The Psychometric Testing of the Diabetes Health Promotion Self-Care Scale

Transcription

1 The Journal of Nursing Research h VOL. 20, NO. 2, JUNE 2012 The Psychometric Testing of the Diabetes Health Promotion Self-Care Scale Ruey-Hsia Wang 1 * & Li-Ying Lin 2 & Chung-Ping Cheng 3 & Min-Tao Hsu 1 & Chia-Chan Kao 4 1 RN, PhD, Professor, College of Nursing, Kaohsiung Medical University & 2 RN, MSN, Supervisor, Department of Nursing, Kaohsiung Veterans General Hospital & 3 PhD, Associate Professor, Department of Psychology, National Cheng Kung University & 4 RN, PhD, Assistant Professor, Department of Healthcare Administration, I-Shou University. ABSTRACT Background: Health-promoting behavior is an important strategy to maintain and enhance health of patients with Type 2 diabetes. Few instruments have been developed to measure health promotion self-care behavior of patients with Type 2 diabetes. Purpose: Developing and psychometric testing of the Chinese version of the Diabetes Health Promotion Self-Care Scale (DHPSC) for patients with Type 2 diabetes. Methods: Four hundred and eighty-nine patients with Type 2 diabetes were recruited from endocrine clinics in four hospitals in Kaohsiung City in southern Taiwan. Exploratory and confirmatory factor analyses were used to assess the construct validity of the scale. Correlations between the DHPSC and the satisfaction subscale of Diabetes Quality of Life, Diabetes Empowerment Scale, and HbA1c were calculated to evaluate concurrent validity. Internal consistency and testyretest reliability were used to assess the reliability of the scale. The study was conducted in 2007 and Results: A proposed second-order factor model with seven subscales and 26 items fit the data well. The seven subscales were interpersonal relationships, diet, blood glucose self-monitoring, personal health responsibility, exercise, adherence to the recommended regimens, and foot care. The DHPSC statistically significantly correlated with the satisfaction subscale of Diabetes Quality of Life and the Diabetes Empowerment Scale. HbA1c only statistically significantly correlated with the subscale of health responsibility. Reliability was supported by acceptable Cronbach s alpha (range,.78y.94) and testyretest reliability (range,.76y.95). Conclusions/Implications for Practice: The DHPSC has satisfactory reliability and validity. Healthcare providers can use the DHPSC to comprehensively assess the health promotion self-care behaviors of patients with Type 2 diabetes. KEY WORDS: diabetes health promotion self-care, instrument development, Type 2 diabetes. Introduction Diabetes is a health problem affecting more than 180 million people worldwide, and this number is expected to more than double by 2030 (World Health Organization, 2009). In the United States, there are 23.6 million individuals with diabetes, accounting for 7.8% of the total population (American Diabetes Association, 2010). In Taiwan, the prevalence rate of diabetes is 9.2%, affecting 2 million people (Department of Health, 2009), nearly 98% of whom have Type 2 diabetes (Department of Health, 2009). Diabetes is thus an important health issue for Taiwanese people. Type 2 diabetes requires many self-care measures to control its progression and complications (Funnell et al., 2011). Self-care behaviors have been defined by Orem (1995) as actions taken by individuals to care for themselves within their environmental conditions. Health promotion and maintenance are recognized as outcomes achieved through selfcare (Orem, 1995). Traditionally, the goal of self-care behaviors in diabetes patients places more emphasis on maintaining good blood glucose levels and preventing complications (Weinger, Butler, Welch, & La Greca, 2005). Most of these self-care tactics emphasize regulating eating habits, engaging in physical activity, following a medication regimen, and self-monitoring blood glucose levels (Toobert, Hampson, & Glasgow, 2000; Weinger et al., 2005). These behaviors highlight activities related to illness control that contribute primarily to physical outcome. Facing complex and demanding activities for disease control, patients with diabetes often experience emotional distress and a low quality of life (Furler et al., 2008). Thus, psychosocial issues must be considered in addition to physical issues (Paul, Smith, Whitford, O Kelly, & O Dowd, 2007). Accepted for publication: March 5, 2012 *Address correspondence to: Ruey-Hsia Wang, No. 100, Shih-Chuan 1st Road, Kaohsiung City 80708, Taiwan, ROC. Tel: +886 (7) ext. 2641; Fax: +886 (7) ; wrhsia@kmu.edu.tw doi: /jnr.0b013e318254eb47 122

