Consultation

20, AVENUE APPIA – CH-1211 GENEVA 27 – SWITZERLAND – TEL CENTRAL +41 22 791 2111 – FAX CENTRAL +41 22 791 3111 – WWW.WHO.INT Ref.: C.L.49.2022 The World Health Organization (WHO) presents its compliments to Member States and has the honour to consult them on joint estimates of WHO and the International Labour Organization (ILO) of the work-related burden of disease and injury (WHO/ILO Joint Estimates). In the spirit of the Sustainable Development Goals, 1 WHO and ILO produce the WHO/ILO Joint Estimates. 2 The organizations have systematically selected pairs of occupational risk factors and health outcomes. For each pair, the existing evidence has been systematically reviewed and synthesized. Databases have been established on exposure to the risk factors and their health effects. For pairs with the required evidence and data, estimates have been produced. … The attached guidance note details the data sources and methods used. Prior to publication planned for the first quarter of 2023, WHO would like to share the following WHO/ILO Joint Estimates that were developed and compiled globally: (1) proportion of the population exposed to solar ultraviolet radiation; (2) number of deaths due to non-melanoma skin cancer attributable to occupational exposure to solar ultraviolet radiation; and (3) number of disability-adjusted life years due to non-melanoma skin cancer attributable to occupational exposure to solar ultraviolet radiation. WHO will share with each Member State a data sheet with the WHO/ILO Joint Estimates for the country on request. Member States are invited to provide feedback to the finalization of the WHO/ILO Joint Estimates. Focal points designated by the Ministries of Health and/or Labour in 2020 (following C.L.8.2020) may want to comment. Requests for data sheets with the estimates for the country and other information on these estimates can be sent to Dr Frank Pega, [email protected], Department of Environment, Climate Change and Health no later than 21 December 2022. The World Health Organization avails itself of this opportunity to renew to Member States the assurance of its highest consideration. GENEVA, 9 November 2022 70th UNGA. (2015). Transforming our world: the 2030 Agenda for Sustainable Development. New York, 1 NY: UN. 2 WHO, ILO (2021). WHO/ILO joint estimates of the work-related burden of disease and injury, 2000-2016: global monitoring report. Geneva: WHO, ILO. ... ENCL.: (1) Guidance note to facilitate country consultation on WHO/ILO Joint Estimates of the burden of non-melanoma skin cancer attributable to occupational exposure to solar ultraviolet radiation, for the years 2000, 2010 and 2019 November 2022 Department of Environment, Climate Change and Health, World Health Organization (WHO) Headquarters Background WHO and the International Labour Organization (ILO) produce interagency estimates of the work- related burden of disease (WHO/ILO Joint Estimates) (WHO, ILO 2021). These comprise estimates of exposure to occupational risk factors and burden of disease attributable to these occupational exposures. To establish the evidence base for producing these estimates, WHO and ILO specifically conducted systematic reviews and compiled input data in databases. Objectives We aimed to estimate the proportion of the population occupationally exposed to solar ultraviolet radiation (UVR) and the burden of non-melanoma skin cancer (NMSC) attributable to occupational exposure to solar UVR. Objectives of the country consultation Through this country consultation WHO invites feedback from countries on its estimates. Results Feedback is invited on the following WHO/ILO Joint Estimates: 1. Proportion of the population occupationally exposed to solar UVR; 2. Number of deaths due to NMSC attributable to occupational exposure to solar UVR; and 3. Number of disability-adjusted life years (DALYs) due to NMSC attributable to occupational exposure to solar UVR. These estimates are produced for three years (2000, 2010, 2019) and disaggregated by sex (3 categories: both female and male, female, male) and age group (18 categories: ≥15, 15-19, … , 90-94, ≥95 years). Data sources The estimates were produced using of six sets of input data described below. Input Data 1: