A 3-Factor Model for the FACIT-Sp Reference: Canada, Murphy, Fitchett, Peterman, Schover. Psycho-Oncology. Published Online: Dec 19, 2007; DOI: 10.1002/pon.1307. Copyright John Wiley & Sons Ltd. Investigators Andrea L. Canada, Patricia E. Murphy, George Fitchett Rush University Medical Center, Chicago, IL Amy H. Peterman UNC Charlotte, Charlotte, NC Leslie R. Schover UT M. D. Anderson Cancer Center, Houston, TX This study was funded in part by grants from the National Institutes of Health, R21 CA106958 (Canada) and K08 AG020145 (Fitchett). 1
INTRODUCTION A growing number of studies document that many people with a diagnosis of cancer turn to religion and spirituality (R/S) to help them cope with the illness and its treatment. One explanation for the pervasive use of R/S coping in the face of cancer is as follows: Religious and spiritual beliefs provide patients with a framework for ascribing meaning to the illness and answers to existential questions (Holland et al., 1999; p. 23). Among investigators who study the role of R/S in coping with cancer, it is now widely recognized that R/S represents a broad construct with a number of subdimensions, including beliefs, public and private devotional practices, experiences, and relationships (Fetzer/NIA Working Group, 1999; Hill & Pargament, 2003). Good measures of R/S are a prerequisite for advancing our understanding of the role of R/S in coping with cancer (Stefanek et al., 2005). An important feature of such good measures is clarity about the dimensions of R/S that are assessed. 2
The Functional Assessment of Chronic Illness Therapy- Spiritual Well-Being scale (FACIT-Sp) was developed and validated with 1,617 oncology (breast, colon, lung, or head and neck cancer) and HIV patients (Peterman et al., 2002). A principal components analysis with Varimax rotation was conducted on the 12 items and revealed the presence of two factors: 1. Meaning/Peace contains 8 items assessing a sense of meaning, peace, and purpose in life; and 2. Faith contains 4 items measuring several aspects of the relation between illness and one's faith and spiritual beliefs. Recently, we re-examined the items in the FACIT-Sp, specifically the 8 items in the Meaning/Peace subscale. In light of the importance of clarity about the specific dimensions of R/S that are measured by a given instrument, we questioned whether these 8 items actually formed two separate factors: 1. Meaning assessing a cognitive dimension of R/S; and 2. Peace assessing an affective dimension of R/S. No changes to the Faith factor were warranted. 3
ORIGINAL 2- FACTOR SOLUTION Item 2 reason for living PROPOSED 3- FACTOR SOLUTION Item 3 life been productive Item 5 sense of purpose MEANING MEANING /PEACE Item 8r life lacks meaning and purpose Item 1 feel peaceful Item 4r trouble feeling peace of mind Item 6 deep inside feel comfort PEACE Item 7 sense of harmony Item 9 comfort in faith FAITH Item 10 strength in faith Item 11 difficulties strengthen faith FAITH Item 12 whatever happens, okay Study Aims The primary aim of the present study was to employ confirmatory factor analysis to test for evidence of a 3-factor solution for the FACIT-Sp scale. The secondary aim was to examine whether the 3-factor solution advanced our understanding of the relationships between R/S and QoL/emotional well-being in a sample of cancer survivors. 4
Sample METHODS Participants were part of a larger study designed to assess quality of life in long-term female cancer survivors. Participants were recruited from The University of Texas M. D. Anderson Cancer Center (UTMDACC) institutional tumor registry. Consenting women were asked to complete a 60- to 90- minute structured telephone interview consisting of demographic items and a number of standardized questionnaires. Participants were a minimum of 5 years post-diagnosis (invasive cervical cancer, breast cancer, Hodgkin s disease, or non-hodgkin s lymphoma) at the time of study participation. Measures Religion/Spirituality (R/S) was assessed with the Functional Assessment of Chronic Illness Therapy Spiritual Well-Being Scale (FACIT-Sp). Health-related quality of life (QoL) was measured using the Medical Outcomes Study 12-item Short Form (SF-12) Physical (PCS) and Mental Health Component Scores (MCS) (Gandek et al., 1998). Emotional distress was assessed using the Brief Symptom Inventory 18 (BSI-18) (Derogatis, 2001; Zabora et al., 2001), which measures Somatization, Depression, and Anxiety, as well as a Global Severity Index (GSI) or summary score. 5
