World Health Organization Regional Office for Europe

(1)

WHO European Centre for Environment and Health

Platz der Vereinten Nationen 1, D-53113 Bonn, Germany

Tel.: +49 228 815 0400 The WHO Regional

Office for Europe

The World Health Organization (WHO) is a specialized agency of the United Nations created in 1948 with the primary responsibility for international health matters

each with its own programme geared to the particular health conditions of the countries it serves.

Member States Albania Andorra Armenia Austria Azerbaijan Belarus Belgium

Bosnia and Herzegovina Bulgaria

Croatia Cyprus Czechia Denmark Estonia Finland France Georgia Germany Greece Hungary Iceland Ireland Israel Italy Kazakhstan Kyrgyzstan Latvia Lithuania Luxembourg Malta Monaco Montenegro Netherlands

Romania

Russian Federation San Marino Serbia Slovakia Slovenia Spain Sweden Switzerland Tajikistan North Macedonia Norway

Poland Portugal

Republic of Moldova

World Health Organization Regional Office for Europe

UN City, Marmorvej 51, DK-2100 Copenhagen Ø, Denmark

Tel.: +45 45 33 70 00 Fax: +45 45 33 70 01 E-mail: euwhocontact@who.int

Website: www.euro.who.int The WHO Regional

The World Health Organization (WHO) is a specialized agency of the United Nations created in 1948 with the primary responsibility for international health matters and public health. The WHO Regional Office for Europe is one of six regional offices throughout the world, each with its own programme geared to the particular health conditions of the countries it serves.

Romania

Russian Federation San Marino Serbia Slovakia Slovenia Spain Sweden Switzerland Tajikistan Turkey Turkmenistan Ukraine United Kingdom North Macedonia Norway

Poland Portugal

Republic of Moldova

World Health Organization Regional Office for Europe

Romania

Russian Federation San Marino Serbia Slovakia Slovenia Spain Sweden Switzerland Tajikistan Turkey Turkmenistan Ukraine United Kingdom Uzbekistan North Macedonia Norway

Poland Portugal

Republic of Moldova

World Health Organization Regional Office for Europe

Romania

Poland Portugal

Republic of Moldova

World Health Organization Regional Office for Europe

Romania

Poland Portugal

Republic of Moldova

World Health Organization Regional Office for Europe

Romania

Poland Portugal

Republic of Moldova

Health impact assessment of air pollution: AirQ+ multiple- area data input

World Health Organization Regional Office for Europe

Romania

Poland Portugal

Republic of Moldova WHO/EURO:2020-1558-41309-56211

(2)

(3)

Health impact assessment of air pollution: AirQ+

multiple-area data input

December 2020

(4)

Address requests about publications of the WHO Regional Office for Europe to:

Publications

WHO Regional Office for Europe UN City, Marmorvej 51

DK-2100 Copenhagen Ø, Denmark

Alternatively, complete an online request form for documentation, health information, or for permission to quote or translate, on the Regional Office website (http://www.euro.who.int/pubrequest).

Abstract

AirQ+ is a software tool for quantifying the health burden and impact of air pollution developed by the WHO Regional Office for Europe. AirQ+ includes methodologies to assess the impacts of short- and long-term exposure to ambient air pollution.

The main methodologies use evidence generated by epidemiological cohort studies showing a relationship between average long-term air pollution concentration levels and the mortality risks in exposed populations. Assessing the impact of air pollution is suggested when evaluating the consequences of policies and interventions or of hypothetical scenarios. AirQ+

should always be used with the support of an epidemiologist or air pollution impact assessment expert. To facilitate users in their analyses, AirQ+ comes with manuals that require increasing levels of expertise. This manual introduces routine users of AirQ+ to the analysis of the impact of air pollution on public health when considering data from multiple areas. The new multiple-area input functionalities of AirQ+ are useful when analysing not just a single area but many areas (for example, several countries, regions, subregions, cities) for a comprehensive assessment.

Keywords

AIR POLLUTANTS PARTICULATE MATTER

AIR POLLUTION – adverse effects AIR POLLUTION – exposure AIR POLLUTION – health impacts

© World Health Organization 2020

Some rights reserved. This work is available under the Creative Commons Attribution- NonCommercial-ShareAlike 3.0 IGO licence (CC BY-NC-SA 3.0 IGO; https://creativecommons.

org/licenses/by-nc-sa/3.0/igo).

Under the terms of this licence, you may copy, redistribute and adapt the work for noncommercial purposes, provided the work is appropriately cited, as indicated below. In any use of this work, there should be no suggestion that WHO endorses any specific organization, products or services. The use of the WHO logo is not permitted. If you adapt the work, then you must license your work under the same or equivalent Creative Commons licence. If you create a translation of this work, you should add the following disclaimer along with the suggested citation:

“This translation was not created by the World Health Organization (WHO). WHO is not responsible for the content or accuracy of this translation. The original English edition shall be the binding and authentic edition: Health impact assessment of air pollution: AirQ+ multiple-area data input.

Copenhagen: WHO Regional Office for Europe; 2020”.

Any mediation relating to disputes arising under the licence shall be conducted in accordance with the mediation rules of the World Intellectual Property Organization. (http://www.wipo.int/

amc/en/mediation/rules/)

Suggested citation. Health impact assessment of air pollution: AirQ+ multiple-area data input.

Copenhagen: WHO Regional Office for Europe; 2020. Licence: CC BY-NC-SA 3.0 IGO.