Cross-sectional data on proportion of survey participants in occupations classified as exposed to solar UVR WHO and ILO established the WHO/ILO Global Cross-sectional Occupations Database. Microdata were sourced on occupation, measured using 4-digit International Classification of Occupations (ISCO) codes (ILO 2012). The database comprises 166 million observations from 763 Labour Force Surveys collected by statistical offices in 96 countries/areas between 1 January 1996 and 31 December 2021 (Table 1). The microdata were extracted from the source surveys, which countries, areas or Eurostat shared with the ILO Department of Statistics and/or the WHO Department of Environment, Climate Change and Health. The data were aggregated by population defined by country/area, year, sex and age group. Raw data were not modified in any way. Each survey participant’s exposure status was assigned via proxy of occupation (i.e., ISCO code) into two categories: (1) any (or high) occupational exposure to solar UVR, or (2) no (or low) occupational exposure to solar UVR. The occupation-exposure matrix is available from WHO on request. The database is not publicly available to protect the confidentiality of the official survey data it contains. 1 Table 1: Coverage of surveys and countries/areas in the WHO/ILO Global Cross-sectional Occupations Database Region (defined as per WHO classification) World Africa Americas South- Europe Eastern Western East Asia Mediter- Pacific ranean Number of 50 53 11 60 22 37 234 countries/areas Surveys (N) 69 168 49 391 41 45 763 Countries/areas 18 15 8 33 9 13 96 with ≥1 survey (N) (36·0%) (28·3%) (72·7%) (55·0%) (40·9%) (35·1%) (41·0%) (% of countries/areas) Footnote: Areas are assigned to regions following the WHO Coronavirus (COVID-19) Dashboard (WHO n.d.). Input Data 2: Longitudinal data on proportion of survey participants in occupations classified as exposed to solar UVR WHO and ILO also established the WHO/ILO Global Longitudinal Occupations Database. It comprises microdata with repeated measures of 4-digit ISCO codes on 247 year-on-year transitions from quarterly Labour Force Surveys collected by national statistical offices in 31 countries between 1 January 2000 and 31 December 2020 (Table 2). Countries shared these data with Eurostat or the ILO Department of Statistics. These surveys use sample rotation to ensure sample overlaps, repeatedly measuring the same survey participants over consecutive years. Because the microdata did not include individual participant identifiers, we probabilistically linked data longitudinally using matching by household number, household sequence number, sex and birth year. The microdata on occupations were extracted, harmonized into the two exposure status categories using the occupation-exposure matrix, and aggregated by population defined by country, year, sex and age group. Raw data were not modified. To protect confidentiality the database is unpublished. Table 2: Coverage of year-on-year transitions and countries in the WHO/ILO Global Longitudinal Occupations Database Region (defined as per WHO classification) World Africa Americas South- Europe Eastern Western East Asia Mediter- Pacific ranean Number of 47 35 11 53 21 27 194 countries Transitions (N) 21 58 7 143 11 7 247 Countries with 2 5 2 20 1 1 31 ≥1 survey wave (4.3%) (14.3%) (18.2%) (37.7%) (4.8%) (3.7%) (16.0%) (N) (% of countries) Input Data 3: Estimates of the total number of population Estimates of the total number of population by country/area, year, sex and age group for the years 1950-2019 were sourced from the United Nations global population estimates (UN 2019). Input Data 4: Estimates of probability of death Estimates of probability of death by country/area, year, sex and age group were sourced from the WHO life tables (WHO 2020). Input Data 5: Estimates of total number of deaths and disability-adjusted life years Estimates of total number of deaths and disability-adjusted life years (DALYs) for NMSC for the years 2000, 2010 and 2019 were sourced from the WHO Global Health Estimates (WHO 2020). 