Skip to main content
Download PDF
Download ePub

Self-Efficacy to Manage Chronic Disease (SEMCD) scale: translation and evaluation of measurement properties for a swedish version

Archives of Public Health volume 81, Article number: 2 (2023) Cite this article

Abstract

Background

Reinforcing self-efficacy in patients is important in person-centered care; therefore, reliable and valid measures of a person’s self-efficacy is of clinical relevance. A questionnaire suitable for self-efficacy and patient engagement that is not limited to a particular condition is the Self-efficacy to Manage Chronic Disease (SEMCD). This study aims to evaluate the measurement properties of a Swedish translation of the SEMCD with a Rasch analysis.

Methods

The translation and cultural adaptation of the SEMCD was performed according to the International Society for Pharmacoeconomics and Outcomes Research (ISPOR) recommendations. Self-reported data was collected from two cohorts: patients with pituitary tumors (n = 86) and patients on sick leave due to common mental disorders (n = 209). Measurement properties were evaluated with a Rasch analysis in RUMM2030.

Results

The original six-item SEMCD did not fit to a unidimensional scale. Two items, item 5 and item 6, deviated both statistically and conceptually and were removed. A four-item solution, the SEMCD-4 with collapsed thresholds for mid-range response options, showed good targeting and unidimensionality, no item misfit, and a reliability of 0.83.

Conclusion

In a Swedish context with a mix of patients with pituitary tumors or common mental disorders, SEMCD-4 showed satisfactory measurement properties. Thus, SEMCD-4 could be used to identify patient self-efficacy in long-term illnesses. This knowledge about patient self-efficacy may be of importance to tailor person-centered support based on each patient´s resources, needs and goals.

Peer Review reports

Background

Prevention and management of long-term illness is a major health challenge and a global priority [1]. There is a need for future care in which each person living with a long-term illness is more involved, takes greater responsibility for and manages to a greater extent symptoms and other consequences in everyday life [2]. Person-centered care (PCC) emphasizes the importance of knowing the patient as a person, and aims to engage patients as active partners in their care and treatment [3, 4]. By acknowledging the patient’s unique experiences, capabilities, resources and needs, PCC can support a patient’s self-management and coping [3]. PCC coupled with health promotion aiming to improve self-management abilities has been suggested as a way to provide sustainable, affordable, high quality health care [5].

A concept closely related to PCC is self-efficacy, which refer to a person’s confidence in their capacity to influence events that affect their life [6]. A person’s self-efficacy beliefs have important implications for managing chronic and long-term illness [7], as patients who feel more confident in performing activities are more likely to reach treatment goals [8]. PCC seeks to reinforce self-efficacy in patients rather than to simply educate and convince them of the importance of such activities [3, 9]. There is a relationship between self-efficacy to manage chronic disease and health-related quality of life (HRQOL); patients with multi-morbidity in primary care who report higher self-efficacy also have higher HRQOL than those who report lower self-efficacy [10].

Although self-efficacy is primarily considered task-specific, meaning a person´s belief in their ability to deal with a specific situation at hand [11], it may also be perceived as a general concept, which instead refers to a person´s global confidence in their ability to perform across a variety of situations [12]. Several PCC interventions, addressing various conditions and performed in different health care contexts, have been shown to increase both illness-specific [13, 14] and general levels of self-efficacy [15, 16].

A person’s self-efficacy is typically measured by self-report questionnaires. A particularly suitable questionnaire for self-efficacy and patient engagement is the Self-Efficacy to Manage Chronic Disease (SEMCD) [17]. The SEMCD is not limited to a particular condition, rather it is intended to be useful in a broad range of situations involving individuals with varying long-term illnesses. Examples of previous applications are cardiovascular diseases, diabetes, arthritis, cancer, asthma and neurological conditions [18,19,20,21,22,23,24]. Furthermore, SEMCD was used in a Swedish study of graft rejection after lung transplantation [25], but without psychometric evaluation. Recently, another Swedish study translated the SEMCD and tested its psychometric properties in a population of patients with systemic sclerosis [26].

To the best of our knowledge, previous validation studies of the SEMCD have used Classical Test Theory (CTT) [17,18,19, 26, 27]. CTT does not adequately handle ordinal data and separates item and person attributes [28, 29]. To overcome those shortages, a measurement science perspective is needed to analyze the measurement properties of the SEMCD. According to Rasch [30], data are evaluated against a mathematical measurement model to guide the construction of stable linear measures for item and person attributes from raw data, such as questionnaire responses. By examining the extent to which observed data accord with the expected values that are defined by the measurement model, the Rasch analysis assesses if requirements for internal validity and invariant measurement across items and persons are met.

Hence, this study aims to evaluate the measurement properties of a Swedish translated version of the SEMCD with a Rasch analysis.

Methods

Study design and setting

The translation and cultural adaptation of the SEMCD was performed according to international recommendations [31]. The data used in this study derived from baseline assessments from two different interventional studies:

  1. i.

    a single-center quasi-experimental study evaluating a person-centered practice after surgery for patients with pituitary tumors (PT) [32]; hereafter, this sample is referred to as the PT-cohort; and

  2. ii.

    a randomized controlled study evaluating the effects of a person-centered eHealth intervention (combined digital platform and phone support) for patients on sick leave due to common mental disorders (CMD) [16, 33]; hereafter, this sample is referred to as the CMD-cohort.

Participants and data collection

Table 1 summarizes the participant characteristics from the PT- and CMD-cohorts. The PT-cohort consists of 86 patients with pituitary tumors (aged 24–87) included in the intervention group between February 2018 and December 2020. All consecutive patients over the age of 18 and planned for neurosurgery due to a pituitary tumor were asked to participate. Exclusion criteria were conditions that might restrict understanding of the study or the ability to adhere to the protocol (e.g., cognitive impairments or drug addiction). Detailed information about the data collection can be found in Jakobsson et al. [32].