Cataloguing-in-Publication (CIP) data. CIP data are available at http://apps.who.int/iris.

Sales, rights and licensing. To purchase WHO publications, see http://apps.who.int/bookorders. To submit requests for commercial use and queries on rights and licensing, see http://www.who.int/

about/licensing.

Third-party materials. If you wish to reuse material from this work that is attributed to a third party, such as tables, figures or images, it is your responsibility to determine whether permission is needed for that reuse and to obtain permission from the copyright holder. The risk of claims resulting from infringement of any third-party-owned component in the work rests solely with the user.

General disclaimers. The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of WHO concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted and dashed lines on maps represent approximate border lines for which there may not yet be full agreement.

The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed or recommended by WHO in preference to others of a similar nature that are not

Document number: WHO/EURO:2020-1558-41309-56211

(5)

The WHO Regional Office for Europe gratefully acknowledges Magali Corso (Santé publique France), Sylvia Medina (Santé publique France), Michal Krzyzanowski (King’s College London, United Kingdom of Great Britain and Northern Ireland) and Joseph Spadaro (Spadaro Environmental Research Consultants, Philadelphia, Pennsylvania, United States of America) for their comments and suggestions. We are grateful to Tibor Málnási (National Public Health Center, Hungary) for providing input when initially developing this publication. Thanks also goes to Dorota Jarosinska (WHO European Centre for Environment and Health, WHO Regional Office for Europe) for her comments in finalizing the publication. The AirQ+ project was partially financed by the German Federal Ministry for the Environment, Nature Conservation and Nuclear Safety.

(7)

Introduction

The aim of this report is to introduce the users of AirQ+ to the analysis of the impact of air pollution on public health when considering data from multiple areas. This is useful when users are analysing not just a single area but have to consider many areas (for example, several countries, regions, subregions, cities) for a comprehensive assessment.

To facilitate this kind of analysis, the user of AirQ+ can save time by entering air pollution data for multiple areas in comma-separated values (csv) files that use the semicolon (;) as the separator character, and running a simultaneous analysis instead of separate analyses area by area. Data should be prepared in the csv format:

the first row should include the names of the fields (for example, Subregion; Date; Daily Mean PM_2.5), followed by rows of data from area 1 or region 1, rows of data from area 2 or region 2, etc.

To import data files, select the Air Quality Data tab in the Analysis Properties window and then use the Import Data button to activate the Import Air Quality Data window (Fig. 1).

Fig. 1. AirQ+ Import Air Quality Data window

(8)

The results of a multi-area analysis include the usual output of Estimated Attributable Proportion, Estimated number of Attributable Cases and Estimated number of Attributable Cases per 100 000 Population at Risk for each single area analysed. The Population Attributable Proportion is calculated with a population-weighted mean of area-specific results. The results also include summary total amounts based on the sum of single- area results.¹ Additionally, the Detailed Results window displays results using a population-weighted air pollution mean concentration value for all analysed areas. The user can consider all-area results when the mean value calculation makes sense, for example in the case of regions that are part of one country, or districts of a city (see examples below).

AirQ+ comes with the “testData” folder, which contains five example data files for testing. Table 1 shows the mortality, population and air quality data in each file.

Table 1. Example test files included in AirQ+

Example File name Description

A

02AreasTestPMDailyValues.csv Contains daily mean PM_2.5 concentration data from two areas, Subregion 1 and Subregion 2, for three years from 2004 to 2006

B

03AreasTestPMMinMaxDaysValues.csv Contains air quality data for PM_2.5 between 10 and 190 µg/m³ from three areas: Subregion 1, Subregion 2 and Subregion 3; air pollution data include frequency of days within a specific range of PM_2.5 values, step 10.²

C 12AreasTestPMDailyValues.csv Contains daily PM_2.5 concentration data from 12 areas

D

13AreasTestPMYearlyMeans.csv Contains data from 13 areas with different yearly PM_2.5 concentration mean values, ranging from 12 µg/m³ to 40 µg/m³

E

647Areas_with Incidence and Population.csv Contains data from 647 areas with different yearly PM_2.5 concentration mean values, ranging from 2.74 µg/m³ to 16.37 µg/m³; data are modelled with populations ranging from 1 to 24 408, with different mortality rates.

The examples with two and three areas (examples A and B) use different mortality rates for exposed populations that have comparable sizes. The example with 12 areas (Example C) uses data on exposed populations that range between 1 091 660 and 1 610 406, and mortality rates that vary from 450 per 100 000 population to 1200 per 100 000 population. The example with 13 areas (Example D) uses data on exposed populations ranging from 100 000 to 5 000 000, but with the same mortality rate.

1 For details on key terms, please refer to the AirQ+ Glossary, which is accessible on the main welcome window.

2 The category 0–10 includes all air pollution concentration values equal to or greater than 0 and less than 10 [x, x+10[.

(9)

Issues and challenges

There are two important points to consider. First, the examples are theoretical and are not based on data from a specific country or city. Second, the relative risks of air pollution should be applied to areas that have sufficiently large populations. There is no quantitative threshold to define how large a population should be.

When dealing with areas that include small and extremely small populations (e.g. census blocks or zip codes), for example fewer than 1000 inhabitants, there are some issues to take into account. One issue is related to the comparability (in respect of age, sex etc. distribution) of the small size of the study population with the population that generated the risk estimates used in the assessment. A second issue is related to the availability of incidence data that are difficult to obtain for a very small scale. A third issue is that having many different subareas from one single area has to be justified e.g. by distinct exposure patterns. It is advised to check the usual overall estimate of the impacts for the area as a whole.

