Water safety frameworks in developing countries: science-policy linkages

1. Water safety frameworks in developing countries: science-policy linkages Dr Guy Howard, DFID

2. Policy and science • Policy requires evidence – the role of science • BUT other factors – social, ethical, political also important • SO policy often evidence-informed rather than evidence-based • Policy works through formal and informal processes • don’t get fixated on policy documents

3. Policy and science – fraught with mis-understanding • Scientists think policy-makers do not follow evidence • Policy-makers think scientists narrow technicians • Further complicated when science is • Limited in its development in-country • Small number of ‘voices’ dominate • Limited operational research of policy relevance

4. Water safety framework – science and policy in action Water safety framework comprises (WHO 2004): • Health-based targets • Water safety plans • Surveillance • Driven by improving public health • Also allows transparent trade-offs • Applied in Bangladesh

5. Health-based targets and quantitative risk assessment

6. Bangladesh context • In early 1990s reached 97% coverage in rural areas (WHO & Unicef 2000) – contribution to reduced diarrhoea • 1993 arsenic first detected • Survey (1999-2000) indicates 27% tubewells >50µg/l and 46% >10 µg/l (BGS& DPHE 2001) • Blanket testing shows 29% > 50µg/l (about 20% of country total)

7. Policy context • Technologies identified for use in mitigation • Dug wells • Rainwater harvesters • Pond sand filters • Deep tubewells • Arsenic-removal technologies • Surface water use prioritised • Priority given to 1st 3 technologies • Others only to be used when these failed

12. Risk substitution • These policy choices equate to ‘Specified technology Health-based targets’ • BUT no consideration given to risk substitution • Pathogens • Cynaobacterial toxins • Other chemicals (natural or anthropogenic) • QHRA undertaken to estimate impact of these risks and benefits of policy choices

13. RAAMO method (abbreviated) • Representative set of main technologies tested monsoon & dry season using cluster sampling • Model developed with output in DALYs for microbial & arsenic risks • 3 ‘reference’ pathogens • Composite model bacteria, cryptosporidium and rotavirus • Input data from indicator organisms - derived relationships to pathogens • Arsenic disease burden input data direct arsenic measurement and focus on main health outcomes

14. Model architecture (Howard et al 2007) [E. coli]in sewage [Pathogen] in sewage Measured[TTC]for option Volume ofwater consumedunboiled locally Ratio[pathogen]:[E. coli] Measured[TTC]locally Ratio[TTC]:[E. coli] Measured[E. coli]locally Dose of pathogens Predicted[pathogen]for option Predicted[E. coli]for option Inputs to or outputs from the model Infection withRotavirus, Shigella,Cryptosporidium Cancer of lung, bladder and skin Measured[Arsenic]for option Model process steps Disease Burden for option

15. Results: Microbial DALYs

16. Results: arsenic DALYs

17. WHO reference level of risk • GDWQ suggests reference levels of risk can be used in setting health-based targets • Suggested 10-6 DALYs – broadly equivalent to 10-5 lifetime cancer risk used in chemical guideline derivation • Conceptually elegant, but poses practical problems • E.g real-life considerations of impact of diarrhoea & cancer

18. RAAMO findings reference risk • Microbial contaminants • Reference risk not achievable for any technologies year-round • Arsenic standard already much higher DALY score than WHO reference risk • Not clear how useful the concept of ‘reference risk’ is in such circumstances • Better to use comparative measures & aim for best achievable result

19. Changes in practice & policy • Chlorination on dug wells and PSFs increasingly standard – overcame previous resistance • Led to re-emphasis by implementing agencies on technologies with lower health risk • Review of National Policy and Implementation Plan now initiated

20. Water safety plans and surveillance

21. Water safety plans • Identified as critical to improving water safety from RAAMO • O&M failures particularly highlighted in relation to microbial quality • In principle agreement by all major players that should be followed • BUT needed locally relevant WSPs with evidence in-country of their effectiveness

22. The process that was followed • National conference on water quality • Government, donors & NGOs agree need for pilot projects • Workshop to develop a set of draft WSPs for rural technologies • Govt, donors, NGOs involved • Pilot projects to apply WSPs • Consolidation of experience and revision of WSPs

23. Generic WSPs developed Cover all water supplies with particular technology Developed using proformas & existing knowledge and experience Systematic assessment of: hazards, degree of risk, control measures, monitoring, validation and verification Actions plans developed Developing the WSPs

24. Protected dug wells Pond sand filters Rainwater harvesters Deep tubewells Shallow tubewells Piped systems From tubewell From surface water after multi-stage filtration Gravity-fed (minor) Later Arsenic removal systems Technologies considered

25. Pilot projects • 3 NGOs covering 82 communities across Bangladesh (also Unicef/GoB) • Every major technology addressed • Baseline, intermediate and final water quality assessments undertaken • Community caretakers supported with pictorial monitoring tools

26. Key findings • Reductions in sanitary risks for all technologies • Microbial quality improved • Water handling hygiene practices improved • 12% reduction in diarrhoeal disease in one pilot project • Greater caretaker accountability • Pictorial tools found useful

28. Water quality risk grading scheme

31. Evidence to practice • All major water rural programmes committed to WSPs • Moved into 2nd edition – regular interactions • Capacity support offered via ITN-BUET (training, consultancy) • See www.buet.ac.bd/itn

32. Key lessons • Bangladesh showed possible to get wide stakeholder buy-in • Important to get all major players agreed on single set of products and approaches • Need to develop more standardised approaches • More cost-effective • Consistency and coherence

33. Surveillance • Surveillance protocol developed and approved by GoB • Limited but practical • Emphasis on cost-effective survey methods • Very difficult to implement • Funding for monitoring scarce and very limited

34. Implications and conclusions

35. Implications • RAAMO & WSPs showed could get change in policy and practice • BUT also highlighted informal as well as formal policy processes • Formal policies relative recent so no immediate change likely • Water & Sanitation Program now leading process of formal policy review

36. Future challenges • Climate change and impacts on water supply and sanitation • DFID/WHO funding joint work • Decadal forecasting to 2020 and 2030 • Assessment of technology susceptibility • Focus on ‘hotpsots’ and longer-term policy implications

37. Conclusions • Bangladesh water safety framework • Good illustration of policy-science interactions • Uncertainty could be managed provided transparent and quantified • Evidence based practice important • Highlighted value of informal as well as formal policy processes • This type of approach needs wider replications