Atypical organophosphate poisoning and a successful case of prolonged intubation in a low-resource newly developed intensive care unit in rural Zambia

Intensive care unit

We present background information on our intensive care unit (ICU) resources at the time of this case presentation, to allow our readers to consider how our learning points could apply to their local critical care unit.

The six-bed unit was established in September 2019. Initially intended as a higher acuity ward for improved monitoring (assigned two nurses per shift), the unit subsequently obtained ventilators in March 2020. From then, the hospital has been ventilating medical, surgical and obstetric patients with a dedicated nursing team, but in the absence of a trained critical care physician. In mid-2021, the unit began the use of non-invasive ventilation, sedation (propofol or ketamine, via syringe driver) and inotropic support (epinephrine or dopamine when available, peripherally administered via infusion pump).

We were not able to provide dialysis on-site, or perform central venous access, invasive blood pressure monitoring, end-tidal CO2, blood gases or electrolyte tests. Opiates were unavailable and the supply of all medications could be disrupted at times. Masks, airway equipment and circuit tubing were cleaned and reused. Due to regular electricity outages, the hospital relied on a backup generator, however, there were still periods without electricity. We were reliant on concentrators and cylinder oxygen, however, since late-2021, an oxygen plant has been established at the hospital (supported by solar panels), supplying piped oxygen to some wards.

Of the 12 nurses on the ICU rota, 4 had previously worked in an ICU. In 2020, the whole team received two training sessions covering the new ventilators. This was provided by the ministry of health and the central teaching hospital in Lusaka. New nurses joining since then have been trained in turn by the nursing head of the department.

Case presentation

A man in his 30s was admitted after an intentional ingestion of approximately 100 mL of organophosphate pesticide 3 hours prior to admission. He displayed acute cholinergic symptoms with vomiting, diarrhoea, urinary incontinence, sweating and high-volume respiratory secretions. The specific compound he ingested was dichlorvos (2,2-dichlorovinyl dimethyl phosphate), a class 1B ‘highly hazardous’ chemical as per the WHO, sales of which are restricted or banned in many countries.11

On arrival to the ward, he was unconscious with a glasgow coma scale (GCS) of 3/15, copious airway secretions and a poor respiratory effort, bradycardic (central pulse 50–60 bpm) with an unrecordable blood pressure and a blood sugar of 2.5 mmol/L. Family removed his soiled clothes and washed him. He received ventilation via a bag- valve- mask with 10L of oxygen to achieve 77% saturations (maximum oxygen capacity via our oxygen concentrator), incremental atropine and epinephrine boluses, intravenous fluids and dextrose. He was transferred to the ICU and intubated for airway protection and respiratory failure via rapid sequence induction. A 1.5 mg/kg dose of suxamethonium (100 mg) was used, in the absence of an alternative muscle relaxant.

He was the second patient to present with an OP that evening, and as such the hospital atropine supply was depleted. Despite 70 mg of atropine, we achieved suboptimal atropinisation and he required an epinephrine infusion for haemodynamic stability. In the absence of invasive monitoring, his blood pressure was monitored with a manual sphygmomanometer, shared between ICU patients. He also received 20 mmol of intravenous magnesium over 4 hours within the first 24 hours and was commenced on metronidazole and ceftriaxone due to aspiration.

Twelve hours postadmission, his GCS and haemodynamics improved and he no longer required inotropes. However, within hours, he developed resistant status epilepticus, which did not terminate despite 40 mg intravenous diazepam, followed by nasogastric phenobarbitone loading (our only available anticonvulsants). Additionally, he had worsening high-volume respiratory secretions, bradycardia and hypotension. More atropine was sourced by family and the hospital, and he was commenced on an infusion titrated to end points of a dry chest, mean arterial pressure >60 mm Hg and pulse >80 bpm. Despite an escalating dose of atropine, he had pinpoint pupils, respiratory secretions, high-volume diarrhoea, fluctuating GCS and intermittent seizure activity with no spontaneous breaths on minimal sedation.

