The impact of ambient air pollution on hospital admissions

Abstract

Ambient air pollution is the environmental factor with the most significant impact on human health. Several epidemiological studies provide evidence for an association between ambient air pollution and human health. However, the recent economic literature has challenged the identification strategy used in these studies. This paper contributes to the ongoing discussion by investigating the association between ambient air pollution and morbidity using hospital admission data from Switzerland. Our identification strategy rests on the construction of geographically explicit pollution measures derived from a dispersion model that replicates atmospheric conditions and accounts for several emission sources. The reduced form estimates account for location and time fixed effects and show that ambient air pollution has a substantial impact on hospital admissions. In particular, we show that \({\text{SO}}_{2}\) and \({\text{NO}}_{2}\) are positively associated with admission rates for coronary artery and cerebrovascular diseases while we find no similar correlation for PM10 and \({\text{O}}_{3}\). Our robustness checks support these findings and suggest that dispersion models can help in reducing the measurement error inherent to pollution exposure measures based on station-level pollution data. Therefore, our results may contribute to a more accurate evaluation of future environmental policies aiming at a reduction of ambient air pollution exposure.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2

Notes

  1. 1.

    For instance, the study by He et al. [18] relies on Chinese city-level mortality data. Because a city in China can be relatively large and heterogeneous, the estimated level of pollution exposure may be significantly different to the true level of pollution exposure. Schlenker and Walker [26] conduct their analysis using zip code level data for California. The average size of a zip code in California is above 37,000 inhabitants, ranging between 11,000 and more than 100,000 inhabitants.

  2. 2.

    We perform a number of robustness checks to ensure that the reassignment method does not affect the identification. For instance, we use gridded housing data from the Swiss land register to accomplish the recoding of location information. These results are similar to the estimates obtained with our baseline specification.

  3. 3.

    The main topographical areas are the Swiss plateau (North of the Alps), the Basel region with specific climate conditions due to the Rhine valley, the Alpine region (valley floors of all alpine valleys), and the remaining parts. Additional information on these regions are provided in Heldstab et al. [19].

  4. 4.

    Note that there are exceptions since some papers based on the inverse distance method can also find high correlation between observed and predicted pollutant levels (e.g., [7]).

  5. 5.

    Other moments of the pollution distribution function (e.g., annual median, minimum and maximum) could be also relevant for hospital admissions. However, the use of other moments of pollution exposure is limited by a data protection agreement between the FOEN and external data providers.

  6. 6.

    As for possible border effects, note that the dispersion model already accounts for these effects by construction, since it considers emission sources in adjacent regions.

  7. 7.

    Ideally, we would account for access to hospital care with a measure of distance to the nearest hospital. However, such data are not available for the entire study period.

  8. 8.

    Another approach would be to include spatial effects in the regression model. For this reason, we also estimate spatial lag panel models for count data (see, e.g., [4]). The spatial estimates are very similar to the results of our main model and indicate that spatial lags are of minor relevance for most causes of hospital admissions when using the dispersion model pollution measures.

  9. 9.

    Switzerland is a federal state of 26 cantons. Each canton has its own constitution, legislature, and government. Among others, the cantons are responsible for healthcare services, welfare, law enforcement, and public education.

  10. 10.

    For instance, Switzerland has recently introduced a new hospital financing system to promote cost-effective health care. Although the system was simultaneously introduced in all cantons, the reimbursement rates for medical treatment are different between cantons.

  11. 11.

    We are aware that several studies in the health economics literature use admissions per population as the outcome variable. However, the absolute number of admissions is more appropriate in this context because we can use a count-data model that reflects the data generating process of hospital admissions due to pollution exposure.

  12. 12.

    We do not report the estimates of the control variables because of space limitations. The table shows the estimates of 14 (\(7 \times 2\)) regressions. The estimates including all covariates are available upon request from the authors.

  13. 13.

    Note that the effects of different pollutants are comparable since they are all measured in \(\upmu \hbox {g}/\hbox {m}^{2}\).

  14. 14.

