Pneumomediastinum- a single centre case series from 2010-2020

<bold>Introduction:</bold> The incidence of pneumomediastinum is not well known. We set up a retrospective study to look at recorded events over an 11 year period and the outcomes associated with Pnm excluding the covid-19 pandemic. <bold>Methods:</bold> A coding search for “pneumomediastinum” was performed for all patients attending NHCT inclusive from 2010 to 2020 with local Caldicott approval. All results were then manually searched: 35 events were collated over the time-period and analysed. <bold>Results:</bold> 21 were male (60%) with a median age of 51 (IQR 53.5). Median length of stay in hospital was 2 days (IQR 4). 10 were identified as having a smoking history – current or ex – with 9 recorded as non-smoking (25.7%). The aetiology was perforated oesophagus in 8 (34.3%) , vomiting in 4, breath holding 1, cocaine misuse 1, trauma 1, heavy lifting 1. 5 of the oesophageal perforations went for surgery. The rest 19, 55% were deemed spontaneous with no trigger. 10 (52.6%) were male, median age was 32 years (IQR 56.5). Median length of stay was 1 day (IQR 5.5), 5 were current or ex smokers (unknown in 9, never in 5). There was background lung disease in 7 (36.8%), and 7 had an associated pneumothorax with 1 requiring drainage and 18 managed conservatively. Overall, 29 were managed conservatively (82.9%). 5 patients died with pneumomediastinum deemed to have caused/contributed to the death (14.3%). <bold>Conclusions:</bold> This analysis would support previous data that pneumomediastinum is a rare condition and mostly managed conservatively. The dataset is limited (incomplete notes, lacking in follow up of those transferred to surgeons, frailty, reliance on reporting of radiographs and coding). A multicentre dataset would help delineate trends in the condition and differences between spontaneous vs secondary.