2 Diabetes Health Promotion Self-Care Scale VOL. 20, NO. 2, JUNE 2012 Health promotion emphasizes lifestyles that promote physical and psychosocial well-being (Pender, 1996). The practice of health-promoting behaviors has been acknowledged as an important strategy for maintaining and enhancing the quality of life in persons with chronic disability conditions (Mendoza-Núñez, Martínez-Maldonado, & Correa-Muñoz, 2009). Incorporating health promotion into self-care is important for patients with diabetes (Anthony, Odger, & Kelly, 2004). Understanding health-promoting self-care behaviors in patients with diabetes will help nurses design health promotion intervention programs. However, few instruments have been developed to measure the healthpromoting self-care behaviors of diabetes patients. Many scales have been developed to measure the self-care behaviors of diabetes. The Summary of Diabetes Self-Care Measures (SDSCA) and the Self-Care Inventory-Revised (SCI-R) are supported as well-validated scales for measuring diabetes self-care behaviors (Eigenmann, Colagiuri, Skinner, & Trevena, 2009). The SDSCA is a self-report instrument (Aikens, Bingham, & Piette, 2005; Deakin, Cade, Williams, & Greenwood, 2006; Xu, Pan, & Liu, 2010). The latest revised 11-item SDSCA assesses diet, blood glucose testing, foot care, and smoking behaviors (Toobert et al., 2000). The SCI-R is a 14-item scale that evaluates individuals perceptions of their adherence to treatment prescriptions (Weinger et al., 2005). The SCI-R only addresses diet, glucose monitoring, medication, administration, exercise, low glucose levels, and the preventive/routine aspects of self-care. Both SDSCA and SCI-R measure illness control self-care behaviors but ignore measuring self-care behaviors relevant to psychosocial health. Health-promoting behavior is considered as a multidimensional pattern of self-initiated actions and perceptions that serve to maintain or enhance the wellness level (Walker & Hill-Polerecky, 1996). Health responsibility and interpersonal relations are important for health promotion (Walker & Hill-Polerecky, 1996). Health responsibility entails an active sense of accountability for one s own well-being (Walker & Hill-Polerecky, 1996). More generally, people require personal commitment and responsibility for acquiring the knowledge and skills necessary for improving and maintaining their health (Simmons, 1990). Furthermore, personal health responsibility is a fundamental aspect of patient empowerment for diabetes patients (Wong, Stewart, & Furler, 2009). Interpersonal relations entail utilizing communication to achieve a sense of intimacy and closeness in meaningful relationships with others. Harmonious relationships between the self and others are important for Chinese individuals (Wong et al., 2009). Harmonious relationships are important for a good health-related quality of life in patients with diabetes (Choe, Padilla, Chae, & Kim, 2001). Physical activity, nutrition, blood glucose self-monitoring, adherence to recommended regimens, foot care, personal health responsibility, and interpersonal relationships can be complementary in improving the physical and psychosocial health of patients with diabetes. In this study, diabetes health promotion self-care behavior was conceptualized as a set of behaviors performed by patients with diabetes to improve their physical and psychosocial well-being. Interpersonal relationships were defined as behaviors for maintaining meaningful relationships with others. Personal health responsibility was defined as behaviors actively involved in health-benefitting actions. The aim of this study was to develop and test the psychometric properties of the Chinese version of Diabetes Health Promotion Self-Care Scale (DHPSC) for patients with Type 2 diabetes. Methods Item Generation The DHPSC included seven behavior dimensions: diet, exercise, blood glucose self-monitoring, adherence to recommended regimens, foot care, interpersonal relationships, and personal health responsibility. Items for the first five components were developed primarily from a Chinese version of a diabetes self-care behavior scale (Wang, Wang, & Lin, 1998) that had satisfactory validity and reliability in the Chinese population. Interpersonal relationship and personal health responsibility items were also based on previous empirical studies (Walker, Sechrist, & Pender, 1995; Wang & Hsu, 1997). Diet (three items), exercise (three items), blood glucose self-monitoring (five items), the adherence to recommended regimens (three items), foot care (two items), interpersonal relationships (seven items), and personal health responsibility (five items) were included in the DHPSC. The preliminary DHPSC scale was a 28-item scale with seven behavior dimensions. The scale was a 5-point scale, which ranged from always do (5 points) to never do (1 point). Higher