Validation of the Energy Conservation Strategies Inventory (ECSI)

Transcription

1 606 Journal of Pain and Symptom Management Vol. 43 No. 3 March 2012 Original Article Validation of the Energy Conservation Strategies Inventory (ECSI) Young Ho Yun, MD, PhD, Soo Hyun Kim, RN, PhD, A. Jin Yang, MS, Eunmi Ahn, MD, Sook Hyun Kim, RN, Dong Ok Shin, RN, Jung Sook Sun, RN, Soon Ok Kim, RN, Kwang Mi Lee, RN, Keon Suk Lee, RN, Sun Hwa Baik, RN, Suk Kyung Kim, RN, and Hae Suk Seo, RN Cancer Management Branch (Y.H.Y., A.J.Y., E.A.), Research Institute, National Cancer Center, Goyang; Department of Nursing, College of Medicine (S.H.K.), Inha University, Incheon; and Hospital (Y.H.Y., A.J.Y., E.A., S.H.K., D.O.S., J.S.S., S.O.K., K.M.L., K.S.L., S.H.B., S.K.K., H.S.S.), National Cancer Center, Goyang, Gyeonggi, Republic of Korea Abstract Context. In applying good energy conservation strategies to relieve cancerrelated fatigue, it is critical to first identify cancer patients who are at a high risk for poor energy conservation. However, instruments have not been developed to evaluate energy conservation strategies in an oncology setting. Objectives. The aim of this study was to validate an instrument that cancer patients may use to evaluate energy conservation strategies to overcome cancer-related fatigue. Methods. The questionnaire development followed a four-phase process: 1) item generation and reduction, 2) construction, 3) pilot testing, and 4) field testing. Using relevant and priority criteria, as well as pilot testing, we developed a 25-item questionnaire. After field testing, five items were discarded. Finally, 20 items were included in the Energy Conservation Strategies Inventory (ECSI). Factor analysis, multitrait scaling analysis, and Cronbach s a were used to determine the construct validity and reliability. Results. Factor analyses of data from 140 cancer patients resulted in the ECSI, which covers activities related to planning, overcoming distractions, labor saving, burden reducing, and comfort. All subscales (Cronbach s a range, 0.69e0.78) and total scores (Cronbach s a ¼ 0.87) were found to possess acceptable internal consistency. Conclusions. The good psychometric properties of the ECSI instrument show that it may be useful for measuring the frequency of energy conservation strategies used by cancer patients. J Pain Symptom Manage 2012;43:606e613. Ó 2012 U.S. Cancer Pain Relief Committee. Published by Elsevier Inc. All rights reserved. Key Words Energy conservation, validation, cancer-related fatigue Address correspondence to: Young Ho Yun, MD, PhD, Seoul National University College of Medicine, 103 Daehak-ro, Jongno-gu, Seoul , Republic of Korea. lawyun08@gmail.com Ó 2012 U.S. Cancer Pain Relief Committee Published by Elsevier Inc. All rights reserved. Accepted for publication: April 12, /$ - see front matter doi: /j.jpainsymman

2 Vol. 43 No. 3 March 2012 Energy Conservation Strategies Inventory 607 Introduction As the number of cancer patients continues to grow, it has become evident that fatigue is one of the major debilitating symptoms affecting them. 1e6 Therefore, attention has been increasingly directed at reducing fatigue in cancer patients. Studies have shown that fatigue is a symptom that is experienced during treatment in 90% of cancer patients, and more than 50% experience fatigue after treatment has been completed. 1,7,8 Furthermore, cancer-related fatigue (CRF) has profound negative effects on quality of life 1,6,9,10 and may cause delays in treatment. 10 Several strategies, such as energy conservation, have been advanced to reduce fatigue in cancer patients. 11 Energy conservation encompasses a common sense approach that helps patients prioritize and pace activities, as well as delegate less essential activities to others. 11 In applying good energy conservation strategies, it is critical to first identify cancer patients who are at high risk for poor energy conservation and then focus on outcomes that are amenable to changes. These outcomes may emerge out of processes that have an effect on energy conservation. Although a reliable and valid way to measure the impact of energy conservation strategies is still needed, certain measures, such as the Energy Conservation Strategies Survey (ECSS) 12 and Self-Efficacy for Performing Energy Conservation Strategies Assessment (SEPECSA), 13 which have satisfactory psychometric properties, were developed for multiple sclerosis. 