To summarize the issue of dimension of areas, Briggs et al. (2007:p. 357)³ discusses four objectives of environmental epidemiology studies that are usually considered:

(a) to provide a uniform basis for mapping, both to aid visual representation and interpretation of the data, and also to facilitate analysis of spatial patterns;

(b) to have a zone system that is sufficiently fine to reflect local variations in exposures and rates of disease, especially in urban areas where such gradients may be steep;

(c) to achieve reasonably large and consistent denominator populations in all zones, to avoid the so- called ‘small number problem’, which can lead to highly unstable estimates of risk and large variations in uncertainty between zones;

(d) to minimize the need for spatial transformation of data between different spatial units, since this invariably involves some degree of approximation and is thus a further source of error.

The spatial and temporal information related to adopting small areas for analysis has to be carefully taken into account because the adequacy of estimates is to be judged relative to the study objective and the degree of inaccuracy associated with small areas counts (Elliot et al., 2000).⁴

It is recommended to always use AirQ+ with the support of an epidemiologist or air pollution impact assessment expert.

3 Briggs D, Fecht D, De Hoogh K (2007). Census data issues for epidemiology and health risk assessment: experiences from the Small Premature deaths in Area Health Statistics Unit. J R Stat Soc Ser A Stat Soc. 170(2):355–78. doi:https://doi.org/10.1111/j.1467- 985X.2006.00467.x.

4 Elliot P, Wakefield JC, Best NG, Briggs DJ, editors (2000). Spatial epidemiology: methods and applications. Oxford: Oxford University Press.

(10)

Impact evaluation using multi- area data input

Example A. Analysis of two areas with three years of daily ambient air pollution data: PM

_2.5

– long-term – adult mortality

Question to be addressed: How many deaths (out of the total number of deaths from natural causes) are attributable to long-term exposure to PM_2.5exceeding the 2005 WHO air quality guidelines (AQG)⁵ level (10 µg/m³) in two different areas of a city?

Data input

Example A uses the input data file “02AreasTestPMDailyValues.csv”. The data comprise a sample of daily mean PM_2.5concentrations in two areas, Subregion 1 and Subregion 2, for three years from 2004 to 2006.

Line 1 describes the names of the fields.

Lines 2–2193 contain the name of the area, date (in “day.month.year” format) and PM_2.5 mean value.

Example data are:

line 1: Subregion;Date;Daily Mean PM_2.5

line 8: Subregion 1;07.01.2004;29

line 1184: Subregion 2;27.03.2004;25.

This means the daily average PM_2.5concentration in Subregion 1 was 29µg/m³on 7 January 2004, and the daily average PM_2.5concentration in Subregion 2 was 25 µg/m³ on 27 March 2004.

In the New Impact Assessment window (Fig. 2), select ambient pollution, long-term effects and the pollutant PM_2.5in the desired location and click OK.

5 WHO Regional Office for Europe (2006). Air quality guidelines: global update 2005: particulate matter, ozone, nitrogen dioxide, and sulfur dioxide. Copenhagen: WHO Regional Office for Europe (http://www.euro.who.int/en/health-topics/environment- and-health/air-quality/publications/pre2009/air-quality-guidelines.-global-update-2005.-particulate-matter,-ozone,-nitrogen- dioxide-and-sulfur-dioxide, accessed 12 November 2019).

(11)

Fig. 2. AirQ+ New Impact Assessment window for ambient air pollution – PM_2.5– long-term

In the Analysis Properties window (Fig. 3) enter the total population “4246375”. AirQ+ does not use the total population for the analysis, so the user can consider this as additional descriptive information.⁶ Since the input file contains daily data, select the Input Air Quality Data option.

Fig. 3. AirQ+ Analysis Properties window for ambient air pollution – PM_2.5 – long-term – adult mortality (data input: daily PM_2.5 data from Subregion 1 and Subregion 2, 2004–2006)

Click the Enter Air Quality Data button (Fig. 3) to go to the Import Air Quality Data window (Fig. 4).

6 AirQ+ uses the total number of adults (≥ 30 years) exposed to the pollutant, a value the user enters when creating a new Impact Analysis, explained later in this section.

(12)

Fig. 4. Data import of ambient air pollution – PM_2.5 – long-term – adult mortality (data input: daily PM_2.5 data from Subregion 1 and Subregion 2, 2004–2006)

After importing the air quality data, AirQ+ will calculate that the combined areas (subregions 1 and 2) produced an average PM_2.5concentration of 27.50 µg/m³for the years 2004–2006 (Fig. 5). With the information below, the user can enter exposed population values to produce a population-weighted concentration that is used for subsequent calculations.

After creating a new Impact Evaluation (button at the bottom of the Analysis Properties window), enter the following data (Fig. 5):

mortality incidence for adults (≥ 30 years), all natural causes, per 100 000 population in:

Subregion 1: 940

Subregion 2: 830

the total number of adults (≥ 30 years) exposed to the pollutant⁷ in:

Subregion 1: 1 156 588

the default relative risk values for all-cause mortality: 1.062 (95% CI 1.040–1.083)⁸

the default cut-off value recommended by the 2005 WHO AQG: 10 µg/m^3.

7 The total population at risk in both subregions is 2 547 825 (displayed in the Air Quality Data Statistics box of the Detailed Results window, data not shown).