scores indicated better health-promoting self-care behaviors. Content Validity and Item Analysis Seven experts, including two endocrine physicians, one diabetes educator, and four nursing professors, examined the content validity of the preliminary DHPSC. The experts were asked to use a 4-point rating scale to assess relevance of each item. The rating was as follows: 1 = not relevant;2=unable to assess relevance without item revision; 3=relevant but needs minor revision; and4=very relevant and succinct. The number of items rated as 3 or 4 by experts divided by the total numbers of items was calculated as the content validity index of the scale. The content validity index for the DHPSC was.94, which indicated that all items were acceptable for further use (Waltz, Strickland, & Lenz, 2005). The item total correlations of each item were between.30 and.70, except for three items from the adherence to recommended regimens subscale and one item from the foot care subscale, which were all lower than.30. Considering the comprehensiveness of the DHPSC and the number of items in each subscale, the four items were retained in further analysis. 123

3 The Journal of Nursing Research Ruey-Hsia Wang et al. Construct Validity Exploratory factor analysis and confirmatory factor analysis Exploratory factor analysis (EFA) was first used to examine the construct validity of the 28-item DHPSC. The number of factors retained was determined by the researchers substantive consideration, the eigenvalue-greater-than-one rule (Kaiser s rule) and the scree test. The principal axis factor method was used as the extraction method. Because we expected that the subfactors would be interrelated (Costello & Osborne, 2005), an oblique promax was selected as the rotation method. A minimum factor-loading coefficient of.50 was selected as a criterion for retaining an item in the scale. To test how the data fit the factor structure proposed by the EFA and the hierarchical relationship of factors, confirmatory factor analysis (CFA) was performed with a structural equation-modeling program by EQS (version 6.1; Multivariate Software, Inc., Los Angeles, CA, USA). A second-order CFA model was also proposed to assess whether all first-order factors were loaded on a higher-order factor (i.e., health promotion self-care). The maximum likelihood method was employed to estimate the parameters for factor loading, factor variancey covariance, and uniqueness matrices of the CFA. The # 2 /df ratio, the comparative fit index (CFI), the non-normed fit index (NNFI), and the root mean square error of approximation (RMSEA) were used to examine the model fit. If the # 2 /df ratio was lower than 3, the CFI was greater than.90, the NNFI was greater than.90, and the RMSEA was lower than.08, then a good model fit was indicated (Arbuckle, 1997; Joreskog & Sorbom, 1996). Sample For a data set following multivariate normal distribution, 10 participants are required per parameter (Bentler & Chou, 1987). Because structural equation modeling was used to analyze construct validity, the sample size was also determined by a power analysis based on the RMSEA (MacCallum, Browne, & Sungwara, 1996). Given an alpha level of.05, the power was.80, the RMSEA was set at.05 and.08 for null and alternative models, respectively, and the minimum required sample size was 100. In summary, a sample size of 200 was considered acceptable for this study. Because we hoped to cross-validate the factor structure proposed by the EFA on the CFA for different samples, we recruited twice the size of the estimated sample numbers. Early intervention was more cost-effective in preventing and controlling diabetes complications (Li, Zhang, Barker, Chowdhury, & Zhang, 2010). The authors decided to recruit patients diagnosed with Type 2 diabetes for less than 10 years in this preliminary study. Other inclusion criteria included were the abilities to self-care and to communicate in Mandarin or Taiwanese. We recruited 489 patients with Type 2 diabetes from outpatient endocrine clinics in four hospitals in Southern Taiwan by convenience sampling. The full sample (N = 489) was divided randomly into two groups. There were 245 patients (Group 1) and 244 patients (Group 2) for EFA and CFA analysis, respectively. We also collected demographic data for the participants, including age, gender, marital status, educational level, duration of diabetes, whether the patient had complications, and diabetes treatment methods. Concurrent Validity Better self-care behavior leads to a better quality of life (Maddigan, Majumdar, & Johson, 2005). Empowerment was a helping process that enables people to choose to control or make decisions about their lives (Rodwell, 1996). Empowerment can promote