 and 0.75 were provided as examples by Waltz et al, 3 and not as strict criteria. For our review the standard was that moderate correlations of at least 0.45 were taken as evidence of validity between measures of similar constructs. This standard was applied consistently across the four instruments. Second, Dr Oldridge and colleagues stated we challenged the responsiveness of their instrument in the Hillers et al 4 study. In fact, we actually concluded that the instrument was more responsive to change than the other tools. Third, we did err in concluding that the Lim et al 5 study found no statistically different scores between the experimental and the control groups and was therefore not responsive. Seven of 12 (58%) comparisons were statistically significant, indicating that quality of life was better in the patients with no history of myocardial infarction or bypass surgery.
We concluded that more research was needed to establish the validity of the QLMI, because our search identified only a few studies presenting psychometric data. In their letter Drs Oldridge, Lim, and Guyatt cite new data supporting the validity of the instrument. This is good news for cardiovascular researchers whose challenge is to find the appropriate instrument for their population of patients.
Holli A. DeVon, PhD, RN
Carol Estwing Ferrans, PhD, RN
References