8 The relative risk values and the 95% confidence interval (CI) come from the results of the meta-analysis of 13 cohort studies by Hoek G, Krishnan RM, Beelen R, Peters A, Ostro B, Brunekreef B, et al. (2013). Long-term air pollution exposure and cardio- respiratory mortality: a review. Environ Health. 12(1):43. doi:10.1186/1476-069X-12-43.

(13)

The Impact Evaluation window shows that the combined areas (subregions 1 and 2) produced an average PM_2.5 concentration of 27.50 µg/m³for the years 2004–2006 (created when the user imported air quality data).

After entering the exposed population values, AirQ+ will produce a population-weighted value of 27.69 µg/m³ (Fig. 5) and use it for subsequent calculations (see the Results window in Fig. 6).

AirQ+ generates two different numbers for the average concentration:

a population-weighted mean concentration over the three-year period 2004–2006, displayed in the Impact Evaluation window; and

an arithmetic mean of daily values over the three-year period 2004–2006, displayed in the Detailed Results window.⁹

The Detailed Results window shows that the arithmetic mean air pollution is 25.47 µg/m³in Subregion 1 (Fig. 7) and 29.53 µg/m³ in Subregion 2 (data not shown).

Fig. 5. AirQ+ Impact Evaluation window for ambient air pollution – PM_2.5 – long-term – adult mortality (data input: PM_2.5 daily frequency data)

Click the Calculate button in the Impact Evaluation window (Fig. 5) to go to the Results window (Fig. 6).

9 The software works with the population-weighted mean concentration and the user should consider the arithmetic mean only for educational or descriptive purposes.

(14)

Results

AirQ+ calculates the total number of cases¹⁰ as the sum of each output, area by area (Cumulative All Areas).

The results in Subregion 1 and Subregion 2 indicate that 966 and 1280 premature deaths (central estimate) are attributable to long-term exposure to PM_2.5, respectively (Fig. 6). These results indicate the benefit of a reduction from current exposure to WHO AQG. AirQ+ also displays a cumulative sum of area-specific results (Cumulative All Areas). The sum of attributable cases indicates 2246 premature deaths.¹¹ These impacts could be prevented if concentrations of PM_2.5 would not exceed 10 µg/m³. This window displays area-specific results as the Estimated Attributable Proportion based on the area population-weighted mean and area-specific incidence rates.

Fig. 6. AirQ+ results for ambient air pollution – PM_2.5 – long-term – adult mortality

The Detailed Results window gives results for each subregion as well as for all areas. Cumulative results are calculated based on population-weighted mean exposure as well as population-weighted incidence. The user can select the Lower and Upper tabs to see the estimates calculated with, respectively, lower and upper confidence interval (CI) limits of the relative risk. This range is the 95% CI based on the uncertainty in the relative risk values for all-cause mortality that the user entered when creating the Impact Evaluation.

Fig. 7 shows the Detailed Results window with the graph of excess cases of mortality and relative risk. By selecting the “Area” drop-down menu of the Detailed Results window, the user can see (i): the specific results in each area or (ii): the burden in all areas (treated as a single region).

10 AirQ+ can show the total number of cases as a single calculation performed on a single region comprising all areas in the Detailed Results window, which may be of interest to expert users.

11 The population- and incidence-weighted result indicates that 2263 premature deaths are caused by long-term exposure to PM_2.5 (in the Detailed Results window, data not shown). The difference between this value and the value obtained by adding the results area by area is due to approximations used in the calculations.

(15)

Fig. 7. AirQ+ detailed results for ambient air pollution – PM_2.5 – long-term – adult mortality (graph of excess cases in Subregion 1)

In addition, depending on the cut-off value, each subregion has different results. Table 2 shows the results for different cut-off values. In these cases, the benefits from reduction of air pollution on mortality are related to 2844 deaths for a cut-off equal to 5 µg/m³, 2002 deaths for a cut-off equal to 12 µg/m³ and 342 deaths for a cut-off equal to 25 µg/m³.¹²

Table 2. Results for different cut-offs

Metric Cut-off 5 µg/m³ Cut-off 12 µg/m³ Cut-off 25 µg/m³

Subregion 1 Subregion 2 Subregion 1 Subregion 2 Subregion 1 Subregion 2 Estimated

Attributable Proportion (%)

11.59 (7.72–15.06)

13.72 (9.17–17.77)

7.79 (5.15–10.19)

10.01 (6.64–13.05)

0.28 (0.19–0.38)

2.69 (1.76–3.55) Estimated number of

Attributable Cases

1260 (839–1638)

1584 (1059–2051)

846 (560–1107)

1156 (767–1506)

31 (20–41)

311 (203–410) Estimated number

of Attributable Cases per 100 000 Population at Risk

108.92 (72.53–141.58)

113.87 (76.14–147.46)

73.18 (48.38–95.75)

83.07 (55.15–108.28)

2.67 (1.74–3.54)

22.32 (14.62–29.45) Note: numbers in parenthesis are 95% CI values.

12 The population- and incidence-weighted impacts are estimated as 2860 excess cases for a cut-off equal to 5 μg/m³, 2019 excess cases for a cut-off equal to 12 μg/m³ and 360 excess cases for a cut-off equal to 25 μg/m³.

(16)

Example B. Analysis of three areas with one year of ambient air pollution data by range: PM

_2.5

– long-term – adult mortality

Question to be addressed: How many deaths (out of the total number of deaths from natural causes) are attributable to long-term exposure to PM_2.5 exceeding the WHO AQG level (10 µg/m³) in three different regions?

