Background Thirteen episodes of oesophageal food impaction (EFI) per 100 000 people present to a medical setting each year. Several pharmacological interventions meant to relieve such impactions have been explored; none have proven superior.

Objectives Perform a single-arm feasibility study of oral nitroglycerin solution for EFI.

Methods Twenty adult patients presenting to a single urban tertiary medical centre thought to have EFI were given up to three doses of 0.4 mg nitroglycerin solution orally and evaluated for resolution of symptoms, new symptoms and vital signs. Patients with intractable vomiting, haemodynamic instability, airway compromise, oesophageal perforation, coronary ischaemia or presentation delayed greater than 12 hours were excluded.

Results 17 of 20 enrolled subjects received the intervention. The average duration of symptoms prior to intervention was 285 min (SD=187). Four subjects did not tolerate the intervention (inability to swallow or headache). Two of 17 (11.8%) subjects obtained temporally proximal symptom resolution: 11 min after the second dose, and 7 min after the third dose. Seven also received glucagon during their visit, with 0% temporally proximal symptom resolution. Fifteen underwent endoscopy, with food bolus identified in 12. One subject had brief and mild hypotension with spontaneous resolution. Two subjects developed a headache after nitroglycerin administration. The median length of stay for those who found relief without endoscopy was 195 min (range 129–261) vs 374 min (range 122–525) among those with endoscopy.

Conclusion The observed rate of relief after oral nitroglycerin solution for EFI is disappointing but comparable to previous glucagon, benzodiazepines and effervescent beverage studies, and that of placebo. Oral nitroglycerin solution appears to be well tolerated among those able to swallow, although in our sample several subjects were unable to tolerate swallowing entirely.