IntroductionAlthough formal patient registries in complex home ventilation have long been recommended1, recent review found few centres maintain accurate records of activity2. However, the ability to identify patient requirements such as dependence on home ventilation is essential for appropriate care and service commissioning2. In the local setting, the regional ventilation service established a combined risk assessment and treatment registry in 2018. The Pro-VISO tool was developed with examples from national network. Four aspects of treatment risk were included (ventilation, V; interface, I; secretions, S, and oxygen, O) alongside diagnosis, equipment and settings. This study aimed to examine the home ventilation registry for possible change in cohort characteristics before and after the advent of COVID-19 in the UK.MethodsRetrospective analysis of patient registry for ventilation method (invasive, non-invasive), ventilator dependence2 (>14 h/night) and survival was undertaken. Comparison was made of three timepoints, one preceding and two after the first UK wave of COVID-19. Subgroup analysis was performed for mortality amongst non-invasive and invasive users with and without ventilator dependence.ResultsData was available for timepoints T1 (26.2.20), T2 (3.3.21) and T3 (7.6.22). Total caseload at T1, T2 and T3 respectively was n=1042/1094/1139 (figure 1), mean age 59.3 (SD 16.0)/58.9 (SD 16.2)/59.0 (SD 15.8), male 51.7/52.7/52.3%. Dynamic changes to the patient population were seen, with new referrals (T1–2, n=182/T2–3, n=285) and deaths (T1–2, n=132/T2–3, n=234). Subgroup analysis revealed an association between non-invasive ventilator dependence and mortality in T2–3 (n=19/52, 36.5%; p<0.0005). There was no association with mortality for this subgroup at T1–2, or any other group at any time.ConclusionsThe apparent association between home non-invasive ventilator dependence and increased mortality in the second year of COVID-19 in the UK warrants investigation of unmet need in this patient group, compared with the invasively ventilated. Targeted review is planned in the local setting, facilitated by utilisation of home ventilation registry as a method of population surveillance.ReferencesLloyd-Owen SJ, et al. Patterns of home mechanical ventilation use in Europe: results from the Eurovent survey. ERJ 2005; 25: 1025–1031. Allen M. Respiratory Medicine: GIRFT Programme National Specialty Report. London: GIRFT, 2021.

Introduction There remains a paucity of data comparing ambulatory initiation of home non-invasive ventilation (NIV) with a model requiring inpatient admission.1 In our institution, a Quality Improvement (QI) project was performed where an ambulatory model for NIV initiation was developed and evaluated. Methods Ambulatory pathways were formulated for NIV initiation in the outpatient setting, alongside outreach and initiation of NIV for inpatients referred within regional hospitals. The primary outcome measure was 'compliance with NIV' defined as NIV use ≥4 hours/night for ≥75% of nights.2 Results Between 6.1.20 and 1.7.20, 76 referrals for home NIV were assessed within the ambulatory model. Of these, NIV was not indicated in 3 cases and contraindicated in 1 case, while 2 trialled NIV and declined it, leaving 70 patients who commenced home NIV (n=36 following COVID-19 ward 'closure'). Neuromuscular disease was the principal diagnosis in 41% (29/70) with MND comprising 20/29 neuromuscular cases; see table 1. Ventilator interaction data was available for 68 patients where mean NIV use was 5.21 (SD 3.98) hours/night. Of those established by ambulatory pathway, 62% (42/68) were deemed 'compliant' with NIV in comparison to previous data reporting compliance in 62% (56/90) of subjects established through inpatient admission2. It was calculated that delivery of the ambulatory pathway resulted in a cost saving of £197,967 for this period, achieved principally by admission avoidance based on previous length of stay data and Level 2 bed costings. Conclusions An ambulatory model for initiation of home NIV appears to be as effective in achieving compliance as inpatient admission, while carrying health economic benefits. Ambulatory treatment pathways enabled us to deliver service continuity during the COVID-19 pandemic. References Mandal S, Arbane G, Murphy P. et al. Medium-term cost-effectiveness of an automated non-invasive ventilation outpatient set-up versus a standard fixed level non-invasive ventilation inpatient set-up in obese patients with chronic respiratory failure: a protocol description. BMJ Open. 2015; 5; e007082. Ward K, Chakrabarti B, Ashcroft H. et al. Compliance with non-invasive ventilation (NIV) at 90 days may be associated with use of NIV in the first 2 weeks: investigation of telemonitoring data. AJRCCM. 2018; 197; A1463.