Table 1 Participant characteristics from the pituitary tumors (PT)- and common mental disorders (CMD)-cohorts and total for evaluation of the Swedish version of Self-Efficacy to Manage Chronic Disease Scale
Full size table

The CMD-cohort consists of 209 patients (aged 18–65) on sick leave due to common mental disorders (stress and adjustment disorders, depression, or anxiety) who were recruited between February 2018 and June 2020 from primary care centers and randomly assigned to the intervention or control group. Patients were eligible if they were currently employed or studying at least part-time during the past nine months, and if their current sick leave period had not exceeded 30 days. Exclusion criteria were: sick leave > 14 days due to depression, anxiety disorders, stress reactions and disorders during the past 3 months; severe impairments impeding use of the eHealth intervention; ongoing alcohol or drug abuse; severe disease with an expected survival of < 12 months or that could interfere with follow-up or if the intervention was assessed as a burden; or the patient was participating in a conflicting study. Details about the data collection can be found in Cederberg et al. [16, 33].

Measurements

The SEMCD is a questionnaire developed for the assessment of a patient’s confidence in their ability in certain tasks related to self-management (see Table 3 in the Results section for items). With the SEMCD the patient is asked to rate his or her confidence related to six tasks on a 1 to 10 scale (1 anchored to “not at all confident” and 10 to “totally confident”) [17].

The SEMCD was developed in the U.S. by Lorig and co-workers [17, 21] and has been translated into Spanish [21, 24], German [18], French [34], Chinese [35, 36], Turkish [19], Portuguese [37], Persian [38], Arabic [39] and Swedish [26]. A comparison of the translations in the other Swedish study and the present translation shows that the content is the same, but the word-order is different.

In this study, the recommendations from the International Society for Pharmacoeconomics and Outcomes Research (ISPOR) [31] were followed to translate and adapt the SEMCD. The process is shown in Table 2.

Table 2 Translation and adaptation process for the Swedish version of Self-Efficacy to Manage Chronic Disease (SEMCD) Scale
Full size table

The General Self-Efficacy Scale (GSE) is a questionnaire for the assessment of a patient’s confidence in their ability to cope with and adapt to difficult problems and challenges [40]. With the GSE the patient is asked to rate his or her confidence for ten challenges on a four-point scale (1 = “not at all true”; 2 = “hardly true”; 3 = “moderately true”; and 4 = “exactly true”). The GSE was originally developed in Germany and has been translated to and used in more than 30 languages, including Swedish, and is broadly used internationally [12].

Data analysis

The item and person attributes are so-called coupling attributes [41]. In this study the questionnaire item attribute measures range from easy to difficult tasks to manage, hereafter referred to as item difficulty measures. The person attribute measures range from lower to higher confidence in self-management ability, hereafter referred to as self-efficacy measures. An easy task to manage corresponds to an item with high probability of success while a difficult task to manage corresponds to an item with low probability of success. Consequently, a person with a lower self-efficacy score has a lower probability of success with self-efficacy tasks that are difficult to manage.

For all analyses the software Rasch Unidimensional Measurement Model 2030 (RUMM2030) was used and a partial credit model was applied, i.e., each item was allowed to have its own rating scale structure. Correlations were done in SPSS version 27. The Rasch analysis was structured around the fundamental measurement aspects of the Rasch model [29, 42].

Targeting of persons and items

Comparing mean person location with mean item location (i.e., 0 logits) indicates whether or not the sample is off-centered with respect to the items [42]. Moreover, person fit residuals should ideally lie within the -2.5 to + 2.5 range [29]. Together with the fit statistics, examining how the items are distributed along the continuum is crucial for deciding whether a measurement scale can successfully be constructed or not [29]. This means that one must consider if the item distribution is consistent with clinical or theoretical expectations [43].

Threshold ordering

To evaluate the monotonicity of items the threshold orders were evaluated; the ratings on one item should be consistent with the metric estimate of the underlying construct. Collapsing categories was considered when disordered thresholds occurred [42].

Item model fit

Items were assessed according to fit residuals, chi-squared tests and the Item Characteristic Curve (ICC). The following guidelines were applied iteratively: i) mean fit residuals should be close to zero (0) and have standard deviations (SD) close to 1; ii) the individual item fit residuals should be between -2.50 and + 2.50; iii) the chi-squared values should not be statistically significant (Bonferroni corrected); and iv) the dots of the class intervals should follow the ICC to support good fit [29].

Differential Item Functioning (DIF)

For DIF analysis both main effects and interaction effects were taken into account; both were expected to be non-significant. Because of the multiple tests, Bonferroni correction was applied. DIF was tested for cohort, gender, age group and education accordingly (see participant characteristics in Table 1). When DIF was observed, item splits were evaluated in repeated analyses starting with the item with largest DIF [44].

Local dependency

To evaluate local dependency among items, residual correlations were compared against a relative cut off; residual correlations > 0.20 above the average correlations indicated local dependency [45].

Unidimensionality

Having items in a questionnaire measure only a single construct is a fundamental requirement for the Rasch model; unidimensionality should therefore be evaluated. Smith’s [46] method for testing unidimensional is recommended. First, the patterning of residuals is evaluated in a Principal Component Analysis (PCA); second, the first residual factor is used to define two subsets of items by dividing positively and negatively correlated items; and third, the person estimates for each subset are compared by using an independent t-test. To support unidimensionality, the percentage of tests outside the range -1.96 to 1.96 should not exceed 5%.

