Acute phosphine poisoning on board a bulk carrier: analysis of factors leading to a fatal case

History and medical care

On 7th February, 2008 at 18:11, a Romanian bulk carrier (Figure 1) navigating 80 km north of the island of Ouessant, sailing from Rouen in France to Cairo in Egypt, with a load of approximately eighty-thousand tonnes of peas, contacted the Maritime Rescue Coordination Centre (MRCC) Corsen to evacuate of one of its crew members for medical reasons (patient n° 1). This 56-year old man, discretely over weighted, onboard electrician, had received orders from the captain to rest in his cabin since the previous evening as he had been suffering from increasingly severe digestive disorders in the form of acute abdominal pain accompanied with dizziness. His was given emergency medical care and was taken by helicopter to Brest Hospital, France. No information was given by the officers on the vessel as to possible poisoning. No other seafarer was reported ill in the first report.

The first medical examination found the patient conscious and oriented. The Glasgow Coma Scale (GCS) score was 15 at this time. He complained of abdominal and chest pain similar to symptoms of an acute coronary syndrome associated with high blood pressure and reduction of the diastolic-systolic time interval. It was not clear from the ECG done in the helicopter whether there was myocardial infarction or not. In the Accident and Emergency waiting room, the patient’s respiratory condition with cyanosis deteriorated. Suddenly, he convulsed, showed signs of unconsciousness which deteriorated rapidly justifying sedation and intubation. An emergency chest x-ray showed an oedema spreading throughout the two lungs and the heart. An electro-mechanical dissociation was found. Closer inspection of the ECG readings excluded myocardial infarction. The patient went rapidly into shock: his pulse and blood pressure were impossible to take, despite volume expansion and administration of adrenaline, dobutamine and atropine. Biological results were the following: CBC normal, PT at 61% with no other coagulation abnormality, lactic acidosis (anionic hole 39, lactates 25.5 mmol/L) decompensated by an alkaline reserve down to 7 mmol/L), and hyperglycaemia 4.6 g/L, CPK and troponine were both normal. No abnormalities were found in the hepatic or pancreatic results either.

The seafarer died at 21:35 as a result of these major acid-base abnormalities. Metabolic acidosis was suspected to have been caused by an external factor with no further details. The MRCC was informed of the death.

The following morning at 7 :23, the cargo ship (having continued its route to Egypt), and at that time 60 kms from the coast, made another call to the authorities to evacuate a second seafarer (patient n° 2), 41 years old, who was also had symptoms of dizziness with repeated vomiting and abdominal pain. The MRCC requested medical evacuation by helicopter. Collective poisoning was by then suspected.

The medical team found the man to be conscious and oriented. GCS score was 15 at this time, 96% saturation in ambient air. He was complaining of abdominal pain, vomiting and dizziness when standing up and his condition had been deteriorating since the previous evening. Symptoms improved considerably when the seafarer left his bunk. His medical history was the following: active smoker, appendectomy, frequent abdominal pain.

At the same time, the Marine Prefecture services requested the cargo ship to head for the coast for inspection. A further conversation with the ship’s captain concerning the chemical products on board lead the authorities to suspect collective phosphine poisoning: phosphine had been used as an insecticide on board the vessel in the holds containing peas. Clinical examination of the second patient was normal and the biological results were completely normal. Moreover, there was no acid-basic imbalance or any other gasometric abnormality, or higher cardiac enzyme or hepatic readings. The ECG showed a rise in the ST segment in V₁ V₂ associated with a right bundle branch block, characteristic of a type 1 Brugada syndrome. This patient was non-Asian and had no history of sudden death in the family. The seafarer, at that time, with no symptoms, was admitted to hospital, into cardiac intensive care for monitoring for a documented risk of total atrioventricular block or heart rhythm disorders such as ventricular tachycardia. Three days later, he left hospital with stable ECG readings. Onboard the ship, while waiting for the chemical diagnosis, the seafarers were kept out of danger on deck.

As there was no risk of explosion as initially suggested, the ship was authorized to go into the bay of Brest and berth in the port. The eleven other crew members were taken to hospital for medical examination.

Amongst these eleven seafarers aged between 34 and 59 years old, nine said nothing was wrong, one complained of weakness and the other complained of chest pain which improved when taken away from the contaminated area. As well as a standard clinical examination, each patient had a biological examination, chest x-ray and ECG. Specific clinical information being looked for in these consultations and examinations were the following: any medical history, presence or not of neurological signs (headaches, dizziness and abnormalities in the neurological examination); digestive signs (pain, nausea, diarrhoea…); difficulties breathing and chest pain. The biological examination included renal, hepatic, troponine, lactatemia and plasma ionogram testing.

