Collecting research data in emergency situations: the example of the Fort McMurray wildfires. Nicola Cherry St John s OEMAC June 12 th 2017

Transcription

1 Collecting research data in emergency situations: the example of the Fort McMurray wildfires Nicola Cherry St John s OEMAC June 12 th 2017

2 Problems for disaster epidemiology After a disaster: People disperse People are upset/distraught Research just another intrusion Response rates likely to be low Volunteers, if any, likely to be very non-random Baseline, pre-disaster, health indices difficult/impossible to obtain. Earliest days post-disaster may provide crucial information: the researcher needs to move quickly but the disaster area still in disarray

3

4 The Fort McMurray Fire City of some 88,000, hub of the Northern Alberta oil and gas industry. Total evacuation 3 rd May One road out north to mining work camps. south to large conurbations (?) firefighters from Alberta more national/international. Firefighting within the city most intense in early mid May: continued burning away from the city until late August

5 Myriad research questions: rapid decisions Which (working) population? First responders. City population Work camp residents Which exposures? Environmental (particulates, chemical) Psychological (fear, disruption) Evacuation/events since Which outcomes?

6 What is feasible? Evacuation on May 3 rd : city shut down. Conditional return of evacuation from June 1 st. No access for researchers. How do you find your population?

7 Serendipity: Study1 3 days before the fire started we had completed recruiting a cohort of 151 manual workers in Fort McMurray for a study of occupational injury in inter-provincial workers. We also had participants in Fort McMurray from a national cohort of workers in the welding and electrical trades Could these cohorts be useful?

8 Why do existing cohorts matter? Identified before residents are widely dispersed following evacuation Decision to take part is independent of health status after the event Commitment to the research before the disaster enhances participation Collection of pre-event health indicators helps analysis of change and susceptibility.

9 Study 1: research question What was the effect of the events surrounding the fire and evacuation on the health and mental well-being of people working in/around Fort McMurray?

10 Participation 130 competed the post-fire questionnaire (75% of injury cohort plus 16 from trades cohort). 21 not Fort McMurray on May 3 rd. All but 6 of the 109 present were evacuated So evacuation cohort N=103; non-evacuated N=27. Follow-up contact (by mail, on line or by telephone) was, on average, 102 days post fire.

11 Health issues reported Did you have health problems cause or made worse by the fire During or immediately after the fire? Respiratory 15.6% Mental ill-health 13.8% Now (when completing the questionnaire) Respiratory 1.8% Mental ill-health 7.3%

12 Effects on mental health Anxiety and depression scores both significantly higher in those evacuated. 17% of those evacuated had scores consistent with moderate or severe anxiety or depression at the time of completing the questionnaire

13 Relation to work Depression was a function only of financial loss due to lack of work. 1 in 4 had not worked since the evacuation and < half had yet returned to Fort McMurray. >90% planned to do so.

14 Conclusion from study 1 Convenience sample of workers recruited before the fire show persisting ill-health post fire was not widespread. This study, and those from previous disasters, suggests that most will go on to successfully reestablish their lives and will contribute to the economic re-establishment of Fort McMurray

15 Serendipity:study2 Delivery of the MobLab

16

17 Mobile clinical laboratory CFI funded MobLab, delivered on the day Fort McMurray evacuated. Testing rooms set up for spirometry. Clinical laboratory equipped with -80 freezer, centrifuge for biological specimens University vehicle pool pre-prepped to provide insurance, maintenance, driver

18 Study population Funding protocol had identified firefighters as a hard to reach occupational group of interest. Alberta Labour indicated interest in funding a study of the effects of the fire on first responders. Exceptionally approachable local fire chief (Strathcona)

19 Within 2 weeks.. Had MobLab licensed and on the road A protocol written and put through University ethics board We had devised a questionnaire on firefighter practices, exposures, PPE and respiratory and mental-heath Van equipped for collection of clinical samples

20 Early field work First firefighters recruited at Strathcona as they returned from deployment to Fort McMurray on 16 th May, 14 days after the start of the fire. Phase 1 (May-September) recruited 355 firefighters from 13 fire services including Fort McMurray itself (after access to the city had reopened) and industrial fire fighters from Syncrude, both in the thick of the fire from day 1.