Analyses Confirmatory factor analysis (CFA) with LISREL 8.8 (Jöreskog & Sörbom, 2006) was used to compare goodness-of-fit of both the 2- and 3-factor models to our data. SPSS partial correlations were used to compute the associations of QoL/emotional well-being measures with the total FACIT-Sp score, and with each factor of both the 2- and 3-factor models. Results Table 1. Demographics of Long-Term Female Survivors (N=240) Variable Age Time Since diagnosis Ethnicity Range = 21 53 yrs M = 43.5 yrs SD = 6.6 Range = 6 15 yrs M = 9.8 yrs SD = 1.8 n % Caucasian 192 80.0 African-Amercian 14 5.8 Hispanic 18 7.5 Other 16 6.7 Marital Status Education Income Diagnosis Married 190 79.2 Divorced/Separated /Widowed 22 9.3 Never Married 28 11.7 < High School 41 17.1 Some College 73 30.4 > Bachelor s Degree 126 52.6 < $25,000 18 7.5 $26,000 75,000 89 37.1 > $75,000 121 50.4 N/A 12 5.0 Lymphoma 96 40.0 Breast Cancer 123 51.3 Cervical Cancer 21 8.8 6
Table 2. Results of Confirmatory Factor Analysis of the FACIT-Sp Model Satorra -Bentler a df p Scaled Δχ 2 df p b CFI c SMRM d χ 2 Model 1 2 Factors 176.48 53.000.97.100 Model 2a 3 Factors 104.12 51.000 72.36 2.000.99.091 Model 2b adjusted model 2a with correlated error covariance for 77.13 50.008 26.99 1.000.99.086 reverse worded items Model 2c adjusted model 2b with correlated error covariance for items 1 and 4 66.09 49.052 38.03 2.000 1.00.085 Model 2d adjusted model 2c with path from Peace to item 12 added 59.11 48.13 6.98 1 <.01 1.00.054 Values for good fit a p>.05; b p<.05; c cutoff <.95; d cutoff >.09.30 Item 2 reason for living.84.33.27 Item 3 life been productive Item 5 sense of purpose.82.85 MEANING.47 Item 8r life lacks meaning and purpose.73.23.56.17.53 Item 1 feel peaceful Item 4r trouble feeling peace of mind.66.69 PEACE.74.36.30.10 Item 6 deep inside feel comfort Item 7 sense of harmony.83.95.30.42.05 Item 9 comfort in faith.98.02.33.57 Item 10 strength in faith Item 11 difficulties strengthen faith Item 12 whatever happens, okay.99.82.47 FAITH Figure 1. Standardized Coefficients for Final Model 7
Table 4. FACIT-Sp and QoL: Partial Correlations QoL Measure FACIT-SP Original 2 Factors Proposed 3 Factors Total Meaning/Peace Faith Meaning Peace Faith BSI 18 Somatization -.28** -.44***.18** -.28*** -.20**.17** Depression -.52** -.65***.19** -.32*** -.45***.19** Anxiety -.49** -.62***.15* -.22** -.48***.15** SF 12 Global Severity Index Mental Health Component Physical Health Component -.50** -.63***.22** -.29*** -.45***.22**.14*.63*** -.15*.17**.53*** -.17*.50**.22** -.09.18**.05 -.09 *p <.05; **p <.01; ***p <.001. bivariate correlation DISCUSSION Research about R/S and living with cancer requires thoughtful consideration about the dimensions of R/S that are relevant for a given study and selection of the best measures of those dimensions. In the present study we have demonstrated that the FACIT-Sp, a popular measure for research about R/S and coping with cancer, has 3 factors that measure 3 different dimensions of R/S, namely, meaning, peace, and faith. Strengths of the FACIT-Sp are its reasonably strong psychometric properties, its ability to assess these 3 dimensions, and use of language about R/S that is acceptable to a wide population. 8
The findings from this study must be interpreted in light of several limitations: This sample consisted of only relatively young women (treated for cancer between the ages of 14 and 40 years) who were primarily married, highly educated, from the Bible belt region of the United States, and an average of 10 years post-diagnosis. There is evidence to suggest a differential role of R/S in people from different ethnic or age groups (Moadel et al., 2007; Simon et al., 2007). As such, meaning, peace, and faith may have different associations with QoL and distress in populations unlike our sample. Another limitation of our study is that, given the requirements of LISREL, our sample of 240 survivors is relatively small. As such, further research should build upon this work, validating the 3-factor structure stability across a more diverse survivorship population. 9
Conclusion What is the best approach to the measurement of R/S in health-related research? Some investigators have criticized the FACIT-Sp as measuring an outcome of R/S versus a constitutive element of R/S (Koenig, McCullough, Larson, 2001). Others have questioned the value of measures of generic spirituality as compared to measures rooted in the worldview of specific faith traditions (Glicksman, 2002; Hall et al., 2004, 2008). We believe there may be a variety of useful approaches to the measurement of R/S in health-related research. The key to good research in this, as in any, field, is the thoughtful development of study aims and the selection of psychometrically sound measures appropriate to those aims. Where study aims call for a brief measure of the dimensions of meaning, peace, and faith, we believe the FACIT-Sp is a measure worthy of consideration. 10