Background Patients sustaining a Spinal Cord Injury (SCI) may require long-term mechanical ventilation via a tracheostomy. Little UK data exists regarding outcomes of such patients following hospital discharge. We aimed to define the characteristics and chart the outcomes of adult SCI patients discharged with tracheostomy ventilation from a tertiary spinal injuries unit. Methodology The records of patients discharged with long-term tracheostomy ventilation from the Northwest Regional Spinal Injuries Centre were retrospectively analyzed with comorbidity defined using ICD-10 coding. Results The records of 47 patients (Age 51 years (Range 66 years), LOS 366 days (Range 1738 days), 72% male) with SCI discharged with long-term tracheostomy ventilation between1982 and 2019 were available for analysis. 83% (39/47) were classified as sustaining a Traumatic SCI with the level of injury on discharge being C0–1 in 15%, C2–4 in 62% and C5–6 in 15%. 68% (32/47) and 17% (8/47) were classified as ASIA-A and ASIA-B respectively on discharge. 68% (32/47) were exclusively on a normal diet/fluids whilst 23% (11/47) were exclusively fed by a gastrostomy tube. 53% (25/47) were discharged on 24 hour ventilation whilst 47% (22/47) were discharged on a minimum of nocturnal ventilation but less than 24 hour ventilation. 72% (34/47) were discharged to their own place of residence whilst 28% (13/47) were discharged to Institutional Care. 9% (4/47) of subjects had died 12 months post hospital discharge increasing to 17% (8/47) who had died at 3 years post hospital discharge and 21% (10/47) who had died by 5 years post discharge. A coded diagnosis of underlying Pulmonary Disease was associated with death at 12 months (p=0.04) but did not appear to be a significant adverse prognostic factor by 3 or 5 years post discharge. Advanced age was associated with death at 5 years (64 (11) years v 41 (20) years). The level of injury, ASIA classification, length of stay and degree of ventilator dependence did not appear to be linked to survival. Conclusion Patients diagnosed as SCI with long-term tracheostomy ventilation have favourable outcomes following hospital discharge. A coded diagnosis of pulmonary disease predicts early mortality in this group.

Introduction: Intermittent overnight oximetry is used to monitor the effectiveness of home NIV. Aim: The aim of this study was to determine if patients change their ventilator use on the night they have oximetry. Objectives: To compare NIV use when having oximetry to 7 days later. Methods: The Lumis ventilator comes with remote access for 3 months. Consecutive patients issued with a Lumis ventilator were identified who had oximetry within the first 3 months of issue. NIV Use on the night of oximetry was compared with 7 days later using paired t-tests. Results: 168 patients were issued a Lumis of which 25 (15%) had oximetry within 3 months. 16/25 (64%) received NIV for single aetiology respiratory failure (OHS 9, copd 3, skeletal 3, neuromuscular 2) and 9 (36%) for overlap (ohs/copd 6, ohs/skeletal 1, OHS/central apnoea/drugs 2).20 received bi-level ST mode, 8 dynamic modes e.g. IVAPS. 7 had oxygen. Oximetry results –median (range) were ODI 7.8 (0.3-51), Median SpO2 94 (84-99), time <90 40mins (0-399). Using a combined measure of optimal oximetry (ODI<5, Median SpO2>92%, Time <90% <30mins). 5/25 were optimal, 6 failed on one criteria, 9 on 2 and 5 on 3. The mean use of NIV on oximetry was 380 (SD 206) vs 319 mins (232) 7 days later. As a group there was no stat difference between the two nights (mean 60 mins more on oximetry , 95%CI 43 less to 165 more, p=0.24). However in some cases large differences in use were seen in either direction- range 377 less to 647 more. This is clinically important when assessing adequacy of NIV. Conclusion: Oximetry is best interpreted with NIV compliance data as there can be large and and unpredictable differences in use between nights when oximetry is and is not used.