21 Data collected Exposure and health questionnaire all Spirometry all Consent to follow-up, access to prior lung function testing and linkage to Alberta administrative health databases all Follow-up data at 3 months just Strathcona Blood: Strathcona (early/late) and Fort McMurray Urine: Strathcona, FMM, Syncrude

22 Key comparisons 1) Strathcona firefighters at 2 weeks and weeks. Did effects reverse? 2) Strathcona at weeks and Fort McMurray city fire fighters at weeks. Were the higher exposure for the Fort McMurray fire services reflected in worse outcomes at the same point in time? 3) Syncrude and Fort McMurray firefighters? Similar engagement: same outcomes?

23 Estimation of exposures Invaluable collaborations with Alberta Environment: Measured/extrapolated concentration of PM2.5 every day (May-June) for each area of the city. Information on composition of the smoke

24 Particulate (PM2.5)concentrations May 1-June 30th (Alberta Environment)

25 Exposure composition: organic

26 Exposure concentration PAHs

27 Estimated exposure to individual firefighters We have (from detailed questionnaires): When they were (first) deployed What hours they worked each day Where they worked What tasks they did What respiratory protection they wore Their report of smoke intensity Alberta Environment estimates for each location each day

28 Locations

29 Cartridge Respirators Used in Firefighting Full Face or Half Face? Filters and Cartridges P100 filters are generally pink or purple: Full face respirator: covers the entire face and has built-in eye protection Half face respirator: covers the nose and mouth only Filters may be combined with a gas or vapour cartridge. The colour of the cartridge label depends on the gas or vapour targeted. Other Products Organic vapour cartridge (black) and P100 filter Multiple gas and vapour cartridge (olive) and P100 filter WHIFFS wrap-around mask with moist gel filter insert Respro FB-1 mask with dry filter insert P100 filters with nuisance organic vapour relief N95 or P95 filters are generally white and may also be combined with gas and vapour cartridges

30 Smoke Level Rating Guide Light smoke level Medium smoke level (difficult to see beyond 100 yards) Heavy smoke level Very heavy smoke level From Reinhardt and Ottmar, Smoke Exposure at Western Wildfires, USDA, 2000

31 Exposure estimates for individual firefighters by start of deployment

32 Cumulative weighted exposure by fire service

33 Highest mean particulate exposure

34 Compound exposure index Cumulative exposure index based on Albert Environment PM2.5 Task estimates: based on respondents ratings of smoke density (actively attacking versus patrolling) Protection factor: based on self-reports of type of RPE, how often used, how often filters changed. High score=poor protection. Compound exposure index= Cumulative index x Task index x RPE index

35 Health outcomes: respiratory Own reports of respiratory ill health caused or made worse by the fire immediately after the fire and now Own reports using a visual analogue scale of being bothered by: cough; phlegm: wheeze: breathlessness; chest tightness before the fire, immediately after the fire and now Spirometry (before, at recruitment, later)

36 Health outcomes: mental health Own reports of mental ill-health caused or made worse by the fire immediately after the fire and now Scores on the health checklist (Hospital Anxiety and Depression Scale) completed as part of the questionnaire

37 Results: Reported ill-health (N=355) Respiratory Mental ill-health Immediately after the fire 40.2% 4.2% Now 19.4% 3.1%* * 15% by anxiety and depression scale

38 Reported respiratory ill-health: caused or made worse by the fire by fire service Immediately after At questionnaire Strathcona 40.6% 21.8% Fort McMurray 61.0% 32.9% Other structural 28.7% 11.3% Syncrude %

39 Cumulative weighted exposure by fire service

40 Chest symptoms before and immediately after May 3rd Cough not at all x x Very bothered Phlegm not at all x x Very bothered Undue breathlessness not at all x x Very bothered Wheezing or whistling in your chest not at all x x Very bothered A sensation of chest tightness not at all x x Very bothered

41 Respiratory symptoms immediately post fire Cough Phlegm Breathless Wheezing Tightness Strathcona Fort McMurray Other structural Syncrude p=

42 Relation of respiratory outcomes to compound exposure index Self-report of respiratory ill-health very strongly related (P<0.001) to compound exposure index. Visual analogue reports of respiratory symptoms: Immediately post fire all 5 symptoms highly correlated with compound exposure index. Only breathlessness, wheeze and tightness related to exposure at time of the questionnaire.