14 However, both the ECSS and the SEPECSA have not been validated for cancer patients. Furthermore, they do not include subscales to evaluate energy conservation strategies in greater detail, which could be tailored for cancer patients. Therefore, we developed a tool to assess energy conservation strategies for cancer patients. We report the development and initial validation of the Energy Conservation Strategies Inventory (ECSI) (Appendix). Methods Study Design Development and validation of the ECSI comprised four phases: 1) item generation and reduction, 2) scale construction, 3) pilot testing, and 4) field testing. The study was approved by the Institutional Review Board of the National Cancer Center, Korea. Eligible patients signed informed consent forms. Phase I: Item Generation and Reduction. The aim of Phase I was to compile a list of relevant energy conservation issues that covered the areas of interest. We performed an extensive literature review using PUBMED, MEDRIC (Korean), and other databases, searching for the key words energy conservation, cancer, fatigue management, and activity, and generated a list of 66 potentially relevant items. We discussed these 66 items in semistructured interviews with 14 health care professionals (four medical oncologists, one radiation oncologist, two surgical oncologists, two family medicine physicians, two nurses, and three social workers) and 30 cancer patients (seven patients before treatment, 12 during the treatment, and 11 after treatment) in August 2007 and January We asked each participant to evaluate the relevance of the 66 items on a four-point scale ( not at all, a little, quite a bit, and very much ), with four being the most relevant, and the priority for inclusion of 20 items decided with a dichotomous scale (yes or no). After completion, participants were asked to provide any additional information they considered applicable. A $10 coupon was given to each participant. Items that met at least three of the five criteria 1) mean score $ 1.5, 2) prevalence ratio $ 30%, 3) range $ 2 points, 4) priority patients $ one-third, and 5) priority consultants $ one-third) were retained in the list; items meeting two or less of the criteria were deleted. These decision rules for adaptation are based on the guidelines for developing questionnaire modules by the European Organization for Research and Treatment of Cancer Quality-of-Life Group. 15 Based on the results collected from the participants and after additional discussions, we deleted 37 items and combined seven items into three to eliminate content overlap and shorten the instrument. We adopted strict cutoff criteria to minimize the time and burden required to respond. The preliminary ECSI comprised 25 items grouped into nine

3 608 Yun et al. Vol. 43 No. 3 March 2012 domains. The domains included planning, delegation, ergonomic positioning of the body, use of labor-saving devices, pacing oneself, physical labor, using efficient body mechanics, paying attention, and psychological effort. Phase II: Construction. We constructed a list of items for the ECSI from the 25 energy conservation issues. For the rating format, we selected a four-point Likert scale. Phase III: Pilot Testing. The purpose of pilot testing was to identify potential problems with administration (e.g., poor phrasing of questions) and determine whether questions needed to be added or eliminated. Before the pilot testing, 18 cancer patients completed the ECSI, a debriefing questionnaire, and a structured interview. The interview results helped us decide which items to include in the test. The debriefing form helped to identify which items were confusing, difficult to answer, or upsetting to the patient and captured any relevant comments. A coupon valued at $10 was given to each participant. Using input from patients and experts, we decided to use a four-point Likert scale (with zero being not at all ) to minimize inconvenience and cognitive burden. When we pilot tested this questionnaire on 18 cancer patients, we did not encounter any difficulties. Phase IV: Field Testing. We field tested the questionnaire to determine its reliability and validity. We enrolled patients from the National Cancer Center in Korea. Eligible patients were required to 1) be 18 years or older, 2) have