Data input

Example B uses the data file “03AreasTestPMMinMaxDaysValues.csv”. It contains sample PM_2.5 data for an impact analysis using one year of concentration data from three areas: Subregion 1, Subregion 2 and Subregion 3.

Line 1 describes the names of the fields.

Lines 2–58 contain the name of the area, minimum interval value for PM_2.5 concentrations, maximum interval value for PM_2.5 concentrations and number of days.

Example data are:

line 1: subregion;min;max;days

line 9: Subregion 1;70;80;6

line 27: Subregion 2;60;70;6.

This means that there were 6 days with PM_2.5 concentrations between 70 and 80 µg/m³ in Subregion 1, and 6 days with PM_2.5 concentrations between 60 and 70 µg/m³ in Subregion 2.

In the New Impact Assessment window (Fig. 2), select ambient pollution, long-term effects and the pollutant PM_2.5 in the desired location.

In the Analysis Properties window, enter the total population for the example data: “6746375”. AirQ+ does not use the total population for the analysis, so the user can consider this as additional information. Since the input file contains daily data, select the Input Air Quality Data option (Fig. 3).

Click the Enter Air Quality Data button (Fig. 3) to go to the Import Air Quality Data window (Fig. 8).

After importing the air quality data, AirQ+ will calculate that the combined areas (subregions 1–3) produced an average PM_2.5 concentration of 30.85 µg/m³ (data not shown). With the information below, the user can enter exposed population values to produce a population-weighted concentration that is used for subsequent calculations.

(17)

After creating a new Impact Evaluation, enter the following data:

Subregion 1: 940

Subregion 2: 830

Subregion 3: 910

the total number of adults (≥ 30 years) exposed to the pollutant¹³ in:

the default relative risk values for all-cause mortality: 1.062 (CI 1.040–1.083)

After entering the exposed population values, AirQ+ will produce a population-weighted mean concentration value of 30.19 µg/m³ and use it for subsequent calculations; results are displayed in the Detailed Results window (figure not shown). In this example, the daily value is a three-year average of available daily values.

When using a dataset with minimum and maximum values, select the input format “Aggregated” (highlighted in a red box in Fig. 8). Consider that the mean air pollution value is 38.42 µg/m³ for Subregion 1, 31.82 µg/m³ for Subregion 2 and 22.32 µg/m³ for Subregion 3 (data not shown).

Fig. 8. Data import of ambient air pollution – PM_2.5 – long-term – adult mortality (data input: daily PM_2.5 concentrations in Subregion 1, Subregion 2 and Subregion 3 for one year)

13 The total population at risk in all subregions is 4 047 825 (displayed in the Air Quality Data Statistics box of the Detailed Results window, figure not shown).

(18)

Results

The results in Subregion 1, Subregion 2 and Subregion 3 indicate that 1709, 1421 and 975 premature deaths are caused by long-term exposure to PM_2.5, respectively (Fig. 9). These results indicate the benefit of a reduction from current exposure to WHO AQG. The sum of these results indicates 4104 premature deaths.¹⁴ These impacts could be prevented if the concentration of PM_2.5 would not exceed 10 µg/m³.

The Detailed Results window gives results for each area as well as for all areas. Cumulative results are calculated based on population-weighted mean exposure as well as population-weighted incidence. The values in the Lower and Upper columns correspond to the estimates calculated with, respectively, lower and upper confidence interval limits of the relative risk. The interval between the lower and upper values represents some of the uncertainty associated with the estimation. Using a drop-down list, users can display graphs of excess cases of mortality and relative risk for each area or for all areas.

Fig. 9. AirQ+ results for ambient air pollution – PM_2.5 – long-term – adult mortality

14 The population- and incidence-weighted results indicate that 4124 premature deaths are caused by long-term exposure to PM_2.5. (displayed in the Air Quality Data Statistics box of the Detailed Results window, data not shown).

(19)

Example C. Analysis of 12 areas with one year of daily ambient air pollution data: PM

_2.5

– long-term – adult mortality

Question to be addressed: How many deaths (out of the total number of deaths from natural causes) are attributable to long-term exposure to PM_2.5 exceeding the WHO AQG level (10 µg/m³) in 12 different areas?

Data input

Example C uses the input data file “12AreasTestPMDailyValues.csv”. It contains one year of PM_2.5 concentration data from 12 areas aggregated by range.

In the New Impact Assessment window, select ambient pollution, long-term effects and the pollutant PM_2.5, as in the previous examples.

In the Analysis Properties window, enter the total population: “26843985”. AirQ+ does not use the total population for the analysis, so the user can consider this as additional descriptive information. Since the input file contains daily data, select the Input Air Quality Data option.

Following the instructions in the previous examples, use the Import Air Quality Data window (Fig. 10) to import data; AirQ+ will calculate that the 12 areas combined produced an average PM_2.5 concentration of 28.32 µg/

m³. After entering exposed population values, AirQ+ will produce a population-weighted mean concentration value of 28.33 µg/m³ and use it for subsequent calculations; results are in the Detailed Results window (figure not shown).

