Measurement of the quality of life in cancer survivors.

A QOL instrument was developed to measure the specific concerns of long term cancer survivors. The QOL-CS is based on previous versions of the QOL instrument developed by researchers at the City of Hope National Medical Centre (Grant, Padilla, and Ferrell). This instrument was revised over a one year pilot by Hassey-Dow and Ferrell. The revised instrument included 41 items representing the four domains of quality of life incorporating physical, psychological, social, and spiritual well being. The present study was conducted as a mail survey to the membership (n = 1,200) of the National Coalition for Cancer Survivorship with 686 subjects responding to the survey. This survey included a Demographic tool, QOL-CS and the FACT-G tool developed by Cella. Psychometric analysis, performed on 686 respondents, included measures of reliability and validity. Two measures of reliability included test-retest and internal consistency. The overall QOL-CS tool test-retest reliability was 0.89 with subscales of Physical r = 0.88, Psychological r = 0.88, Social r = 0.81, and Spiritual, r = 0.90. The second measure of reliability was computation of internal consistency using Cronbach's alpha coefficient as a measure of agreement between items and subscales. Analysis revealed an overall r = 0.93. Subscale alphas average ranged from r = 0.71 for spiritual well being, r = 0.77 for physical, r = 0.81 for social, and r = 0.89 for psychological. Several measures of validity were used to determine the extent to which the instrument measured the concept of QOL in cancer survivors. The first method of content validity was based on a panel of QOL researchers and nurses with expertise in oncology. The second measure used stepwise multiple regression to determine factors most predictive of overall QOL in cancer survivors. Seventeen variables were found to be statistically significant accounting for 91% of the variance in overall QOL. The fourth measure of validity used Pearson's correlations to estimate the relationships between the subscales of QOL-CS and the subscales of the established FACT-G tool. There was moderate to strong correlation between associated subscales including QOL-CS physical to FACT physical (r = 0.74), QOL-CS Psych to FACT Emotional (r = 0.65), QOL Social to FACT Social (r = 0.44). The overall QOL-CS correlation with the FACT-G was 0.78. Additional measures of validity included correlations of individual items of the QOL-CS tool, factor analysis, and construct validity discriminating known groups of cancer survivors. Findings demonstrated that the QOL-CS and its subscales adequately measured QOL in this growing population of cancer survivors.