43 Spirometry post fire (% predicted) FEV1 FEV1/FVC MEF75 PEF MMEF Strathcona Fort McMurray Other structural p=

44 Clinical obstruction post fire? FEV1/FVC% <80 80< N n % n % n % Strathcona Fort McMurray Other structural chi sq= 9.96 p=0.041

45 Relation* between compound exposure index and spirometry (percent predicted) post fire FEV1 MEF75 PEF B p= B p= B p= Compound Index Ever smoked * Adjusted for smoking

46 Conclusions so far on respiratory health Exposures during the fire are strongly related to: symptoms, self-reported ill-health spirometry (to some degree).

47 Mental ill-health Self reports of mental ill-health caused or made worse by the fire. Almost entirely confined to the Fort McMurray fire service 15% immediately post fire. 11% at the time of the questionnaire.

48 Score on HADS scales (anxiety or depression) % with N high score Strathcona Fort McMurray Other structural Syncrude

49 Reporting of worst moment Worst moment coded as Psychological* Physical Strathcona 34.7% 68.3% Fort McMurray 78.0% 41.5% Other structural 36.0% 58.0% Syncrude 72.7% 36.4% *Relates strongly to mental-ill health

50 Conclusions so far on mental ill-health Fire-fighting in your own backyard combined with very long rotations is related to increased risk. Protective factors for Syncrude yet to be explored (high risk/low effect)

51 What evidence do we have of reversibility? Self-reports of effects immediately post fire and now. Repeat measure in Strathcona at 2 weeks and weeks Time between last day of last deployment and symptoms/signs at weeks in Strathcona and Fort McMurray

52 Recovery made from respiratory problems caused or made worse by the fire % reporting Ill-health Immediately after At 2 weeks At weeks Strathcona 39.4% 25.4% 12.7% Fort McMurray 61.0% 32.9%

53 Recovery of chest symptoms - Strathcona Prior 2 weeks weeks p= (late v prior) Cough 7.7 p= p= p= 0.46 Phlegm 7.4 p= p= p= 0.64 Breathlessness 3.1 p= p= p=0.01 Wheeze 2.9 p= p= p=0.00 Tightness 3.3 p= p= p=0.00

54 Stability of spirometry results (matched pairs) Strathcona (N=71) 2 weeks weeks p= FVC FEV FEV1/FVC MEF PEF MMEF

55 Spirometry prior to fire: repeated measure at Strathcona (N=28) p= Prior 2 weeks weeks FVC FEV FEV1/FVC PEF MMEF

56 Lack of reversibility of scores on HADS scale (mental ill-health) % with high HADS 2 weeks weeks Strathcona 2.8% 5.6% Fort McMurray 34.5%

57 Conclusions so far on reversibility Although respiratory symptoms improve with time there is still an excess of reported breathlessness, wheezing and chest tightness even in Strathcona, with the lowest exposures. FVC and FEV1 appeared to get marginally worse at weeks No evidence of improvement in mental health.

58 Inflammatory markers Measured in plasma from blood taken at Strathcona (early and weeks) and Fort McMurray only. 42 markers reduced to 6 factors by component analysis One factor related closely to exposure: we think that high exposures caused the inflammation reflected in this measure. Does this inflammation reverse with time?

59 Mean Predicted Factor-3 On Last Day of First Deployment Observed and extrapolated biomarker score from plasma 2.5 Rate of Decline Factor Number of Days From Last Day of First Deployment to Day of Sampling Fort-McMurray Predicted Strathcona Predicted Fort-McMurray Observed Strathcona Observed Linear (Fort-McMurray Predicted) Linear (Strathcona Predicted)

60 Questions this marker can potentially answer 1) Is this internal measure of exposure a better predictor of (chronic) ill-health than external exposures estimates? On effects to date, compound index does better 2) Can values on this biomarker be used to validate our RPE protection index. Extrapolated biomarker correlates positively (but weakly: r=0.19; p=0.085) with poor respiratory protection

61 Conclusions from firefighters The Fort McMurray fire was associated with poorer respiratory health during the first 4 months post fire. High exposure and long hours of work were associated with both early respiratory effects and poorer mental health. Evidence from follow-up at Strathcona suggests that early effects on the respiratory system and blood markers do reverse in most people (but still present in some) Assessment of longer term health outcomes is important. Phase 2 is now underway : ALL firefighters deployed to Fort McMurray are invited to join online.

62 Conclusions on collecting data in emergency situations: lessons learned You need: Luck (serendipity) Uncommitted start-up funds Highly engaged community (ethics, insurance, approved driver, funding agency. ) Experienced, flexible, multi-tasking research team (with no family, no interest in time off.) A completely empty diary: this will all be extra

63