Management of hypoxaemic respiratory failure in a Respiratory High-dependency Unit

There are limited data on outcomes of hypoxaemic respiratory failure (HRF), especially in non-intensive care unit (ICU) settings.

To assess outcomes in HRF (without multi-system disease and not requiring early intubation) of patients directly admitted to a Respiratory High-dependency Unit (R-HDU).

This is a retrospective cohort study of HRF compared to hypercapnic respiratory failure (HCRF) in a R-HDU (2007-2011). Patient characteristics (age, gender, pre-morbid status, diagnoses) and outcomes (non-invasive ventilation (NIV) use, survival, ICU admission) were assessed.

There were 1207 R-HDU admissions in 2007-2011, 205 (17%) with HRF and 495 (41%) with HCRF. The proportion with HRF increased from 12.2% in 2007 to 20.1% in 2011 (P < 0.05). HRF patients were younger, more often male and had better pre-morbid performance. Compared to HCRF, HRF was more frequently associated with lung consolidation (61% vs 15%, P < 0.001), interstitial lung disease (12% vs 1%, P < 0.001) and pulmonary hypertension (7% vs 0%, P < 0.001) and less frequently with chronic obstructive pulmonary disease (24% vs 65%, P < 0.001) and obstructive sleep apnoea (8% vs 26%, P < 0.001). Fewer patients with HRF were treated with NIV (28% vs 87%, P < 0.001), but NIV was discontinued early more often (28% vs 7%, P < 0.001). A total of 18% with HRF was transferred to ICU compared to 6% with HCRF (P = 0.06). More patients with HRF died (19.5% vs 12.3%, P = 0.02). Interstitial lung disease, consolidation, shock, malignancy and poorer pre-morbid function were associated with increased mortality.

Initial R-HDU management is an effective option in selected HRF to reduce ICU demand, although mortality and clinical deterioration despite NIV are more common than in HCRF.