Introduction The occurrence of Spinal Cord Injury (SCI) is often complicated by the need for assisted ventilation. This study charts the weaning outcomes of mechanically ventilated SCI subjects admitted over a 10 year period to a regional Spinal Injuries unit. Methodology Acute SCI subjects with Tetraplegia admitted from April 2007–2017 to the Northwest Regional Spinal Injuries Centre (NWRSIC) were identified. Only those presenting with all 3 criteria: a) admission injury level C1-C6 b) admission ASIA score A-C and c) need for mechanical ventilation on arrival to the NWRSIC were included in the final analysis. Results The cohort consisted of 84 subjects (mean age 57 (SD 18) years; 76% male; 81 surviving to discharge). On admission, the level of injury was C1–3 in 28% (C1–3 ASIA A 20%; C1–3 ASIA B 2%; C1–3 ASIA C 6%) and C4–6 in 72% (C4–6 ASIA A 38%; C4–6 ASIA B 17%; C4–6 ASIA C 17%). On admission, 86% (72/84) were tracheostomy ventilated 24 hours/day, 12% (10/84) tracheostomy ventilated at night only and 2% (2/84) using NIV. By discharge, 13% (11/81) were tracheostomy ventilated 24 hours/day (including 2 Phrenic nerve paced), 13% (11/81) tracheostomy ventilated at night only, 7% (6/81) prescribed nocturnal NIV with 65% (53/81) breathing independently. Thus, when taking the entire cohort, 63% (53/81) achieved complete Ventilatory liberation, 12% (10/81) weaned to nocturnal tracheostomy ventilation only and 6% (5/81) were weaned to NIV whilst no further weaning was possible in 16% (13/81). The ability to breathe independently by discharge was found to correlate with level of injury on admission (CC 0.39; p<0.001), level of injury on discharge (CC 0.47; p<0.001) and non-significant trend with improvement in neurological function during admission (CC 0.21; p=0.06) but not age or gender. Conclusion Our data demonstrates that in a cohort of consecutive SCI patients requiring mechanical ventilation on admission to a regional Spinal injuries unit, weaning from mechanical ventilation was possible in 84% of subjects with 63% being liberated completely from Ventilatory support by discharge. The use of NIV in the SCI cohort appears to be an emergent strategy during the weaning process.

A 45-year-old man presented with a one-year history of confusion and change in sleep pattern. He was obese and in ventilatory failure with episodes of stridor, sleep disordered breathing, central cyanosis, dysarthria, dysphagia myoclonus and gait ataxia. EEG showed generalised slowing and CSF was inflammatory. Polysomnography revealed increased sleep drive but reduced NREM sleep and a complete absence of REM sleep. An extensive workup for an autoimmune cause was negative and no cancer was identified. In view of inflammatory CSF, he was treated with IVIG, steroids, plasma exchange and cyclophosphamide, with dramatic improvement. Serum and CSF antibodies to IgLON5 returned positive. HLA genotyping confirmed HLA-DQB1*05:01 and HLA-DRB1*10:01 alleles. The few cases in literature (<20) of this recently described autoimmune syndrome have all been fatal, but our patient maintains improvement at 2 years from onset. His behaviour has normalised and sleep pattern improved with the return of dreams, and resolution of neck pain and dysphagia. This is the first report of a recently recognised rare autoimmune disorder in the UK that has shown sustained improvement with treatment suggesting at least a proportion of cases responds to treatment.

A 45-year-old man was seen with a history of confusion and disorientation for 1 year, during which time he was unable to identify relatives. He was unable to cope at his work as a plasterer, noticed neck pain, dysphagia, and unexplained weight gain. His family reported that his sleep pattern had changed, describing his sleep as disturbed with episodes consistent with stridor, myoclonus (the video at [Neurology.org/nn][1]), and semipurposeful movements. Acknowledgment: The authors thank Edward Jackson and the staff in the Neurosciences Laboratories with the help processing and providing the immunofluorescence results. [1]: /lookup/doi/10.1212/NXI.0000000000000383

Background: Little guidance exists on patient selection for home non-invasive ventilation (NIV) after acute NIV. The role of home NIV in pure COPD is inconclusive (Struik et al. Thorax 2014). Aims: Evaluation of the effect of a referral proforma (RPF) on patient characteristics and outcomes. Methods: The RPF was developed based on current evidence. Data collection pre- and post-RPF included diagnosis, length of stay and survival (Oct 912–Feb 915). Results: Pre-RPF, 55 referrals were received (3.7/month); 8 transfers were not given NIV. Post-RPF, 62 referrals were received (4.8/month), of whom only 25 (40.3%) were referred by RPF. All post-RPF transfers were given NIV; 2 declined. Gender and age varied little (Table One). No significant difference in length of stay, 6- and 12-month survival was seen (pre-RPF v. post-RPF; and pre-RPF v. post-RPF by RPF only). Fewer COPD patients were referred post-RPF (33.9% v. 47.3%, p=0.14). Those referred by RPF were significantly less likely to have COPD than pre-RPF (20.0% v. 47.3%, p=0.02), with increased OHS (obesity hypoventilation; 35.1% v. 21.8%, p=0.04). Referrals without RPF showed similar COPD rates to pre-RPF (47.3% v. 43.2%, p=0.71). Conclusion: While home NIV in COPD is debated, an RPF may improve practice while highlighting the challenges of changing existing behaviours (Stoller, Respiratory Care 2010).