In this example, the mean value is an average of available daily mean values in 12 areas in 2006. The mean PM_2.5 values for the 12 areas are:

Area 01: 28.21 µg/m³ Area 05: 28.11 µg/m³ Area 09: 28.22 µg/m³ Area 02: 28.41 µg/m³ Area 06: 28.44 µg/m³ Area 10: 28.45 µg/m³ Area 03: 28.14 µg/m³ Area 07: 28.01 µg/m³ Area 11: 28.28 µg/m³ Area 04: 29.07 µg/m³ Area 08: 28.41 µg/m³ Area 12: 28.12 µg/m³

Area 01: 939 Area 05: 730 Area 09: 1200

Area 02: 830 Area 06: 462 Area 10: 450

Area 03: 564 Area 07: 752 Area 11: 700

Area 04: 670 Area 08: 530 Area 12: 620

(20)

the total number of adults (≥ 30 years) exposed to the pollutant¹⁵ in:

Area 01: 1 156 588 Area 05: 1 112 430 Area 09: 1 402 235 Area 02: 1 391 237 Area 06: 1 091 660 Area 10: 1 350 700 Area 03: 1 427 400 Area 07: 1 481 370 Area 11: 1 610 406 Area 04: 1 501 320 Area 08: 1 330 210 Area 12: 1 250 835

the default relative risk values for all-cause mortality: 1.062 (95% CI 1.040–1.083)

Fig. 10. Data import of ambient air pollution – PM_2.5 – long-term – adult mortality (data input: daily mean PM_2.5 concentrations in 12 areas in one year)

15 The total population at risk is 16 106 391 (displayed in the Air Quality Data Statistics box of the Detailed Results window, figure not shown).

(21)

Results

The results indicate that a total of 11 871 excess deaths in the 12 areas considered could be “avoided” if WHO AQG for PM_2.5 are respected (Fig. 11).¹⁶ The number of premature deaths in each area is:

Area 01: 1127 Area 05: 838 Area 09: 1747

Area 02: 1211 Area 06: 529 Area 10: 638

Area 03: 832 Area 07: 1144 Area 11: 1174

Area 04: 1090 Area 08: 739 Area 12: 801

Fig. 11. AirQ+ results for ambient air pollution – PM_2.5 – long-term – adult mortality in 12 areas

The Detailed Results window shows the graphs of excess cases of mortality and relative risk in the 12 areas in 2006. By selecting the “Area” drop-down menu, the user can see specific results in each area or the cumulative burden in all areas (treated as a single region).

16 The population- and incidence-weighted results indicate that 11 880 premature deaths are caused by long-term exposure to PM_2.5 (displayed in the Air Quality Data Statistics box of the Detailed Results window, data not shown).

(22)

Example D. Analysis of 13 areas with one year of daily ambient air pollution data: PM

_2.5

– long-term – adult mortality

Question to be addressed: How many deaths (out of the total number of deaths from natural causes) are attributable to long-term exposure to PM_2.5 exceeding the WHO AQG level (10 µg/m³) in 13 different areas?

Data input

Example D uses the input data file “13AreasTestPMYearlyMeans.csv”. The file contains one year of PM_2.5 concentration data for 13 areas aggregated by range.

In the Analysis Properties window, enter the total population: “39333333”. AirQ+ does not use the total population for the analysis, and the user can consider this as additional descriptive information. Since the input file contains daily data, select the Input Air Quality Data option.

Following the instructions in the previous examples, use the Import Air Quality Data window (Fig. 12) to import data; AirQ+ will calculate that the 13 areas combined produced an average PM_2.5 concentration of 27.46 µg/m³. After entering the exposed population values, AirQ+ will produce a population-weighted mean concentration value of 21.06 µg/m³ and use it for subsequent calculations; results are displayed in the Detailed Results window (figure not shown). When using a data set with minimum and maximum values, select the input format

“Aggregated” (highlighted in a red box in Fig. 12).

In this example, the mean value is an average of available daily mean values in 13 areas in 2006. The mean PM_2.5 values for the 13 areas are:

Area 01: 12 µg/m³ Area 06: 26 µg/m³ Area 11: 36 µg/m³ Area 02: 15 µg/m³ Area 07: 28 µg/m³ Area 12: 38 µg/m³ Area 03: 20 µg/m³ Area 08: 30 µg/m³ Area 13: 40 µg/m³ Area 04: 22 µg/m³ Area 09: 32 µg/m³

Area 05: 24 µg/m³ Area 10: 34 µg/m³

(23)

Area 01: 900 Area 06: 900 Area 11: 900

Area 02: 900 Area 07: 900 Area 12: 900

Area 03: 900 Area 08: 900 Area 13: 900

Area 04: 900 Area 09: 900

Area 05: 900 Area 10: 900

the total number of adults (≥ 30 years) exposed to the pollutant¹⁷ in:

Area 01: 5 000 000 Area 06: 1 750 000 Area 11: 500 000 Area 02: 4 000 000 Area 07: 1 500 000 Area 12: 250 000 Area 03: 3 000 000 Area 08: 1 250 000 Area 13: 100 000 Area 04: 2 500 000 Area 09: 1 000 000

Area 05: 2 000 000 Area 10: 750 000

the default relative risk values for all-cause mortality: 1.062 (95% CI 1.040–1.083)

Fig. 12. Data import of ambient air pollution – PM_2.5 – long-term – adult mortality (data input: minimum and maximum PM_2.5 concentrations with exposure days in a year in 13 areas)

17 The total population at risk is 23 600 000 (displayed in the Air Quality Data Statistics box of the Detailed Results window, figure not shown).