Ninety-six hours postingestion, he became alert, could protect his airway and ventilate spontaneously. We noted muscle fasciculations, however, given the risks of ongoing ventilation in a resource-limited setting, we attempted his first extubation. He rapidly developed progressive descending weakness and was commenced on bilevel positive airway pressure (BIPAP) with a likely diagnosis of IMS. He was stable and alert for 7 hours, he then lost consciousness and was re-intubated without suxamethonium.

Over the next 13 days, he had ongoing nicotinic symptoms of fasciculations, deep tendon areflexia and near global paralysis except for minimal eye (from day 7), head, toe and finger movement (from day 8), as well as prolonged muscarinic symptoms of high-volume diarrhoea and respiratory secretions. He intermittently required atropine (until day 11) to control the respiratory secretions, however, when administered he displayed signs of toxicity (tachycardia and fever). He was actively cooled, and we used the lowest effective dose of atropine infusion to avoid fever. This would be stopped when secretions became manageable. His bloods returned WCC (white cell count) 11.4 x10⁹/L, Hb (heamaglobin) 12.9 g/dL, AST (aspartate aminotransferase) 55 U/L, ALT (alanine transaminase) 31 U/L and creatinine 98 μmol/L. A malaria rapid diagnostic test and COVID-19 lateral flow test were negative. He remained alert (E4VTM6), with low-dose propofol sedation in the day and sedation to allow sleep at night. He was managed mostly on volume-control synchronised intermittent mandatory ventilation. Our focus during this period was on reducing complications. He received early bolus enteral nutrition to prevent ileus, maintain gut integrity and as stress ulcer prophylaxis12; omeprazole; 4 days of heparin prophylaxis (as available); physiotherapy; analgesia; regular mouth care to reduce ventilator-associated pneumonia; and was managed head up with rotation to reduce pressure sores and to interact with staff and family. With ongoing fluid losses and in the absence of electrolytes we used a four-lead ECG to screen for significant potassium derangement, monitored urinary specific gravity and did our best to normalise electrolytes with judicious use of oral rehydration solution (ORS), a tactic we have used for several patients in this rural ICU who likely had profound electrolyte abnormalities.

We commenced the first ventilation weaning at this hospital with stepwise reduced ventilatory support. Prolonged respiratory paralysis due to IMS and the absence of blood gases made this challenging, but vital to reduce ventilator-associated muscle weakness and complications. Ten days after intubation, he had a gradual ascending resolution of paralysis, though respiratory muscle fasciculations and weakness persisted. This was complicated by a likely ventilator-associated pneumonia, common in this setting of reused equipment (diagnosed clinically due to lack of mobile imaging).

We attempted extubation again on day 11 as he could generate spontaneous breaths of 5–6 mL/kg and had a cough reflex. He was initially stable for 3 hours on BIPAP, before becoming haemodynamically unstable and unconscious, and was reintubated without suxamethonium. Over the following 5 days, his respiratory muscle power and fasciculations improved.

On day 15 of ventilation, after a challenging 24 hours of oxygen failures and electricity cuts, sometimes lasting several hours, during which the nurses would hand-ventilate the patient through the endotracheal tube, he was successfully extubated onto oxygen. We approximated intensive chest physiotherapy by encouraging him to inflate a surgical glove to reduce alveolar collapse and encourage respiratory muscle work out (figure 1).

He had ongoing physiotherapy, nutrition and psychiatric involvement. He was discharged on antidepressants after 21 days in the hospital.

Outcome and follow-up

The patient was reviewed at 3 months in his home in a rural village 3 km from the hospital. He sat with his wife and four young children around him. He described himself as ‘strong’ with no long-lasting deficits. He had stopped antidepressants of his own accord after 1 month but was well in mood. On examination, there was no neurological deficit noted. He had returned to the intensive farming that provided his living.