The Self-Care in Chronic Obstructive Pulmonary Disease Inventory: Development and Psychometric Evaluation

Abstract

This study developed two instruments, the Self-Care in Chronic Obstructive Pulmonary Disease (COPD) Inventory (SC-COPDI) and the COPD-Self-Care Self-Efficacy Scale (SCES), and tested their psychometric properties on a convenience sample of 498 patients from Northern, Central, and Southern Italy. First, the domains and the items of the SC-SCOPDI were generated based on the middle-range theory of self-care of chronic illness, comprising the dimensions of self-care maintenance, self-care monitoring, and self-care management, and the SCES-COPD was developed accordingly. Second, we assessed the content validity of each scale. Third, we conducted a multicenter cross-sectional study to test their structural validity, convergent and discriminative validity, internal consistency, and test–retest reliability. The theoretical dimensions of the two instruments were confirmed through confirmatory factor analysis. Convergent validity was demonstrated by the correlation among the three self-care scales and the Self-Efficacy Scale, and discriminative validity by higher self-care scale scores in individuals with greater COPD severity and poorer health status. The global reliability index ranged from .78 to .92 for all scales. The intraclass correlation coefficients were higher than .70. Further studies are needed to confirm the psychometric properties of the two instruments in different COPD populations and countries to extend their use in clinical practice.

Keywords

chronic obstructive pulmonary disease instrument development psychometrics self-care self-efficacy

Chronic obstructive pulmonary disease (COPD) is a nonreversible lung disease with a prevalence of 11.7% in people aged 30 and older globally, and it is expected to increase due to exposure to COPD risk factors and the aging of the population (Adeloye et al., 2015). COPD is the third leading cause of death worldwide (World Health Organization, 2018). In Europe, the estimated total direct costs from COPD are about 38.6 billion euros, 6% of the total health-care expenditure; in the United States, the estimated direct costs are 32 billion dollars, whereas the indirect costs are 20.4 billion (Global Initiative for Chronic Obstructive Lung Disease [GOLD], 2018). The economic burden of COPD is due primarily to disease exacerbations that lead to emergency department visits and hospital admissions (GOLD, 2018).

People with COPD experience disabling somatic symptoms including dyspnea, coughing, sputum, and fatigue caused by progressive airflow limitation associated with a gradual reduction of oxygen supply. The physiological symptoms are accompanied by psychological manifestations, such as depression and anxiety, which deeply affect the quality of life of people with COPD (Yohannes & Alexopoulos, 2014). Together, these physical and psychological effects impair patients’ social lives (Giacomini, DeJean, Simeonov, & Smith, 2012). COPD requires a complex therapeutic regimen and daily self-care behaviors directed at preventing, controlling, and managing the physical, emotional, and social consequences of the disease (Clari, Matarese, Ivziku, & De Marinis, 2017). Self-care behaviors of people with COPD vary over time in response to the person’s physical and psychological condition, disease-related cognition, and social support received, and they are intensified as the disease progresses and the disease burden increases (Pinnock, Steed, & Jordan, 2016). Therefore, people with COPD must continually adapt their self-care behaviors to the new circumstances they face. The most common self-care behaviors are aimed at limiting the physical impact of COPD (Clari et al., 2017); they include physical exercise, breathing techniques, respiratory muscle training, and dyspnea management skills. Bronchial hygiene and secretion clearance are also used to reduce breathlessness (Stoilkova, Janssen, & Wouters, 2013) as well as energy conservation techniques to maintain disease stability (Velloso & Jardim, 2006). To cope with psychological issues, relaxation and stress management techniques are advocated (Stoilkova et al., 2013). Other self-care behaviors usually recommended by health-care professionals include the effective use of inhalers, adherence to medications, and prevention of infections through vaccinations and hygienic habits (Jolly et al., 2016). Evidence has shown that effective self-care can improve health-related quality of life and reduce respiratory-related hospital admissions (Lenferink et al., 2017; Zwerink et al., 2014) and the utilization of health-care services (Murphy et al., 2017).

Further research into the effectiveness of self-care behaviors in COPD is hampered by the paucity of reliable and valid instruments. A review identified three instruments developed to assess self-care in COPD (Clari, Matarese, Alvaro, Piredda, & De Marinis, 2016): the Alberto COPD Behavior Inventory, the Self-Care Behavior Scale for COPD patients, and the COPD Self-Management Scale (CSMS). The Alberto COPD Behavior Inventory consisted of 24 items and was not based on theory; only content validity and internal consistency were tested in two studies conducted in the United States (Alberto, 1993) and Turkey (Kaşıkçı & Alberto, 2007). The Self-Care Behavior Scale for COPD patients is a 32-item scale, grounded in Orem’s grand theory of self-care, comprising two domains: behaviors for meeting universal self-care requisites and health deviation self-care requisites. It was tested for content validity and internal consistency in one study (Xiaolian et al., 2002). The CSMS is a 51-item instrument comprising five dimensions: symptom management, daily life management, emotion management, information management, and self-efficacy (Zhang et al., 2013). Again, a theory was not used to guide the development of the CSMS. The content and structural validity, hypothesis testing, internal consistency, and test–retest reliability of the CSMS were tested in a study conducted on a Chinese sample. The review concluded that evidence on the quality of the measurement properties of these three instruments was limited or unknown, and none of them was based on a self-care theory of chronic disease (Clari et al., 2016).

For these reasons, a new instrument evaluating self-care of COPD, the Self-Care in Chronic Obstructive Pulmonary Disease Inventory (SC-COPDI), was developed and tested and its measurement properties are presented in this article. The SC-COPDI was based on the middle-range theory of self-care of chronic illness (Riegel, Jaarsma, & Strömberg, 2012). In that theory, self-care is defined as a process of maintaining health through health-promoting and managing illness practices. Self-care is performed in both healthy and ill states, with specific behaviors captured in three core concepts: self-care maintenance, self-care monitoring, and self-care management. Self-care maintenance focuses on the need to keep oneself physically and emotionally healthy, self-care monitoring is done to detect changes in signs and symptoms, and self-care management addresses the behaviors undertaken when symptoms of a chronic illness occur. These three concepts have been shown to be relevant in several different conditions, and the theory has already been used to support the development of self-care instruments in diabetes (Ausili et al., 2017), chronic diseases (Riegel, Barbaranelli, et al., 2018), and inflammatory bowel disease (Wickman et al., 2019).

We also developed and tested an instrument to measure self-efficacy of self-care, the Self-Care Self-Efficacy Scale in COPD (SCES-COPD), to evaluate the confidence of people with COPD in their ability to perform disease-related self-care behaviors despite possible barriers (Riegel et al., 2012). Although not a component of self-care, self-efficacy has been repeatedly shown to be an essential predictor of self-care behaviors (Caruso et al., 2018; Vellone et al., 2015; Vellone, Pancani, Greco, Steca, & Riegel, 2016). To our knowledge, no specific instrument exists to measure self-care self-efficacy in COPD. Two instruments were identified in the literature to assess self-efficacy in COPD (Clari et al., 2016); however, one of them, the COPD Self-Efficacy Scale, designed by Wigal, Greer, and Kotses (1991), assesses confidence in the management of breathing difficulties when performing specific activities, whereas the other, the Exercise Self-Regulatory Efficacy Scale, developed by Davis, Figueredo, Fahy, and Rawiworrakul (2007), evaluates confidence in the ability to exercise regularly. Therefore, we conducted a study aimed at developing and testing the psychometric properties of two new instruments measuring self-care behaviors and self-care self-efficacy in people with COPD.

Method

A three-phase process was used to develop the SC-COPDI and the SCES-COPD and to test their measurement properties. First, we identified the domains and generated items based on the theory. Second, we assessed the content validity of the identified items. Third, we conducted a multicenter cross-sectional study of people with COPD to test the psychometric properties of the instruments.