(24)

Results

The results indicate that the sum of all impacts equals 13 479 excess deaths for all 13 areas that could be

“avoided” if WHO AQG for PM_2.5 are respected (Fig. 13).¹⁸ The number of premature deaths in each area is:

Area 01: 538 Area 06: 1445 Area 11: 652

Area 02: 1067 Area 07: 1385 Area 12: 349

Area 03: 1576 Area 08: 1275 Area 13: 149

Area 04: 1567 Area 09: 1116 Area 05: 1454 Area 10: 907

The Detailed Results window illustrates the graph of excess cases of mortality and relative risk in 13 areas.

By choosing an option (drop-down box), specific results in each area and the cumulative burden in all areas are displayed on the screen.

18 The population-weighted result is 13 670 excess cases of mortality.

(25)

Example E. Analysis of 647 areas with one year of daily ambient air pollution data: PM

_2.5

– long-term – adult mortality

Question to be addressed: How many deaths (out of the total number of deaths from natural causes) are attributable to long-term exposure to PM_2.5 exceeding 5 µg/m³ using data from 647 different areas?

Premise

In this example, the user imports only annual mean data from multiple areas. Prior to importing data into AirQ+, the user needs to create a csv data file with four fields: area, air pollution, incidence and population at risk. AirQ+ will then create the relevant pollution distributions, which is analogous to the case where the user inputs a single annual mean value from one single area in the Analysis Properties window, but this is done with multiple areas. Similar to the other multi-area import types, unless new air quality data are imported, the created analysis will only work with the areas that were imported.

The same file used to import air pollution data can be used to import incidence and population at risk data.

First, input area and annual mean air pollution data. Then, import the incidence and population at risk data to generate the desired calculation configuration for the evaluation.

AirQ+ needs to check the validity of the csv file. For example, it could be that rows are discarded because the area name was not previously imported in the Import Data panel in the Air Quality Data window. In the Impact Evaluation tab, the user cannot manually change the incidence and population values once data are imported;

to change these data, edit the (four-field) data input file and then reimport the new data to AirQ+.

Data input

Example E uses the data file “647Areas_with Incidence & Population.csv”. It contains one year of average PM_2.5 concentration data from 647 areas.

In the Analysis Properties window, enter the total population “1372989”. AirQ+ does not use the total population for the analysis, so the user can consider this as additional descriptive information. Since the input file contains air pollution data for multiple areas, select the Input Air Quality Data option.

(26)

Select Annual Means as the data input format (highlighted in a red box in Fig 14) as the other options are not allowed. In this example, the mean value is an average of all yearly mean values from the 647 areas. The annual mean PM_2.5 values for the first 10 and the last five areas are:

Area 01: 16.37 µg/m³ Area 07: 15.38 µg/m³ Area 644: 3.85 µg/m³ Area 02: 15.98 µg/m³ Area 08: 15.29 µg/m³ Area 645: 3.56 µg/m³ Area 03: 15.92 µg/m³ Area 09: 15.19 µg/m³ Area 646: 2.94 µg/m³ Area 04: 15.85 µg/m³ Area 10: 14.99 µg/m³ Area 647: 2.74 µg/m³ Area 05: 15.62 µg/m³ …

Area 06: 15.39 µg/m³ Area 643: 4.09 µg/m³

Fig. 14. Data import of ambient air pollution – PM_2.5 – long-term – adult mortality (data input: annual mean PM_2.5 data from 647 areas)

After creating a new Impact Evaluation (Fig. 15), the user needs to import the incidence and population at risk data selecting the appropriate columns in the Import Incidence Values windows (Fig. 16).

(27)

Fig. 15. AirQ+ Impact Evaluation window with the results of the import of data for incidence and population at risk for ambient air pollution – PM_2.5 – long-term – adult mortality (data input: annual mean PM_2.5 data from 647 areas)

Users can choose the Incidence and Population at Risk by using a drop box in Fig. 16 (highlighted in a red box).

Fig. 16. AirQ+ Impact Incidence Values window, incidence and population at risk for ambient air pollution – PM_2.5 – long-term – adult mortality (data input: annual mean PM_2.5 data from 647 areas)

(28)

Additional data for this analysis are:

Area 01: 1005 Area 09: 815 Area 642: 919

Area 02: 882 Area 10: 1020 Area 643: 898

Area 03: 882 … Area 644: 1106

Area 04: 929 Area 637: 1106 Area 645: 929

Area 05: 732 Area 638: 923 Area 646: 898

Area 06: 1020 Area 639: 929 Area 647: 1110

Area 07: 942 Area 640: 945

Area 08: 1030 Area 641: 945

total number of adults (≥ 30 years) exposed to the pollutant in:

Area 01: 31 Area 08: 77 Area 642: 10

Area 02: 2 Area 09: 9 Area 643: 12

Area 03: 106 Area 10: 109 Area 644: 6

Area 04: 23 … Area 645: 74

Area 05: 27 Area 638: 37 Area 646: 2

Area 06: 7 Area 639: 134 Area 647: 5

Area 07: 37 Area 640: 38

Area 08: 6 Area 641: 39

the default relative risk values for all-cause mortality: 1.08 (95% CI 1.060-1.10)

the cut-off value: 5 µg/m^3.

This example works with higher risks than usual, and given the low levels of exposure, with a cut-off lower than the WHO AQG value for PM_2.5. Using higher relative risks than that suggested by the HRAPIE project,¹⁹ the assumption is that the example population exposed to low levels of PM_2.5 might be more susceptible to premature mortality risks than the population exposed to medium levels of concentrations of PM_2.5.