2 Input Data 6: Estimate of risk ratio A specifically conducted WHO/ILO systematic review and meta-analysis reported a pooled risk ratio of 1.60 for incident NMSC among persons with any (or high) occupational exposure to solar UVR, compared with those with no (or low) occupational exposure to solar UVR (95% confidence interval 1.21-2.11, “moderate quality evidence”) (WHO 2021). It concluded there is “sufficient evidence of harmfulness” of occupational exposure to solar UVR for NMSC (WHO 2021). Methods The estimation strategy comprised modelling the input data (i.e., Input Data 1-6) in four models that consecutively built on each other (Models 1-4). Model 1: Multilevel model to estimate proportion of exposed population at each year For each year between 1980-2016, for each population cohort defined by country/area, sex and age group, we produced estimates of the proportion of the population in each of the two exposure categories (i). We modelled Input Data 1 using the following multilevel model (Model 1): 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑖𝑖 = 𝐴𝐴𝑖𝑖 (𝑠𝑠𝑠𝑠𝑠𝑠, 𝑎𝑎𝑎𝑎𝑎𝑎, 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐) + 𝐵𝐵𝑖𝑖 (𝑠𝑠𝑠𝑠𝑠𝑠, 𝑎𝑎𝑎𝑎𝑎𝑎, 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐) ∗ 𝑦𝑦𝑦𝑦𝑦𝑦𝑦𝑦 where proportioni is the proportion of the population in the exposure category i in a given population cohort; year is the survey year; and sex, age and country are the sex, age group and country/area of the population. The intercept Ai(sex, age, country) and slope Bi(sex, age, country) of the year dependence of proportioni are calculated with a multilevel model with sex and age as fixed effects and sex and age as random effects, nested in the country within the region (with regions treated independently). Because proportioni was non-linearly dependent on age, we linearized age by 5th order orthogonal polynomials to prevent collinearity. Model 2: Model of transition probabilities between exposure status categories For each population cohort, we, along with Eurostat, estimated the probability (𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑗𝑗 ) of transitioning from exposure category i in yeart to exposure category i in yeart+1 (Eurostat 2020). The j denotes one of four possible transitions from one exposure category in yeart to another in yeart+1. Using Input Data 2, the survey weights for the target year (yeart+1) were scaled to represent the correct exposure status by population cohort for the initial year (yeart) and the target year. We modelled Input Data 2 using the following multinomial logit regression model (Model 2): 𝑒𝑒 𝛽𝛽𝑗𝑗𝑋𝑋𝑗𝑗 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑗𝑗 = 1 + ∑𝛼𝛼 𝑒𝑒 𝛽𝛽𝛼𝛼𝑋𝑋𝑗𝑗 where 𝛽𝛽𝑗𝑗 is the set of regression coefficients describing the longitudinal weights associated with the transition j; 𝑋𝑋𝑗𝑗 is a set of explanatory variables (sex and age as a fractional polynomial with maximal permitted degree of 4 associated with the transition j); and the summation (index α) goes through all possible transitions j. With Model 2, we derived transition probabilities for the 11 countries with data in Input Data 2. Eurostat derived transition probabilities modelling quarterly European Union Labour Force Surveys covering 20 countries and shared these transition probabilities with WHO and ILO. For population cohorts for whom the required longitudinal data were unavailable, probabilityj was imputed as the mean of all transition probabilities of the population defined by the same sex and age in the region, weighted by the number of observations contributing to the transition probabilities. Model 3: Microflow model to estimate exposed population over time window For each population cohort, we estimated the proportion (𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑘𝑘 ) of the population in each exposure category (k) over the time window of exposure. We defined k as the exposure category i in 3 any year in the time window. The approach assumed a 20-year latency between exposure and occurrence of clinical disease, and a 20-year exposure time window (centred on exposure year). We simulated microflows of worker cohorts grouped by exposure status, through consecutive age groups, and censored throughout the time window based on modelled probability of death (Input Data 4). 