Instrument Development

Initially, we analyzed the quantitative and qualitative evidence on COPD self-care together with the contents of other COPD self-care instruments to identify self-care behaviors performed in the various stages of the disease. The self-care behaviors derived from the literature were grouped into three categories to reflect the constructs of the middle-range theory of self-care of chronic illness (Riegel et al., 2012): self-care maintenance, self-care monitoring, and self-care management behaviors.

The COPD self-care maintenance behaviors identified in the literature (Online Appendix: Supplemental Table 1-A) included those behaviors aiming at (1) preventing lung inflammation, such as avoiding people affected by flu or environments with lung irritants or tobacco smoke; (2) promoting effective breathing, such as using respiratory techniques and resting during physical activities; (3) maintaining physical and social activities, including doing regular physical exercise by walking or exercising the arms, and pursuing social activities; and (4) adhering to prescribed therapies including medications, medical visits, and vaccinations.

The self-care monitoring behaviors discussed in the literature (Online Appendix: Supplemental Table 1-B) were grouped in behaviors related to (1) monitoring the early manifestations of respiratory infection, such as changes in amount or color of sputum and increased breathlessness and (2) checking for extra-respiratory symptoms, including sleep disturbances caused by breathlessness, activity intolerance, and side effects of inhaled medications.

The self-care management behaviors identified in the literature (Online Appendix: Supplemental Table 1-C) included (1) modifying therapy in response to exacerbation, as recommended by a health-care professional, or initiating contact with the health-care professional; (2) consulting the health-care professional when symptoms occur or worsen; and (3) reducing exertions during daily activities in the presence of dyspnea, for example, sitting while showering or doing chores. In total, 46 items were developed based on the identified behaviors: 21 for the Self-Care Maintenance, 10 for the Self-Care Monitoring, and 15 for the Self-Care Management Scales.

The development process of the SCES-COPD was similar to that described for the SC-COPDI. Nine items were selected after the item generation process representing two aspects considered relevant in the COPD self-care literature (Online Appendix: Supplemental Table 1-D): (1) confidence in adhering to the treatment regimen, such as following therapeutic advice, checking for symptoms, and taking medications properly; and (2) confidence in managing symptoms: recognizing the symptoms of an exacerbation, relieving symptoms, and evaluating the effectiveness of actions.

Content Validity

Content validity of the items developed was assessed to measure the extent to which the items reflected the construct to be measured. Items were assessed for relevance, comprehensibility, and comprehensiveness (Terwee et al., 2018) by a multidisciplinary team of experts in COPD (n = 20), including pneumologists (n = 8), nurses (n = 8), physiotherapists (n = 4), and individuals with COPD (n = 20). The COPD subjects had different stages of disease (from mild to very severe), gender (male = 10), and ages (from 55 to 85 years). We used the method suggested by Lawshe (1975) as it permits consideration of the level of agreement among experts occurring by chance. The relevance of each item was evaluated by asking the experts to indicate whether they considered it as essential, useful but not essential, or unnecessary. The item content validity ratio (I-CVR) was computed using a formula (N_e – N/2/N/2), in which N_e is the number of experts indicating the item is essential and N is the total number of experts. The I-CVR can range from 1, meaning that all judges assessed the item as essential, to −1, indicating that all the judges scored the item as nonessential. The critical value of .50 is suggested when 20 judges are involved (Ayre & Scally, 2014). The I-CVR for the 46 items of SC-COPDI ranged from .90 to −.50. Twelve items did not reach the value of .50 in either expert or patient groups and were eliminated. The new version of the instrument had 34 items: 15 items for the Self-Care Maintenance Scale, 8 for the Self-Care Monitoring Scale, and 11 for the Self-Care Management Scale. A Scale Content Validity Index (S-CVI) was calculated for each scale as the mean of I-CVR scores of the items retained (Lawshe, 1975). The S-CVI for the Self-Care Management Scale was .70, for the Self-Care Monitoring Scale .50, and for the Self-Care Management Scale was .52.

Finally, we asked the two groups of experts (patients and health-care professionals) to assess each item for comprehensibility (yes/no) and to suggest other relevant self-care behaviors that were not listed (comprehensiveness). Two researchers reviewed the list of self-care behaviors suggested, and no new items were added.

Item content was refined and formatted with a 5-point summated rating scale indicating how frequently the behaviors were performed (1 = never to 5 = always). Also, to accommodate the different clinical manifestations of COPD and different prescriptions, a “not applicable” option was added for specific items. For example, inhalers may not be prescribed to all COPD patients so the item asking about side effects of inhalers would not always be answered.

To mirror the instrument developed to assess self-care by individuals with chronic illness, which was also based on the middle-range theory of self-care of chronic illness (Riegel, Barbaranelli, et al., 2018), a question was added to the Self-Care Self-Monitoring Scale to determine whether respondents experienced symptoms and their rapidity in recognizing them as symptoms of COPD (0 = not recognized to 5 = quickly recognized). Asymptomatic COPD patients skip the self-Care Management Scale because they cannot respond to questions about symptom management if they do not have symptoms.

The I-CVR for the SCES-COPD ranged from .90 to −.50 and 7 items of the 9 reached the critical value of .50. The S-CVI was .75. Respondents were asked to indicate their level of confidence in their ability to perform self-care on a 5-point summated rating scale (not confident = 1 to extremely confident = 5). The two instruments were pilot tested on five COPD patients. Minor revisions were made to facilitate understanding of the instructions and the answer options.

Scoring for each of the three self-care scales and the SCES-COPD is calculated by summing the items’ scores and standardizing the sum to 100 to make the scores comparable among the four scales. Doing so facilitates comparisons between scales with a different number of items. For each scale, higher scores indicate better self-care.

Psychometric Testing

After item generation and content validity testing, the psychometric properties of the SC-COPDI and SCES-COPD were tested. In particular, we evaluated the structural validity, convergent and discriminative validity, internal consistency, and test–retest reliability of the four scales.

Sample

A convenience sample of outpatients and inpatients with COPD was recruited from different health-care settings (hospital units, outpatient clinics, primary care practices, and rehabilitation centers) in Northern, Central, and Southern Italy. To be included, individuals had to be 18 years of age or older and diagnosed with COPD at least 1 year earlier. COPD patients with diagnosis of dementia and those unable to understand or read Italian were excluded. The sample size recommended to test dimensionality, and internal consistency is generally of 10 participants for each item of the scale (De Vet, Terwee, Mokkink, & Knol, 2011). As each scale (Self-Care Maintenance, Self-Care Monitoring, Self-Care Management, and Self-Care Self-Efficacy) was validated separately, a minimum of 150 participants was sought. However, the sample size was increased to 500 to represent different social conditions, disease stages, levels of education, and ages.

Data collection

Trained research assistants identified eligible COPD patients in the health-care settings and asked their informed consent to participate in the study. After signing the consent form, participants filled in the instruments autonomously or were helped by the research assistants if they had difficulty with vision or writing.

Measures

The severity of the airflow limitation caused by the disease was assessed using the GOLD (2018) grading system that classifies the disease in four grades based on spirometric values: GOLD 1 = mild, GOLD 2 = moderate, GOLD 3 = severe, and GOLD 4 = very severe. The spirometric values were gathered from the participant’s medical records, when available. Further clinical data, including the length of time with COPD and number of hospitalizations for COPD in the previous year, were self-reported by subjects. The impact of COPD on patient life was evaluated using the modified Medical Research Council Dyspnea Questionnaire (mMRC) and the COPD Assessment Test (CAT™). The mMRC (Mahler & Wells, 1988) is a self-report questionnaire measuring the impact of dyspnea on patients’ daily activities on a scale of 0–4, where 0 represents no limitation and 4 very severe limitation in daily activities. The CAT™ (Jones et al., 2009) is an 8-item unidimensional instrument that assesses the disease impact on health status as determined by symptom burden and activity limitation; it considers coughing, phlegm, chest tightness, breathlessness going up hills/stairs, activity limitations at home, confidence in leaving home, sleep, and energy. The scores range from 0 to 40, where values <10 denote low impact of the disease on the patient’s life, 11–20 a medium impact, 21–30 a high impact, and >30 a very severe impact (Jones, Tabberer, & Chen, 2011). It was administered to a subset of 166 COPD subjects. The Italian version of the instrument downloaded from the website (www.catestonline.org) was used. Sociodemographic data collected included age, gender, education level, living conditions (alone or with family), and employment status (retired or working).

