METHODS

Deaths at sea are not usually registered with the local registrars of deaths and are not included in routine national occupational mortality statistics, as published by the Office of Population Censuses and Surveys (OPCS) and the Office for National Statistics (ONS). Thus standardised mortality ratios for fishermen and seafarers, as reported by these sources, are often grossly underestimated.^5,11 This is because deaths at sea are registered separately at the Registrar General for Shipping and Seamen (RGSS). In this study, details of all deaths in British fishing were identified from official mortality files held at the RGSS, and at the principal British marine investigative authority, the Marine Accident Investigation Branch (MAIB). The causes and circumstances of deaths were determined from postmortem examination reports, death certificates and registers, reports of coroners’ inquisitions, procurator fiscals’ fatal accident inquiries, official marine inquiries, log books, and other official documents.

Inclusion criteria

Included in this study were all deaths among commercial fishermen in British registered fishing vessels between 1976 and 1995, occurring at work or after being landed ashore as a result of accident or illness at work, provided that the death occurred within 30 days of discharge ashore. Deaths among leisure fishermen and non-fishermen such as passengers and stowaways were excluded, as were deaths in merchant ships, pleasure craft, and other non-fishing vessels. It is expected that almost all deaths that satisfied the inclusion criteria were identified from the study data sources. The study period was the 20 years from 1 January 1976 to 31 December 1995 for deaths in British commercial fishing and for the comparison of fatal accident rates with other British industries.

Certification of deaths

Of 616 fatalities in British fishing between 1976 and 1995, 77 (13%) were due to natural causes. At least 46 of the diagnoses (60%) were based on postmortem examination; 11 other diagnoses (14%) were obtained from Scottish deaths registers which, for legal reasons, usually denoted the underlying cause of death only and did not indicate whether a postmortem examination had been performed. Fifteen diagnoses (19%) were obtained from other medical sources, and five through official marine inquiries or trawling skippers (unknown source of certification, 6%). Of the 539 deaths due to non-natural or unknown causes, 246 fishermen were lost at sea. The results of a postmortem examination were available for at least 175 of the other 293 deaths (60%), and diagnoses were provided by other medical sources in at least a further 58 cases (20%). Of the 246 fishermen who were lost at sea, 221 (90%) were subject to MAIB investigation or other official marine inquiries. Information on the other 25 deaths was obtained from procurator fiscals’ fatal accident inquiries in 13 cases, coroners’ inquisitions (nine), and other sources (three).

Population at risk

The numbers of commercial fishermen employed annually in UK fishing were obtained from official government statistics, published by the former Ministry of Agriculture, Fisheries and Food (MAFF) from 1976 to 1995.¹² The fishermen were conventionally counted as if they had been employed for the whole of each year, yielding a total population of 440 355 fishermen-years at risk. The annual numbers of fishermen remained fairly stable throughout the study period, reaching a peak of 23 990 in 1981, before declining to 19 928 in 1995. Over the entire study period, 74.6% of the fishermen were full time employed and 25.4% part time; annual percentages of full time fishermen were also fairly stable between 1976 (76.4% full time) and 1995 (80.7%).

Official published statistics show that there were a total of 112 558 fishing vessel-years in the active British fishing fleet from 1976 to 1990; 79 888 of these were for vessels under 12 metres in length, 28 992 for vessels measuring 12–23.9 metres, and 3778 for vessels of 24 metres and above.^13–15 From 1991 the official figures were no longer collected in this format by MAFF but included all registered British fishing vessels rather than just the active fleet; a total of 52 657 vessel-years from 1991 to 1995 referred to, respectively, 42 306, 7970, and 2381 vessel-years for vessels of under 12, 12–23.9, and 24 metres or more in length.

Comparison of fatal accident rates with other workers

Comparison was made with all workers in Great Britain, and with those working in five main industrial sectors, using fatal accident statistics published annually by the Health and Safety Executive (HSE) for the period 1976 to 1995.¹⁶ Numerators were all fatal accidents occurring at the workplace, excluding non-work related traffic accidents and self inflicted poisoning by alcohol or drugs, and denominators were the populations of workers at risk.