the self-care behaviors of patients with diabetes (Skinner et al., 2006). With better selfcare behaviors, patients with diabetes had better HbA1c control (Sousa, Zauszniewski, Musil, Price Lea, & Davis, 2005). Correlations of the DHPSC with quality of life, perceptions of empowerment, and HbA1c level were examined to assess the concurrent validity of the DHPSC. The satisfaction subscale of the Chinese version of the Diabetes Quality of Life (DQoL) Scale (Huang et al., 2008) was used to assess how satisfied subjects were with their lives as it related to diabetes. The satisfaction subscale of the DQoL is a 15-item scale with a 5-point Likert scale ranging from 1 (very dissatisfied) to5(very satisfied). Higher scores indicate higher levels of quality of life. The Cronbach s alpha and testyretest reliability were.88 and.75, respectively, in a previous study conducted with Taiwanese patients with diabetes (Liu, Tai, Hung, Hsieh, & Wang, 2010). In this study, Cronbach s alpha for the satisfaction subscale of the DQoL was.87. Perceptions of empowerment were measured by a Chinese version of the Diabetes Empowerment Scale-Short Form (DES-SF; Lin, 2005). The DES-SF is an eight-item scale that measures empowerment by diabetes-related psychosocial self-efficacy; it is a 5-point Likert scale ranging from 1 (strongly disagree) to 5(strongly agree). Higher scores indicate higher perceptions of empowerment. A previous study indicated that the DES-SF had satisfactory reliability, with a Cronbach s alpha of.92 and a testyretest reliability of.88 (Liu et al., 2010). In this study, Cronbach s alpha for DES-SF was.90. Data from the satisfaction subscale of the DQoL and the DES-SF were collected when the DHPSC was administered to participants. The latest values of HbA1c levels as a measurement of long-term glycemic control were also collected from hospital medical records. Reliability Reliability was calculated for the overall DHPSC and for each subscale. To assess the internal consistency, we calculated Cronbach s alpha from the data of the 489 participants. A convenience sample of 25 diabetic patients 124

4 Diabetes Health Promotion Self-Care Scale VOL. 20, NO. 2, JUNE 2012 who satisfied the inclusion criteria were given the questionnaire over a 2-week interval. The intraclass correlation coefficient (Bolli et al., 2009) was calculated to assess the testyretest reliability. Ethical Considerations The institutional review board of the hospital affiliated with Kaohsiung Medical University approved this study. All participants were informed that there was no penalty for refusing to participate, and participants had the right to withdraw from the study at any time. Every participant signed a consent form after agreeing to participate. Results Participant Demographics The demographic data of the 489 participants are shown in Table 1. Their mean age was 58.1 (10.8) years. Approximately half (42.5 %) had been diagnosed with diabetes within the past 3 years. Most (88.8%) had received oral hypoglycemic agents. EFA The EFA was first conducted on Group 1 (n = 245). KaiserYMeyerYOlkin measure of sampling adequacy (.86) and a Bartlett s test of sphericity (# 2 = 7,621.77, df = 378, p G.001) consistently showed that the data set was suitable for factor analysis. On the basis of a scree test and Kaiser s rule, seven factors were extracted as expected. The eigenvalues of the retained factors ranged from 7.30 to The seven factors accounted for 69.31% of the total variance. Furthermore, each item loaded on its constructed corresponding factor (Table 2). All loadings were above.50, and half were above.80. The largest factor loading of items cross-loaded on other factors was.20. CFA To confirm the factor structure produced by EFA and whether seven factors were loaded on a second-order factor, a secondorder CFA was conducted on Group 2 (n = 244), according to the factor structure produced by EFA in Group 1. We found that all 28 items were significantly loaded on their constructed corresponding factors. However, the fit indices were # 2 = , df = 398, p G.001, # 2 /df =2.005, NNFI =.885, CFI =.894, RMSEA =.064 (90% CI: 0.058, 0.071), indicating that these were not good fits. Because the factor loading should be higher than.50 (Bagozzi & Yi, 1988), we deleted one item from the blood glucose selfmonitoring subscale and one item from the personal health responsibility subscale (Table 2). The factor loadings of two items of the foot care subscale were below.5; however, we TABLE 1. Demographic Data of Participants (N = 489) Variable n % Gender Male Female Marriage Married Unmarried Educational level Primary school and below Junior and senior high school Junior college and above Unknown Duration of diabetes 1Y3 years Y6 