Data Analysis

Descriptive statistics were computed for all variables (frequency, percentage, mean, standard deviation [SD], kurtosis, and skewness coefficients where appropriate). Missing data were assessed at the variable and item level. The sociodemographic and clinical missing data were deleted pairwise, and the missing items on the self-care scales were handled with maximum likelihood estimation.

As the instrument dimensionality must be measured before choosing a method for estimating reliability, we started the analysis by testing the scale dimensionality and subsequently reliability (Barbaranelli, Lee, Vellone, & Riegel, 2015). We used a confirmatory factor analysis (CFA) approach to test the structural validity, as the instruments were theory based. Factor loadings >|.30| were considered adequate (Comrey & Lee, 1992; Tabachnick & Fidell, 2007). As items were not normally distributed, we used robust maximum likelihood method for parameter estimation. Model fit was examined using the following fit indices: χ² statistics, comparative fit index (CFI), Tucker and Lewis index (TLI), root mean square error of approximation (RMSEA), and standardized root mean square residual (SRMR; Byrne, 2006; Meade, Johnson, & Braddy, 2008; Vandenberg & Lance, 2000). The model goodness of fit was assessed using the following criteria: CFI and TLI values of .90–.95 indicate acceptable fit, and values >.95 indicate good model fit (Hu & Bentler, 1999); RMSEA values ≤.05 indicate a well-fitting model, .05–.08 a moderate fit (Browne & Cudek, 1992); values of SRMR ≤.08 indicate good fit (Hu & Bentler, 1999). The χ² statistics, although computed, were not used in interpreting model fit because they are influenced by sample size. Each of the three scales of the SC-COPDI and the SCES-COPD were tested individually.

The sample was randomly split into two subsamples using SPSS Version 24 (IBM, Armonk, NY). One subsample was used to develop the model (e.g., eliminating items with inadequate loadings) and the second subsample to validate the trimmed solution obtained from the first subsample (Anderson & Gerbing, 1988; Hoyle & Panter, 1995). The final replicated trimmed model was tested using the total sample to obtain parameter estimates that were more stable in dimensionality and reliability testing.

The reliability was estimated with Cronbach’s α coefficient, factor score determinacy (Muthén & Muthén, 1998–2017), and with composite reliability (Fornell & Larcker, 1981) or the omega coefficient (McDonald, 1999) to consider the multidimensionality of the scales (Barbaranelli, Lee, Vellone, & Riegel, 2014). Values ≥.70 were considered as adequate (Bagozzi & Yi, 2012). Item discrimination was estimated with corrected item-total correlation coefficient (ITC; Crocker & Algina, 1986), considering values ≥.20 as adequate (Kline, 1986).

To evaluate the stability of the SC-COPDI, test–retest reliability was assessed for each scale, readministering the scales within 2 weeks to a subset of 50 patients with stable COPD. The intraclass correlation coefficients (ICCs) through two-way random effects model were calculated for the scores of each scale. A sample of 50 subjects was required to reach an ICC = .80 (95% confidence interval [CI] ± 0.1) with two repeated measures (De Vet et al., 2011). Also, weighted κs with quadratic weights were computed to analyze individual items’ level of agreement. An ICC ≥.70 (Nunnally & Bernstein, 1994) and a weighted κ ≥ .40 (Landis & Koch, 1977) were considered acceptable.

Since a valid and reliable COPD self-care instrument is not available in the literature, we could not assess criterion validity. We verified the construct validity by posing several hypotheses. First, we hypothesized that people with more severe COPD would score higher in the self-care scales since the increasing airflow obstruction leads to worsening symptoms, which motivate greater self-care behaviors. Second, we hypothesized that people with a more severe impact of dyspnea on daily activities and poorer health status would score higher in the self-care scales based on the rationale that more impact motivates more self-care (Riegel, Jaarsma, Lee, & Strömberg, 2018). Therefore, we expected higher scores (>10 points) on the self-care scales in patients with higher GOLD grades, mMRC levels, and CAT™ classes compared to patients with lower COPD grades, dyspnea levels, and CAT™ classes. Third, we hypothesized moderate correlations among the Self-Care Maintenance, Self-Care Monitoring, and Self-Care Management Scales, as they represent interrelated albeit different aspects of the self-care construct as described in the middle-range theory of self-care of chronic illness (Riegel et al., 2012) and demonstrated in other self-care instruments derived from the theory (Ausili et al., 2017; Riegel, Barbaranelli, et al., 2018). Finally, we predicted a moderate correlation between the three self-care scales and the SCES-COPD, as self-efficacy has been found to be a predictor of self-care (Caruso et al., 2018; Vellone et al., 2016). To measure the correlation between self-care scales, we used the Pearson’s product-moment correlation coefficients. Correlations of .10–.29 were considered as small, .30–.49 as moderate, and ≥.50 as strong (Cohen, 1988). To compare mean scores, across groups of COPD patients who differed in the identified clinical measures one-way analysis of variance using post hoc Tukey’s honestly significant difference test was conducted.

The level of significance was set at ≤.05. Statistical analyses were performed using SPSS Version 24 except for CFA, which was performed using Mplus Version 8.1 (Muthén & Muthén, 1998–2017).

Ethical Considerations

The study was carried out in accordance with the Declaration of Helsinki. Approval was obtained from a University Ethics Committee. The directors of health-care institutions where patients were recruited gave their authorization to conduct the study. All participants were fully informed about the study by research assistants, and all provided written informed consent. Patient anonymity was ensured by numerically coding each participant’s survey.

Results

Of 656 eligible people with COPD identified in the clinical settings, 500 (76.21%) were recruited and 156 (23.79%) declined participation. The main reasons for refusing were lack of time, no interest in the research, and feeling too sick. As two subjects left the SC-COPDI completely blank, they were excluded, leaving a final sample of 498 participants from eight regions in Italy. The Self-Care Maintenance Scale was completed by 498 patients, the Self-Care Monitoring Scale by 486 (12 questionnaires were not completed due to misunderstanding of the scale instructions), the Self-Care Management Scale was completed by 290 symptomatic patients, and the Self-Care Self-Efficacy Scale was completed by 398 patients, as this scale was added in a later phase of instrument testing. The self-care scales had no missing values.

The majority of the participants were enrolled in pneumological clinics (50.20%) and hospital units (32.33%). Most respondents were elderly (74.12, SD = 9.23), retired (82.92%), living with family members (85.10%), and poorly educated (70.35%). The mean duration of COPD was 9.39 (SD = 8.02) years, ranging from 1 to 40 years, and 70.32% of participants were affected by moderate and severe COPD grades (Table 1).

Table 1.

Sociodemographic and Clinical Characteristics of the Study Participants.

