The study sample included 36,000 individuals living in Ontario, Canada and who were randomly selected to participate in the National Population Health Survey in 1996 and 24,000 individuals in Ontario who were randomly selected to participate in the Canadian Community Health Survey in 2001. Design: A cross sectional and longitudinal analysis was employed.

Methods: This study used individual level data. We adjusted for the endogeneity by defining the key independent variables, variation in hospital and community nursing services use, as the degree to which each survey respondent used more or less nursing services than expected given their level of “need.”

1996 data was used to estimate regression models of the effect of proxies of need on the use of nursing services. We employed a double-hurdle model whereby nursing services use was modeled by estimating two regression equations for each dependent variable. To model the effect of variation in nursing service use on health status in the context of individual differences in needs, regression models were employed using the 2001 data.

Findings: In predicting self-report health status and HUI scores, the difference between expected and actual use is statistically significant (with a very small p-value), but not practically significant (with a very small coefficient). Self-reported health status was generally lower for older patients and females were less likely than males to report better health. Respondents not living in metropolitan areas were more likely to report better health than respondents living in urban non-metro areas. Rspondents with a university degree were most likely to be healthy than those with less education.

Back to Health Human Resource Planning
Back to 15th International Nursing Research Congress
Sigma Theta Tau International
July 22-24, 2004