Assessment of environmental health risks is feasible by secondary use of administrative registers

Abstract

National register systems include detailed individual-level information. In the Nordic countries, these data sources include personal identification number, which can be used for linkages between registers. A case study on the effects of possible hazardous waste on the former and current residents of Myllypuro in Helsinki, Finland, is presented to assess if the utilisation of pre-collected official health register information is feasible in environmental health research. National register information was used as the primary data source, since large-scale health examination studies are seldom feasible and the use of surveys may result in biased information. The exposure data were based on residence information from the Central Population Register and the outcome data came from three national population and health registers. No evidence of health problems was found. The use of administrative register data was feasible, since the main prerequisites of epidemiological studies – the enumeration of exposed population without selection bias, the tracking of exposed population without loss to follow-up and the formation of different exposure measures – were reached. The small size of study sample and the rarity of several outcome measures impeded the analysis and the evaluation of clinical and public health aspects of the main findings.

Keywords:

• Administrative registers
• birth outcomes
• data linkage
• environmental risk
• epidemiology
• register-based research

Acknowledgements

I thank Dr Annukka Ritvanen (THL) and Dr Antti Pönkä (City of Helsinki) for their valuable help for this study.

Declaration of Interest: The author reports no conflict of interest.

An earlier version of this paper was presented in the International Statistical Institute (ISI) and International Association for Official Statistics (IAOS) conference in Sydney, Australia, in April 2005.

The article was submitted to Journal of Official Statistics in 30 May 2008, accepted for publication in 10 February 2010. The article was lost during the editorial process, and the new Editors-of-Chief of the journal rejected the paper in 27 April 2012.