Reliability

To evaluate the internal consistency reliability, the Person Separation Index (PSI) was used. The PSI is equivalent to Cronbach’s alpha and interpreted as follows: zero (0) indicates all errors/noices and one (1) no error.

Group differences

ANOVA was used to evaluate group differences for cohort, gender, age group and education [29].

In addition, correlation between self-efficacy measures from SEMCD and GSE was evaluated. We expected the correlation to be positive but not strong as the SEMCD and GSE assesses different dimensions of self-efficacy. Assessment of the measurement properties of GSE is beyond this study. However, to perform statistical analyses (both creating a total “GSE measure” and correlations) we applied the Rasch formula to derive linear GSE person measures. In short, the GSE did not have any misfitting items, response options worked as intended, and although the PSI was > 0.90 it was slightly positively skewed (mean location 0.683; SD 2.063).

Results

The results are presented in four sections. First results from the Rasch analysis of the full scale SEMCD are reported. Second, rationales and suggestions for modifications are presented. Third, results from the Rasch analysis of the modified SEMCD are reported. Finally, we report the correlations between the modified SEMCD and the GSE.

Initial analysis – SEMCD full scale

Targeting of persons and items

Mean self-efficacy measures were close to zero (-0.0049) indicating a well targeting between the sample and the items. Furthermore, 5 (1.7%) persons had fit residuals > 2.5 and 18 (6.1%) persons had fit residuals < 2.5.

Threshold ordering

Disordered thresholds were present for item 1 [Confidence in keeping fatigue from interfering] and item 3 [Confidence in keeping emotional distress from interfering]. For the remaining items thresholds were ordered but narrow in the middle of the response options.

Item model fit

Mean fit residuals were close to zero (0.075), but the SD was larger than desired (2.874). All items except item 6 [Confidence in reduce the illness effects] showed fit residuals within the desired range of ± 2.5 and non-significant chi-square values (item 6 chi-square 5.798, p < 0.001). Table 3 summarizes all fit statistics.

Table 3 Summary of fit statistics for the Swedish version of the Self-Efficacy to Manage Chronic Disease (SEMCD) scale showing full scale and modified versions. The SEMCD item 1-4 where response option 4-5-6 is collapsed for item 1 is further reported in the section on SEMCD-4.
Full size table

DIF

Item 6 [Confidence in reduce the illness effects] showed significant DIF for all comparisons (cohort, gender, age group and education). Item 3 [Confidence in keeping emotional distress from interfering] showed significant DIF for cohort, gender and age group comparisons. Item 1 [Confidence in keeping fatigue from interfering] showed significant DIF for comparisons between cohorts and age groups.

Local dependency

The average residual correlation was -0.173, giving a relative cut-off of 0.027. Three residual correlations were above the cut-off: item 1 [Confidence in keeping fatigue from interfering] and item 2 [Confidence in keeping pain/physical discomfort from interfering] (0.090); item 1 [Confidence in keeping fatigue from interfering] and item 3 [Confidence in keeping emotional distress from interfering] (0.175); and item 5 [Confidence in reduce need to the doctor] and item 6 [Confidence in reduce the illness effects] (0.099).

Unidimensionality

The positively loaded item residuals in the first residual factor were item 5 [Confidence in reduce need to the doctor] and 6 [Confidence in reduce the illness effects]. The remaining four item residuals were negatively loaded. The comparison of the person self-efficacy measures from those subsets, i.e., items 1–4 and items 5–6, had 7.85% persons outside the desired range of ± 1.96.

Reliability

The PSI without extremes was 0.834 and with extremes 0.862.

Group differences

There was no statistically significant difference in self-efficacy measures based on education (p = 0.243). For the other group comparisons: the PT-cohort showed higher self-efficacy compared to the CMD-cohort (p < 0001), older (> 65 years) showed higher self-efficacy compared to younger (< 35 years) (p < 0001), and men showed higher self-efficacy compared to women (p < 0001).

Rationale for modification of the SEMCD

Item 6 [Confidence in reduce the illness effects] had several drawbacks in terms of individual item fit and effect on unidimensionality. It is also conceptually different from all other items in the SEMCD. As described previously [17], there is a difference between items 5 and 6 compared to items 1–4 in terms of area of confidence assessed. Items 5 and 6 concern confidence to do things to manage the disease, whereas items 1–4 concern confidence to keep aspects of the disease from interfering with things a person wants to do [17]. When removing item 6, item 5 [Confidence in reduce need to the doctor] showed minor misfit. Due to its conceptual difference from items 1–4 as well as its somewhat misleadingly referring only to the doctor and no other profession, item 5 was also removed. This action corresponds well with the suggested argument for changing ‘see a doctor’ to ‘meet health care’ suggested by Mattsson et al. [26]. Further, we iteratively assessed response category function and DIF. After these iterative analyses the optimum result was obtained when removing both item 5 and item 6 as well as collapsing the middle (4–5-6) response options for item 1 [Confidence in keeping fatigue from interfering], giving SEMCD-4.

SEMCD-4

Targeting of persons and items

Mean person location was -0.294 (SD 1.592), indicating good sample-item targeting. However, as shown in the person-item threshold histogram in Fig. 1, item thresholds are not covered by the person thresholds. Specifically, there is a gap in item thresholds between 1.936 and 4.844 logits, where the higher threshold – alone – belongs to item 1 [Confidence in keeping fatigue from interfering].

Fig. 1
figure 1

Person-Item threshold distributions for the swedish version of Self-Efficacy to Manage Chronic Disease (SEMCD) scale. Top histogram (pink bars) showing sample distribution of person self-efficacy measures and bottom histogram (blue bars) distribution of location of item thresholds in the SEMCD-4. Green line shows information curve

Full size image

In total, 1 (0.3%) person had fit residual > 2.5, i.e., an unexpected irregular response pattern, and 19 (6.5%) persons had fit residuals < 2.5, i.e., low variations in their response patterns.