Statistical methods

The statistical methods used include cause specific mortality rates, relative risks, and Spearman’s rank correlation coefficients. Ninety five per cent confidence intervals for mortality rates and relative risks were adjusted for multiple fatalities arising through the same accident; for example, when a fishing vessel founders with the loss of several lives.

RESULTS

Of 616 fatalities in British fishing between 1976 and 1995, 77 deaths (13%) were caused by illness, 454 (74%) by accidents, and 82 (13%) by other non-natural causes (table 1). In the remaining three cases the medical diagnosis was not specific enough to determine the cause of death.

Demographics

All but one of the 616 deceased were men, 577 (94%) were British nationals, followed by Spanish (18), Danish (8), and Dutch (6). The mean age of the deceased was 37.5 years (SD 13.1); 20% were aged 16–24 years, 49% were 25–44, 30% were 45–64, and 1% were aged 65 and over. Forty five per cent of the deceased British nationals had been resident in Scotland, 18% in Humberside, 8% in Cornwall, and 29% in other locations.

Deaths due to illness

Of 77 deaths from natural causes, 62 (81%) were caused by cardiovascular disease. Fifty five of the 62 (89%) were taken ill at sea, and seven in port. Only four of the deceased (6%) were evacuated from their trawlers before death and all four were dead on arrival at hospital. Most of the 62 were found dead or died within minutes of acute heart attack. The mean age at death from cardiovascular disease was 53.9 years (SD 7.3).

Deaths due to accidents

Deaths from accidents have been split into two categories. These are firstly, casualties to vessels resulting from an accident or incident involving the fishing vessel such as a collision, capsize, grounding, or fire; and secondly, personal accidents directly affecting individual fishermen at work. Casualties to vessels accounted for 270 (44%) and personal accidents 184 (30%) of all deaths.

Thirty seven of the 102 vessel casualties, and 115 deaths, were caused by fishing vessels foundering or capsizing as a result of hazardous weather and sea state conditions (table 2). All but two of the 37 vessels were less than 24 metres in length, and 21 of these were under 12 metres. The median gross tonnage of the trawlers was 11.2 (interquartile range 4.3–41.2), the mean age was 13.3 years (SD 10.2), and most were constructed from wood or fibreglass rather than steel.

Fourteen vessels foundered after grounding on rocks and harbour entrances (38 deaths), usually when returning to port in darkness or poor visibility as watchmen were fatigued or had fallen asleep, or because of faulty radar or engine failure. Ten trawlers were involved in collisions (29 deaths) and 12 capsized after snagging trawling nets on shipwrecks and other underwater obstructions (26 deaths).

Table 3 shows the causes of the 184 personal accidents, and the work duties or activities of the deceased at the time. The most common causes were falls overboard (57 deaths, 31%), struck by trawling equipment (38, 21%) and entanglement in equipment (24, 13%). Most accidents (145, 79%) arose at sea; 82 of these (57%) occurred during operations to “shoot-out” or “haul-in” trawling nets, and weather conditions during these manoeuvres were gale force 8 or worse in 14 of 50 reported cases.

The fatal accident rates for casualties to vessels were respectively 6.8, 44.3, and 52.9 per 10 000 vessel-years at risk for vessels of lengths under 12, 12–23.9, and 24 metres or over in the active British fishing fleet from 1976 to 1990. The corresponding fatal accident rates for personal accidents were 2.6, 25.3, and 151 per 10 000 vessel-years.

The 454 fatal accidents in fishing refer to drowning, disappearance at sea, or hypothermia (394, 87%), injury (43, 9%), asphyxiation or poisoning by fumes (16, 4%) and other causes (one). Only nine (2%) were evacuated and admitted to hospital before death, although a further seven (2%) were dead on arrival. The mean age at death from accidents was 34.3 years (SD 11.8).