Sample Characteristics	n (%)	Mean ± SD (Range)
Sociodemographic characteristics
Setting
Pneumological clinic	250 (50.20)
Hospital unit	161 (32.33)
Rehabilitation center	33 (6.62)
Family physician office	54 (10.85)
Sex
Male	278 (55.80)
Female	220 (44.20)
Education
No education	15 (3.01)
Elementary	186 (37.30)
Middle school	150 (30.04)
High school	99 (19.90)
University	48 (9.60)
Living status
Alone	74 (14.90)
Family	424 (85.10)
Employment status
Retired	413 (82.92)
Employed	85 (17.08)
Age		74.12 ± 9.23 (45–100)
Clinical characteristics
GOLD grade^a
1—Mild	56 (19.78)
2—Moderate	101 (35.69)
3—Severe	98 (34.63)
4—Very severe	28 (9.90)
Not reported	215
CAT class (n = 166)
0–10—Low impact	15 (9.04)
11–20—Moderate impact	54 (32.53)
21–30—High impact	71 (42.77)
31–40—Very High impact	26 (15.66)
mMRC grade
0—No limitation	50 (10.79)
1—Mild limitation	138 (29.81)
2—Moderate limitation	78 (16.82)
3—Severe limitation	104 (22.48)
4—Very severe limitation	93 (20.10)
Not reported	35
COPD years		9.28 ± 7.90 (1–40)

Note. n = 498. SD = standard deviation; COPD = chronic obstructive pulmonary disease; mMRC = modified Medical Research Council dyspnea questionnaire; CAT = COPD Assessment Test; GOLD = Global Initiative for Chronic Obstructive Lung Disease; FEV₁ = forced expiratory volume in one second; FVC = forced vital capacity.

^aGOLD 1 = FEV₁ ≥ 80%; GOLD 2 = 50% ≤ FEV₁ < 80%; GOLD 3 = 30% ≤ FEV₁ < 50%; GOLD 4 = FEV₁ < 30%.

Times for completion of the SC-COPDI and SCES-COPD ranged between 10 and 15 min for symptomatic people with COPD and between 7 and 10 min for asymptomatic ones. No difficulties in understanding the item meanings were reported by research assistants. In some cases, participants needed to be helped in reading the questionnaires due to their vision impairments or lack of glasses, with assistants reading the scale items for them.

Self-Care Maintenance Scale

Dimensionality

The scale was hypothesized to comprise four dimensions. We anticipated that the items related to avoiding people with cold/flu, avoiding tobacco smoke, avoiding polluted environment, and protecting the mouth/nose in cold air would cluster together in a factor reflecting disease prevention behaviors. Likewise, we expected that the items concerned with using deep breathing, pausing during daily activities, and using breathing techniques would cluster in a factor assessing improving breathing behaviors. In addition, items related to doing physical activities, exercising the arms, engaging in social activities, doing relaxing activities, and sleeping a sufficient amount would cluster in a factor describing physical activity promotion behaviors. Finally, we expected that items addressing annual flu vaccination, taking medicines, and regularly visiting a health-care provider would cluster together, reflecting the treatment adherence behaviors.

The CFA performed on the first subsample tested a four-factor solution based on the postulated dimensions. Two items, doing relaxing activities and sleeping sufficiently, had low factor loadings (≤.30) and were removed from further testing. Without these items, the model goodness-of-fit indices were good: χ²(59, N = 249) = 64.17, p = .30, CFI = .99, TLI = .99, RMSEA = .02 (90% CI [.00, .04]), p = .98, SRMR = .05. This four-factor model was replicated on the second subsample, obtaining the following fit indices: χ²(59, N = 249) = 87.52, p < .01, CFI = .94, TLI = .92, RMSEA = .04 (90% CI [.02, .06]), p = .68, SRMR = .05. The items addressing pausing during activities and doing physical activities had significantly correlated residuals. The correlation was theoretically justified since physical activities entail periods of rest to regulate breathing and prevent breathlessness. For this reason, the model was respecified, freeing this parameter. This new parameter was replicated in the second subsample. This final trimmed model was retested on the total sample of 498 subjects obtaining good fit indices (Table 2). All factor loadings were significant with values ranging from .37 to .82 (Table 3).

Table 2.

Fit Indices From Confirmatory Factor Analysis for the Self-Care Scales (Robust Maximum Likelihood Estimator).

Scale	χ²	df	p(χ²)	CFI	TLI	SRMR	RMSEA	RMSEA, 90% CI	RMSEA, p < .05
Self-care maintenance (n = 498)	102.4	58	<.001	.95	.94	.04	.04	[.03, .05]	.92
Self-care monitoring (n = 486)	24.38	12	.02	.99	.98	.03	.05	[.02, .07]	.56
Self-care management (n = 290)	69.41	31	.001	.96	.93	.05	.06	[.04, .09]	.10
Self-care self-efficacy (n = 389)	22.64	11	.02	.99	.98	.02	.05	[.02, .08]	.41

Note. χ² = Chi-square; df = degree of freedom; p = probability; CFI = comparative fit index; TLI = Tucker and Lewis index; SRMR = standardized root mean square residual; RMSEA = root mean square error of approximation; CI = confidence interval.

Table 3.

Factor Loadings and Corrected Item-Total Correlations for the Self-Care Scales.

Item	Loading		ITC
Self-Care Maintenance Scale
Factor 1: Disease prevention behaviors
(1) Avoid people with cold or flu	.65		.55
(2) Move away from places where someone is smoking	.73		.57
(3) Avoid contact with sprays, paints, solvents, and dust	.72		.58
(12) Protect mouth/nose in cold air	.45		.38
Factor 2: Improving breathing behaviors
(4) Keep lungs free by coughing or with deep breathing	.62		.45
(5) Pause during daily activities	.42		.34
(6) Use abdominal or blurred lips breathing	.72		.48
Factor 3: Physical activities promotion behaviors
(7) Exercise regularly (walking, cycling, swimming, etc.)	.56		.45
(8) Do arm exercises	.82		.46
(9) Engage in social activities	.37		.31
Factor 4: Treatment adherence behaviors
(10) Get a flu vaccine every year	.45		.26
(11) Take the medicines as prescribed	.47		.30
(13) Make regular visits to health-care provider for checkups	.53		.31
Self-Care Monitoring Scale	Specific Factors	General Factor	ITC
Factor 1: Respiratory symptom monitoring behaviors
(1) Monitor the increase of phlegm	.70	.68	.72
(2) Monitor the color change of phlegm	.63	.65	.73
(3) Monitor the increase of coughing	.26	.77	.71
(4) Monitor for an increase in breathlessness or whistles	–.21	.85	.73
Factor 2: Extra-respiratory symptom monitoring behaviors
(5) Monitor waking up during the night because of trouble breathing	.68	.55	.73
(6) Check struggle to fall asleep due to trouble breathing	.68	.47	.68
(7) Monitor getting tired when doing something	.28	.59	.73
(8) Check for side effects of inhaled medications	.49	.40	.57
Self-Care Management Scale
Factor 1: Autonomous behaviors
(1) Talk to health-care provider for problems with prescriptions	.64		.56
(2) Go to health-care provider for health problems	.89		.59
(8) Modify therapy as recommended when symptoms worsen	.41		.43
Factor 2: Consulting behaviors
(3) Speak to health-care provider if breathlessness increases	.92		.78
(4) Speak to health-care provider if coughing increases	.92		.84
(5) Speak to health-care provider if sputum increases	.71		.83
(6) Speak to health-care provider if sputum changes color	.67		.81
(7) Speak to health-care provider for side effects of inhaled medications	.37		.35
Factor 3: Problem-solving behaviors
(9) Sit doing housework when breathlessness occurs	.81		.49
(10) Sit when showering in case of breathlessness	.64		.49
Self-care self-efficacy
Factor 1: Symptom management
(1) Prevent the onset of symptoms	.49		.44
(5) Recognize the symptoms of an exacerbation	.75		.64
(6) Do something to relieve symptoms	.82		.78
(7) Assess the effectiveness of behaviors performed to relieve symptoms	.79		.74
Factor 2: Treatment adherence	.67		.73
(2) Follow the therapeutic advice	.89		.62
(3) Continue to check symptoms (4) Take medicines following the instructions given	.61		.69

Note. ITC = corrected item-total correlation coefficient.

Correlations among factors ranged from .01 to .62, with an average correlation of .33, indicating a relevant association among the different factors of self-care maintenance, which can be considered as interrelated albeit different aspects of a more general “self-care maintenance” dimension. The correlation between residuals of the items related to pausing during activities and doing regular physical activities was −.20, representing the inverse relation between rest and activity (Online Appendix: Supplemental Figure 1).