Threshold ordering

As response categories were collapsed to resolve disordered thresholds, all thresholds became ordered.

Item model fit

Mean fit residuals were 0.501 (SD 0.765) and item fit residuals were within the desired range of ± 2.5 with non-significant chi-square values (Table 3), supporting unidimensionality.

We did not have any a priori understanding of easier or more difficult items (i.e., the probability of success), but the analyses showed that item 2 [Confidence in keeping pain/physical discomfort from interfering] was the easiest, followed by item 4 [Confidence in keeping other symptoms from interfering], then item 3 [Confidence in keeping emotional distress from interfering] and then item 1 [Confidence in keeping fatigue from interfering] (location column in Table 3).

DIF

Item 3 [Confidence in keeping emotional distress from interfering] showed significant DIF between the cohorts (Fig. 2a) as well as for gender. Item 1 [Confidence in keeping fatigue from interfering] showed significant DIF for comparisons between age groups (Fig. 2b). The DIF on item 3 was successfully resolved by splitting the item for cohort, but not gender. The item split showed that for item 3 there were approximately 4 logits difference between the cohorts, where it was easier for the PT-cohort compared to the CMD-cohort (i.e., the PT-cohort had a higher probability of success managing emotional stress). The DIF on item 1 was successfully resolved by splitting the item for the oldest cohort members (over 65 years); the older cohort members endorsed item 1 1.2 logits higher. However, splitting items did not change the person self-efficacy measures significantly.

Fig. 2
figure 2

a-b Differential item functioning curves for the swedish version of Self-efficacy to Manage Chronic Disease scale. Differential item functioning curves for a item 3 [Confidence in keeping emotional distress from interfering] for the cohort groups (pituitary tumors-cohort in blue and common mental disorders-cohort in red) and b item 1 [Confidence in keeping fatigue from interfering] for the different age groups

Full size image

Local dependency

The average residual correlation was -0.317, giving the relative cut-off of -0.117. The item residual correlation between item 1 [Confidence in keeping fatigue from interfering] and item 3 [Confidence in keeping emotional distress from interfering] was slightly above the average cut-off (-0.083).

Unidimensionality

Item 1[Confidence in keeping fatigue from interfering] and item 2 [Confidence in keeping pain/physical discomfort from interfering] had positively correlated item residuals in the first residual factor, and item 3 [Confidence in keeping emotional distress from interfering] and item 4 [Confidence in keeping other symptoms from interfering] were negatively correlated. When comparing the person estimates from the subsets, less than 5% (4.44%), were outside the desired range of ± 1.96. These findings support unidimensionality.

Reliability

Despite removing items as well as collapsing response options, the high reliability was not affected. The PSI without extremes was 0.829 and with extremes 0.868.

Group differences

The same statistically significant and non-significant differences as for the full scale were shown. These were: no difference for education (p = 0.129); a difference in which the PT-cohort showed higher self-efficacy compared to the CMD-cohort (p < 0001); older (> 65 years) showed higher self-efficacy compared to younger (< 35 years) (p < 0001); and men showed higher self-efficacy compared to women (p < 0001).

Correlation between SEMCD-4 and GSE

As expected, there was a positive correlation between the person self-efficacy measures from SEMCD-4 and GSE (Pearson correlation coefficient 0.424, p < 0.001). For subgroups, the lowest correlation was observed among 36–45-year-old persons (Pearson correlation coefficient 0.155; p = 0,243) and the highest correlation was observed among 56–65-year-old persons (Pearson correlation coefficient 0.537; p < 0.001).

Discussion

This article reports the translation of the SEMCD and an evaluation of its measurement properties. Our findings propose that a four-item solution, SEMCD-4, fits well to the measurement model in a Swedish context, which implies satisfactory measurement properties.

Removing items has, at least, two critical implications. First, the question arises as to whether parts of the initial item set could be used to measure the same attribute or if only one aspect of the attribute is being measured. Previous work has shown difficulties fitting the six items together into one dimension [17], and the results in this study confirm this difficulty. Recently, the PROMIS®-network has provided an item bank for measuring self-efficacy in chronic conditions [47]. Based on a Delphi panel they have divided the construct and proposed five different dimensions of self-efficacy management: daily activities, medications and treatment, symptoms, emotions, and social interactions. The symptom dimension showed a strong correlation (r = 0.76) with the SEMCD, which is not surprising as it comprises four specific symptom items [47]. The major difference between our proposed SEMCD-4 and the PROMIS® symptom dimension item bank is that SEMCD-4 measures four general symptom management items compared to the PROMIS® item bank with 28 items addressing confidence in symptom management, symptom management in different settings, and in keeping symptoms from interfering with daily activities [47]. Thus, taking previous work and the results of this study into consideration, the unidimensionality of the construct self-efficacy to manage chronic disease is questionable. However, were there greater equality in the number of items covering the different aspects of self-management (there is a need to add items to SEMCD), it might be possible to create a higher ordered measure of self-efficacy for self-management in long-term illnesses. This hypothesis, however, needs further investigation. A second critical implication is that comparability might be lost in removing items. From the perspective of the CTT, this conjecture is correct. The Rasch model, applied in this study, offers a logical progression from the CTT paradigm to a more sophisticated model [48], namely, a specific metrological approach to human-based measurements [41], to overcome the shortages with ordinality and lack of separation between item and person attributes. A key factor for comparability, independent of which items are included, is traceability to metrological references, which is enabled with calibrated items [41, 49, 50]. Thus, applying the Rasch model ensures improved possibilities for metrological comparability. Specifically, sets of items (e.g., SEMCD and SEMCD-4) can be linked to a common self-efficacy measures by means of co-calibration of item parameters [50].