been diagnosed with cancer (confirmed by an oncologist), 3) have not been diagnosed as having chronic fatigue syndrome, 4) be able to read and understand Korean and fill out the questionnaire, and 5) provide written informed consent. We confirmed the eligibility criteria by reviewing medical records and through an interview. Because previous researchers have suggested a ratio of about five to 10 subjects per item for validation of an instrument, 16,17 the 140 subjects included in the field test was appropriate. Statistical Analysis To examine construct validity, factor analysis was conducted using principal methods with orthogonal varimax rotation. We assessed the adequacy of the data set for the factor analyses using the Bartlett test of sphericity and the Kaiser-Meyer-Olkin measure of sampling accuracy. Also, we used multitrait scaling analysis to examine the extent to which the ECSI items could be combined into a more limited multi-item set. We evaluated the convergent validity of ECSI items by examining the correlation between an item and its own scale (a correlation of $0.4, corrected for overlap, was evidence of validity). For divergent validity, we compared the magnitude of the correlation of an item with its own scale to other scales. Scaling successes were defined as those cases in which an item was correlated significantly more with its own scale (corrected for overlap) than with a different scale. 18 To examine the internal consistency of the ECSI, item-to-total correlations and Cronbach s a coefficients were used. Item-to-total analysis demonstrates a correlation between the respective items and the summated scale score (without the respective item). Items with an item-to-total correlation coefficient less than 0.3 were, therefore, discarded to improve scale homogeneity. 19 A low Cronbach s a value suggests that some items either have a very high variability or that the items are not measuring the same thing. As recommended, we sought a Cronbach s a of 0.70 or greater as the minimum criterion for internal consistency. 19 SPSS version 17.0 software (SPSS Inc., Chicago, IL) was used for these analyses. Results Pilot Testing The interview results helped us determine which items to include in the test. According to those results, we increased the ECSI s clarity and ease of response, and we used a debriefing form to note which items were confusing, difficult to answer, or upsetting to the patient, along with noting relevant comments. Participant Characteristics Of the eligible patients, 140 cancer patients remained in the field study after we reviewed the completed questionnaires. Table 1 shows the demographic and clinical characteristics of

4 Vol. 43 No. 3 March 2012 Energy Conservation Strategies Inventory 609 Table 1 Sociodemographic and Clinical Characteristics of the Study Sample (n ¼ 140) Characteristics n (%) Age, years (mean [SD] 51.6 [11.4]) <45 41 (29.3) 45e64 79 (56.4) $65 20 (14.3) Gender Men 75 (53.6) Women 65 (46.4) Education <High school education 49 (35.0) $High school education 91 (65.0) Occupation Employed 57 (40.7) Unemployed 83 (59.3) Marital status With spouse 115 (82.1) Without spouse 25 (17.9) Cancer site Stomach 25 (17.9) Breast 23 (16.4) Liver 22 (15.7) Colon 10 (7.2) Head and neck 7 (5.0) Uterus 3 (2.1) Others 50 (35.6) Stage 0eI 21 (23.1) II 20 (22.0) III 13 (14.2) IV 37 (40.7) Missing 21 (15.0) Metastasis No 113 (80.7) Yes 27 (19.3) Type of treatment a Surgical 87 (62.6) Chemotherapy 70 (50.0) Radiotherapy 43 (30.7) Status of treatment Pretreatment 30 (21.4) During treatment 80 (57.1) After treatment 30 (21.4) ECOG PS 0 ¼ Fully active 70 (50.0) 1 ¼ Restricted but ambulatory 64 (45.7) 2 ¼ Ambulatory, capable of self-care 6 (4.3) 3 ¼ Capable of only limited self-care 0 (0) 4 ¼ Completely disabled 0 (0) ECOG PS ¼ Eastern Cooperative Oncology Group performance status. Valid percentages are reported in cases where data were missing. a Multiple responses were possible. the participants. The mean age of the subjects was 51.6 years (standard deviation ¼ 11.4), and the proportion of men and women was almost equal. Most patients (82.1%) were married, and 65% had at least a high school education. The majority (95.7%) had a good functional status (an Eastern Cooperative Oncology Group scale score of zero or one). Factor Analysis Based on item-to-total analysis, four items (two items for delegation, one item for setting the priorities, and one item for operation of adaptive tools and equipment ) were eligible to be discarded (correlation coefficients <0.30). However, all the researchers agreed that the item for setting the priorities should be included in the total items because of its conceptual importance. 