    The negative parameter estimates for some types of pollution could also be caused by insufficient variation in the measures of exposure. For instance, since there are only a few \({\text{SO}}_{2}\) monitoring sites, the within variation in pollution exposure is limited, inducing collinearity between the fixed effects and the measures of pollution exposure.

  15. 15.

    As discussed in the literature, in this case the measurement error is likely non-classical. This implies that the covariance tends to be negative (see, e.g., [2]).

  16. 16.

    As a measure of explanatory power, we use the Pseudo R-squared value, which is defined as the squared correlation between predicted and observed count outcome.

References

  1. 1.

    Beatty, T.K., Shimshack, J.P.: Air pollution and children’s respiratory health: a cohort analysis. J. Environ. Econ. Manag. 67(1), 39–57 (2014)

    Article  Google Scholar 

  2. 2.

    Bound, J., Krueger, A.B.: The extent of measurement error in longitudinal earnings data: do two wrongs make a right? J. Lab. Econ. 9(1), 1–24 (1991)

    Article  Google Scholar 

  3. 3.

    Cameron, C., Miller, D.: A practitioner’s guide to cluster-robust inference. J. Hum. Resour. 50(2), 317–372 (2015)

    Article  Google Scholar 

  4. 4.

    Cameron, C., Trivedi, P.: Regression analysis of count data. Cambridge University Press, Cambridge (2013)

    Book  Google Scholar 

  5. 5.

    Cesar, A.C.G., Nascimento, L.F.C., de Carvalho, J.A., Jr.: Association between exposure to particulate matter and hospital admissions for respiratory disease in children. Rev. Saude Publica 47(6), 1209–1212 (2013)

    Article  PubMed  PubMed Central  Google Scholar 

  6. 6.

    Chay, K.Y., Greenstone, M.: The impact of air pollution on infant mortality: evidence from geographic variation in pollution shocks induced by a recession. Q. J. Econ. 118(3), 1121–1167 (2003)

    Article  Google Scholar 

  7. 7.

    Currie, J., Neidell, M.: Air pollution and infant health: what can we learn from California’s recent experience? Q. J. Econ. 120(3), 1003–1030 (2005)

    Google Scholar 

  8. 8.

    Currie, J., Neidell, M., Schmieder, J.F.: Air pollution and infant health: lessons from New Jersey. J. Health Econ. 28(3), 688–703 (2009)

    Article  PubMed  PubMed Central  Google Scholar 

  9. 9.

    Dehbi, H.-M., Blangiardo, M., Gulliver, J., Fecht, D., de Hoogh, K., Al-Kanaani, Z., Tillin, T., Hardy, R., Chaturvedi, N., Hansella, A.L.: Air pollution and cardiovascular mortality with over 25 years follow-up: a combined analysis of two British cohorts. Environ. Int. 99, 275–281 (2017)

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. 10.

    Deryugina, T., Heutel, G., Miller, N., Molitor, D., Reif, J.: The mortality and medical costs of air pollution: evidence from changes in wind direction. NBER Working Paper 22796 (2016)

  11. 11.

    European Environmental Agency: Air Quality in Europe—2015 Report. Publication Office of the European Union, Luxembourg (2015)

    Google Scholar 

  12. 12.

    Federal Council: Ordinance on air pollution control (2016). https://www.admin.ch/opc/en/classified-compilation/19850321/index.html. Accessed date 15 Nov 2016

  13. 13.

    FOEN: Air pollution maps (2016). http://www.bafu.admin.ch/luft/luftbelastung/schadstoffkarten. Accessed date 15 Nov 2016

  14. 14.

    FSO: Area statistics (2016a). https://www.bfs.admin.ch/bfs/de/home/statistiken/raum-umwelt/erhebungen/area.html. Accessed date 15 Nov 2016

  15. 15.

    FSO: Medical statistics of the hospitals (2016b). https://www.bfs.admin.ch/bfs/de/home/statistiken/gesundheit/erhebungen/ms.html. Accessed date 15 Nov 2016

  16. 16.