Deaths due to other causes

Of four suicides, three jumped overboard and one was found hanged in a cabin, two were homicide victims when beaten to death ashore, and two died from alcohol intoxication. Official investigation was unable to establish the circumstances of 74 deaths from non-natural causes; 29 disappeared at sea and, apart from seven who had shown symptoms of psychiatric disturbance, most likely fell overboard. Another 44 also went missing, 10 at sea and 34 in port, but were subsequently found drowned. Twenty nine of the 34 were known to have been ashore, and most had been drinking, before failing to return to their vessels. Although it is likely that they accidentally fell into docks when returning to trawlers alone, the lack of evidence led coroners’ inquisitions to record open verdicts. Finally, the circumstances of death were not known for one fishermen who died from non-natural causes, and the cause of death was not specific enough to determine whether three deaths were due to natural or non-natural causes.

Comparison of fatal accident rates with international fishing mortality studies

Table 4 shows a summary comparison with fatal accident rates calculated from previous studies. For improved comparability with the other studies, unexplained drowning and disappearances at sea have been classified as accidents, even though British coroners’ inquisitions normally record open verdicts for unexplained drowning and are not usually held, in the absence of a body for forensic examination, for disappearances at sea.

The fatal accident rate from this study, 120 per 100 000 fishermen-years (95% CI 104 to 136), was lower than those calculated from earlier studies of British fishing, in 1958–67 and 1961–80. The highest fatal accident rates were calculated from studies of Alaskan and New Zealand fishermen, although the extremely high Alaskan rate of 415 per 100 000 (1980–88) subsequently declined to 116 (1990–98). The British rate was broadly comparable to those for Danish, Australian, and Icelandic fishermen, and higher than in Canadian (Atlantic provinces), American, and Polish fishing.

Comparison of fatal accident rates with other industries in Great Britain

Between 1976 and 1995, the mortality rate for fatal accidents occurring at work for all workers in Great Britain was 2.0 per 100 000 worker-years (table 5). The fatal accident rate in British fishing (103.1 per 100 000; 95% CI 87.5 to 118.7) was 52.4 times higher (95% CI 42.9 to 63.8) than for all workers in Great Britain.

The fatal accident rate in British fishing was 10 times higher than for workers in the energy and water supply industries, 12 times that for the construction industry, and 46 times greater than in manufacturing. If unexplained drowning and disappearances at sea are included as accidents, mortality in fishing (120 per 100 000) would be 60.8 times higher (95% CI 50.5 to 73.2) than for all workers.

Trends in fatal accident rates

The annual fatal accident rate in British fishing fluctuated considerably with no overall upward or downward trend between 1976 and 1995 (Spearman’s rank correlation coefficient (SR) −0.16). During the same period, the fatal accident rate for all workers in Great Britain declined sharply (SR −0.93) as did those in the construction (−0.87) and manufacturing industries (−0.95).

Table 6 shows the fatal accident rate in British fishing in five year time periods from 1976–80 to 1991–95. Although the fatal accident rate fluctuated from 135 per 100 000 fishermen-years to 77, 105, and 96 during the four 5-year periods, there were no significant changes in the rate. The relative risk of a fatal accident in fishing, compared with all workers, decreased from 53 (in 1976–80) to 36 (1981–85), but increased to 55 (in 1986–90) and to 77 (1991–95).

DISCUSSION

This study, with a population of over 440 000 fishermen-years at risk between 1976 and 1995, represents one of the largest studies of work related mortality in commercial fishing. Fishing was established as an extremely hazardous occupation in Britain during the 1960s and, after the introduction of additional safety legislation in 1970, the fatal accident rate did not decline during the subsequent decade.⁹ Although the fatal accident rate was highest in the late 1970s, and increased again during the late 1980s, the present study found no significant change between 1976 and 1995.

The fatal accident rate was lower than in earlier studies of British fishing, and was much lower than in smaller studies of Alaskan and New Zealand fishing in the 1970s and 1980s. However, after legislation aimed largely at the heavy loss of life through casualties to vessels,²⁷ the Alaskan fatal accident rate in the 1990s declined to a level comparable with British fishing. The British rate was broadly similar or higher than those in the other fishing fleets. Although many fatal accidents occurred from small fishing vessels of under 12 metres in length, in terms of vessel-years at risk, the corresponding fatal accident rates were much lower in these vessels. However, an increase in the proportion of small vessels over time, from 63% of all active British fishing vessels in 1976 to 76% in 1990, did not reduce the fatal accident rate.

