The impact of fuel poverty upon self-reported health status among the low-income population in Europe

Background Cross-sectional studies with binary outcomes analyzed by logistic regression are frequent in the epidemiological literature. However, the odds ratio can importantly overestimate the prevalence ratio, the measure of choice in these studies. Also, controlling for confounding is not equivalent for the two measures. In this paper we explore alternatives for modeling data of such studies with techniques that directly estimate the prevalence ratio. Methods We compared Cox regression with constant time at risk, Poisson regression and log-binomial regression against the standard Mantel-Haenszel estimators. Models with robust variance estimators in Cox and Poisson regressions and variance corrected by the scale parameter in Poisson regression were also evaluated. Results Three outcomes, from a cross-sectional study carried out in Pelotas, Brazil, with different levels of prevalence were explored: weight-for-age deficit (4%), asthma (31%) and mother in a paid job (52%). Unadjusted Cox/Poisson regression and Poisson regression with scale parameter adjusted by deviance performed worst in terms of interval estimates. Poisson regression with scale parameter adjusted by χ2 showed variable performance depending on the outcome prevalence. Cox/Poisson regression with robust variance, and log-binomial regression performed equally well when the model was correctly specified. Conclusions Cox or Poisson regression with robust variance and log-binomial regression provide correct estimates and are a better alternative for the analysis of cross-sectional studies with binary outcomes than logistic regression, since the prevalence ratio is more interpretable and easier to communicate to non-specialists than the odds ratio. However, precautions are needed to avoid estimation problems in specific situations.

Health planners and policy makers are increasingly asking for a feasible method to identify vulnerable persons with the greatest health needs. We conducted a systematic review of the association between a single item assessing general self-rated health (GSRH) and mortality. Systematic MEDLINE and EMBASE database searches for studies published from January 1966 to September 2003. Two investigators independently searched English language prospective, community-based cohort studies that reported (1) all-cause mortality, (2) a question assessing GSRH; and (3) an adjusted relative risk or equivalent. The investigators searched the citations to determine inclusion eligibility and abstracted data by following a standardized protocol. Of the 163 relevant studies identified, 22 cohorts met the inclusion criteria. Using a random effects model, compared with persons reporting "excellent" health status, the relative risk (95% confidence interval) for all-cause mortality was 1.23 [1.09, 1.39], 1.44 [1.21, 1.71], and 1.92 [1.64, 2.25] for those reporting "good,"fair," and "poor" health status, respectively. This relationship was robust in sensitivity analyses, limited to studies that adjusted for co-morbid illness, functional status, cognitive status, and depression, and across subgroups defined by gender and country of origin. Persons with "poor" self-rated health had a 2-fold higher mortality risk compared with persons with "excellent" self-rated health. Subjects' responses to a simple, single-item GSRH question maintained a strong association with mortality even after adjustment for key covariates such as functional status, depression, and co-morbidity.