Papers

Later effects of grounding of tanker Braer on health in Shetland

BMJ 1994; 309 doi: http://dx.doi.org/10.1136/bmj.309.6957.773 (Published 24 September 1994) Cite this as: BMJ 1994;309:773
  1. D Campbell,
  2. D Cox,
  3. J Crum,
  4. K Foster,
  5. A Riley
  1. Scottish Centre for Infection and Environmental Health, Ruchill Hospital, Glasgow G20 9NB Shetland Health Board, Brevick House, Lerwick, Shetland ZE1 0RB Grampian Health Board, Aberdeen AB9 1RE Lothian Health Board, Edinburgh EH8 9RS
  1. Correspondence to: Dr Campbell.
  • Accepted 16 June 1994

The public health response to the oil spillage from the tanker Braer in January 1993 included a cohort study to examine its effects on health.1 The first phase has been reported.2 We report longer term effects.

Subjects, methods, and results

In June 1993 participants in the first phase were given an extended questionnaire. A systematic sample of non-participating exposed people was surveyed by telephone. Participants aged 16 and over completed the general health questionnaire (28 questions).3 Weight, peak expiratory flow, urine analysis haematology, and liver and renal function were measured. For categorical data X2 values or Fisher's exact tests with two tailed probability were computed; testing of hypotheses was performed by determining odds ratios and confidence intervals by using discordant pairs when applicable. For continuous data, means and mean differences for paired data and their 95% confidence intervals were calculated.

Overall, 344 of the original 420 subjects (80%) and 77 of the 92 controls (84%) participated. The distribution of sex and age and the proportions exposed inside and outside were no different from the first phase. Sixteen of the 76 (21%) non-responders surveyed did not differ in distributions of sex and age from participants. Of exposed people, 7% (25/334) perceived their health to be poor compared with none of the 77 controls (X2=8.05, df=3 P<0.05). Significantly more considered it to have altered since the incident, 16% (54/334) reporting deterioration and 2% (7/334) improvement whereas only 3% (2/77) of controls described deterioration (X2=11.80, df=1, P<0.001) Of exposed people, 6% (21/334) attributed the change to the Braer, with 3% (9/334) reporting both poor health and a decline in health; the remainder, though reporting deterioration, still considered it to be good.

Comparison of the symptoms of exposed people in the two weeks before this phase with their presence immediately after the incident showed more tiredness (odds ratio 1.86, 95% confidence interval 1.19 to 2.92) and fever (2.25, 1.14 to 4.44), fewer throat (0.45, 0.31 to 0.66), skin (0.38, 0.20 to 0.70), and eye irritations (0.18, 0.10 to 0.32), and fewer headaches (0.55, 0.36 to 0.83). More controls had tiredness (4.00, 1.13 to 14.17) and headaches (5.50, 1.22 to 24.81) in June than after the event. In June exposed people were more likely to report throat irritation (3.37, 1.33 to 9.05) and breathlessness on exertion (4.81, 1.09 to 29.92) in the previous 14 days. A greater proportion of those exposed reported weakness (9.18, 1.32 to 182.96) at some time over the five months (table). Eighty six per cent (31/36) declared these symptoms to be persisting, with one third describing their onset in January. There were no differences in perception of health and onset of symptoms between the non-responders and participants.

Numbers of people reporting symptoms with onset since Braer incident by type of exposure

View this table:

The mean general health questionnaire score of those exposed (2.92, 2.33 to 3.50) was significantly greater than that of controls (0.83, 0.25 to 1.41). A greater proportion of exposed subjects (60/254, range 0-25) was above the threshold score compared with controls (2/59, range 0-13) (X2=9.15, df=1, P<0.003). The mean scores for those exposed inside (2.71, 2.05 to 3.37) or outside (3.42, 2.27 to 4.61) did not differ. Those exposed had greater overall scores compared with controls for somatic symptoms (1.15 (0.92 to 1.38) v 0.29 (0.10 to 0.48)) and anxiety and insomnia (0.83 (0.63 to 1.03) v 0.19 (0.02 to 0.36)) but not for personal dysfunction and severe depression. Among the biological markers examined only aspartate aminotransferase activity was different between the two phases (0.40 (−0.38 to 1.18) v −179 (−2.83 to −0.76)) but not between type of exposure.

Comment

Any overreporting of self reported symptoms would tend to magnify possible health effects. Although the drop out rate between the two studies was appreciable, the telephone survey showed no differences in demography, symptoms, or exposure. The health of exposed participants and controls was similar, though probable upper respiratory infection was detected. A greater proportion of exposed people considered their health to have deteriorated, but only nine reported it as poor because of the incident.

The finding that both groups were more tired in June and the greater proportion of exposed people reporting weakness are unexplained. Whether a physical abnormality is being measured is arguable as only four people reported problems with work. The few women who were pregnant or conceived during the incident delivered without adverse outcomes (M Hunter, C Rowlands, personal communication).

Human disasters are recognised as causes of long term mental illness and community disruption.4 The general health questionnaire detects psychiatric disorders in community settings.3 Of those exposed, 24% scored above the level at which a subject could be considered a case compared with 3% of controls. This was not related to potential level of exposure and may be in response to strains on the fabric of this community.

The results of this phase of the study were presented to the people of the South Mainland of Shetland at a public meeting held on 13 September 1993.

We thank of the staff of the department of laboratory services, Gilbert Bain Hospital, Lerwick. Without the continuing co-operation of the people of Shetland this study could never have occurred. This work was supported by funding form the Secretary of State for Scotland. The views expressed are ours.

Members of the Shetland Health Study Group were: Mr C Cumming, Gilbert Bain Hospital, Shetland; Drs P Christie, D Brewster, Common Services Agency, Scottish Health Service, Edinburgh; Dr V S G Murray, National Poisons Unit, London; Dr J S Oliver, Glasgow University, Glasgow; Dr A Proudfoot, Edinburgh Royal Infirmary, Edinburgh; Mrs M Stove, Mrs M Robertson, Miss K Hudson, Shetland Health Board, Shetland Dr M D Hunter, Dr C M Rowlands, Mrs S Mann, Levenwick Medical Practice, Shetland; Dr A Min, Mrs S Thompson, Hillswick Medical Practice, Shetland, M R Hall, Shetland Islands Council, Shetland; and Dr G I Forbes, Environmental Health (Scotland) Unit, Glasgow.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.