HEARING LOSS IN AIRCRAFT MAINTENANCE TECHNICIANS

1
HEARING LOSS IN AIRCRAFT MAINTENANCE TECHNICIANS
Maya Guest1, May Boggess2 1 Faculty of Health,
University of Newcastle, Australia. 2 Department
of Statistics, Texas AM University, College
Station, USA
2
F111, a flying fuel tank
3
Deseal, then reseal a fuel tank

• The fuel tanks were a confined space
• They were very cramped, with tradespeople
  crawling around the braces

4
Health Concerns

• Concerns about various symptoms experienced by
  workers were raised in early 1999 with symptoms
  including
• memory loss
• fatigue
• neurological problems eg. colour vision
• Deseal/reseal activities ceased in 2000
• The SHOAMP Study

5
The SHOAMP Study

• Aims
• compare series of general health, medical and
  neurophysiological outcomes between F-111
  deseal/reseal personnel and appropriate
  comparisons
• Research Question
• Is there an association between adverse health
  status and an involvement in F-111 deseal/reseal
  activities?
• Study Design
• Retrospective cohort postal questionnaire exam

6
Study Group

• Exposed group
• in Deseal/Reseal program at Amberley
• N 616
• Two comparison groups
• same time, but in non-technical at Amberley
• N406
• same time, but in technical at Richmond
• N516
• Total health exams
• N1538

7
Measuring Hearing Thresholds

• Pure-tone audiometry at the frequencies of 0.5,
  1, 2, 3, 4, 6, 8 kHz for air conduction
• Australian Standard AS1269.4.1998 by trained
  nurses
• Measures threshold dB (smaller is better)

8
Hearing Threshold result one person
9
Hearing Threshold result one person
10
How to compare groups?

• Treat each frequency separately
• Do 95 confidence intervals for each group
  overlap?

11
Mean (95 CI) Hearing Thresholds
12
Distribution of Hearing Thresholds lower
frequencies
13
Distribution of Hearing Thresholds higher
frequencies
14
Problems

• 16 observations on single person will be
  correlated
• Distribution heavily skewed
• Multiple test correction (eg. Bonferroni) needed
  to control overall error rate
• Other factors need to be controlled for
• AGE!

15
The ISO-7029

• The ISO-7029 statistical distribution of hearing
  thresholds as a function of age provides by
  gender the expected median value of hearing
  thresholds relative to the median threshold at
  the age of 18 years and the statistical
  distribution above and below the median value for
  the range of audiometric frequencies from 125 Hz
  to 8000 Hz for populations of otologically normal
  persons of given age between 18 and 70 years

16
ISO 7029 healthy pop. lower frequencies

17
ISO 7029 healthy pop. higher frequencies
18
Quantile model?

• Mean regression coefficients estimated by
  minimizing the sum of the squares of the
  residuals
• Quantile regression coefficients estimated by
  minimizing the sum of the absolute values of the
  residuals

19
Quantile model?

• Mean regression 1821 Gauss showed it was ML,
  least variance IF residuals are normal.
• Quantile regression 1818 Laplace showed it had
  smaller variance than mean for certain
  distributions with long tails.
• Central Limit Theorem is not a cure all.

20
Statistical Analysis quantile model to compare
to normal population

• Response hearing threshold (dB)
• Explanatory variables
• Frequency, Age
• Posting category, Rank category
• Alcohol consumption category, Smoking status
• Diabetes status
• SSRIs (anti-depressants), malaria medication
• Ringing in the ears
• Exposure group, civilian solvent exposure
• Bootstrap standard errors correlation within
  person.

21
Statistical Analysis quantile model to compare
to normal population

• Statistically significant explanatory variables
• Frequency, Age
• Smoking status, Diabetes status
• SSRIs (anti-depressants)
• Ringing in the ears
• Exposure group
• Clinically significant variables
• Frequency
• Age

22
Result table

• --------------------------------------------------
  ----------------------------
• Coef. Std. Err. z
  Pgtz 95 Conf. Interval
• -------------------------------------------------
  ----------------------------
• _Ifrequen10 -1.051308 .3152126 -3.34
  0.001 -1.669113 -.4335024
• _Ifrequen15 -.6029429 .3331257 -1.81
  0.070 -1.255857 .0499715
• _Ifrequen20 -1.881937 .3809262 -4.94
  0.000 -2.628539 -1.135336
• _Ifrequen30 .8669362 .4485159 1.93
  0.053 -.0121388 1.746011
• _Ifrequen40 7.536735 .4582282 16.45
  0.000 6.638624 8.434846
• _Ifrequen60 8.997418 .3901236 23.06
  0.000 8.23279 9.762046
• _IfreXg40_2 -2.720623 .9044762 -3.01
  0.003 -4.493363 -.9478817
• _IfreXg60_2 -2.208105 .8227658 -2.68
  0.007 -3.820696 -.5955134
• _IfreXg80_2 -3.32165 .7481134 -4.44
  0.000 -4.787925 -1.855374
• age -.4541155 .2376783 -1.91
  0.056 -.9199565 .0117254
• age2 .006715 .002729 2.46
  0.014 .0013662 .0120638
• frequency .2533864 .1252859 2.02
  0.043 .0078305 .4989422
• fa -.0225578 .0058333 -3.87
  0.000 -.0339908 -.0111248
• fa2 .0004074 .000067 6.08
  0.000 .0002761 .0005387
• _Ismoke_ca3 1.7394 .5696168 3.05
  0.002 .6229713 2.855828

23
Predicted hearing median threshold
24
Predicted hearing median threshold
25
Predicted hearing median threshold
26
Predicted hearing median threshold
27
Conclusion

• Need to reconsider noise exposure limits if
  workers are additionally exposed to chemicals
• Need to reconsider the efficacy of hearing
  protectors in combined exposures

28
Take home message

• No one-size-fits all in statistics
• Central Limit Theorem is not a cure-all

29
The TUNRA Study Team

• Principal Investigators
• Catherine DEste, Associate Professor in
  Biostatistics, Centre for Clinical Epidemiology
  Biostatistics, The University of Newcastle.
• John Attia, Senior Lecturer in Epidemiology,
  Centre for Clinical Epidemiology Biostatistics,
  The University of Newcastle Academic Consultant,
  Hunter Area Health Service
• Anthony Brown, Director of Primary Health Care
  and Population Health, Macquarie Area Health
  Service Conjoint Associate Professor,
  Environmental and Occupational Health, The
  University of Newcastle.
• Julie Byles, B.Med, PhD, Professor and Director,
  Centre for Research and Education in Ageing
  (CREA), Faculty of Health, The University of
  Newcastle.

• Associate Investigator
• Robert Gibberd, Associate Professor, Centre for
  Clinical Epidemiology Biostatistics, The
  University of Newcastle.
• CEO of TUNRA Ltd
• Soozy Smith, PhD, TUNRA Ltd, The University of
  Newcastle.
• Project Support
• Meredith Tavener, Project Manager.
• Richard Gibson, Associate Lecturer in
  Biostatistics (Research), Centre for Clinical
  Epidemiology Biostatistics, The University of
  Newcastle, Project Statistician.
• Maya Guest,. Research Higher Degree candidate,
  PhD Fellow for SHOAMP.

30
Questions/Thank you etc.