With regards to the item hierarchy of SEMCD-4, it was evident that is it easier being confident managing physical symptoms than symptoms dominated by psychological components. That is, there is a higher probability of scoring higher on the item for managing physical symptoms or pain compared to the items for managing emotional stress or fatigue. This difference may be due to the symptomatology of the cohorts. Symptoms of fatigue and being more sensitive to stressful situations are prevalent in CMD [51, 52], and fatigue is found to be among the most frequent and bothersome complaints reported by patients treated for pituitary disease [53, 54]. Another explanation could be that physical symptoms are more closely related to clinical signs (e.g., feeling hot or flushed, body aches, and weakness are symptoms of fever), which are still often emphasized in clinical assessments [55] and thereby presumably given more attention and intervention. However, that possible explanation is undercut by the fact that medically unexplained physical symptoms and pain are health problems that are increasing in the western world [56]. We would therefore encourage further investigation into the item hierarchy and the relation between different aspects of self-efficacy in different clinical settings.

Because of the need for self-report questionnaires to evaluate self-efficacy in managing chronic diseases, two Swedish versions of the SEMCD have been developed in parallel: the SEMCD-4 evaluated in this present study and the recently investigated SEMCD-Swe [26]. As mentioned above, the SEMCD-Swe and our work show comparable content but different word-order, which might not be a major issue. Furthermore, respondents in the Mattsson et al. [26] study noted that items 5 [Confidence in reduce need to the doctor] and 6 [Confidence in reduce the illness effects] were difficult to understand. They had to allow for covariance between those items to fit a single-factor structure, but choose to keep them in their final version. Different length questionnaires can be confusing, but the benefit of the Rasch analysis in this study is that is showed that only the four symptom items can be combined with validity, something that is supported by previous work [17].

In this cross-sectional study, we showed positive moderate correlations between person measures from the SEMCD and GSE. This corroborates other studies that evaluated correlations between GSE and specific self-efficacy scales [12, 18, 57]. Considering that both specific self-efficacy scales and the GSE originate from the concept of self-efficacy, such positive correlations are expected. However, the correlation between the SEMCD and GSE was not strong enough to consider them to be measuring the same thing. Hence, it is possible to exhibit higher general self-efficacy and at the same time lower self-efficacy to manage a chronic disease, and vice versa. However, their relationship over time after an intervention of person-centered care is unexplored.

The selection of the two cohorts in this study was based on their clinical differences (e.g., etiology, symptom expression, treatment, different care levels) but also on their similarity in both being long-term illnesses for which self-management is a major part of care. In line with the original idea of the SEMCD, it should not be restricted to a specific diagnosis or setting but rather used to target self-efficacy for self-management in general [17]. At this stage, however, generalizability of our results to other long-term illnesses is limited. There was also a practical reason for the choice of cohorts as in both cases the interventions are evaluating the effects of PCC and both use self-efficacy as an outcome measure. The studies were not specifically designed to assess the measurement properties of SEMCD. Shortcomings include not having appropriate sample sizes and not equally balanced cohorts in terms of diagnostic group. However, the possibility of including SEMCD in intervention studies in the future presents opportunities for further analysis of sensitivity to change and item stability over time.

This study has implications for clinical practice given the global burden of long-term and chronic illness and the concomitant need for person-centered interventions supporting patients’ self-management. The SEMCD-4 is well suited to identify patients with lower levels of self-efficacy and thus probably in higher need of support to manage symptoms in long-term illness, with that support tailored to individual needs. The SEMCD-4 also has a low respondent burden and worked well in our two cohorts; it can therefore possibly be used in research and clinical practice to evaluate interventions aiming at increased self-management in different patient groups and clinical settings.

Conclusion

In a Swedish context with patients with pituitary tumors or common mental disorders, the SEMCD-4, a four-item solution that includes management of symptoms, shows satisfactory measurement properties. For the clinical groups studied here, and perhaps beyond, the SEMCD-4 could be used to identify patient self-efficacy for managing long-term illness. It could also be used to tailor person-centered support based on each patient´s resources, needs and goals.

We encourage further work to explore a higher ordered measure of self-efficacy for self-management in long-term illness. A higher-ordered measure is warranted when including more self-efficacy related activities beyond symptom management, requiring investigations of the item hierarchy and relations between different self-efficacy items across different clinical settings.

Availability of data and materials

The datasets used during the current study are available from the corresponding author on reasonable request.

Abbreviations

CMD:

Common mental disorders

CTT:

Classical Test Theory

DIF:

Differential Item Functioning

GSE:

General Self-Efficacy Scale

HRQOL:

Health-related quality of life

ICC:

Item Characteristic Curve

ISPOR:

International Society for Pharmacoeconomics and Outcomes Research

PCA:

Principal Component Analysis

PCC:

Person-Centered Care

PSI:

Person Separation Index

PT:

Pituitary tumors

SD:

Standard deviations

SEMCD:

Self-efficacy to Manage Chronic Disease

References

  1. Global Action Plan for the Prevention and Control of NCDs 2013–2020. Available from: https://www.who.int/publications-detail-redirect/9789241506236

  2. Lloyd HM, Ekman I, Rogers HL, Raposo V, Melo P, Marinkovic VD, et al. Supporting innovative person-centred care in financially constrained environments: the we care exploratory health laboratory evaluation strategy. Int J Environ Res Public Health. 2020;17:3050.