10,20,21 Thus, a total of 22 items were included for factor analysis. Factor analysis initially extracted six factors, but one factor was found to have a very poor internal consistency (Cronbach s a ¼ 0.34); this factor comprised the items change in the placement of equipment and tools and using home appliances. Therefore, we discarded these two items. All the researchers agreed to the deletion of the items for the sake of brevity of the questionnaire. The second factor analysis extracted five factors, accounting for 60.1% of the total variance, and all the eigenvalues were >1.0. The Bartlett test of sphericity yielded a statistically significant result (P < 0.001), indicating that the items share a common variance (i.e., item correlations do not form an identity matrix), and the Kaiser-Meyer-Olkin measure of sampling adequacy was 0.830, suggesting that the variables measured more than one component. Table 2 lists the item-to-factor loadings for the 20 items and five factors from the varimax rotation factor analysis. The scores for the total ECSI and subscales were calculated as the sum of each item, where a higher score represents a better energy conservation strategy. The range of total ECSI scores was 20e80. Reliability Table 3 presents the descriptive statistics and reliability measures for the ECSI subscales. Internal consistency (Cronbach s a) for all the ECSI subscales emerged as moderate to high (0.69e0.78), and the total for over 20 items was Multitrait Scaling The ECSI showed relatively good convergent and divergent validity (Table 3). All items from

5 610 Yun et al. Vol. 43 No. 3 March 2012 Table 2 Varimax Rotated Factor Analysis of Patient Response From the Energy Conservation Strategies Inventory Subscales No. Item FAC1 FAC2 FAC3 FAC4 FAC5 16 Nature Enjoyable activities Control life with rhythm Balance of work and rest Mind control Music and meditation Using a chair with the back and arm Ergonomic positioning Comfortable clothes and footwear Reducing mental exhaustion One thing at a time Avoiding the rush hour Identification of work level Setting the priorities Make a list of purchases Regular life Elimination of heavy tasks Take a nap Avoiding heavy objects Simplify the tasks Eigenvalue Variance explained, % Total variance explained, % 60.1 FAC ¼ factor. Significance <0.05 for all the bold values. the Distraction, Comfort, and Burden Reducing scales correlated at least 0.4 with their own hypothesized scales, corrected for overlap, and no scaling errors were seen with these items. However, two items from Planning and Labor Saving scales did not meet the 0.4 criterion (r ¼ 0.37 and r ¼ 0.38, respectively), and one item showed a scaling error. Overall, the analysis confirmed that the ECSI met the recommended psychometric standards. Table 3 shows the five subscales: Planning (Items 1e4), Labor Saving (Items 5e8), Comfort (Items 9e11), Distraction (Items 12e17), and Burden Reducing (Items 18e20). Discussion We report the development and validation of a questionnaire to assess energy conservation strategies for CRF, which was shown to possess satisfactory psychometric properties. We included only 20 items because brevity is critical for patients with fatigue. There is no specific treatment for CRF such as morphine in the case of pain or selective serotonin reuptake inhibitors for depression. Therefore, self-care is a critically important aspect to the management of CRF. The National Comprehensive Cancer Network (NCCN) says that energy conservation is a useful self-care Table 3 Descriptive Statistics, Reliability, and Multitrait Scaling of the Energy Conservation Strategies Inventory Subscales Range of Scores Mean (SD) Cronbach s a Item-Own Scale Correlation a Item-Other Scale Correlation Scaling Success Scaling Error (%) Distraction (Items 12e17) 6e (3.9) e e /30 0 Comfort (Items 9e11) 3e (2.2) e e /15 0 Burden Reducing (Items 18e20) 3e (2.1) e e /15 0 Planning (Items 1e4) 4e (2.6) e e /20 0 Labor Saving (items 5e8) 4e (2.6) e e /20 5 SD ¼ standard deviation. a Corrected for overlap.