years Y10 years Complications of the disease Yes No Treatment Diet and exercise Oral hypoglycemic agent Insulin Oral hypoglycemic agent and Insulin Age (M T SD) 58.1 T 10.8 Range (years) 19Y87 still retained these items in the 26-item DHPSC because foot care is emphasized in patients with diabetes. A second-order CFA was performed again on the 26-item scale. The results showed that all 26 items were significantly loaded on their constructed corresponding factors. Each first-order factor was also significantly loaded on the second-order factor, termed as health promotion selfcare. The fit indices were acceptable, with # 2 = , df = 292, p G.001, # 2 /df = 1.90, NNFI =.915; CFI =.924; RMSEA =.061 (90% CI: 0.053, 0.068). These results indicated a satisfactory model fit. Proportions of the secondorder factor explained the seven first-order factors, ranging from 9% to 69%. The factor loadings of each item to a firstorder factor and a first-order factor to a second-order factor are shown in Table 2. Concurrent Validity Pearson s correlation coefficients for the 26-item DHPSC and satisfaction subscales of DQoL, perceptions of empowerment, and HbA1c levels are shown in Table 3. It can be observed that the satisfaction subscale of the DQoL was 125

5 The Journal of Nursing Research Ruey-Hsia Wang et al. TABLE 2. Factor Analysis of the 28-Item Diabetes Health Promotion Self-Care Scale (N = 489) Item Factor Loadings of EFA (n = 245) Factor Loadings and Percentages of Explained Variance in a Second-Order CFA (n = 244) Interpersonal relationships.46 a 21.30% Spend time with intimate friends Find ways to satisfy the needs of intimate relationships Keep in touch and communicate with people about whom you are concerned Maintain meaningful relationships with other people Express care, love, and warmth to others Praise the strength and virtues of other people Resolve conflicts with others through discussion and negotiation when having Personal health responsibility.52 a 27.04% Discuss personal health issues with healthcare providers (not only limited to diabetes) Report healthcare providers any unusual signs or symptoms (not only limited to diabetes) Discuss with healthcare providers about their suggestions when not clearly understood Consult healthcare providers about methods of practicing self-care (not only limited to diabetes) Actively visit the doctor when blood glucose is not well controlled Diet.63 a 39.69% Follow the rules of diet control for diabetes when having a meal out or in an unfamiliar setting Follow the rules of diet control when having meals Make sure to substitute foods within the same category Exercise.62 a 38.44% Keep exercising regularly Push self to exercise even though feeling lazy Find time to exercise even on busy schedule Foot care.83 a 68.89% Wear appropriate shoes and socks when out or traveling Perform self-examination of feet or apply foot cream according to suggestions of healthcare providers Blood glucose self-monitoring.41 a 16.81% Measure blood glucose or urine glucose according to the suggestions of healthcare providers Increase the frequency of testing blood glucose or urine glucose when feel uncomfortable Write down the results of each blood glucose or urine glucose testing Check the blood glucose or urine glucose as usual even when away from home Prevent hypoglycemia according to suggestions of healthcare providers when exercising Adherence to the recommended regimens.30 a 9.30% Take diabetic medicine on time even when out Take medications at the prescribed times Take prescribed doses of medication Note. Factor loadings of retained item are in bold. EFA = exploratory factor analysis; CFA = confirmatory factor analysis. a Factor loadings of the first-order factor to the second-order factor. 126

6 Diabetes Health Promotion Self-Care Scale VOL. 20, NO. 2, JUNE 2012 TABLE 3. Internal Consistency, Test Retest Reliability, and Correlations With Satisfaction Subscale of Quality of Life Scale, Perceptions of Empowerment Scale, and HbA1c of the 26-Item Diabetes Health Promotion Self-Care Scale (N = 489) Correlation Coefficient Subscale Mean SD Cronbach_s! ICC (n = 25) Satisfaction QoL Perceptions of Empowerment HbA1c Interpersonal relationships **.40** j.09 Personal health responsibility **.23** j.10* Diet **.40**.01 Exercise **.28** j.08 Foot care V.89.13**.30**.06 Blood glucose self-monitoring ** j.03 Adherence to the ** recommended regimens Total score **.49** j.08 Note. QoL = quality of life; ICC = intraclass correlation coefficient; V : not available because of two items. *p G.05. **p G.01. significantly correlated with the overall scale and each subscale of the 26-item DHPSC, except for the blood glucose self-monitoring subscale. The perceptions of empowerment