Reliability

The Cronbach’s α coefficient for the entire scale was .73. However, α assumes that the items satisfy a unidimensional structure. As four dimensions were represented in the Self-Maintenance Scale, we computed the global reliability index for multidimensional scales (Raykov, 2013). This coefficient was .78. The factor score determinacy coefficients for the four dimensions ranged from .71 to .89. All items had adequate discrimination, presenting corrected item-correlation coefficients higher than .20. The ICC coefficient for the total score of the Self-Care Maintenance Scale was .88 (95% CI [.77, .93]). Weighted κ varied from .57 to .93.

Self-Care Monitoring Scale

Dimensionality

The scale was hypothesized to have two dimensions. We expected items about monitoring sputum quantity, color of sputum, increasing of coughing, and increasing of breathlessness to cluster together in a factor measuring respiratory symptom monitoring behaviors. Similarly, we expected items about monitoring night awakening, checking difficulties falling asleep, noting increasing fatigue, and monitoring side effects of inhaled medications to cluster together in a factor measuring the extra-respiratory symptoms monitoring behaviors.

We performed a CFA on the first subsample, positing two different factors. The fit of this model was inadequate: χ²(19, N = 243) = 84.84, p < .001, CFI = .88, TLI = .82, RMSEA = .12 (90% CI [.09, .15]), p < .001, SRMR = .09. An examination of the modification indices indicated that allowing cross-loadings or covariances between residuals of items within the same factor would improve model fit. However, these modifications were not justifiable theoretically. Moreover, an examination of the factor loadings revealed a pattern of substantially high values, ranging from .48 to .96. Finally, the two factors were substantially correlated at .49. This pattern suggested the presence of a “bifactor” structure (Morin, Arens, & Marsh, 2016). This means that the items measuring self-care monitoring behaviors as an overarching construct use statements that refer to specific dimensions of this construct (respiratory symptom and extra-respiratory symptom monitoring behaviors); thus, items may reflect two different sources of true variability: one related to the overarching global construct and the other related to the corresponding specific dimensions. These two sources of multidimensionality can be accounted for with the bifactor framework. Accordingly, a bifactor model was specified, positing a general factor, named general self-monitoring behaviors, with all items loaded and two specific factors representing the two clusters of items defined above. All three factors were orthogonal. This model had an excellent fit: χ²(12, N = 243) = 20.36, p = .06, CFI = .99, TLI = .96, RMSEA = .05 (90% CI [.00, .09]), p = .40, SRMR = .05, and was fully replicated in the second subsample: χ²(12, N = 243) = 16.68, p = .16, CFI = .99, TLI = .98, RMSEA = .04 (90% CI [.00, .08]), p = .60, SRMR = .03. When the bifactor model was rerun on the total sample of 486 subjects, the fit indices were excellent (Table 2).

Table 3 presents the parameter estimates for the bifactor solution, which reveal the presence of a global factor that is well defined by all items. In a bifactor solution, factor loadings for specific factors tend to be smaller in magnitude than those for the general factor because each item is associated with at least two factors (the global factor and the specific factor). Therefore, the factor loadings of specific factors revealed that extra-respiratory symptom monitoring was relatively well defined by all its four items, while respiratory symptom monitoring was adequately defined by items monitoring the increasing of sputum and sputum changing color. Items monitoring the increasing of coughing and increasing of dyspnea contribute only marginally to this factor (Online Appendix: Supplemental Figure 1).

Reliability

Cronbach’s α coefficient was .87. However, as we identified three dimensions in the Self-Care Monitoring Scale, we retested reliability with the global reliability index for multidimensional scales, obtaining a coefficient of .92. The factor score determinacies for the two factors were .92 and .86. The corrected item-correlation coefficients for all the items ranged from .57 to .73, showing adequate discrimination. The ICC coefficient for the total score was .84 (95% CI [.74, .91]). Weighted κ varied from .66 to .81.

Self-Care Management Scale

Dimensionality

The scale was hypothesized to have three dimensions. We expected that items addressing pursed-lip breathing, talking to health-care providers, visiting health-care providers for problems, and modifying prescriptions to cluster together in a factor measuring autonomous behaviors. Similarly, we expected that items addressing calling the provider for dyspnea, coughing, increased sputum, a change in sputum color, and medication side effects would cluster together in a factor measuring consulting behaviors. Finally, we expected that items addressing sitting when working or showering would cluster together in a factor measuring problem-solving behaviors.

Using CFA, we tested a three-factor solution on the entire sample of symptomatic patients (n = 290). The first solution had inadequate fit indices: χ²(41, N = 290) = 264.35, p < .001, CFI = .75, TLI = .67, RMSEA = .14 (90% CI [.12, .15]), p < .001, SRMR = .08. The misfit was due primarily to a very high correlation (.90) between 2 items addressing related behaviors: calling the health-care provider for increased sputum, and a change in sputum color, as these two manifestations often occur simultaneously. Although the items had impressively high factor loadings of .95 and .93 on the same factor, the observed correlation between them was not totally captured by the single factor. This was evidenced by a high modification index (177) associated with the correlation between residuals. Together, these issues increased the correlation between items. To account for these issues, the model was reestimated by freely estimating the covariance between residuals of the 2 items. Since the item addressing breathing with pursed lips in a panic attack had a poor loading (<.20 in the completely standardized solution), it was omitted from the subsequent analysis. When the model was respecified, the fit indices were adequate (Table 2).

All factor loadings were significant with values ranging from .37 to .92 (Table 3). The correlations among factors were .28, .35, and .63, indicating a relevant association among the different factors representing self-care management behaviors. These factors can be considered as interrelated although representing different aspects of a more general “self-care management” dimension. The correlation between residuals of items regarding speaking to health-care providers for increased sputum and a change in sputum color was .82, but the 2 items still maintained very high factor loadings (.71 and .67, respectively; Online Appendix: Supplemental Figure 1).

Reliability

The Cronbach’s α coefficient for the entire scale was .83; the global reliability index for multidimensional scales was .87. The factor score determinacy coefficients were .93, .97, and .82. All items presented adequate discrimination, with ITCs higher than .20. The ICC coefficient for the total score was .77 (95% CI [.62, .86]). Weighted κ varied from .51 to .99.

Self-Care Self-Efficacy Scale

Dimensionality

The scale was hypothesized to include two dimensions. We expected items addressing therapeutic advice, checking symptoms, and taking medicines to cluster together in a factor measuring confidence in treatment adherence. Similarly, we expected items addressing preventing symptoms, recognizing symptoms, relieving symptoms, and evaluating intervention effectiveness to cluster together in a factor measuring confidence in symptom management.

Using CFA, we tested a two-factor solution on the sample of 389 subjects. The first solution had inadequate fit indices: χ²(13, N = 389) = 89.53, p < .001, CFI = .92, TLI = .87, RMSEA = .12 (90% CI [.10, .15]), p < .001, SRMR = .06. The modification indices demonstrated high, significant indices associated with the covariances between residuals of items regarding taking medications and following therapeutic advice and of items regarding relieving symptoms and effectiveness evaluation. This can be explained by the fact that the first 2 items both refer to confidence in following therapeutic recommendations provided by health-care personnel, and of the second pair of items, one is the consequence of the other as confidence in an action taken to alleviate a symptom involves confidence in evaluating the effectiveness of the action. After the model was respecified, the fit improved significantly (Table 2).

All factor loadings were significant with values ranging from .49 to .89 (Table 3). The correlation between the two factors was .80, indicating a strong association between the factors, which can be considered different aspects, albeit interrelated, of a more general “self-care self-efficacy” dimension. The correlation between the residuals of items related to following therapeutic advice and taking medicines was .50, and the correlation between residuals of items related to relieving symptoms and effectiveness evaluation was .57 (Online Appendix: Supplemental Figure 1).

