Module 4: Assessment & prediction of the health impacts of climate change

1. Module 4: Assessment & prediction of the health impacts of climate change

2. Key messages in Module 4 • Observational studies are based on the time- & space- specific relationship between health effect & climate factor • Time series studies & spatial studies are the principal methods of analyzing climate-relatedness of a specific health outcome • Disease burden estimates model health impacts

3. Key messages in Module 4 • Weather-health relationship analysis is a basic step for predicting climate-related health effects, but it does not necessarily represent the climate effect on health • Modelling is based on the established relationship between climate factors & a specific health effect • Modelling is a useful tool for predicting future, but not without limits

4. Module 4 outline 1 3 4 2 Uncertainties in analysis & modelling Types of analysis Changing vulnerability Modelling

5. 1 Types of analysis of climate-related health effects

6. Climate change Climate Seasonality Weather Decades Years Months Days FUTUREIMPACTS(mid century onwards) HISTORICALEVIDENCE(recent past) Conventional epidemiology,observation Models, synthesis, ‘triangulation’

7. Types of analysis OBSERVATIONAL • Episodes or event analysis: heat wave, flood, drought, cyclone, El Niño… • Time-series analysis: mortality/morbidity vs. temperature/precipitation • Seasonality: diarrhoea, aero-allergens, vector-borne diseases • Changes in geographical distribution: temperature/precipitation vs. vector borne diseases (VBDs)

8. Types of analysis MODEL-BASED • Health burdens: risk assessments • Decision analysis of health impact of policy options

9. Short-term changes: two approaches • Episode analysis - transparent - risk defined by comparison to local baseline • Regression analysis of all days of year (time-series) - uses full data set - requires fuller data & analysis of confounders - can be combined with episode analysis

10. Meanannual birth weight 1979-1986, Tari, Southern Highland, Papua New Guinea SOI El Niño Source: Allen (2002)

11. Seasonality Cases of diarrhoeal disease Current distribution Distribution under global warming? Date of year

12. Seasonalityof cholera in Bangladesh Source: Hashizume et al. (2009)

13. Inter-annual variation: example of dengue epidemics in the South Pacific, 1970-1998 10 2.0 9 1.5 8 La Niña years SOI (Southern Oscillation Index) 1.0 7 0.5 6 SOI number of epidemics 5 0.0 4 -0.5 3 -1.0 2 El Niño years -1.5 1 0 -2.0 1982 1984 1992 1994 1980 1986 1988 1990 1998 1970 1972 1974 1976 1978 1996 Source: Hales & Woodward (1999) 13

14. Smooth function of date Triangle: attributable morbidity/ mortality Smooth function of date with control for influenza Period of heat Influenza ‘epidemic’ Principles of episode analysis No. of disease/deaths/day January December Date

15. Daily deaths in Seoul, heat wave 1994 Source: Cheong (2015)

16. Daily deaths vs. temperature in Seoul, 1994 Source: Cheong (2015)

17. Episode analyses: interpretation • Common sense, transparent • Relevant to PH warning systems • But • How to define ‘episode’? - relative or absolute threshold - duration - composite variables • Uses only selected part of data • Most sophisticated analysis requires same methods as for regression of all days of year

18. Time-series regression • Short-term temporal associations • Daily/weekly • Suitable for episodes or effects of local fluctuations in meteorological parameters • U- or V-shape of temperature-response function • Different lags

19. Lags • Heat impacts short: 0-2 days • Cold impacts long: 0-21 days • Vary by cause-of-death - CVD: prompt - respiratory death: slow • Should include terms for all relevant lags

20. Summary of time-series • Provide evidence on short-term associations of weather & health • Robust design • Repeated finding of direct heat + cold effects • Some uncertainties over PH significance • Uncertainties in extrapolation to future(No historical analogue of climate change)

21. Changes in geographical distribution of disease (1) BIOLOGICAL MODELS • Use of (laboratory derived) biological evidence (2) STATISTICAL MODELS • Analyses of disease prevalence or vector abundance in relation to geographical factors

22. Scrub typhus & temperature

23. ‘Transmission potential’ for malaria Bites per day Incubation period (days) P(S) per day

24. Estimated population at risk of dengue fever: (A) 1990, (B) 2085 Source: Hales et al. (2002) (A) (B)

25. Conclusions • Most methods of ‘climate’ attribution based on analysis of weather-health associations: episode analysis, time-series, seasonality, inter-annual variations • Relevance to climate change limited by uncertainties over multiple effect-modifiers – changes in vulnerability of population & health • Modelling intrinsic to assessment of likely future burdens & the effect of adaptation options, but entails many uncertainties

26. 2080 2050 2020 Modelled temperature change 2 Modelling the health impacts of climate change

27. 2080s 2050s 2020s 20 Modelled temperature change 19 18 Selected scenario of temperature change to 2100 17 16 15 14 13 1860 1900 1950 2000 2050 2100 Year

28. Estimating future health impacts of climate change • Expert judgment • Simple extrapolation • Mathematical/statistical modeling • Bivariate • Multivariate • Fully integrated