scale significantly correlated with the overall scale and each subscale of the 26-item DHPSC, except for the adherence to recommended regimens subscale. The HbA1c levels only significantly correlated with the health responsibility subscale. Internal Consistency and Reliability Cronbach s alpha for the 26-item DHPSC overall scale and each subscale ranged from.78 to.94 (Table 3). The intraclass correlation coefficient overall scale and each subscale of 26-item DHPSC ranged from.76 to.95 (Table 3). Discussion The CFA results indicated that the DHPSC had satisfactory construct validity. Health promotion self-care behaviors are multidimensional. Assessing each component separately rather than to combine scores across components is necessary (Johnson, 1992). Because the second-order factor analysis model was supported, the seven subscales of the DHPSC can be examined individually or summed together into one scale (Noar, 2003) to represent a unified construct of health promotion self-care. The DHPSC can enable nurses to assess specific dimensions of health promotion self-care behaviors and provide specific interventions for patients with diabetes. Factor loadings of items for diet, exercise, blood glucose self-monitoring, and adherence to recommended regimens subscales were all larger than.5. This indicated that these items were indicators of their corresponding behaviors. However, factor loadings of foot care items were lower than.5. Although we retained these items in the DHPSC, we must revise the items in the foot care subscale in the future. Interpersonal relationships and personal health responsibility were important subscales of the DHPSC in this study. The results indicate that Taiwanese patients with Type 2 diabetes considered interpersonal relationships and personal health responsibility important in their health promotion self-care behavior. Stress management is also important for psychosocial health (Walker & Hill-Polerecky, 1996). Further studies are required to include stress management in the DHPSC. There are strong ceiling effects (a mean score of 4.59) and low explained variances (9.3%) for the adherence to recommended regimens subscale. This might be because participants were recruited from hospitals that may have good medication adherence. However, we found that behaviors involved in the adherence to recommended regimens in participants with junior and senior high school educational levels were significantly worse than those with a primary school educational level or below (4.61 T 0.69 vs T 0.73, df = 410, t =1.98,p =.048). This result is consistent with a previous study (Xu et al., 2010), indicating that, although the adherence to recommended regimens subscale had a strong ceiling effect, it still had contrasting group validity. The latest version of SDSCA also did not include any adherence to medication items because of the strong ceiling effects of medication taking (Toobert et al., 2000). The necessity for including the adherence to recommended regimens subscale in the DHPSC should be considered in the future. Patients can experience needle sticks when self-monitoring blood glucose, which may be frustrating. This may impact the DQoL. As expected, we found that the blood glucose self-monitoring subscale did not significantly correlate with the satisfaction subscale of the DQoL. However, the scores of other subscales and the overall scale of the DHPSC significantly correlated with the DQoL satisfaction subscale, supporting the construct validity of the DHPSC. 127

7 The Journal of Nursing Research Ruey-Hsia Wang et al. Perceptions of empowerment positively correlated with the DHPSC total score, supporting the concurrent validity of the DHPSC. Perceptions of empowerment were also significantly correlated with each subscale of the DHPSC, except for the adherence to recommended regimens subscale. The adherence to recommended regimens may indicate that patients have less control in making decisions. It is reasonable that no significant correlation between perceptions of empowerment and an adherence to recommended regimens was found. No statistically significant correlation was found between the overall DHPSC and the HbA1c level. Although better self-care is generally believed to result in better glycemic control, adherence to self-care does not always lead to good metabolic control (Toljamo & Hentinen, 2001). Therefore, it is reasonable that the overall DHPSC did not significantly correlate with the HbA1c level. However, personal health responsibility was the only subscale of the DHPSC that was significantly correlated with the HbA1c level. This result is consistent with a previous study (Wong et al., 2009). The strong internal consistency of the DHPSC and its subscales