Reliability

Cronbach’s α coefficient was .86. The global reliability index for multidimensional scales was .87, showing adequate internal coherence of the Self-Care Self-Efficacy Scale. The score determinacy coefficients were .91, and .90. The corrected item-correlation coefficients for the items of the two factors ranged from .44 to .78. The ICC coefficient for the SCES-COPD was .84 (95% CI [.74, .91]). Weighted κ for the 7 items ranged from .70 to .86.

Construct Validity

The Self-Care Maintenance Scale, Self-Care Monitoring Scale, and Self-Care Management Scale were moderately correlated among themselves, with coefficients ranging from .37 to .42. Each of the self-care scales was also moderately correlated with the SCES-COPD (r = .31–.44; Table 4). In the hypothesis testing, we anticipated that people with more severe COPD and dyspnea, and poorer health status, would score higher in the self-care scales. Subjects with GOLD 4 presented higher mean scores on the Self-Care Monitoring and Self-Care Management Scales than subjects with GOLD Grades 1 and 2. Higher mean scores were also identified for all the three self-care scales in individuals who presented more severe impact of dyspnea on daily activities (mMRC 3) and worst health status (CAT™ 4) compared to individuals with less dyspnea (mMRC 0) and low or moderate impact of the disease on their health status (CAT™ 1 or 2). Differences in self-management scores were also identified between individuals of class CAT™ 1 and all other CAT™ classes (Table 4).

Table 4.

Analysis of Variance and Correlations Among Self-Care Scales and Clinical Variables.

	Self-Care Maintenance Scale	Self-Care Monitoring Scale	Self-Care Management Scale
	Mean ± SD	Mean ± SD	Mean ± SD
GOLD grade
1—Mild	56.23 ± 18.69	54.21 ± 25.40^a	51.36 ± 25.56^a
2—Moderate	56.98 ± 17.55	62.38 ± 20.73	49.04 ± 21.89^a
3—Severe	56.60 ± 17.35	63.23 ± 21.96	53.22 ± 20.90
4—Very severe	58.49 ± 18,35	65.94 ± 23.53^b	62.56 ± 21.71^b
mMRC grade
0—No limitation	46.88 ± 17.47^a	60.35 ± 29.19^a	50.85 ± 24.52
1—Mild limitation	53.27 ± 17.79	64.89 ± 26.32	47.74 ± 24.54^a
2—Moderate limitation	53.92 ± 15.55	69.02 ± 25.43	51.69 ± 19.44
3—Severe limitation	56.95 ± 16.94^b	70.09 ± 23.91^b	53.50 ± 21.55
4—Very severe limitation	54.69 ± 17.12^b	68.93 ± 24.93	57.52 ± 24.53^b
CAT class
1—Low impact	48.33 ± 19.69^a	56.73 ± 24.27^a	35.00 ± 20.55^a
2—Moderate impact	55.48 ± 19.22	58.19 ± 27.11^a	52.38 ± 25.27^b
3—High impact	56.98 ± 17.68	64.81 ± 18.47	53.71 ± 20.13^b
4—Very high impact	64.27 ± 18.16^b	73.66 ± 20.94^b	50.22 ± 23.86^b
SCES-COPD	0.31**	0.44**	0.40**
Self-Care Maintenance Scale	—	0.37**	0.38**
Self-Care Monitoring Scale		—	0.42**

Note. Means in a column with different superscript letter are significantly different at p < .05 level. SD = standard deviation; COPD = chronic obstructive pulmonary disease; mMRC = modified Medical Research Council dyspnea questionnaire; CAT = COPD Assessment Test; GOLD = Global Initiative for Chronic Obstructive Lung disease; SCES-COPD = Self-Care Self-Efficacy Scale in COPD.

**p < .001.

Discussion

The aim of this study was to develop two new instruments and evaluate their psychometric properties. Overcoming the limitations of other self-care instruments, the SC-COPDI is based on a middle-range theory that addresses self-care in chronic illness, ensuring the theoretical foundation of the instrument. To our knowledge, this is the first study to apply the middle-range theory of self-care of chronic illness to COPD. The SC-COPDI encompasses three scales representing the theoretical dimensions of self-care: self-care maintenance, self-care monitoring, and self-care management. In each scale, the theoretical factors expressing the different sets of behaviors that characterize the practice of self-care in this disease are confirmed.

The Self-Care Maintenance Scale measures behaviors aimed at maintaining the disease stable through the prevention of respiratory infection and adherence to prescriptions and at promoting health through the maintenance of respiratory functionality and physical activities. These behaviors are not related among themselves as shown by the low correlations we found. For example, people with COPD may take their medications as prescribed by physicians but not engage in physical activity or use breathing techniques. This inconsistency in self-care maintenance behaviors has been found in other studies, showing that self-care maintenance is a complex and many-sided dimension of illness (Ausili et al., 2017; Riegel et al., 2012).

The Self-Care Monitoring Scale presents two factors describing different clusters of symptoms that people with COPD monitor: the triads of respiratory symptoms (dyspnea, coughing, and sputum), and the additional disease manifestations that derive from respiratory symptoms, such as fatigue and sleep disturbance, and from the side effects of medications, including palpitations, tremor, insomnia, dry mouth, and difficulty urinating (GOLD, 2018). In our study, a bifactor model is identified showing that all of the disease symptoms reflect both the action of a general factor of symptoms monitoring and also of two specific factors linked to the monitoring of respiratory versus extra-respiratory symptoms.

The Self-Care Management Scale includes three factors describing the behaviors that people with COPD perform autonomously, those that they perform after consulting health-care providers and those carried out to solve common daily problems. A strong correlation is found between 2 items measuring the management of two cardinal symptoms in COPD: change of color and increasing of sputum. This is justified by the fact that these two manifestations frequently occur together and require similar management, although sometimes only the increased sputum or only the color change may characterize an exacerbation of the disease (GOLD, 2018). The factors related to consulting and autonomous behaviors present a strong correlation (r = .65), showing that these sets of behaviors are intercorrelated, in accordance with other studies (Ausili et al., 2017; Riegel, Barbaranelli, et al., 2018), whereas weak correlations are found with problem-solving behaviors.

The theoretical hypotheses are confirmed, supporting the construct validity of the self-care scales. Higher self-care monitoring and self-care management scores are found in people with more severe airflow limitation, demonstrating that the scales are able to detect an increase in self-care behaviors due to progressive airflow obstruction. By contrast, the self-care maintenance behaviors are not influenced by airway limitations, showing that such behaviors are not always correlated with clinical manifestations of the disease (GOLD, 2018). Greater self-care maintenance, self-care monitoring, and self-care management behaviors were found in people with a more severe impact of the disease on their daily lives. These results are consistent with prior studies conducted in diabetes (Ausili et al., 2017) and heart failure (Dickson, Buck, & Riegel, 2011), showing that people with more compromised health status and more severe disease symptoms are most likely to perform self-care. Consistent with the middle-range theory of self-care of chronic illness, the three self-care scales are found as correlated, illustrating that they are related manifestations of the same general construct (Riegel et al., 2012).

The SCES-COPD was also developed and tested in this study to measure self-efficacy in COPD self-care. It is essential to assess self-efficacy because it can predict the self-care behaviors that an individual with COPD will perform or avoid (Bourbeau, Nault, & Dang-Tan, 2004; Wigal, Greer, & Kotses, 1991). The scale presents two factors that describe the confidence of people with COPD in their ability to adhere to prescriptions and to manage symptoms. The posited hypothesis that people with higher self-care self-efficacy would present higher self-care maintenance, self-care monitoring, and self-care management scores is confirmed in our sample.

Furthermore, we demonstrated the reliability of the three self-care scales and the Self-Care Self-Efficacy Scale measured as internal consistency and test–retest reliability. In these scales, the global reliability index, which is used to measure internal coherence in multidimensional instruments, was more than adequate, ranging from .78 to .92. As the internal coherence of the multidimensional scales was demonstrated, it is justified to calculate a global score for each scale even though they are made up of more than one dimension (Barbaranelli et al., 2014). Test–retest reliability performed on a subsample of participants demonstrated adequate stability for all four scales.