    Article  Google Scholar 

  3. Ekman I, Swedberg K, Taft C, Lindseth A, Norberg A, Brink E, et al. Person-centered care — ready for prime time. Eur J Cardiovasc Nurs. 2011;10:248–51.

    Article  Google Scholar 

  4. Leplege A, Gzil F, Cammelli M, Lefeve C, Pachoud B, Ville I. Person-centredness: Conceptual and historical perspectives. Disabil Rehabil. 2007;29:1555–65.

    Article  Google Scholar 

  5. Swedberg K, Cawley D, Ekman I, Rogers HL, Antonic D, Behmane D, et al. Testing cost containment of future healthcare with maintained or improved quality—The COSTCARES project. Health Sci Rep. 2021;4:e309.

    Article  Google Scholar 

  6. Bandura A. Self-efficacy: Toward a unifying theory of behavioral change. Psychol Rev. 1977;84:191–215.

    Article  CAS  Google Scholar 

  7. Lorig KR, Holman HR. Self-management education: History, definition, outcomes, and mechanisms. Ann Behav Med. 2003;26:1–7.

    Article  Google Scholar 

  8. Bandura A. Self-efficacy: The exercise of control. New York, NY, US: W H Freeman/Times Books/ Henry Holt & Co; 1997. p. ix, 604.

    Google Scholar 

  9. Bandura A. The anatomy of stages of change. Am J Health Promot AJHP. 1997;12:8–10.

    Article  CAS  Google Scholar 

  10. Peters M, Potter CM, Kelly L, Fitzpatrick R. Self-efficacy and health-related quality of life: a cross-sectional study of primary care patients with multi-morbidity. Health Qual Life Outcomes. 2019;17:37.

    Article  Google Scholar 

  11. Frei A, Svarin A, Steurer-Stey C, Puhan MA. Self-efficacy instruments for patients with chronic diseases suffer from methodological limitations - a systematic review. Health Qual Life Outcomes. 2009;7:86.

    Article  Google Scholar 

  12. Luszczynska A, Scholz U, Schwarzer R. The general self-efficacy scale: multicultural validation studies. J Psychol. 2005;139:439–57.

    Article  Google Scholar 

  13. Ali L, Wallström S, Ekman I, Swedberg K, Fors A. Effects of person-centred care via telephone on self-efficacy in patients with chronic obstructive pulmonary disease: Subgroup analysis of a randomized controlled trial. Nurs Open. 2020;8:927–35.

    Article  Google Scholar 

  14. Fors A, Taft C, Ulin K, Ekman I. Person-centred care improves self-efficacy to control symptoms after acute coronary syndrome: a randomized controlled trial. Eur J Cardiovasc Nurs. 2016;15:186–94.

    Article  Google Scholar 

  15. Fors A, Ekman I, Taft C, Björkelund C, Frid K, Larsson ME, et al. Person-centred care after acute coronary syndrome, from hospital to primary care - a randomised controlled trial. Int J Cardiol. 2015;187:693–9.

    Article  Google Scholar 

  16. Cederberg M, Alsén S, Ali L, Ekman I, Glise K, Jonsdottir IH, et al. Effects of a Person-Centered eHealth Intervention for Patients on Sick Leave Due to Common Mental Disorders (PROMISE Study): open randomized controlled trial. JMIR Ment Health. 2022;9:e30966.

    Article  Google Scholar 

  17. Ritter PL, Lorig K. The English and Spanish self-efficacy to manage chronic disease scale measures were validated using multiple studies. J Clin Epidemiol. 2014;67(11):1265–73.

  18. Freund T, Gensichen J, Goetz K, Szecsenyi J, Mahler C. Evaluating self-efficacy for managing chronic disease: psychometric properties of the six-item Self-Efficacy Scale in Germany. J Eval Clin Pract. 2013;19:39–43.

    Article  Google Scholar 

  19. Incirkus K, Özkan NN. Validity and reliability study of the Turkish version of the self-efficacy for managing chronic disease 6-item scale. Turk J Med Sci. 2020;50:1254–61.

    Article  Google Scholar 

  20. Kephart G, Packer TL, Audulv Å, Warner G. The structural and convergent validity of three commonly used measures of self-management in persons with neurological conditions. Qual Life Res. 2019;28:545–56.

    Article  Google Scholar 

  21. Lorig KR, Sobel DS, Ritter PL, Laurent D, Hobbs M. Effect of a self-management program on patients with chronic disease. Eff Clin Prac. 2001;4(6);256–62.

  22. Lorig K, Ritter PL, Plant K, Laurent DD, Kelly P, Rowe S. The South Australia health chronic disease self-management internet trial. Health Educ Behav. 2013;40:67–77 SAGE Publications Inc.

    Article  Google Scholar 

  23. Lorig K, Laurent DD, Plant K, Krishnan E, Ritter PL. The components of action planning and their associations with behavior and health outcomes. Chronic Illn SAGE Publications. 2014;10:50–9.

    Article  Google Scholar 

  24. Lorig KR, Ritter PL, González VM. Hispanic chronic disease self-management: a randomized community-based outcome trial. Nurs Res. 2003;52:361–9.

    Article  Google Scholar 

  25. Forsberg A, Nilsson M, Jakobsson S, Lennerling A, Kisch A. Fear of graft rejection 1–5 years after lung transplantation-A nationwide cohort study. Nurs Open. 2018;5:484–90.

    Article  Google Scholar 

  26. Mattsson M, Sandqvist G, Hesselstrand R, Olsson D, Kwakkenbos L, Nordin A, et al. Validity and reliability of the Swedish version of the Self-Efficacy for Managing Chronic Disease scale for individuals with systemic sclerosis. Scand J Rheumatol. 2021;0:1–10 Taylor & Francis.