Commercial fishing is an extremely hazardous occupation when compared with general workforces that include both men and women. The relative risk of 52 in this study is similar to that of 53 for Danish fishing between 1988 and 1992.²⁰ Relative risks of about 50 and 12 were identified for Alaskan fishing during 1980 to 1988, in comparison to all workers in the USA and Alaska respectively,¹⁷ although by 1991–98 the relative risk for the USA had declined to 26.²² A lower relative risk of 18 was found for Australian fishing.²¹ However, the fatal accident rates for the general workforces in these countries were higher than in Britain.

The fatal accident rate in British fishing, between 1959 and 1968, was five times that for the construction industry and 20 times greater than in manufacturing.¹¹ The corresponding relative risks from this study, 12 and 46, are substantially higher. Fatal accident rates in fishing are also typically higher than those from recent studies of merchant shipping,^10,28 the occupation historically regarded as the most dangerous. In Britain, fishing has been the most hazardous of all occupations in recent decades.¹⁰

Although previous studies of British fishing did not calculate a relative risk comparison with all shore based workers, the fatal accident rate for all employees in Great Britain was 5.6 per 100 000 worker-years in 1961.²⁹ Fatal accident rates for all workers are typically similar or slightly higher than for all employees and would indicate a relative risk of 35 to 40, for British fishing during 1959 to 1968, which compares with 52.4 in this study. The present study also identified relative risks that increased from the early 1980s up to 77 in 1991–95 and show that, in relative terms, fishing has remained at least as hazardous as before.

During the study period, there were structural labour market changes in the British workforce; in particular, reductions in the numbers employed in the energy and water supply and construction industries which predominantly employ men and which had high fatal accident rates among the shore based industries, and a large increase in the service industry that increasingly employed women and which had by far the lowest fatal accident rate. Especially since the fatal accident rate in fishing did not alter greatly between 1976 and 1995, these changes would contribute to the increased risk in fishing compared to the general workforce.

Cardiovascular disease was the main cause of fatal illness. Few of these deceased were evacuated from their vessels before death, and most were found dead or died within minutes of onset. Fishermen aged 18 and over, unlike merchant seafarers, are not presently subject to medical examination in order to work at sea. However, as high levels of smoking and obesity have been found in European fishermen,^30,31 primary prevention and health promotion may offer some benefits. The introduction of health screening may save some lives, but unemployment would most likely have a detrimental effect on fishermen’s health ashore.

As found in previous studies,^9,17,21,22 casualties involving fishing vessels were the cause of most fatal accidents, and foundering or capsize of trawlers in hazardous weather conditions was the most common type of casualty. Most of these vessels were relatively small in size, made of wood or fibreglass, and were often not suited to operating in the prevailing weather conditions. Although some vessels were returning to port, or sailing for shelter amid deteriorating weather conditions, improvements in weather forecasting and evaluation would be beneficial.

There were often contributory factors affecting the foundering of fishing vessels in adverse conditions. Economic pressures had led some fishermen to overload the vessels with equipment, creating stability difficulties during hauling-in operations; some vessels were found to be unseaworthy, and others were unsuited to the prevailing conditions—the exhaustion of previously lucrative fishing grounds has brought increased pressures for fishermen to go further to sea in small and unsuited vessels.^32,33 Classification of fishing vessels according to the weather conditions they are suited for, and closer attention to stability limits, would be important preventative measures.

The capsize of trawlers after nets became snagged on underwater obstructions was another major cause of death in British fishing. Since several of these obstructions were considered to be shipwrecks, the maintenance of maps or guides with the accurate locations of shipwrecks around the British Isles would be of benefit. Most fatal collisions involved trawlers being “run down” by flag of convenience or other foreign flagged merchant ships, whose crews were often outside the legal jurisdiction of British marine investigation, and raises the important matter of loss of life through watch keeping negligence and criminal prosecution in international merchant shipping.