The patient complaining of weakness (patient n°3, 46 years old) was the one who confirmed the death of his colleague. At that time, he was wearing a protection mask. Clinical examination in Accident and Emergency was completely normal. Oxygen saturation was 95% in ambient air with spontaneous breathing even though a smoker. Para-clinical examination was also normal. He was kept under observation for twenty-four hours.

The patient complaining of occasional chest pain on board the vessel (patient n° 4, 38 years old) improved considerably in Accident and Emergency. Symptoms including retrosternal chest pain had lasted for several minutes the previous day. Clinical examination was normal and the ECG showed a banal right bundle branch block. He was kept in for observation for twenty-four hours.

Another patient (patient n°5, 55 years old and ship’s cook), had no symptoms, but showed some abnormalities when examined. Chest examination found crackling at the base of the lungs and the ECG showed up a right bundle branch block. His biological results were completely normal and the chest x-ray didn’t show any signs of overload. He was kept under observation for twenty-four hours. At the end of the observation period, the branch block had disappeared and the chest examination had become normal. He was advised to consult again quickly should palpitations, difficulties breathing or chest pain reoccur in the following few days.

For the eight other patients, clinical examinations, biological results, chest x-rays and ECGs were normal. They were, however, kept for observation for twenty-four or forty-eight hours (according to a calculated exposure risk), except one of them who refused to go to hospital and for another whose presence on board was deemed necessary and compatible with his state of health.

Medical follow-up

Out of the 11 seafarers on board, 10 are still alive and in good health 5 years after this tragedy.

An autopsy was carried out on the deceased seafarer on 13 February 2008, 6 days after his death. A dilated heart condition was detected, a healthy coronary network, absence of lung tissue abnormalities and pleuropericardial effusion. The liver, spleen, kidneys (other than an isolated cyst), encephalon and alimentary canal were normal. Moreover, there was no evidence of trauma. Blood, stomach, kidney, brain, heart and lung samples were taken for anatomical, pathological and toxicological analysis.

From a histological point of view, the tissues analysed were within the normal range. No traces of medication or narcotics were found. Phosphorous compounds were tested using a technique of gas phase analysis associated with mass spectrometry (detection from 0.1 mg/l). The Musshoff [6] technique was used for phosphine calibration. In all tissues, phosphine characterisation was negative. It was extremely likely that the phosphine had been eliminated from the samples considering its high volatility [7] or due to spontaneous ignition [8], bearing in mind that it was tested for 6 days after death and that it was absorbed by inhalation and not ingestion. To conclude, when exposure and clinical symptoms are well identified, chemical analysis for phosphine in blood or urine is not recommended as phosphine is rapidly oxidised into phosphite and hypophosphite [8,9].

Toxicology tests

Phosphine is a gas used for the fumigation of grain on board cargo ships [1]. In this specific case, an external French company was responsible for fumigating the vessel. This company started procedures by offloading the product on the quayside in the port of Rouen, before the vessel’s departure.

Once the grain had been loaded into the cargo holds, it was then treated with aluminium phosphide pellets placed on the surface of the piles of grain in each hold: equivalent of 1 g to 1.5 g of aluminium phosphide per m³. This large quantity of aluminium phosphide pellets reacted in contact with the air, water and moisture to release a phosphine gas thereby killing insects and larva.

The chemical reaction is AlP + 3 H20 → PH3 + Al(OH)3 [6].

In other words, the more humid the weather (which is the case in February in Europe) the faster (and at high levels of concentration) gas is produced.

Phosphine is an expanding gas which spreads through the closed environment of holds reaching pesticide concentration (several hundred ppm). This concentration reaches its peak once all the aluminium phosphide pellets have reacted with the air or water in the holds then it decreases during the voyage, until hardly detectable after 3 weeks. Nevertheless, great care must be taken when handling the open pellet blisters after fumigation in order to control complete dissolution. Protective equipment must be worn in these cases (Figure 2).

The ship’s captain was informed of this fumigation and the external French company gave him MSA colorimetric tubes so that he could test for the presence of this gas onboard should the gas spread from the hold through the ship.

In the case we are presenting, phosphine gas was tested for on 8 February, in other words, the day the second seafarer was poisoned. Testing was positive in the two cabins of the poisoned seafarers. Emergency services were aware of these positive tests and told the ship’s captain to evacuate and ventilate the two cabins.

Marine rescue services arrived onboard 4 hours after the cabins had been ventilated and found 0.24 ppm of phosphine near the holds and a trace of phosphine in the ventilated cabins.