Marc Saez

Methodological issues in studies of air pollution and daily counts of deaths or hospital admissions.

STUDY OBJECTIVE: To review the issues and methodologies in epidemiologic time series studies of daily counts of mortality and hospital admissions and illustrate some of the methodologies. DESIGN: This is a review paper with an example drawn from hospital admissions of the elderly in Cleveland, Ohio, USA. MAIN RESULTS: The central issue is control for seasonality. Both over and under control are possible, and the use of diagnostics, including plots, is necessary. Weather dependence is probably non-linear, and adequate methods are necessary to adjust for this. In Cleveland, the use of categorical variables for weather and sinusoidal terms for filtering season are illustrated. After control for season, weather, and day of the week effects, hospital admission of persons aged 65 and older in Cleveland for respiratory illness was associated with ozone (RR = 1.09, 95% CI 1.02, 1.16) and particulates (PM10 (RR = 1.12, 95% CI 1.01, 1.24), and marginally associated with sulphur dioxide (SO2) (RR = 1.03, 95% CI = 0.99, 1.06). All of the relative risks are for a 100 micrograms/m3 increase in the pollutant. CONCLUSIONS: Several adequate methods exist to control for weather and seasonality while examining the associations between air pollution and daily counts of mortality and morbidity. In each case, care and judgement are required.

Time-series analysis of air pollution and cause-specific mortality

Ten large European cities provided data on daily air pollution as well as mortality from respiratory and cardiovascular mortality. We used Poisson autoregressive models that controlled for trend, season, influenza epidemics, and meteorologic influences to assess the short-term effects of air pollution at each city. We then compared and pooled the city-specific results in a meta-analysis. The pooled relative risks of daily deaths from cardiovascular conditions were 1.02 [95% confidence interval (CI) = 1.01–1. 04] for a 50

Effect on Prices of the Attributes of Holiday Hotels: A Hedonic Prices Approach

g/m3 increment in the concentration of black smoke and 1.04 (95% CI = 1.01–1.06) for an increase in sulfur dioxide levels in western European cities. For respiratory diseases, these figures were 1.04 (95% CI = 1.021.07) and 1.05 (95% CI = 1.03–1.07), respectively. These associations were not found in the five central European cities. Eight-hour averages of ozone were also moderately associated with daily mortality in western European cities (relative risk = 1.02; 95% CI = 1.00–1.03 for cardiovascular conditions and relative risk = 1.06; 95% CI = 1.02–1.10 for respiratory conditions). Nitrogen dioxide did not show consistent relations with daily mortality. These results are similar to previously published data and add credence to the causal interpretation of these associations at levels of air pollution close to or lower than current European standards. (Epidemiology 1998; 9:495–503)

Air pollution and mortality in Barcelona.

To investigate the relationship between antimicrobial use and resistance in our hospital, we collected antimicrobial susceptibility and use data from existing microbiology laboratory and pharmacy databases for the period July 1st, 1991-December 31, 1998. The data was analyzed as time series and autoregressive integrated moving average (Box-Jenkins) and transfer function models were built. By using this method, we were able to demonstrate a temporal relationship between antimicrobial use and resistance, to quantify the effect of use on resistance and to estimate the delay between variations of use and subsequent variations in resistance. The results obtained for two antimicrobial-microorganism combinations: ceftazidime-gram-negative bacilli and imipenem-Pseudomonas aeruginosa, are shown as examples.

Case-Crossover Analysis of Air Pollution Health Effects: A Systematic Review of Methodology and Application

This article examines the effect on price of different characteristics of holiday hotels in the sun-and-beach segment. The effect on price is estimated under the hedonic function perspective by means of random effect models, known also as mixed or panel models. Some 82,000 prices were gathered between 1991 and 1998 from tour operator catalogues. The study reveals huge price differences between 4-star hotels and the rest, coupled with practically no difference between 1-star and 2-star hotels. Other attributes with a significant effect on price are town, hotel size, distance to the beach and availability of parking place. The results can assist hotel managers in shaping pricing and investment strategies.

Use of Poisson regression and Box–Jenkins models to evaluate the short-term effects of environmental noise levels on daily emergency admissions in Madrid, Spain