    Google Scholar 

  27. Riehm KE, Kwakkenbos L, Carrier M-E, Bartlett SJ, Malcarne VL, Mouthon L, et al. Validation of the self-efficacy for managing chronic disease scale: a scleroderma patient-centered intervention network cohort study. Arthritis Care Res. 2016;68:1195–200.

    Article  Google Scholar 

  28. Hobart JC, Cano SJ, Zajicek JP, Thompson AJ. Rating scales as outcome measures for clinical trials in neurology: problems, solutions, and recommendations. Lancet Neurol. 2007;6:1094–105.

    Article  Google Scholar 

  29. Hobart J, Cano S. Improving the evaluation of therapeutic interventions in multiple sclerosis: the role of new psychometric methods. 2009;2009:214.

  30. Rasch G. Studies in mathematical psychology: I. Probabilistic models for some intelligence and attainment tests. Oxford, England: Nielsen & Lydiche; 1960.

    Google Scholar 

  31. Wild D, Grove A, Martin M, Eremenco S, McElroy S, Verjee-Lorenz A, et al. Principles of good practice for the translation and cultural adaptation process for Patient-Reported Outcomes (PRO) measures: report of the ISPOR task force for translation and cultural adaptation. Value Health. 2005;8:94–104.

    Article  Google Scholar 

  32. Jakobsson S, Olsson DS, Andersson E, Hallén T, Krabbe D, Olofsson A-C, et al. Extended support within a person-centered practice after surgery for patients with pituitary tumors: protocol for a Quasiexperimental study. JMIR Res Protoc. 2020;9:e17697.

    Article  Google Scholar 

  33. Cederberg M, Ali L, Ekman I, Glise K, Jonsdottir IH, Gyllensten H, et al. Person-centred eHealth intervention for patients on sick leave due to common mental disorders: study protocol of a randomised controlled trial and process evaluation (PROMISE). BMJ Open. 2020;10:e037515.

    Article  Google Scholar 

  34. Hudon C. The self-efficacy for managing chronic disease scale – French version: a validation study in primary care. Eur J Pers Centered Healthc. 2014;2:533–8.

    Article  Google Scholar 

  35. Chow SKY, Wong FK. The reliability and validity of the Chinese version of the short-form chronic disease self-efficacy scales for older adults. J Clin Nurs. 2014;23:1095–104.

    Article  Google Scholar 

  36. Hu H, Li G, Arao T. Validation of a Chinese version of the self-efficacy for managing chronic disease 6-item scale in patients with hypertension in primary care. ISRN Public Health. 2013;2013:1–6.

    Article  Google Scholar 

  37. Marconcin P, Tomé G, Carnide F, Yázigi F, Campos P, Pais S, et al. Translation, cultural adaptation and validation of the self-efficacy to manage chronic disease 6-item scale for european Portuguese. Acta Reumatol Port. 2021;46:15–22.

    Google Scholar 

  38. Eslami A, Daniali S-S, Mohammadi K, Reisi-Dehkordi N, Mostafavi-Darani F. Cultural adaptation and psychometric properties of the Persian version of self-efficacy in chronic disease patients. Iran J Nurs Midwifery Res. 2017;22:57–61.

    Article  Google Scholar 

  39. Allam MM, El-Zawawy HT, Ibrahim Ismail I, Ghazy RM. Cross-cultural reliability of an Arabic version of the self-efficacy for managing chronic disease 6-item scale in Arab patients with diabetes mellitus. Prim Care Diabetes. 2020;14:305–10.

    Article  Google Scholar 

  40. Schwarzer R, Jerusalem M. Generalized Self-Efficacy scale. In Weinman J, Wright S, Johnston M. Measures in health psychology: A user’s portfolio. Causal and control beliefs. Windsor: NFER-NELSON; 1995. pp. 35–7.

  41. Pendrill L. Man as a Measurement Instrument. NCSLI Meas. 2014;9:24–35.

    Article  Google Scholar 

  42. Tennant A, Conaghan PG. The Rasch measurement model in rheumatology: What is it and why use it? When should it be applied, and what should one look for in a Rasch paper? Arthritis Rheum. 2007;57:1358–62.

    Article  Google Scholar 

  43. Hagell P. Testing rating scale Unidimensionality using the Principal Component Analysis (PCA)/<i>t</i>-test protocol with the Rasch model: the primacy of theory over statistics. Open J Stat. 2014;04:456–65.

    Article  Google Scholar 

  44. Hagquist C, Andrich D. Recent advances in analysis of differential item functioning in health research using the Rasch model. 2017;8.

  45. Christensen KB, Makransky G, Horton M. Critical values for Yen’s Q 3: identification of local dependence in the Rasch model using residual correlations. Appl Psychol Meas. 2017;41:178–94.

    Article  Google Scholar 

  46. Smith RM. A comparison of methods for determining dimensionality in Rasch measurement. Struct Equ Model Multidiscip J Routledge. 1996;3:25–40.

    Article  Google Scholar 

  47. Gruber-Baldini AL, Velozo C, Romero S, Shulman LM. Validation of the PROMIS® measures of self-efficacy for managing chronic conditions. Qual Life Res Int J Qual Life Asp Treat Care Rehabil. 2017;26:1915–24.

    Article  Google Scholar 

  48. Petrillo J, Cano SJ, McLeod LD, Coon CD. Using classical test theory, item response theory, and Rasch measurement theory to evaluate patient-reported outcome measures: a comparison of worked examples. Value Health. 2015;18:25–34.

    Article  Google Scholar 

  49. Cano SJ, Pendrill LR, Melin J, Fisher WP. Towards consensus measurement standards for patient-centered outcomes. Measurement. 2019;141:62–9.

