Background Acute neuromuscular respiratory failure (NMRF) is a life-threatening feature of a variety of neurological conditions that can present in extremis prior to the establishment of a definitive diagnosis, so early clinical decision making is difficult. Population-based data on the frequency, outcome and aetiological spectrum are lacking.

Objective To establish accurate epidemiological descriptive statistics in this patient group.

Methods The regional Intensive Care National Audit and Research Centre (ICNARC) database was searched for patients admitted with acute NMRF from 1/1/2000 to 31/12/2010. Demographics, diagnosis, length of intensive care unit (ICU) stay, follow-up and outcome (modified Rankin score (mRS)) were recorded. A comparison dataset of all non-NMRF neurology patients admitted to ICU was obtained.

Results 55 patients were identified; age 17–88 (median 66 years), M:F ratio 1:1.5, incidence rate (IR) 2.81 (2.12 to 3.66) cases per million person-years and mortality rate (MR) 0.26 (0.08 to 0.60) deaths per million person-years. Causes included inflammatory neuropathy (65%), myasthenia gravis (18%), rhabdomyolysis (2%) and amyotrophic lateral sclerosis (9%), and 5% were undiagnosed. Follow-up ranged from 0 to 7 years (median 500.5 days); long-term mRS 1 (range 0–6). NMRF patients were older (p<0.0001), had longer ICU stay (p<0.0001), but significantly better outcome (p<0.0001) than 93 non-NMRF neurology patients requiring ICU admission.

Conclusion Inflammatory and degenerative neuromuscular conditions can present in acute NMRF. Long-term outcome is good and MR is low, and significantly better than in other neurology patients requiring ICU admission despite longer ICU stay.