29. Mathematical/statistical models • Simplified representation of a more complex, dynamic relationship • Reduce complexities & background noise to a simpler mathematical representation • Necessarily ‘wrong’ (incomplete, simplified), but useful for: • Insights into processes • Indicative estimates of future impacts • Enhancing communication to peers, public & policy-makers

30. Models • Models are useful • Particularly if the relationship is strong or involves a clear threshold above which an outcome event is very likely • Consistent framework for structuring scientific knowledge • Explore interactions & feedbacks • Models do not predict • Limited knowledge of all factors driving an outcome • Policy-makers must understand that models estimate changes in probability • Models are difficult to validate

31. Future burdens: risk assessment • To demonstrate the potential nature & size of health burdens that may arise under climate change • To provide evidence on the measures needed to protect human health • To provide comparative evidence about the possible effect (on health) of alternative adaptation &/or mitigation policies

32. 2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 3 1.7 1.7 1.7 1.7 1.7 1.7 1.7 1 1.0 1.0 1.0 1.0 1.0 1.0 1.0 2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 3 1.7 1.7 1.7 1.7 1.7 1.7 1.7 1.0 1 1.0 1.0 1.0 1.0 1.0 1.0 2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 3 1.7 1.7 1.7 1.7 1.7 1.7 1.7 1 1.0 1.0 1.0 1.0 1.0 1.0 1.0 2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.7 3 1.7 1.7 1.7 1.7 1.7 1.7 Future burdens: risk assessment GHG emissions scenarios defined by IPCC GCM model: Generates series of maps of predicted future distribution of climate variables Health impact model Generates comparative estimates of the regional impact of each climate scenario on specific health outcomes Conversion to GBD ‘currency’ to allow summation of the effects of different health impacts

33. Temperature distribution Heat-related mortality (Delhi) 140 120 Relative mortality (% of daily average) 100 80 0 10 20 30 40 Daily mean temperature /degrees Celsius

34. Scenarios Coherent, internally consistent depictions of pathways to possible futures based on assumptions about economic, ecological, social, political & technological development. • Scenarios include: • Qualitative storylines that describe assumptions about the initial state & the driving forces, events & actions that lead to future conditions • Models that quantify the storyline • Outputs that explore possible future outcomes if assumptions are changed • Consideration of uncertainties

35. Goals of scenarios • To provide policy relevant analyses of possible consequences of mitigation policies • To better understand the potential impacts of climate variability & change • To facilitate the development & implementation of effective & efficient adaptation strategies, policies & measures to reduce negative impacts

36. Representative concentration pathways (RCP)

37. Projected deaths by cause, according to national income level Source: Friel et al. (2011)

38. 3 Uncertainty in analysis & modelling

39. Relative importance of different uncertainties & their evolution in time

40. Sources of uncertainty • Full range of ‘not improbable’ futures captured? • Model uncertainty • Were appropriate models chosen? • Are assumptions & associations likely to remain constant over time? • Rate, speed & regional extent of climate change • Policy uncertainty • Changes in economic development, technology etc. • How populations in different regions will respond • Effectiveness of mitigation & adaptation strategies & policies

41. Uncertainties • EXTRAPOLATION (going beyond the data) • VARIATION (in weather-health relationship -- largely unquantified) • ADAPTATION (we learn to live with a warmer world) • MODIFICATION (more things will change than just the climate)

42. 4 Changing vulnerability to climate change

43. Changing vulnerability • Changes in population - Demographic structure (age) - Prevalence of weather-sensitive disease • Environmental modifiers • Adaptive responses - Physiological habituation (acclimatization) - Behavioural change - Structural adaptation - PH interventions

44. Projected heat-related deaths in adults >65 due to higher mean annual temperatures, Australia 2100 + ???+ ??? High GHG emissions Low GHG emissions + 14,000 + 6,900 + 11,900 Possible synergistic effect of temperature & aging (especially at higher temperatures than previously encountered) + 6,300 Estimated deaths due to very hot days in 2100 Combined (additive) effect of temperature + aging Independent effect of aging +2,100 + 600 Independent effect of temperature Deaths due to very hot days in 2000 Baseline (current) no. of annual deaths related to heat = 1,100 Source: Woodruff et al. (2005)

45. Observed & simulated malaria distribution for five malaria models Source: Caminade et al. (2014)

46. What we covered in Module 4 1 3 4 2 Uncertainties in analysis & modelling Types of analysis Changing vulnerability Modelling

47. Learning from Module 4 • Observational studies are based on the time- & space- specific relationship between health effect & climate factor • Time series studies & spatial studies are the principal methods of analyzing climate-relatedness of a specific health outcome • Disease burden estimates model health impacts

48. Learning from Module 4 • Weather-health relationship analysis is a basic step for predicting climate-related health effects, but it does not necessarily represent the climate effect on health • Modelling is based on the established relationship between climate factors & a specific health effect • Modelling is a useful tool for predicting future, but not without limits

49. What action might you take in your work, given what you learnt in Module 4?

50. Coming up next… Module 7: Vector-borne disease