Table 1

Analysis of methods for quantification in health impact assessment

	Exposure impact assessment	Outcome assessment	Health outcomes
Waste to energy facility (Hallenbeck, 1995)⁶	Pollution levels based upon worst case emission factors derived from existing similar facilities; computerised air pollution dispersal model.	Carcinogens: maximum annual average ground level concentrations were used to compute maximum theoretical cancer risks due to lifetime inhalation, using risk factors from USEPA (that is, quantitative risk assessment).	Lifetime cancer risks resulting from each pollutant separately (for example, maximum theoretical cancer risk due to lifetime exposure to PCBs: 10⁻⁸).
Freeway Melbourne (Dunt, 1995)⁷	Literature study; baseline data on accidents, pollution and noise levels; models predicting traffic flows.	Literature study; model to predict traffic accidents; extrapolation on basis of data on current percentage of road traffic accidents with casualties.	Road traffic accidents (−154) and resulting injuries (−100 to 110 separate injuries).
Law on alcohol and hotel and catering industry (Zwart 1997)⁸	A (separate) economic effect analysis gave an estimate of the change in revenue from sale of alcohol. This was interpreted as a decrease in per capita consumption.	Ledermann formula to estimate the number of excessive drinkers on basis of per capita consumption.	Excessive drinkers (−11000 persons drinking >7 E alcohol/day; −5000 persons with >11 E/day; −3000 persons with >15 E/day).
HIA on tobacco policy (Mooy, 1998)⁹	Economic literature gave an estimate of the change in amount sold as a function of price (price elasticity) of tobacco. This was equated with a corresponding decrease in smoking prevalence.	Computer simulation model (PREVENT), a dynamic multi-state life table based model.	Deaths from lung cancer in absolute numbers prevented per year; life years gained (both graphically depicted as numbers per calendar year).
Extension of waste disposal site (Fehr, 1999)^10,¹¹	Expected emissions based on extrapolation of measurements at other sites, supplemented by data from technical literature; pollution dispersal models for chemicals including eight carcinogens (by external agency); noise from technical statement.	Quantitative risk assessment (QRA) for carcinogens using health risk assessment computer program.	Individual additional lifetime cancer risk (graphically depicted per village and pathway); estimated additional cancer burden (<0.01 additional cases in 70 years).
City bypass road Krefeld (Fehr, 1999)^10,¹²	Changes in traffic flow (including resulting changes in noise) estimated in environmental impact assessment report; air pollution dispersion models.	Quantitative risk assessment (QRA) for two carcinogens using health risk assessment computer program.	Individual additional lifetime cancer risk; additional cancer burden.
Home energy efficiency scheme (Kemm, 2000)¹³	Expert opinion (of authors) to estimate the increase in the home “comfort range” (the range of outside temperatures at which indoor temperature can be kept at or above 16°C, below which mortality starts to rise); CO₂ emission using emissions inventory and estimated energy savings; expert (authors?) estimate of effect on employment on basis of financial investment.	Ecological data on the effect of (outside) temperature on mortality were combined with climate data for Wales after applying an increase in the home comfort range for the number of people benefiting.	Number of deaths per year (1 or 2, on basis of 30000 homes treated).
Alconbury airport (France, 2000)¹⁴	Used estimate of project planner for vehicle movements and planning application estimates for increase in PM₁₀ (2.5% increase; assumed base level at 50 μg/m³)	Application of rates of death/illness resulting from PM₁₀ from WHO study; three methods/models for prediction of road traffic accident injuries on basis of vehicle movements; application of regional data on fatality of road traffic accidents.	Injury-only accidents (1 to 19 per year); RTA induced fatality (1 per 3–57 years); “long term mortality” from air pollution (0.2 deaths/y); respiratory hospital admissions (0.1/y); cardiovascular hospital admissions (0.3/y; chronic bronchitis incidents in adults over 25 (0.2/y); bronchitis in children under 15 (0.7/y); restricted activity days (170/y); asthma attacks in children under 15 (0.4); asthma episodes in adults over 15 (4.3/y).