6 Vol. 43 No. 3 March 2012 Energy Conservation Strategies Inventory 611 strategy for individuals with CRF, with potential benefits for anemia, pain, insomnia, physical activity, distress, and nutrition. 11 There is a misconception that CRF is similar to body fatigue that is responsive to increased rest. Whereas CRF is not caused by overactivity, rest is a means of conserving energy resources, which may enable an individual to continue to participate in his or her most valued activities. 10 The results of our study indicate that the ECSI has certain important psychometric properties, such as acceptable internal consistency and convergent construct validity. Energy conservation strategies commonly include priority setting, delegation, pacing oneself, and planning highenergy activities to take place at times of peak energy. 10 The ECSI consists of five subscales that focus on fatigue and are consistent with previous concepts of energy conservation strategies and suggestions: 1 Planning, Overcoming Distraction, Labor Saving, Comfort, and Burden Reducing activities. Unlike ECSS and SEPECSA, which were developed for multiple sclerosis, 12e14 ECSI has some advantages. Five subscales such as Planning, Overcoming Distractions, Labor Saving, Burden Reducing, and Comfort may evaluate energy conservation strategies in more detail, and they may be tailored for cancer patients. The ECSI has three items for Comfort in lieu of adaptive tools and equipment, or the assignment of tasks to others. In addition, the ECSI does not have any items for Delegation. During Phase IV, the items for Delegation and Operation of adaptive tools and equipment were deleted based on the findings from the factor analysis, and for the sake of brevity of the questionnaire, which is a critical consideration for patients with fatigue. The ECSI does have six items for Distraction, such as enjoyable activities, mind control, meditation, and nature. We found some items that exhibited double loading; mind control loaded equally into the Distraction and Burden Reducing domains; avoiding the rush hour loaded equally into the Comfort and Burden Reducing domains; regular life loaded equally into the Distraction and Planning domains; and simplify the tasks loaded equally into the Burden Reducing and Labor Saving domains. Among them, determining each domain of avoiding the rush hour and regular life was reasonable. However, we had difficulty in determining the domain for mind control and simplify the tasks. As for mind control, because it is related to relaxation, we included it in the Distraction rather than the Burden Reducing domain. In the case of simplify the tasks, it asked whether subjects reduce or skip their tasks. It may be to lessen physical labor. Therefore, we determined its domain as Labor Saving. This study has several limitations. First, because it exclusively relied on cancer patients, most of the concepts measured by the ECSI are specific to cancer patients and the results may not be able to be generalized to patients with other chronic diseases. With some modifications, however, they may be applied to measuring energy conservation strategies in patients with other chronic diseases. Second, because this study was performed in Korea, generalizations to other countries should be made cautiously, with proper cross-cultural validation studies. Third, the development of the ECSI was based on the recommendation of the NCCN fatigue guideline that the energy conservation strategy is one of the nonpharmacologic management strategies for CRF. However, the NCCN recommendation is based on only one randomized controlled trial by Barsevick et al., and the intervention has not yet been established. Fourth, because the usefulness of the ECSI has not been confirmed, intervention studies need to be performed to clarify the clinical usefulness of this scale before implementing it into clinical practice. Finally, we did not include having fatigue in the eligibility criteria and did not describe the degree of fatigue the patients experienced. Further studies need to be performed with the ECSI and other instruments to assess fatigue. Nevertheless, this simple and sensitive assessment tool with good psychometric properties may be clinically useful in developing energy conservation strategies for cancer patients and in evaluating the effects of energy conservation strategies on CRF. Specifically, patients with advanced cancer may be good candidates for energy conservation strategies with the ECSI. Disclosures and Acknowledgments This work was supported by the National Cancer Center Grant The authors

