Estimating the best way forward: Expert and farmer evaluations of environmental interventions to reduce human exposure to E. coli O157

Paul Cross, Dan Rigby & Gareth Edwards-Jones

RELU funded ‘Reducing Escherichia coli O157 risk in rural communities’

Uncertainty & O157 Management

Uncertainty regarding:

How people become infected (sporadic infection)

Effectiveness of measures

Likelihood of measures being adopted

Is there a problem?

Many potential measures+

Absence of hard (e.g. RCT) evidence on measures to reduce risk+

A (perceived) need to act

= a problem

Managing Uncertainty

1. Identify all possible interventions2. Elicit ‘expert’ opinion on interventions

Aim: Identify best candidate interventionsIdeally =

highly effective+highly practical

Managing Uncertainty

Which experts? Experts (effectiveness)

Inter alia; Public Health, Veterinary Microbiology (Food), Microbiology (Agricultural/Environmental/Clinical), Risk Assessment, Business, Land Management

Farmers (practicality) Members of farmers unions in Wales and Scotland. Farmers’ markets

How to elicit their views? Novel method: Best Worst Scaling

Best-Worst Scaling

Market research tool

Possible to carry out over distance; no face to face; anonymous

Multiple choice based Scaled, fine resolution results Allows respondents to rank long lists without the associated

cognitive gymnastics (bite-sized chunks)

Best-Worst Scaling

Mosteffective

Leasteffective

A Encourage Farmers and farm visitors to wash hands

following contact with farm animals.

B Vaccinate cattle to control pathogen colonisation and

faecal excretion of E. coli O157.

C Remove farm animals from proximity of private water

supplies (e.g. at least 50m from well, borehole or other private water supply by fencing-off).

D Prevent children under the age of 11, and other

vulnerable groups, coming into contact with animals at petting, or public visitor farms.

E No application of manure to ready-to-eat crops within 12

months of harvest and 6 months of drilling/planting.

Best-Worst Scaling analysis

Take all the “most effective” & “least effective” choices

5 item set we gain information on 7of the 10 paired combinations

Respondent chooses the two measures with the maximum difference in performance (best and worst)

Maximises the ability to predict peoples choices

Intervention generation

Literature review Non-systematic Published and grey literature included Project members Opportunity for experts to add interventions to list

Respondent sample

Results of the expert elicitationExperts (Effectiveness)Round 1 Contacted 53 experts 31(75%) completed survey Reduced initial list of 99 to 30Round 2 Contacted 70 experts 41 (60%) completed survey of 30 interventions

Famers (Practicality)Round 3 50 in Wales 50 in Scotland

CV scores: a measure of agreement?

Experts had higher CV scores than farmers for ‘effectiveness’ (p<0.001) and ‘practicality’ (p<0.01)

Farmer CV mean scores were very similar between Scotland and Wales.

The mean confidence interval for the practicality assessment was 0.86 for farmers and 1.46 for experts (p<0.001) and for effectiveness was 0.94 and 1.35 respectively (p=0.002).

Double fencing

Vaccination

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 300

1

2

3

4

5

6

7

8

9

10

Daily cleaning of water troughs

Hand washing

Effectiveness scores

Is there consensus?

-6 -4 -2 0 2 4 6 8 10

-10

-8

-6

-4

-2

0

2

4

Practicality

Effectiveness

Vaccination (Intervention 19)

Is there consensus?

-1 0 1 2 3 4 5

-4

-2

0

2

4

6

8

Practicality

Effectiveness

Reducing stocking densities by 50%

Is there consensus?

-6 -4 -2 0 2 4 6

-4

-2

0

2

4

6

8

Practicality

EffectivenessP &E Scores are widely distributed indicating poor agreement amongst respondents

16. HACCP for manure handling

Practicality scores

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 300

1

2

3

4

5

6

7

8

9

Vaccination

Hand washing

Reduce cattle stocking densities by 50%.

Scottish and Welsh farmers (practicality)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 300

1

2

3

4

5

6

7

8

9

Wales

Scotland

Intervention

Mea

n sc

ore

Locate manure >50m

-3 -2 -1 0 1 2 3 4 5

-4

-3

-2

-1

0

1

2

3

4

5

6

1

2 3

4

5

6

7

89

10

11

12

13

14 15

16

17 18

19

20

21

2223

24

25

2627

282930

Practicality

Effectiveness

19: Vaccination of cattle

1: Hand washing

27: Reduce cattle by 50%

12: Septic tank leakage High practicality/

low effectivenessLow practicality/ low effectiveness

Low practicality/ High effectiveness

High practicality/ High effectiveness

Best-Worst Scaling 2 x 2 plots

Method suitability

Best suited to the evaluation of large sets of standalone measures

Best suited to the evaluation of multidimensional measures (effectiveness and practicality)

Smaller multi-level sets of measures, identify the ‘best’ bundles (ACA, CBC)

Future

Applications in other areas E.g. Cryptosporidium, Salmonella, anti-microbial resistant genes

Modelling of interventions % reduction in disease prevalence

Development of MACCs % reduction/cost

Participating institutions

Health Protection Agency; Health Protection Scotland; University of Glasgow: School of Veterinary Medicine; Scottish Agricultural Ccollege; Veterinary Laboratory Agency; Scottish Infection Research Network; HPS Colindale; FSA: Microbiological Safety Division; Bioss Scotland; Wageningen University and Research Centre; Liverpool John Moores University: Centre for Public Health; NFUScotland; NFUCymru; Farmers Union Wales; Office of the Chief Veterinary Officer Wales; Meat Promotion Wales; Quality Meat Scotland; Countryside Council Wales

Acknowledgements

Thank you

Top interventions by effectiveness and practicality

No. Intervention1 Encourage Farmers and farm visitors to wash hands following contact with farm animals.

4 Prohibit recreational activities (such as walking and camping) to land where manure, slurry or abattoir waste have been applied, or animals and faeces present, in the previous four weeks.

5 Monitoring of private water supplies to identify those with either high indicator counts, or those in areas of high risk. These supplies would need to be treated (e.g. by ozonation, chlorination or ultra-violet treatment).

7 Keep livestock and pets out of ready-to-eat crop areas, using fencing for example.

8 No application of manure to land at high risk of direct flow to watercourses (e.g. adjacent to a watercourse, borehole or road culvert, or areas with a dense network of open drains.

9 Locate solid manure heaps and slurry pits at least 50m away from watercourses, field drains and ready-to-eat crops.

13 No slurry or livestock manure to be applied to high risk fields (ie high risk of transport into adjacent areas watercourses (e.g. when soils saturated or frozen, or heavy rain expected).

19 Vaccinate cattle to control pathogen colonisation and faecal excretion of E. coli O157.