In our study, the completion of the self-care scales by the people affected by COPD did not present any difficulty; the questionnaires were easy to understand and quick to fill out. Reading difficulties were the only obstacle reported by study participants, as they often did not have their glasses in the clinical settings; however, the participants showed that they were able to respond when a person read the items for them. Therefore, its use in clinical practice would not represent a burden for people with COPD and their health-care providers.

This study presents a few limitations. First, the sample was recruited in one country and, although it represented different disease stages, ages and genders, it was not representative of the entire COPD population. Further testing in different countries and with broader samples of people with COPD is needed to confirm the psychometric properties of SC-COPDI and SCES-COPD. The responsiveness of the instruments, which is the ability to detect changes over time in the measured construct (De Vet et al., 2011), was not tested in our study. There is a need for longitudinal studies in which interventions promoting self-care are performed. In longitudinal studies, the effect of self-care behavior changes on clinical outcomes and quality of life should be evaluated.

Conclusion

The SC-COPDI is a 32-item instrument developed to measure self-care in people with COPD based on the middle-range theory of self-care of chronic illness. The instrument is composed of three separate scales that reflect the theory from which it was derived. In addition, the 7-item SCES-COPD was developed to measure self-care self-efficacy, which has a powerful influence on self-care in chronic diseases. The SC-COPD and the SCES-COPD showed good psychometric properties in an Italian population. Further testing in different populations is recommended at this point.

Supplemental Material

Supplemental Material, Appendix_Figure_1_and_Appendix_Supplemental_Table_1 - The Self-Care in Chronic Obstructive Pulmonary Disease Inventory: Development and Psychometric Evaluation

Supplemental Material, Appendix_Figure_1_and_Appendix_Supplemental_Table_1 for The Self-Care in Chronic Obstructive Pulmonary Disease Inventory: Development and Psychometric Evaluation by Maria Matarese, Marco Clari, Maria Grazia De Marinis, Claudio Barbaranelli, Dhurata Ivziku, Michela Piredda and Barbara Riegel in Evaluation & the Health Professions

Footnotes

Acknowledgments

The authors wish to thank all the nurses that contributed to the data collection and especially Emanuela Russo, Giulio Fasciolo, Minoosh Rahimzadeh, and Giulia Valente for their enthusiastic collaboration. We are also very grateful to all COPD patients who donated us their time permitting the development and testing of the instruments.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Maria Matarese

Supplemental Material

Supplemental material for this article is available online.

References

Adeloye

Chua

Lee

Basquill

Papana

Theodoratou

.,…Global Health Epidemiology Reference Group. (2015). Global and regional estimates of COPD prevalence: Systematic review and meta-analysis. Journal of Global Health, 5, 1–7. doi:10.7189/jogh.05-020415

Alberto

(1993). The reliability and validity of the COPDSC: A tool that measures self-care behaviour of persons with chronic obstructive pulmonary disease. Perspectives in Respiratory Nursing: A Publication of the Respiratory Nursing Society, 4, 3–4.

Anderson

J. C.

Gerbing

D. W.

(1988). Structural equation modeling in practice: A review and recommended two-step approach. Psychological Bulletin, 103, 411–423.

Ausili

Barbaranelli

Rossi

Rebora

Fabrizi

Coghi

… Riegel

. (2017). Development and psychometric testing of a theory-based tool to measure self-care in diabetes patients: The Self-Care of Diabetes Inventory. BMC Endocrine Disorders, 17, 66. doi:10.1186/s12902-017-0218-y

Ayre

Scally

A. J

. (2014). Critical values for Lawshe’s content validity ratio: Revisiting the original methods of calculation. Measurement and Evaluation in Counseling and Development, 47, 79–86. doi:10.1177%2F0748175613513808

Bagozzi

R. P.

(2012). Specification, evaluation, and interpretation of structural equation models. Journal of the Academy of Marketing Science, 40, 8–34.

Barbaranelli

Lee

C. S.

Vellone

Riegel

(2014). Dimensionality and reliability of the self-care of heart failure index scales: Further evidence from confirmatory factor analysis. Research in Nursing and Health, 37, 524–537. doi:10.1002/nur.21623

Barbaranelli

Lee

C. S.

Vellone

Riegel

(2015). The problem with Cronbach’s alpha: Comment on Sijtsma and van der Ark (2015). Nursing Research, 64, 140–145. doi:10.1097/NNR.0000000000000079

Bourbeau

Nault

Dang-Tan

(2004). Self-management and behaviour modification in COPD. Patient Education and Counseling, 52, 271–277.

10.

Browne

M. W.

Cudek

(1992). Alternative ways of assessing model fit. Social Methods Research, 21, 230–258. doi:10.1177/0049124192021002005

11.

Byrne

B. M.

(2006). Structural equation modeling with EQS: Basic concepts, applications, and programming (2nd ed.). Mahwah, NJ: Lawrence Erlbaum Associates.

12.

Caruso

Rebora

Dellafiore

Fabrizi

Riegel

Ausili

Di Mauro

. (2018). Clinical and socio-demographic determinants of inadequate self-care in adults with type 1 diabetes mellitus: The leading role of self-care confidence. Acta Diabetologica. doi:10.1007/s00592-018-1259-z

13.

Clari

Matarese

Alvaro

Piredda

De Marinis

M. G.

(2016). Measurement properties of instruments evaluating self-care and related concepts in people with chronic obstructive pulmonary disease: A systematic review. Heart & Lung, 45, 441–448. doi:10.1016/j.hrtlng.2016.06.006

14.

Clari

Matarese

Ivziku

De Marinis

M. G.

(2017). Self-care of people with chronic obstructive pulmonary disease: A meta-synthesis. Patient, 10, 407–427. doi:10.1007/s40271-017-0218-z

15.

Cohen

(1988). Statistical power analysis for the behavioral sciences. Hillsdale, NJ: Erlbaum.

16.

Comrey

A. L.

Lee

H. B.

(1992). A first course in factor analysis. Hillsdale, NJ: Erlbaum.

17.

Crocker

Algina

(1986). Introduction to classical and modern test theory. Orlando, FL: Holt, Rinehart and Winston.

18.

Davis

A. H. T.

Figueredo

A. J.

Fahy

B. F.

Rawiworrakul

(2007). Reliability and validity of the Exercise Self-regulatory Efficacy Scale for individuals with chronic obstructive pulmonary disease. Heart & Lung, 36, 205–216. doi:10.1016/j.hrtlng.2006.08.007

19.

De Vet

H. C.

Terwee

C. B.

Mokkink

L. B.

Knol

D. L.

(2011). Measurement in medicine: A practical guide. Cambridge, England: Cambridge University Press.

20.

Dickson

V. V.

Buck

Riegel

(2011). A qualitative meta-analysis of heart failure self-care practices among individuals with multiple comorbid conditions. Journal of Cardiac Failure, 17, 413–419. doi:10.1016/j.cardfail.2010.11.011

21.

Fornell

Larcker

D. F.

(1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18, 39–50. doi:10.2307/3151312

22.

Giacomini

DeJean

Simeonov

Smith

(2012). Experiences of living and dying with COPD: A systematic review and synthesis of the qualitative empirical literature. Ontario Health Technology Assessment Series, 12, 1.

23.

Global Initiative for Chronic Obstructive Lung Disease. (2018). Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease. Retrieved from http://www.goldcopd.org

24.

Hoyle

R. H.

Panter

A. T.

(1995). Writing about structural equation models. Structural equation modeling: Concepts, issues, and applications. Thousand Oaks, CA: Sage.

25.

L. T.

Bentler

P. M.

(1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6, 1–55. doi:10.1080/10705519909540118

26.