7 612 Yun et al. Vol. 43 No. 3 March 2012 declare no conflicts of interest. We thank BioSciEditors for professional editing. References 1. Kim SH, Son BH, Hwang SY, et al. Fatigue and depression in disease-free breast cancer survivors: prevalence, correlates, and association with quality of life. J Pain Symptom Manage 2008;35:644e Geinitz H, Zimmermann FB, Stoll P, et al. Fatigue, serum cytokine levels, and blood cell counts during radiotherapy of patients with breast cancer. Int J Radiat Oncol Biol Phys 2001;51:691e Morrow GR, Hickok JT, Roscoe JA, et al. Differential effects of paroxetine on fatigue and depression: a randomized, double-blind trial from the University of Rochester Cancer Center Community Clinical Oncology Program. J Clin Oncol 2003;21: 4635e Morris T, Greer HS, White P. Psychological and social adjustment to mastectomy: a two-year followup study. Cancer 1977;40:2381e Meyer L, Aspegren K. Long-term psychological sequelae of mastectomy and breast conserving treatment for breast cancer. Acta Oncol 1989;28:13e Weitzner MA, Meyers CA, Stuebing KK, Saleeba AK. Relationship between quality of life and mood in long-term survivors of breast cancer treated with mastectomy. Support Care Cancer 1997;5:241e Broeckel JA, Jacobsen PB, Horton J, Balducci L, Lyman GH. Characteristics and correlates of fatigue after adjuvant chemotherapy for breast cancer. J Clin Oncol 1998;16:1689e Okuyama T, Akechi T, Kugaya A, et al. Factors correlated with fatigue in disease-free breast cancer patients: application of the Cancer Fatigue Scale. Support Care Cancer 2000;8:215e Jacobsen PB, Stein K. Is fatigue a long-term side effect of breast cancer treatment? Cancer Control 1999;6:256e Barsevick AM, Dudley W, Beck S, et al. A randomized clinical trial of energy conservation for patients with cancer-related fatigue. Cancer 2004;100:1302e Berger AM DW, Beck S, Sweeney C, Whitmer K, Nail L. NCCN Practice Guidelines in Oncology. Cancer-related fatigue. In: NCCN, ed., Available from physician_gls/pdf/fatigue.pdf. Accessed January 31, Mallik PS, Finlayson M, Mathiowetz V, Fogg L. Psychometric evaluation of the Energy Conservation Strategies Survey. Clin Rehabil 2005;19:538e Liepold A, Mathiowetz V. Reliability and validity of the Self-Efficacy for Performing Energy Conservation Strategies Assessment for persons with multiple sclerosis. Occup Ther Int 2005;12:234e Matuska K, Mathiowetz V, Finlayson M. Use and perceived effectiveness of energy conservation strategies for managing multiple sclerosis fatigue. Am J Occup Ther 2007;61:62e Blazeby J, Sprangers M, Cull A, Groenvold M, Bottomley A. Guidelines for developing questionnaire modules, 3rd ed. Brussels, Belgium: EORTC Quality of Life Group, Ho HZ, O Farrell S, Hong SH, You S. Developmental research: theory, method, design, and statistical analysis. In: Green JL, Camilli G, Elmore PB, eds. Complementary methods for research in education, 3rd ed. Washington, DC: American Educational Research Association, Cochran WG. Sampling techniques. New York, NY: John Wiley & Sons, Fayers PM. Quality of life: Assessment, analysis and interpretation. West Sussex, UK: John Wiley & Sons, Ltd, Harman HH. Modern factor analysis. Chicago, IL: University of Chicago Press, Barsevick AM, Whitmer K, Sweeney C, Nail LM. A pilot study examining energy conservation for cancer treatment-related fatigue. Cancer Nurs 2002;25:333e Yuen HK, Mitcham M, Morgan L. Managing post-therapy fatigue for cancer survivors using energy conservation training. J Allied Health 2006; 35:121Ee139E.

8 Vol. 43 No. 3 March 2012 Energy Conservation Strategies Inventory 613 Appendix Energy Conservation Strategies Inventory The following questions aim to learn more about the everyday activities that can have an impact on your level of fatigue. Please read each question carefully and check the number which best describes your activity during the past one week. Very Often Often Sometimes Never 1. Before doing something, I make plans regarding the process, the steps, and the amount of time I will spend on it. 2. I set priorities on the things I have to do I make a shopping list before making a purchase I follow a regular pattern in my everyday activities (such as eating, sleeping, etc.). 5. I take just enough naps so that they do not disturb my nighttime sleep I try to avoid activities that consume a lot of energy Whenever possible, I try to skip activities or procedures that are not absolutely necessary. 8. I avoid lifting or moving heavy objects I adjust my posture so that I can do something more efficiently and comfortably I use a chair with a backrest and armrest I wear clothes and shoes that are easy to put on and comfortable to wear I adjust my work and lifestyle to my personal rhythm I include time during the day for rest to maintain a good balance between work and rest. 14. When I am very tired, I take a break by listening to music, meditating, etc I usually do activities that I can have fun doing I sometimes escape from my everyday life and enjoy nature (walks, hiking, traveling, etc.). 17. I always live my life with a relaxed attitude I do things one at a time I try to avoid crowded times when I do something I reduce the amount of attention I pay to things that do not concern me Copyright 2007, NCC Research Institute. All rights reserved.