was supported by a Cronbach s alpha coefficient of.78 or above. TestYretest reliability scores of.76 or above also indicated good testyretest reliability (Rosner, 2006). The DHPSC had satisfactory reliability. There are some limitations to this study. First, we arbitrarily selected participants who had been diagnosed with diabetes for less than 10 years. Participants were also recruited from hospitals in southern Taiwan only. The selection bias may limit the generalization of the DHPSC. Further studies that recruit patients with longer durations of diabetes and from hospitals of different areas are required to further examine the validity and reliability of the DHPSC. Because health promotion behavior is influenced by culture (Tripp-Reimer, Choi, Kelley, & Enslein, 2001); the DHPSC must be tested for its utility in different countries. Further studies are also required to assess whether the DHPSC can predict the physical and psychosocial outcomes of patients with Type 2 diabetes. The DHPSC is a reliable and valid scale for measuring health-promoting behaviors of patients with Type 2 diabetes in Taiwan. Nurses can use the DHPSC to assess the health promotion self-care behaviors of patients with Type 2 diabetes. Nurses can provide adequate interventions to improve the behaviors of patients with Type 2 diabetes. References Aikens, J. E., Bingham, R., & Piette, J. D. (2005). Patientprovider communication and self-care behavior among Type 2 diabetes patients. The Diabetes Educator, 31(5), 681Y690. American Diabetes Association. (2010). Diabetes statistics. Retrieved from Anthony, S., Odgers, T., & Kelly, W. (2004). Health promotion and health education about diabetes mellitus. The Journal of the Royal Society for the Promotion of Health, 124(2), 70Y73. Arbuckle, J. (1997). Amos user s guide V3. 6. Chicago, IL: Small Waters Corporation. Bagozzi, R. P., & Yi, Y. (1988). On the evaluation of structural equation models. Journal of the Academy of Marketing Science, 16(1), 74Y94. Bentler, P. M., & Chou, C. P. (1987). Practical issues in structural modeling. Sociological Methods & Research, 16(1), 78Y117. Bolli, G. B., Songini, M., Trovati, M., Del Prato, S., Ghirlanda, G., Cordera, R., I Noacco, C. (2009). Lower fasting blood glucose, glucose variability and nocturnal hypoglycaemia with glargine vs. NPH basal insulin in subjects with Type 1 diabetes. Nutrition, Metabolism, and Cardiovascular Disease: NMCD, 19(8), 571Y579. Choe, M. A., Padilla, G. V., Chae, Y. R., & Kim, S. (2001). Quality of life for patients with diabetes in Korea, I: The meaning of health-related quality of life. International Journal of Nursing Studies, 38(6), 673Y682. Costello, A. B., & Osborne, J. W. (2005). Best practices in exploratory factor analysis: Four recommendations for getting the most from your analysis. Practical Assessment, Research & Evaluation, 10(7), 1Y9. Deakin, T. A., Cade, J. E., Williams, R., & Greenwood, D. C. (2006). Structured patient education: The diabetes X-PERT programme makes a difference. Diabetic Medicine, 23(9), 944Y954. Department of Health, Executive Yuan, Taiwan, ROC. (2009). 2005Y2008 nutritional and health survey in Taiwan. Retrieved from (Original work published in Chinese) Eigenmann, C. A., Colagiuri, R., Skinner, T. C., & Trevena, L. (2009). Are current psychometric tools suitable for measuring outcomes of diabetes education? Diabetic Medicine, 26(4), 425Y436. Funnell,M.M.,Brown,T.L.,Childs,B.P.,Haas,L.B.,Hosey,G.M., Jensen, B., I Weiss, M. A. (2011). National standards for diabetes self-management education. Diabetes Care, 34(Suppl.1), S97YS104. Furler, J., Walker, C., Blackberry, I., Dunning, T., Sulaiman, N., Dunbar, J., I Young, D. (2008). The emotional context of self-management in chronic illness: A qualitative study of the role of health professional support in the self-management of Type2diabetes. BMC Health Services Research, 8, 214Y222. doi: / Huang, I. C., Liu, J. H., Wu, A. W., Wu, M. Y., Leite, W., & Hwang, C. C. (2008). Evaluating the reliability, validity and minimally important difference of the Taiwanese version of the Diabetes Quality of Life (DQOL) measurement. Health and Quality of Life Outcomes, 6, 87Y99. Johnson, S. B. (1992). Methodological issues in diabetes research. Measuring adherence. Diabetes Care, 15(11), 1658Y1667. Joreskog, K. G., & Sorbom, D. (1996). LISREL 8: User s reference guide. Chicago, IL: Scientific Software International. Li, R., Zhang, P., Barker, L. E., Chowdhury, F. M., & Zhang, X. (2010). Cost-effectiveness of interventions to prevent and control diabetes mellitus: A systematic review. Diabetes Care, 33(8), 1872Y1894. Lin, C. C. (2005). Development and testing of the diabetes selfmanagement instrument (published doctoral dissertation). State University of Michigan, USA. Liu, M. Y., Tai, Y. K., Hung, W. W., Hsieh, M. C., & Wang, R. H. (2010). Relationships between emotional distress, empowerment perception and self-care behavior and quality of life 128