    Gong, G., Samaniego, F.J.: Pseudo maximum likelihood estimation: theory and applications. Ann. Stat. 9(4), 861–869 (1981)

    Article  Google Scholar 

  17. 17.

    Gourieroux, C., Monfort, A., Trognon, A.: Pseudo maximum likelihood methods: applications to poisson models. Econometrica 52(3), 701–720 (1984)

    Article  Google Scholar 

  18. 18.

    He, G., Fan, M., Zhou, M.: The effect of air pollution on mortality in China: evidence from the 2008 Beijing Olympic Games. J. Environ. Econ. Manag. 79, 18–39 (2016)

    Article  Google Scholar 

  19. 19.

    Heldstab, J., Leippert, F., Wüthrich, P., Künzli, T., Stampfli, M.: PM10 and PM2.5 ambient concentrations in Switzerland. FOEN, Bern (2013)

  20. 20.

    Knittel, C.R., Miller, D.L., Sanders, N.J.: Caution, drivers! Children present: traffic, pollution, and infant health. Rev. Econ. Stat. 98(2), 350–366 (2016)

    Article  Google Scholar 

  21. 21.

    Lagravinese, R., Moscone, F., Tosetti, E., Lee, H.: The impact of air pollution on hospital admissions: evidence from Italy. Region. Sci. Urban Econ. 49, 278–285 (2014)

    Article  Google Scholar 

  22. 22.

    Luechinger, S.: Air pollution and infant mortality: a natural experiment from power plant desulfurization. J. Health Econ. 37, 219–231 (2014)

    Article  PubMed  Google Scholar 

  23. 23.

    Moretti, R., Neidell, M.: Pollution, health, and avoidance behavior evidence from the ports of Los Angeles. J. Hum. Resour. 46(1), 154–175 (2011)

    Google Scholar 

  24. 24.

    Newell, K., Kartsonaki, C., Lam, K.B.H., Kurmi, O.: Cardiorespiratory health effects of gaseous ambient air pollution exposure in low and middle income countries: a systematic review and meta-analysis. Environ. Health 17, 41 (2018)

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  25. 25.

    Sanders, N.J., Stoecker, C.: Where have all the young men gone? Using sex ratios to measure fetal death rates. J. Health Econ. 41, 30–45 (2015)

    Article  PubMed  Google Scholar 

  26. 26.

    Schlenker, W., Walker, W.R.: Airports, air pollution, and contemporaneous health. Rev. Econ. Stud. 83(2), 768–809 (2016)

    Article  Google Scholar 

  27. 27.

    WHO: Air Quality Guidelines: Particulate Matter, Ozone, Nitrogen Dioxide and Sulfur Dioxide. WHO Regional Office for Europe, Copenhagen (2005)

    Google Scholar 

  28. 28.

    WHO Regional Office for Europe and OECD: Economic Cost of the Health Impact of Air Pollution in Europe: Clean Air, Health and Wealth. WHO Regional Office for Europe, Copenhagen (2015)

  29. 29.

    Wooldridge, J.: Handbook of applied econometrics: quasi-likelihood methods for count data, Vol. II. Blackwell Publishing Inc, Oxford (1999)

    Google Scholar 

Download references

Acknowledgements

We thank the Swiss Federal Office for the Environment (FOEN) for providing the ambient air pollution data and the Swiss Federal Statistical Office (FSO) for making available the hospitalization data. We are grateful to one anonymous reviewer and the editor for constructive suggestions.

Author information

Affiliations

Authors

Corresponding author

Correspondence to Sandro Steinbach.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary material 1 (PDF 3107 KB)

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Filippini, M., Masiero, G. & Steinbach, S. The impact of ambient air pollution on hospital admissions. Eur J Health Econ 20, 919–931 (2019). https://doi.org/10.1007/s10198-019-01049-y

Download citation

Keywords

  • Ambient air pollution
  • Dispersion model
  • Hospital admissions
  • Count panel data

JEL Classification

  • I10
  • Q51
  • Q53