Monitoring the health effects of disasters and environmental hazards : The primary care database as a tool to inform public decision-making

Introduction

Public leaders are responsible for minimizing the potential negative effects of disasters and environmental hazards on the health and well-being of exposed citizens. In order to decide upon appropriate measures, decision makers depend heavily on the availability of reliable information on the human health consequences of exposure to specific events, substances or agents. Disaster researchers and epidemiologists can play a crucial role in monitoring activities. The objective of this contribution is to describe how the research possibilities of a particular data infrastructure, the Nivel Primary Care Database (PCD), were used in the context of three disasters and four environmental hazards cases in the Netherlands.

Methods

The PCD continuously collects and processes electronic health record data (e.g. healthcare utilization, symptoms, diagnoses and prescriptions; coded according to the International Classification of Primary Care) of (i) more than 500 general practices with 1.6 million registered patients (10% of the Dutch population); and (ii) electronic health records of primary care out-of-hours services with a total catchment area of 11.5 million. The PCD can overcome shortcomings from other data sources like body materials, interviews and questionnaires, and allows longitudinal and cross-classified multilevel analyses with patients nested in GPs and postal codes. The three disaster cases involve exposure to the fireworks disaster in Enschede (2000; 23 people killed, 950 wounded: data range 1999–2005); the Volendam New Year's fire (2001; 14 children killed, 350 wounded: data range 2000–2006); and a shooting incident in a shopping mall in Alphen aan den Rijn (2011; 7 people killed, 17 wounded; data range 2010–2011). Environmental hazards cases include: living nearby livestock (especially goat farms) (data range 2007–2017), living in a fruit production area where pesticides are used (data range 2013–2016), living in the vicinity of an industrial area with chemical companies (data range 2006–2008 and 2013–2015).

Results

The disasters and major crises were accompanied by a significant increase in health care utilization and psychological problems among affected groups and illustrate a pattern of normalization over time. Particular risk groups could be followed through time e.g. non-western immigrants, people with existing psychological problems, and children with or without burns. Industrial and agricultural activity was associated repeatedly with health problems related to airways and lungs (the association between goat farms and pneumonia received special attention). In the pesticides case, PCD data could be combined with a complementary survey providing information on self-perceived symptoms and perceptions, which confirmed that participants mostly did not attribute their health problems to pesticides in fruit production. Almost all of the studies provided information on the prevalence of medically unexplained symptoms and chronic health problems. In one industrial case, municipalities close to the industrial zone could be compared with other communities in the same region as well as a comparison group of patients in a living environment with a similar level of urbanization and industrial activity.

Conclusions

A major challenge for disaster health researchers and epidemiologists is to understand the association between population health on the one hand, and exposure levels and other factors on the other. The plausibility of this association is of crucial interest to decision makers. Several of the cases illustrate the potential of registry-based monitoring for decision-making. Since the fireworks disaster, the PCD's national coverage and the availability of other geographical datasets have increased substantially. Further optimization is an ongoing process that will increase the potential for future monitoring of the health effects of disasters and environmental hazards.