8 Diabetes Health Promotion Self-Care Scale VOL. 20, NO. 2, JUNE 2012 in patients with Type 2 diabetes. The Journal of Nursing, 57(2), 49Y60. (Original work published in Chinese) MacCallum, R. C., Browne, M. W., & Sugawara, H. M. (1996). Power analysis and determination of sample size for covariance structure modeling. Psychological Methods, 1(2), 130Y149. Maddigan, S. L., Majumdar, S. R., & Johnson, J. A. (2005). Understanding the complex associations between patientprovider relationships, self-care behaviours, and health-related quality of life in Type 2 diabetes: A structural equation modeling approach. Quality of Life Research, 14(6), 1489Y1500. Mendoza-Núñez, V. M., Martínez-Maldonado Mde, L., & Correa- Muñoz, E. (2009). Implementation of an active aging model in Mexico for prevention and control of chronic diseases in the elderly. BMC Geriatrics, 9, 40. Noar, S. M. (2003). The role of structural equation modeling in scale development. Structural Equation Modeling, 10, 622Y647. Orem, D. E. (1995). Nursing: Concepts of practice (5th ed.). St. Louis, MO: C.V. Mosby. Paul, G., Smith, S. M., Whitford, D., O Kelly, F., & O Dowd, T. (2007). Development of a complex intervention to test the effectiveness of peer support in Type 2 diabetes. BMC Health Services Research, 7, 136. doi: / Pender, N. J. (1996). Health promotion in nursing practice. Norwalk, CT: Appleton & Lange. Rodwell, C. M. (1996). An analysis of the concept of empowerment. Journal of Advanced Nursing, 23(2), 305Y313. Rosner, B. (2006). Fundamentals of biostatistics (6th ed.). Belmont, CA: Thomson-Brooks/Cole. Simmons, S. J. (1990). The health-promoting self-care system model: Directions for nursing research and practice. Journal of Advanced Nursing, 15(10), 1162Y1166. Skinner, T. C., Carey, M. E., Cradock, S., Daly, H., Davies, M. J., Doherty, Y., I Oliver, L. (2006). Diabetes education and selfmanagement for ongoing and newly diagnosed (DESMOND): Process modelling of pilot study. Patient Education and Counseling, 64(1Y3), 369Y377. Sousa, V. D., Zauszniewski, J. A., Musil, C. M., Price Lea, P. J., & Davis, S. A. (2005). Relationships among self-care agency, self-efficacy, self-care, and glycemic control. Research and Theory for Nursing Practice, 19(3), 217Y230. Toljamo, M., & Hentinen, M. (2001). Adherence to self-care and glycaemic control among people with insulin-dependent diabetes mellitus. Journal of Advanced Nursing, 34(6), 780Y786. Toobert, D. J., Hampson, S. E., & Glasgow, R. E. (2000). The summary of diabetes self-care activities measure: Results from 7 studies and a revised scale. Diabetes Care, 23(7), 943Y950. Tripp-Reimer, T., Choi, E., Kelley, L. S., & Enslein, J. C. (2001). Cultural barriers to care: inverting the problem. Diabetes Spectrum, 14(1), 13Y22. Walker, S., & Hill-Polerecky, D. (1996). Psychometric evaluation of the health-promoting lifestyle profile II. Retrieved from Walker, S. N., Sechrist, K. R., & Pender, N. J. (1995). The form of permission to use health-promoting lifestyle profile. Omaha, NE: University of Nebraska Medical Center. Waltz, C. F., Strickland, O., & Lenz, E. R. (2005). Measurement in nursing and health research. New York, NY: Springer Publishing Company. Wang, J. S., Wang, R. H., & Lin, C. C. (1998). Self-care behaviors, self-efficacy, and social support effect on the glycemic control of patients newly diagnosed with noninsulin-dependent diabetes mellitus. The Kaohsiung Journal of Medical Sciences, 14(12), 807Y815. (Original work published in Chinese) Wang,R.H.,&Hsu,H.Y.(1997).Healthpromotionbehaviorsand related factors among the elderly in the San-Min district of Kaohsiung City. Nursing Research (Taiwan), 5(4), 321Y330. (Original work published in Chinese) Weinger, K., Butler, H. A., Welch, G. W., & La Greca, A. M. (2005). Measuring diabetes self-care: A psychometric analysis of the Self-Care Inventory-Revised with adults. Diabetes Care, 28(6), 1346Y1352. Wong, A. K., Stewart, A. G., & Furler, J. S. (2009). Development and validation of the diabetes management orientation scale (DMOS): Assessing culturally related approaches to diabetes self-management. Diabetes Research and Clinical Practice, 86(1), 24Y30. World Health Organization. (2009). Diabetes statistics. Retrieved from Xu, Y., Pan, W., & Liu, H. (2010). Self-management practices of Chinese Americans with Type 2 diabetes. Nursing & Health Sciences, 12(2), 228Y

9 The Journal of Nursing Research VOL. 20, NO. 2, J UNE * Cronbach s α intraclass correlation coefficient * wrhsia@kmu.edu.tw 130