The fatal grounding of fishing vessels largely arose when returning to port and were mostly due to navigational error. Although faulty radar, fog, and alcohol consumption were cited in some official inquiry reports, many of the watch keepers were known to have fallen asleep or were considered fatigued after long trips at sea. Fishermen typically work long hours at sea, with short rest periods often disrupted by trawlers rolling in heavy seas, engine noise and vibration, and confined sleeping conditions. Fatigue was also cited, or a likely causal factor, for many of the fatal personal accidents, and has previously been identified as a major factor affecting accident rates.^6,11,34 Prevention should be aimed at the implementation of adequate rest periods both at sea and, in between fishing trips, ashore.

A majority of fatal personal accidents at sea (57%) arose during operations to “haul-in” or “shoot-out” trawling nets. These were often conducted in hazardous conditions. These operations have previously been identified as potentially high risk manoeuvres leading to a large proportion of accidents in fishing.^33,35,36 Improved evaluation is required as to weather related risks when deciding to conduct these operations.

Over 40% of fatal personal accidents at sea arose when fishermen were struck by or entangled in trawling lines and other equipment. The deceased were knocked or dragged overboard and drowned, or were dragged into the winch and fatally injured, or died from other injuries. Prevention should focus on the deck layout and the adequate guarding of trawling cables and the winch. Several of the deceased were engaged in manual operations to adjust the lines, and several others died after their clothing became entangled. Emphasis should be placed on methods to minimise the need to handle lines, alertness as to the position of the lines, and the use of close fitting clothing.³⁷

A further 75 fishermen drowned (52% of personal accidents at sea) after they fell or were washed overboard by heavy seas. Prevention should be aimed at maintaining hazard free decks to minimise the risks of slips and falls, and at a more widespread use of personal flotation devices (PFDs). The former should include improved lighting on deck at night, and the regular removal of oil, ice, nets, ropes, wires, and other potential obstacles from trawling decks. The use of boots with adequate grip that are suited to operations on fishing vessels’ decks, the use of anti-skid surfacing on decks, and increased height of railings on small vessels are further important preventative measures. The fishermen who were washed overboard by heavy seas further illustrate the need for better evaluation of weather related risks when deploying crew on deck.

Previous studies have shown that the use of PFDs is effective in improving the survival prospects of fishermen after falls overboard or after abandonment of stricken vessels.^22,25 However, many of the fishermen who drowned in this study were in fishing vessels that capsized or foundered rapidly after a collision, trawling nets snagging an underwater obstruction, or heavy seas swamping the vessel, and may have had insufficient time to retrieve a PFD, launch a life raft, or send distress signals. Others had insufficient time to board life rafts, which had been launched by colleagues who were subsequently rescued. A more widespread use of PFDs among fishermen, particularly those in hazardous situations such as those engaged in trawling net operations on deck, and the maintenance of life rafts and PFDs in easily retrievable locations should be effective in reducing drowning of fishermen. The rapid capsize or foundering of fishing vessels should also be addressed by improved awareness of the vessel’s stability limits and the need to close hatch covers and watertight doors after use.

The British commercial fishing industry has been under increased financial pressures in recent years, particularly with the implementation of catching restrictions. These can lead fishermen to take greater risks by taking under-manned and badly maintained trawlers to sea and to continue fishing in rougher weather.³⁴ Particularly in this climate, the industry is unreceptive to the introduction of additional safety regulations, especially when developed and enforced by those outside the industry. It is therefore important that safety measures should be developed through dialogue with fishermen and the fishing industry. This should be achieved by the use of multidisciplinary teams of experts, including fishermen, who are most able to develop safety solutions that are both practicable and economically affordable to the fishing industry. Further research should also focus on the identification of risk factors that are specific to the different subsectors of international fishing fleets.

Although the fatal accident rate in British fishing has fallen since the 1950s and 1960s, there has been an even greater reduction in the general British workforce. In relative terms, the occupation of commercial fishing remains at least as hazardous as before. Prevention should be directed, most importantly, towards improved safety awareness and reductions in hazardous working practices. In particular, this should be aimed at the unnecessary operation of small vessels and trawling net manoeuvres in hazardous weather and sea state conditions.