19 WHO Regional Office for Europe (2013). Health risks of air pollution in Europe – HRAPIE project. Copenhagen: WHO Regional Office for Europe ((http://www.euro.who.int/en/health-topics/environment-and-health/air-quality/publications/2013/health- risks-of-air-pollution-in-europe-hrapie-project.-new-emerging-risks-to-health-from-air-pollution-results-from-the-survey-of- experts, accessed 25 November 2019).

(29)

Results

The results indicate a benefit of an estimated reduction of 454 deaths (95% CI: 345–560) for the combined 647 areas, summing all attributable death values of each areas, if a value of concentration of 5 µg/m³ of PM_2.5 are reached.²⁰

The premature deaths in each area are:

Area 01: 0.03 Area 05: 0.02 Area 09: 0.05

Area 02: 0.00 Area 06: 0.01 Area 10: 0.01

Area 03: 0.08 Area 07: 0.03 …

Area 04: 0.02 Area 08: 0.00 Areas 638–647: 0.00

The number of premature deaths is 0 when the relative risk is 1, that is, when the level of exposure is below the selected cut-off value. Fig.17 shows the Results window for the analysis.

The number of excess cases of mortality should be checked in the Detailed Results tab. By selecting the

“Area” drop-down menu, specific results in each area can be checked. Detailed results can be exported and analysed using software outside AirQ+.

20 The population- and incidence-weighted impact is equal to 456 deaths.

(30)

WHO European Centre for Environment and Health

Platz der Vereinten Nationen 1, D-53113 Bonn, Germany

Tel.: +49 228 815 0400 The WHO Regional

Romania

Russian Federation San Marino Serbia Slovakia Slovenia Spain Sweden Switzerland Tajikistan North Macedonia Norway

Poland Portugal

Republic of Moldova

World Health Organization Regional Office for Europe

The World Health Organization (WHO) is a specialized agency of the United Nations created in 1948 with the primary responsibility for international health matters and public health. The WHO Regional Office for Europe is one of six regional offices throughout the world, each with its own programme geared to the particular health conditions of the countries it serves.

Romania

Poland Portugal

Republic of Moldova

World Health Organization Regional Office for Europe

Romania

Poland Portugal

Republic of Moldova

World Health Organization Regional Office for Europe

Romania

Poland Portugal

Republic of Moldova

World Health Organization Regional Office for Europe

Romania

Poland Portugal

Republic of Moldova

World Health Organization Regional Office for Europe

Romania

Poland Portugal

Republic of Moldova

Health impact assessment of air pollution: AirQ+ multiple- area data input

World Health Organization Regional Office for Europe

Romania

Poland Portugal

Republic of Moldova WHO/EURO:2020-1558-41309-56211

World Health Organization Regional Office for Europe

WHO European Centre for Environment and Health

World Health Organization Regional Office for Europe

World Health Organization Regional Office for Europe

World Health Organization Regional Office for Europe

World Health Organization Regional Office for Europe

World Health Organization Regional Office for Europe

Health impact assessment of air pollution: AirQ+ multiple- area data input

World Health Organization Regional Office for Europe

Health impact assessment of air pollution: AirQ+

multiple-area data input

December 2020

Abstract

Keywords

Contents

Acknowledgments

Introduction

Issues and challenges

Impact evaluation using multi- area data input

Example A. Analysis of two areas with three years of daily ambient air pollution data: PM

– long-term – adult mortality

Example B. Analysis of three areas with one year of ambient air pollution data by range: PM

– long-term – adult mortality

Example C. Analysis of 12 areas with one year of daily ambient air pollution data: PM

– long-term – adult mortality

Example D. Analysis of 13 areas with one year of daily ambient air pollution data: PM

– long-term – adult mortality

Example E. Analysis of 647 areas with one year of daily ambient air pollution data: PM

– long-term – adult mortality

WHO European Centre for Environment and Health

World Health Organization Regional Office for Europe

World Health Organization Regional Office for Europe

World Health Organization Regional Office for Europe

World Health Organization Regional Office for Europe

World Health Organization Regional Office for Europe

Health impact assessment of air pollution: AirQ+ multiple- area data input

World Health Organization Regional Office for Europe

World Health Organization Regional Office for Europe

WHO European Centre for Environment and Health

World Health Organization Regional Office for Europe

World Health Organization Regional Office for Europe

World Health Organization Regional Office for Europe

World Health Organization Regional Office for Europe

World Health Organization Regional Office for Europe

Health impact assessment of air pollution: AirQ+ multiple- area data input

World Health Organization Regional Office for Europe

Health impact assessment of air pollution: AirQ+

multiple-area data input

December 2020

Abstract

Keywords

Contents

Acknowledgments

Introduction

Issues and challenges

Impact evaluation using multi- area data input

Example A. Analysis of two areas with three years of daily ambient air pollution data: PM

– long-term – adult mortality



























Example B. Analysis of three areas with one year of ambient air pollution data by range: PM

– long-term – adult mortality



























Example C. Analysis of 12 areas with one year of daily ambient air pollution data: PM

– long-term – adult mortality









Example D. Analysis of 13 areas with one year of daily ambient air pollution data: PM

– long-term – adult mortality









Example E. Analysis of 647 areas with one year of daily ambient air pollution data: PM

– long-term – adult mortality









WHO European Centre for Environment and Health

World Health Organization Regional Office for Europe

World Health Organization Regional Office for Europe

World Health Organization Regional Office for Europe

World Health Organization Regional Office for Europe

World Health Organization Regional Office for Europe

Health impact assessment of air pollution: AirQ+ multiple- area data input

World Health Organization Regional Office for Europe