WHO long form scoring, reliability, validity and norms for New Zealand : a thesis presented in fulfilment of the requirements for the degree of Master of Public Health at Massey University, Wellington Campus, New Zealand
Background Self-reported health measures provide information about a wider range of health outcomes than objective measures of health status, such as mortality and hospitalisation rates. National health surveys play a role in monitoring population health. The New Zealand Health Monitor (NZHM) is the organised, co-ordinated and integrated survey programme of the Ministry of Health in New Zealand. The New Zealand Health Survey (NZHS) is one of the chief surveys of the NZHM. One of the categories of information collected in the NZHM is health outcomes, and within this there is the subcategory of health status. The International Classification of Functioning and Disability (ICF) provides the framework to describe the critical elements of non-fatal health outcomes captured by health status instruments. NZHM is to collect data on most if not all of these 21 ICF dimensions. The WHO Long Form was developed as the health module in the WHO Multi-country Survey Study. The WHO Long Form is made up of 20 health domains, some overlapping with the eight SF-36 domains. The WHO Long Form did not have a set scoring system for scales, unlike the SF-36 instrument. The SF-36 has been previously tested and validated in New Zealand in the 1996/97 NZHS. Methods The 2002/03 NZHS used a complex sample design. A total of 12,929 people responded to the survey, with 12,529 respondents being included in the CURF dataset available for research. The health status section of the 2002/03 NZHS measures health-related quality of life (HRQL) covered 16 health and health-related domains. The questions were derived from the SF-36 and the WHO Long Form questionnaire on health status. The health domains covered in the 2002/03 NZHS were general health, vision, hearing, digestion, breathing, pain, sleep, energy and vitality, understanding, communication, physical functioning, self-care. The health-related domains covered in the 2002/03 NZHS were mental health, role-physical and role-emotional (usual activities), and social functioning. There were five key aims specific to the current thesis. First, to group the WHO Long Form items in the 2002/03 NZHS into scales for each health domain and develop standard scoring protocols for each scale. Second, to test the reliability of the scales using standard psychometric tests for the total NZ population and for major population subgroups. Third, to test the validity of the scales using the standard psychometric tests for the total NZ population and for major population subgroups. Fourth, to construct norms for the WHO Long Form scales for the NZ population. And finally, to provide recommendations for the health status component of future NZ health surveys. Results In summary, this thesis developed a method for producing scale scores for domains of health not previously measured in New Zealand Health Surveys, providing greater coverage of domains from the ICF. There were virtually no missing data for all items and subgroups within the questions used to develop the scales. The scaling approach was consistent with that for the SF-36, allowing the new scales to be presented alongside the SF-36 scales. All scales for the total population and major population subgroups met the required criterion for satisfactory psychometric properties, with the exception of digestion and bodily excretions scale. For the digestion and bodily excretions scale, the Cronbach's alpha was lower than that required for between group comparisons. The composite physical functioning and social functioning scales performed no better than the existing SF-36 scales and were highly correlated with these scales. Conclusion Notwithstanding the limitations of this study, key findings of interest are that the new WHO Long Form questions can be used to form scales that cover physical functioning, social functioning, vision, hearing, digestion and bodily excretions, breathing, self-care, understanding, communication and sleep. The majority of the questions and scales work for the NZ population and subgroups. All but one of the scales, digestion and bodily excretions, have satisfactory psychometric properties for the total population and major subpopulation groups of interest. The respondent burden is an important consideration for the NZHS, thus it cannot be argued that enough is gained from adding questions to the physical functioning and Social Functioning domains, thus it would be recommended that the SF-36 scales are used to measure there two domains of health. The new WHO Long Form scales can now be presented alongside the SF-36 scales and used in future analyses looking at interrelationships between factors such as health risk and health status.