The development and initial evaluation of a weight specific quality of life instrument

Abstract

SETTING THE SCENE: Obesity can have debilitating effects on a person’s health-related quality of life (HRQoL). Participant-reported HRQoL should be assessed in addition to Body Mass Index (BMI) to evaluate the effectiveness of obesity interventions.
AIMS: This programme of work aimed to a) assess the need for a new weight-related quality of life (WRQoL) scale, b) develop a WRQoL scale with input from UK samples and c) conduct the initial psychometric evaluation of the new WRQoL scale.
METHODS:
Assessing the need for a WRQoL scale: A systematic review was undertaken. Medline, EMBASE, CINAHL and PsycINFO were searched for literature published between 1974-2018 detailing the development and psychometric evaluation of adult WRQoL scales. Two independent researchers screened the articles by title, abstract and full text to identify relevant papers. Data was extracted from the articles using the COSMIN checklist. Additionally, cognitive debriefing interviews were conducted to test content/face validity of the
current ‘gold standard’ WRQoL measure.
Developing a WRQoL scale: A qualitative approach was used to conduct and analyse one-to-one interviews over two phases (preliminary interviews and item generation interviews). Adults with experience of weight issues were recruited via opportunity sampling at community locations. All participants had their BMI and waist circumference measured. Findings from the
item generation interviews were used to generate items in expert panel meetings. Cognitive interviews assessed the face validity of the new instrument.
Initial evaluation of the new scale: Exploratory factor analysis, internal consistency, known groups comparisons, concurrent validity and test-retest reliability were conducted on the draft scale.
RESULTS:
Assessing the need for a WRQoL scale: The systematic review identified 9886 articles which were screened initially by title, then by abstract (n = 966) and finally by full text (n = 426). Twenty-eight articles contained information regarding the development or psychometric evaluation of 17 WRQoL scales. No instrument had evidence for all psychometric properties, demonstrating the need for a new WRQoL scale. The cognitive debriefing interviews highlighted issues with the content validity of the most used WRQoL scale.
Developing a WRQoL scale: The preliminary interviews (n = 10) enabled the development of an interview schedule for the item generation interviews. Data from the item generation interviews (n = 48) were used to draft a 31-item instrument during discussions with an expert panel. The initial items covered six themes identified in the item generation interviews; physical health, mobility, clothing, food, feeling towards themselves and psychosocial experience.
Initial evaluation of the new scale: The final draft scale contained 29 items covering four domains (confidence with self, getting around, feeling valued and weight stigma). Good internal consistency, test-retest reliability, and concurrent validity was shown, and the new scale was able to discriminate between BMI groups.
CONCLUSION: This programme of work has contributed to knowledge by a) providing a detailed evaluation of existing WRQoL scales; b) providing a clear description of the impact of obesity on everyday life, from people who have experienced weight issues;
c) developing and preliminary evaluating a WRQoL scale with input from the population it is intended for. After further development and psychometric work, the instrument will be able to describe and measure changes in WRQoL in community and clinical populations.