Self-efficacy instruments for patients with chronic diseases suffer from methodological limitations - a systematic review

Background: Measurement of self-efficacy requires carefully developed and validated instruments. It is currently unclear whether available self-efficacy instruments for chronic diseases fulfill these requirements. Our aim was to systematically identify all existing self-efficacy scales for five major chronic diseases and to assess their development and validation process. Methods: We conducted a systematic literature search in electronic databases (MEDLINE, PSYCHINFO, and EMBASE) to identify studies describing the development and/or validation process of self-efficacy instruments for the five chronic diseases diabetes, chronic obstructive pulmonary disease (COPD), asthma, arthritis, and heart failure. Two members of the review team independently selected articles meeting inclusion criteria. The self-efficacy instruments were evaluated in terms of their development (aim of instrument, a priori considerations, identification of items, selection of items, development of domains, answer options) and validation (test-retest reliability, internal consistency reliability, validity, responsiveness) process. Results: Of 584 potentially eligible papers we included 25 (13 for diabetes, 5 for asthma, 4 for arthritis, 3 for COPD, 0 for heart failure) which covered 26 different self-efficacy instrument versions. For 8 instruments (30.8%), the authors described the aim before the scales were developed whereas for the other instruments the aim was unclear. In one study (3.8%) a priori considerations were specified. In none of the studies a systematic literature search was carried out to identify items. The item selection process was often not clearly described (38.5%). Test-retest reliability was assessed for 9 instruments (34.6%), validity using a correlational approach for 18 (69.2%), and responsiveness to change for 3 (11.5%) instruments. Conclusion: The development and validation process of the majority of the self-efficacy instruments had major limitations. The aim of the instruments was often not specified and for most instruments, not all measurement properties that are important to support the specific aim of the instrument (for example responsiveness for evaluative instruments) were assessed. Researchers who develop and validate self-efficacy instruments should adhere more closely to important methodological concepts for development and validation of patient-reported outcomes and report their methods more transparently. We propose a systematic five step approach for the development and validation of self-efficacy instruments.