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑘𝑘 is derived using the following microflow simulation model (Model 3): 𝑆𝑆𝑘𝑘 (𝑡𝑡, 𝑐𝑐) = 𝐼𝐼𝑘𝑘 (𝑡𝑡0) + 𝑆𝑆𝑘𝑘 (𝑡𝑡 − 1, 𝑎𝑎𝑐𝑐 (𝑡𝑡 − 1)) 𝑛𝑛 +� ��𝑆𝑆𝑖𝑖 �𝑡𝑡 − 1, 𝑎𝑎𝑐𝑐 (𝑡𝑡 − 1)� ⋅ 𝑇𝑇𝑖𝑖2𝑘𝑘 (𝑎𝑎𝑐𝑐 (𝑡𝑡))� 𝑖𝑖=0 − �𝑆𝑆𝑘𝑘 �𝑡𝑡 − 1, 𝑎𝑎𝑐𝑐 (𝑡𝑡 − 1)� ⋅ 𝑇𝑇𝑘𝑘2𝑖𝑖 (𝑎𝑎𝑐𝑐 (𝑡𝑡))�� − (𝑆𝑆𝑘𝑘 (𝑡𝑡 − 1, 𝑎𝑎𝑐𝑐 (𝑡𝑡 − 1)) ⋅ 𝑋𝑋(𝑎𝑎𝑐𝑐 (𝑡𝑡))) where S is the stock of population in a specific exposure status at the given point in time for a specific population cohort/age; I is the inflow of population into the specific exposure status at beginning of simulation; T is the transition probability from one exposure category to another as a function of the cohort; X is the probability of amount of population being censored as a function of the cohort; c is the specific cohort at the study year (sex, age, country/area); a is the age of the cohort as a function of time (considering ageing); n is the number of considered exposure categories (here: one); and k is the highest exposure status (i.e., any or high occupational exposure to solar UVR). Model 4: Burden of disease estimation model Using the Comparative Risk Assessment framework (Ezzati 2020), we estimated the proportional reduction in death or disease that would occur if exposure was reduced to a level with a minimum risk (i.e., no or low occupational exposure to solar UVR), while other conditions remain unchanged. Information on the population distribution of exposure to the risk factor was combined with information on the increased risk of incident disease that was caused by exposure to the risk factor (Figure 1). Figure 1: Comparative Risk Assessment method for burden of disease estimation Risk factor distribution in the Effect of the risk factor on population the health outcome Population-attributable Total deaths or DALYs from x health outcome fraction Burden of deaths or DALYs from the health outcome attributable to the risk factor Using estimates outputted from Model 3 and Input Data 5 and 6, we calculated the population- attributable fraction (PAF), the proportion of health loss from the disease seen in each population that can be attributed to exposure to the specific occupational risk factor, using Model 4: ∑𝑛𝑛𝑘𝑘=1 𝑃𝑃𝑘𝑘 (𝑅𝑅𝑅𝑅𝑘𝑘 − 1) 𝑃𝑃𝑃𝑃𝑃𝑃 = 𝑛𝑛 ∑𝑘𝑘=1 𝑃𝑃𝑘𝑘 (𝑅𝑅𝑅𝑅𝑘𝑘 − 1) + 1 4 where Pk is the proportion of the population in exposure status category k (exposed); RRk is the risk ratio for the exposure category k; and n is the total number of exposure categories. Applying this fraction to the total disease burden, gives the attributable disease burden. References Eurostat (2020). Labour market flow statistics explained. Luxembourg, Eurostat. https://ec.europa.eu/eurostat/statistics- explained/index.php/Labour_market_flow_statistics_in_the_EU Ezzati et al (2002). Lancet;360(9343):1347‐60. ILO (2012). ISCO–08: International Standard Classification of Occupations. Geneva, ILO. UN (2019). World Population Prospects 2019. New York, UN. https://population.un.org/wpp/ WHO (n.d.). WHO Coronavirus (COVID-19) Dashboard. Geneva, WHO. https://covid19.who.int/ WHO (2020). Life tables. Geneva, WHO. https://www.who.int/gho/mortality_burden_disease/life_tables/life_tables/en/ WHO (2020). Global Health Estimates 2019. Geneva, WHO. https://www.who.int/data/global- health-estimates WHO (2021). The effect of occupational exposure to solar ultraviolet radiation on malignant skin melanoma and non-melanoma skin cancer: a systematic review and meta-analysis from the WHO/ILO joint estimates of the work-related burden of disease and injury. Geneva, WHO. https://apps.who.int/iris/handle/10665/350569 WHO, ILO (2021). WHO/ILO joint estimates of the work-related burden of disease and injury, 2000- 2016: global monitoring report. Geneva, WHO. https://www.who.int/publications/i/item/9789240034945 5 Saatja: <[email protected]> Saadetud: 11.11.2022 19:38 Koopia: <[email protected]> Teema: WHO: C.L.49.2022 (WHO/ILO country consultation on the work-related burden of disease and injury) Dear Member States of the World Health Organization,Please find attached, in the six official languages, C.L.49.2022 regarding the country consultation to review estimates for the work-related burden of disease and injury.Should you have questions regarding C.L.49.2022 please send an email to Best regards,Department of Governing BodiesWorld Health Organization***DO NOT RESPOND TO THIS E-MAIL. YOU WILL NOT RECEIVE A REPLY***