    Article  Google Scholar 

  50. Salzberger T, Cano S, Abeth-Webb L, Afolalu E, Chrea C, Weitkunat R, et al. Addressing traceability of self-reported dependence measurement through the use of crosswalks. Measurement. 2021;181:109593.

    Article  Google Scholar 

  51. Alsén S, Ali L, Ekman I, Fors A. Facing a blind alley - Experiences of stress-related exhaustion: a qualitative study. BMJ Open. 2020;10:e038230.

    Article  Google Scholar 

  52. Grossi G, Perski A, Osika W, Savic I. Stress-related exhaustion disorder – clinical manifestation of burnout? A review of assessment methods, sleep impairments, cognitive disturbances, and neuro-biological and physiological changes in clinical burnout. Scand J Psychol. 2015;56:626–36.

    Article  Google Scholar 

  53. Andela CD, Scharloo M, Ramondt S, Tiemensma J, Husson O, Llahana S, et al. The development and validation of the Leiden Bother and Needs Questionnaire for patients with pituitary disease: the LBNQ-Pituitary. Pituitary. 2016;19:293–302.

    Article  Google Scholar 

  54. Andela CD, Scharloo M, Pereira AM, Kaptein AA, Biermasz NR. Quality of life (QoL) impairments in patients with a pituitary adenoma: a systematic review of QoL studies. Pituitary. 2015;18:752–76.

    Article  Google Scholar 

  55. Wallström S, Ekman I. Person-centred care in clinical assessment. Eur J Cardiovasc Nurs. 2018;17:576–9.

    Article  Google Scholar 

  56. Busvold KIH, Bondevik H. Medically unexplained physical symptoms, misunderstood and wrongly treated? A semiotic perspective on chronic pain. Physiother Theory Pract Taylor & Francis. 2018;34:411–9.

    Article  Google Scholar 

  57. Fors A, Ulin K, Cliffordson C, Ekman I, Brink E. The Cardiac Self-Efficacy Scale, a useful tool with potential to evaluate person-centred care. Eur J Cardiovasc Nurs. 2015;14:536–43.

    Article  Google Scholar 

Download references

Acknowledgements

This research would not have been possible without patient engagement and the extensive data collection within the two different interventional studies. We acknowledge the patients with pituitary tumors that participated in the study and Gudmundur Johannsson and Ann-Charlotte Olofsson from the Department of Endocrinology and Tobias Hallén and Thomas Skoglund from the Department of Neurosurgery at Sahlgrenska University Hospital responsible for the study including patients with pituitary tumors. Furthermore, we gratefully acknowledge the contributions of patients, the participating primary health care centers, managers and health care professionals to the study including patients on sick leave due to CMD.

Funding

Open access funding provided by University of Gothenburg. The PT-study was supported by grants from the Swedish state under agreement between the Swedish government and the county councils, the ALF-agreement (ALFGBG-719531), and The Healthcare Board, Västra Götaland Region in Sweden. The CMD project was supported by the Swedish Research Council for Health, Working Life and Welfare (2016–07418, 2017–00557, and 2019–01726), and was financed by grants from the Swedish state under the agreement between the Swedish government and the country councils, the ALF agreement (ALFGBG-772191 and ALFGBG-932659). This work was also supported by The University of Gothenburg Centre for Person-centred Care (GPCC), Sweden. GPCC is funded by the Swedish Government’s grant for Strategic Research Areas (Care Sciences) and the University of Gothenburg, Sweden.

Author information

Authors and Affiliations

  1. RISE Research Institutes of Sweden, Division Safety and Transport, Measurement Science and Technology, Gothenburg, Sweden

    Jeanette Melin

  2. Centre for Person-Centred Care (GPCC), University of Gothenburg, Gothenburg, Sweden

    Jeanette Melin, Andreas Fors, Sofie Jakobsson & Ida Björkman

  3. Institute of Health and Care Sciences, The Sahlgrenska Academy, University of Gothenburg, 40530, Gothenburg, SE, Sweden

    Andreas Fors, Sofie Jakobsson & Ida Björkman

  4. Region Västra Götaland, Research, Education, Development and Innovation, Primary Health Care, Gothenburg, Sweden

    Andreas Fors

  5. Institute of Neuroscience and Physiology, University of Gothenburg, Gothenburg, Sweden

    David Krabbe

  6. Neurocare, Sahlgrenska University Hospital, Gothenburg, Sweden

    David Krabbe

Authors
  1. Jeanette Melin
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Andreas Fors
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Sofie Jakobsson
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. David Krabbe
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Ida Björkman
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

I.B. and J.M. designed the present study and I.B. performed the cognitive interviews. A.F. is the project leader for the study including patients with CMD. S.J. coordinated the data collection for the study including patients with pituitary tumors. J.M. analyzed the data and interpreted it together with I.B., A.F., S.J. and D.K. I.B. and J.M. drafted the manuscript with critical input from A.F., S.J. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Ida Björkman.

Ethics declarations

Ethics approval and consent to participate

This study was approved by a Regional Ethical Review Board in Gothenburg, Sweden (approval reference 387–15 and 497–17, T 023–18, T 526–18). All participants received oral and written information and signed an informed consent.

Consent for publication

All authors fulfil the criteria for authorship and have read and approved the final version of this manuscript.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Melin, J., Fors, A., Jakobsson, S. et al. Self-Efficacy to Manage Chronic Disease (SEMCD) scale: translation and evaluation of measurement properties for a swedish version. Arch Public Health 81, 2 (2023). https://doi.org/10.1186/s13690-022-01022-x

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s13690-022-01022-x

Keywords

Archives of Public Health

ISSN: 2049-3258