Adult corrosive ingestions in the Pietermaritzburg Metropolitan Surgical Service

There is a paucity of data around corrosive ingestion in South Africa over the last three decades. As such, we set out to review our experience with adult corrosive ingestion in our tertiary gastrointestinal surgical service.

A retrospective, quantitative review was performed. The parameters analysed were demographics, substance ingestion, ingestion time to first presentation to a healthcare facility, clinical presentation, severity of injury based on endoscopic classification, computed tomography (CT) findings, management and outcomes. Patients presenting within 72 hours with alarm symptoms underwent flexible upper endoscopy and injury severity grading. In patients presenting after 72 hours, a water-soluble contrast study was obtained prior to upper endoscopy. Patients with signs of sepsis, surgical emphysema or physiological instability were referred for urgent CT to exclude oesophageal perforation and mediastinitis.

Between January 2012 and January 2019, a total of 64 patients presented with a history of corrosive ingestion - 40 (31%) were males and 24 (19%) females. The average time from ingestion to presentation was 72 hours. In 78% of patients, the agents were intentionally ingested, whilst 22% claimed accidental ingestion. A quarter of the patients (21%) presented to the unit clinically unstable, requiring emergent cardiorespiratory support. Eight (12%) patients required urgent surgical intervention due to the extent of injury. Nine (14%) patients demised during the acute admission. Of this group, three patients had undergone surgical intervention, and six were managed conservatively. Eighty-five per cent of all patients survived their initial admission.

This paper has highlighted the problem of corrosive ingestion in our setting. It remains a complicated problem to manage associated with significant morbidity and mortality rates. The current trend in the assessment of these patients is increased use of CT scan to assess the extent of transmural necrosis. Our algorithms should change to reflect this contemporary approach.