Woodprocessing plant (Kemm, 2000)¹⁵	Used estimates from environmental impact report.	Application of national accident rates (per 100 million vehicle km).	Deaths due to road traffic accidents (1 per 10 years) and injury only accidents (1 per 2 years).
Finningley airport (Abdel Aziz et al 2000)¹⁶	Data from planning application (employment) and environmental statement (noise, air pollution); estimate air craft risk by MRC environment and health (method not stated).
National Botanic Garden of Wales (Kemm 2000)¹⁷	Business plan (employment); (uncomplicated) economic estimates to estimate income rise; estimates of numbers of visitors travelling by car, and average distance travelled.	Estimation on basis of UK data on income and mortality rates by income decile; application of national accident rates (per 100 million vehicle km).	Reduction in overall mortality (2%) resulting from income rise; annual car user and pedestrian casualties, including deaths and severely injured (for example, 0.063 car user deaths/y).
Port Southampton Dibden Bay (Taylor 2001)¹⁸	Used pollution data (PM₁₀) and noise levels from EIA report; used traffic projections and estimates on amount of employment generated from technical statements.	COMEAP dose-response curves for PM₁₀; traffic projections are combined with present traffic accident rates and literature on effects of mitigation measures; Scott-Samuel’s method of estimating effect of (un-) employment on mortality.	Annual respiratory deaths due to air pollution (0.01); hospital admissions for respiratory problems (one per 3–4 years); personal injury road traffic accidents (17.3 per year, which equals a 12% rise—assuming no mitigation measures are taken. Possibly an additional one due to more heavy traffic). Estimates of effect of speed limits (−4.7% or 1–2 reported personal injury accidents) and street lighting (−9% or 2–3 personal injury accidents/y) are also given, as are premature deaths due to unemployment (just over 3 avoided annually).
HIA Chad-Cameroon oil pipeline project (Jobin, 2003)¹⁹	Estimated truck miles on basis of preliminary project plans; applied data from malaria study in Nigeria to estimated population exposed; used computer simulation model with estimates on HIV seroprevalence and frequency of sexual contacts based on previous epidemiological studies, and estimated change of partner due to project.	Applied US accident data multiplied by 10 to account for road conditions; applied Nigerian malaria death rates, and assumed effects of prevention and curative services; computer simulation model included HIV transmission risk.	Road traffic deaths (2.5 per year); deaths due to malaria (3 per year); deaths due to HIV/AIDS (70 per year).
HIA drinking water privatisation (Fehr, 2003)²⁰	Difference between current levels of six carcinogens in drinking water and maximum permitted levels is calculated; exposure is estimated using population number, and standardised estimates and measured data on tap water intake and body weight.	“Standard methodology of quantitative risk assessment”	Additional cancer cases from a lifetime of exposure (maximum 10938 cases for population of 18 million)
Foot and mouth disease disposal options (UK Department of Health, 2001)²¹	Air emissions calculated for arbitrary numbers of cattle disposed of; data on percentage of BSE infected older cattle.	Dispersal modelling for BSE	Cases of variant Creutzfeldt-Jakob disease as function of %—age older cattle disposed of by pyre burning or burying.
Foresight vehicle initiative (Abrahams et al, 2002)^22,²³	Used consensus panel, email discussion, and other methods with expert groups to estimate changes in determinants due to new technologies.	An epidemiological model (ARMADA)²⁴ was used to estimate consequences of road traffic accidents and air pollution.	Over period 2000–2029: number of deaths (±20) and serious injuries (±2000) (traffic accidents); first hospital admissions for respiratory and cardiovascular disease (±23,000) (air pollution).
Regional planning guidance transport chapter (Pitches and Kemm, 2003)²⁵ and West Midlands local plan	Uses scenarios to estimate effect on physical activity, vehicle miles and air pollutant emissions (not exposure); draft assessment report of Highways Agency for traffic flows.	Estimates of health consequences of physical activity increases; model for traffic accident injuries as function of vehicle miles.	All cause mortality, cardiovascular and colon carcinoma deaths, cardiovascular events (physical activity); slight, serious and fatal injuries (traffic accidents)