A common concentration-response function based on the results applying lags

Abstract Introduction. Estimating the impact of short-term exposure on health outcomes needs knowledge both profile and magnitude relative risks. This motivates constructions practical reliable concentration-response functions (C-RFs). Aim. To define a method finding parametric function whose adjustable parameters can be tuned by data-driven well established routines. Material methods. Mortality data for period from 1987 to 2015 (10,592 consecutive days) in Montreal, Canada, are used illustrative purposes. Exposure ambient ozone measured its concentration levels is considered risk. Concentration-response built using statistical modelling, conditional Poisson regression, natural spline technique, rudimentary hierarchical clustering. The case-crossover design applied fit model C-RF mortality consisting daily counts non-accidental deaths. Results. Log-linear models were computed concentrations cofactors lagged 0 7 days; results statistically significant within this range lags. effectiveness fitting was confirmed tests. Digital routines created perform all computational tasks; software codes (written R platform) included. specifying current responses cumulative several previous days obtained exposures. Conclusions. proposed estimation appears effective producing results. constructed monotonic non-decreasing.