STUDY OBJECTIVES: Studies conducted in Barcelona reported a short term relation between daily air pollutant values and emergency department admissions for exacerbation of chronic obstructive pulmonary diseases and asthma. Air pollution in Barcelona is mainly generated by vehicle exhaust and is below the World Health Organization air quality guidelines. The acute relation between air pollution and mortality was assessed. DESIGN: Daily variations in total mortality, mortality in subjects older than 70 years, and cardiovascular and respiratory mortality were related with daily variations in air pollutants for the period 1985-91, using autoregressive Poisson regression models that allow to control for temperature, relative humidity and variables handling temporal and autoregressive patterns. MAIN RESULTS: Black smoke and sulphur dioxide (SO2) were related to total mortality (relative risks (RR) for 100 micrograms/m3 = 1.07 and 1.13 respectively), elderly mortality (RR = 1.06 and 1.13), cardiovascular mortality (RR = 1.09 and 1.14), and respiratory mortality (RR = 1.10 and 1.13); all the associations being statistically significant, except for respiratory mortality (p = 0.07). The association between SO2 and respiratory mortality was stronger in summer (1.24, p < 0.01) than in winter (1.08, p > 0.1). Oxidant pollutants (nitrogen dioxide and ozone) were positively related with elderly mortality (RR = 1.05 and 1.09, respectively) and cardiovascular mortality (RR = 1.07 and 1.09) during the summer (p < 0.05), but not during the winter. CONCLUSIONS: Current air pollutant levels were related to mortality in Barcelona. These results were consistent with similar studies in other cities and coherent with previous studies with emergency room admissions in Barcelona.

Air pollution and cardiovascular admissions association in Spain: results within the EMECAS project

Background Case-crossover is one of the most used designs for analyzing the health-related effects of air pollution. Nevertheless, no one has reviewed its application and methodology in this context. Objective We conducted a systematic review of case-crossover (CCO) designs used to study the relationship between air pollution and morbidity and mortality, from the standpoint of methodology and application. Data sources and extraction A search was made of the MEDLINE and EMBASE databases. Reports were classified as methodologic or applied. From the latter, the following information was extracted: author, study location, year, type of population (general or patients), dependent variable(s), independent variable(s), type of CCO design, and whether effect modification was analyzed for variables at the individual level. Data synthesis The review covered 105 reports that fulfilled the inclusion criteria. Of these, 24 addressed methodological aspects, and the remainder involved the design’s application. In the methodological reports, the designs that yielded the best results in simulation were symmetric bidirectional CCO and time-stratified CCO. Furthermore, we observed an increase across time in the use of certain CCO designs, mainly symmetric bidirectional and time-stratified CCO. The dependent variables most frequently analyzed were those relating to hospital morbidity; the pollutants most often studied were those linked to particulate matter. Among the CCO-application reports, 13.6% studied effect modification for variables at the individual level. Conclusions The use of CCO designs has undergone considerable growth; the most widely used designs were those that yielded better results in simulation studies: symmetric bidirectional and time-stratified CCO. However, the advantages of CCO as a method of analysis of variables at the individual level are put to little use.

Material deprivation and leading causes of death by gender: evidence from a nationwide small area study

The relationship between environmental factors and hospital admissions has usually been analysed without taking into account the influence of a factor closely related to traffic in big cities, that is, environmental noise levels. We analysed the relationship between environmental noise and emergency admissions, for all causes and specific causes in Madrid (Spain), for the study period 1995–1997, using two statistical methods for the analysis of epidemiological time series data: Poisson autoregressive models and Box–Jenkins (ARIMA) methodology. Both methods produce a clear association between emergency admissions for all and specific causes and environmental noise levels. We found very similar results from both methods for all and circulatory causes, but slightly different for respiratory causes. Around 5% of all emergency admissions can be attributed to high noise levels, with a lower figure for specific causes. Current levels of environmental noise have a considerable epidemiological impact on emergency admissions in Madrid. A reduction of environmental noise levels could be accompanied by a possible reduction in the number of emergency admissions.

The EMECAM project: a multicentre study on air pollution and mortality in Spain: combined results for particulates and for sulfur dioxide

Objective: To evaluate the short term effect of air pollution on cardiovascular admissions in 14 Spanish cities Methods: The period under study was from 1995 to 1999. Daily emergency admissions for all cardiovascular diseases (CVD) and heart diseases (HD) were obtained from hospital records, and the corresponding daily levels of particulates, SO2, NO2, CO, and ozone were recorded. The magnitude of association was estimated using Poisson generalised additive models controlling for confounding and overdispersion. For each cause, lagged effects, up to three days, of each pollutant were examined and combined estimates were obtained. For ozone the analyses were restricted to the warm period. One and two pollutant models were performed. Results: Associations were more consistent in lag 0 (concurrent day) and 1 (lag 0–1), except in the case of ozone where there was a more delayed relation (lag 2–3). For combined estimates an increase of 10 μg/m3 in the PM10 levels in lag 0–1 was associated with an increase of 0.9% (95% CI: 0.4 to 1.5%) in the number of hospital admissions for CVD, and 1.6% (0.8 to 2.3%) for HD. For ozone the corresponding estimates for lag 2–3 were 0.7% (0.3 to 1.0) for CVD, and 0.7% (0.1 to 1.2) for HD. An increase of 1 mg/m3 in CO levels was associated with an increase of 2.1% (0.7 to 3.5%) in CVD admissions, and 4.2% (1.3 to 7.1%) in HD admissions. SO2 and NO2 estimates were more sensitive in two pollutant models Conclusions: A short term association between increases in daily levels of air pollutants and the number of daily admissions for cardiovascular diseases, with specificity for heart diseases, has been described in Spanish cities.