Jolly

Majothi

Sitch

A.J

Heneghan

N. R.

Riley

R. D.

Moore

D. J

… Jordan

R. E

. (2016). Self-management of health care behaviors for COPD: A systematic review and meta-analysis. International Journal of Chronic Obstructive Pulmonary Disease, 11, 305–326. doi:10.2147/COPD.S90812

27.

Jones

P. W.

Harding

Berry

Wiklund

Chen

W. H.

Kline Leidy

(2009). Development and first validation of the COPD Assessment Test. European Respiratory Journal, 34, 648–654. doi:10.1183/09031936.00102509

28.

Jones

P. W.

Tabberer

Chen

W. H.

(2011). Creating scenarios of the impact of COPD and their relationship to COPD assessment test (CAT™) scores. BMC Pulmonary Medicine, 11, 42. doi:10.1186/1471-2466-11-42

29.

Kaşıkçı

M. K.

Alberto

(2007). Family support, perceived self-efficacy and self-care behaviour of Turkish patients with chronic obstructive pulmonary disease. Journal of Clinical Nursing, 16, 1468–1478. doi:10.1111/j.1365-2702.2006.01782.x

30.

Kline

(1986). A handbook of test construction. New York, NY: Methuen.

31.

Landis

J. R.

Koch

G. G.

(1977). The measurement of observer agreement for categorical data. Biometrics, 33, 159–174.

32.

Lawshe

C. H.

(1975). A quantitative approach to content validity 1. Personnel Psychology, 28, 563–575. doi:10.1111/j.1744-6570.1975.tb01393.x

33.

Lenferink

Brusse-Keizer

van der Valk

P. D.

Frith

P. A.

Zwerink

Monninkhof

E. M.

… Effing

T. W

. (2017). Self-management interventions including action plans for exacerbations versus usual care in patients with chronic obstructive pulmonary disease. Cochrane Database of Systematic Reviews. doi:10.1002/14651858.CD011682.pub2

34.

Mahler

D. A.

Wells

C. K.

(1988). Evaluation of clinical methods for rating dyspnea. Chest, 93, 580–586. doi:10.1378/chest.93.3.580

35.

McDonald

R. P.

(1999). Test theory. A unified treatment. Mahwah, NJ: Lawrence Erlbaum Associates.

36.

Meade

A. W.

Johnson

E. C.

Braddy

P. W.

(2008). Power and sensitivity of alternative fit indices in tests of measurement invariance. Journal of Applied Psychology, 93, 568. doi:10.1037/0021-9010.93.3.568

37.

Morin

A. J.

Arens

A. K.

Marsh

H. W.

(2016). A bifactor exploratory structural equation modeling framework for the identification of distinct sources of construct-relevant psychometric multidimensionality. Structural Equation Modeling: A Multidisciplinary Journal, 23, 116–139. doi:10.1080/10705511.2014.961800

38.

Murphy

L. A.

Harrington

Taylor

S. J.

Teljeur

Smith

S. M.

Pinnock

Ryan

. (2017). Clinical-effectiveness of self-management interventions in chronic obstructive pulmonary disease: An overview of reviews. Chronic Respiratory Disease, 14, 276–288. doi:10.1177%2F1479972316687208

39.

Muthén

L. K.

Muthén

B. O

. (1998–2017). Mplus user’s guide (8th ed.). Los Angeles, CA: Author.

40.

Nunnally

Bernstein

I. H.

(1994). Psychometric theory. New York, NY: McGraw Hill.

41.

Pinnock

Steed

Jordan

(2016). Supported self-management for COPD: Making progress, but there are still challenges. European Respiratory Journal, 48, 6–9. doi:10.1183/13993003.00576-2016

42.

Raykov

(2013). Scale construction and development using structural equation modeling. New York, NY: Guilford Press.

43.

Riegel

Barbaranelli

Sethares

K. A.

Daus

Moser

D. K.

Miller

J. L.

… Jaarsma

. (2018). Development and initial testing of the self-care of chronic illness inventory. Journal of Advanced Nursing, 74, 2465–2476. doi:10.1111/jan.13775

44.

Riegel

Jaarsma

Lee

C. S.

Strömberg

(2018). Integrating symptoms into the middle-range theory of self-care of chronic illness. Advances in Nursing Science, 19. doi:10.1097/ANS.0000000000000237

45.

Riegel

Jaarsma

Strömberg

(2012). A middle-range theory of self-care of chronic illness. Advances in Nursing Science, 35, 194–204.

46.

Stoilkova

Janssen

D. J.

Wouters

E. F.

(2013). Educational programmes in COPD management interventions: A systematic review. Respiratory Medicine, 107, 1637–1650. doi:10.1016/j.rmed.2013.08.006

47.

Tabachnick

B. G.

Fidell

L. S.

(2007). Using multivariate statistics. Boston, MA: Pearson Education.

48.

Terwee

C. B.

Prinsen

C. A.

Chiarotto

de Vet

H. C.

Bouter

L.M.

Alonso

… Mokkink

L. B

. (2018). COSMIN methodology for assessing the content validity of PROMs—User manual. Retrieved from http://www.cosmin.nl

49.

Vandenberg

R. J.

Lance

C. E.

(2000). A review and synthesis of the measurement invariance literature: Suggestions, practices, and recommendations for organizational research. Organizational Research Methods, 3, 4–70. doi:10.1177/109442810031002

50.

Vellone

D’Agostino

Buck

H.G.

Fida

Spatola

C. F.

Petruzzo

… Riegel

. (2015). The key role of caregiver confidence in the caregiver’s contribution to self-care in adults with heart failure. European Journal of Cardiovascular Nursing, 14, 372–381. doi:10.1177/1474515114547649

51.

Vellone

Pancani

Greco

Steca

Riegel

(2016). Self-care confidence may be more important than cognition to influence self-care behaviors in adults with heart failure: Testing a mediation model. International Journal of Nursing Studies, 60, 191–199. doi:10.1016/j.ijnurstu.2016.04.016

52.

Velloso

Jardim

J. R.

(2006). Functionality of patients with chronic obstructive pulmonary disease: Energy conservation techniques. Jornal Brasileiro de Pneumologia, 32, 580–586.

53.

Wickman

U. L.

Yngman-Uhlin

Hjortswang

Wenemark

Stjernman

Riegel

Frisman

G. H.

(2019). Development of a self-care questionnaire for clinical assessment of self-care in patients with inflammatory bowel disease: A psychometric evaluation. International Journal of Nursing Studies, 89, 1–7. doi:10.1016/j.ijnurstu.2018.08.016

54.

Wigal

J. K.

Greer

T. L

Kotses

(1991). The COPD self-efficacy scale. Chest, 99, 1193–1196.

55.

World Health Organization. (2018). Global health estimates 2016: Deaths by cause, age, sex, by country and by region, 2000–2016. Geneva, Switzerland: Author. Retrieved from http://www.who.int/healthinfo/global_burden_disease/estimates/en/

56.

Xiaolian

Chaiwan

Panuthai

Yijuan

Lei

Jiping

. (2002). Family support and self-care behavior of Chinese chronic obstructive pulmonary disease patients. Nursing & Health Sciences, 4, 41–49.

57.

Yohannes

A. M.

Alexopoulos

G. S.

(2014). Depression and anxiety in patients with COPD. European Respiratory Review, 23, 345–349. doi:10.1183/09059180.00007813

58.

Zhang

Wang

Cai

Zhang

Wang

(2013). Development and validation of a COPD self-management scale. Respiratory Care, 58, 1931–1936. doi:10.4187/respcare.02269

59.

Zwerink

Brusse-Keizer

van der Valk

P. D.

Zielhuis

G. A.

Monninkhof

E. M.

van der Palen

… Effing

. (2014). Self management for patients with chronic obstructive pulmonary disease. Cochrane Database Systematic Review, 19. doi:10.1002/14651858.CD002990.pub3

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.36 MB