### Principles of mechanical ventilation#### Modes of mechanical ventilationRecommendation 1. Pressure-targeted ventilators are the devices of choice for acute NIV (Grade B).Good practice points #### Choice of interface for NIVRecommendation 2. A full face mask (FFM) should usually be the first type of interface used (Grade D).Good practice points #### Indications for and contra-indications to NIV in AHRFRecommendation 3. The presence of adverse features increase the risk of NIV failure and should prompt consideration of placement in high dependency unit (HDU)/intensive care unit (ICU) (Grade C).Good practice points #### Monitoring during NIVGood practice points #### Supplemental oxygen therapy with NIVRecommendations 4. Oxygen enrichment should be adjusted to achieve SaO2 88–92% in all causes of acute hypercapnic respiratory failure (AHRF) treated by NIV (Grade A).5. Oxygen should be entrained as close to the patient as possible (Grade C).Good practice points

Background: Wireless monitoring and titration of NIV (non-invasive ventilation) has been reported in one European centre (Pinto et al. J Neurol Neurosurg Psychiatry 2010); this is now available in a standard UK device. Aim: Pilot evaluation of the effectiveness of therapy, data transfer and remote modification. Methods: Eligible patients included those new to or already using NIV, excluding patients with likely ventilator dependency; 3 patients were issued Lumis 150 VPAP ST-A devices for trial (Resmed, UK). Results: Ventilator interaction data has been visible since initiation. Patient A: Leak was seen to exceed recommendations for this device and circuit (24LPM), with a median of 59.2LPM. Mask and setting change has been undertaken and monitored remotely, avoiding inpatient titration or 110km travel; median leak is now 5.2LPM. Patient B: Satisfactory compliance (all days >4hours), tidal volume (median 944ml) and minute ventilation (median 13.5LPM) seen; no modifications made. Patient C: Symptomatic sleep disordered breathing and poor compliance with CPAP led to trial of NIV. Remote data shows that NIV was used 19 times in 28 days (average 2.1h/night). Tidal volumes are satisfactory (median 697ml) but continued low compliance has guided input. Conclusion: Remote NIV titration and monitoring is available and effective, allowing targeted input and potential streamlining of initiation and follow-up.

INTRODUCTION:Antimicrobial resistance threatens to undermine treatment of severe infection; new therapeutic strategies are urgently needed. Preclinical work shows that augmented passive immunotherapy with P4 peptide increases phagocytic activity and shows promise as a novel therapeutic strategy. Our aim was to determine ex vivo P4 activity in a target population of patients admitted to critical care with severe infection. METHODS:We prospectively recruited UK critical care unit patients with severe sepsis and observed clinical course (≥3 months postdischarge). Blood samples were taken in early (≤48 h postdiagnosis, n = 54), latent (7 days postdiagnosis, n = 39), and convalescent (3-6 months postdiagnosis, n = 18) phases of disease. The primary outcome measure was killing of opsonized Streptococcus pneumoniae by neutrophils with and without P4 peptide stimulation. We also used a flow cytometric whole blood phagocytosis assay to determine phagocyte association and oxidation of intraphagosomal reporter beads. RESULTS:P4 peptide increased neutrophil killing of opsonized pneumococci by 8.6% (confidence interval 6.35-10.76, P < 0.001) in all phases of sepsis, independent of infection source and microbiological status. This represented a 54.9% increase in bacterial killing compared with unstimulated neutrophils (15.6%) in early phase samples. Similarly, P4 peptide treatment significantly increased neutrophil and monocyte intraphagosomal reporter bead association and oxidation, independent of infection source. CONCLUSIONS:We have extended preclinical work to demonstrate that P4 peptide significantly increases phagocytosis and bacterial killing in samples from a target patient population with severe sepsis. This study supports the rationale for augmented passive immunotherapy as a therapeutic strategy in severe sepsis.

The British Thoracic Society (BTS) published the guideline ‘The use of non-invasive ventilation in acute respiratory failure’ in 2002.1 This was in response to trials that had demonstrated that non-invasive ventilation (NIV) was an alternative to invasive mechanical ventilation (IMV) in life-threatening respiratory acidosis due to acute exacerbations of chronic obstructive pulmonary disease (AECOPD). It drew attention to evidence that, when NIV was used in the less severely unwell patient, it also limited progression to more severe respiratory failure.2 The trial also demonstrated the feasibility, of delivering NIV on general medical or admission wards that had enhanced support and when staff were provided with ongoing training.In subsequent years, NIV has been shown to deliver better rather than equivalent outcomes to invasive ventilation in AECOPD and better evidence has accumulated for the use of NIV in non-COPD disease in the intervening years. Repeated national audits have, however, raised concerns that expected patient benefit is not being delivered and have pointed to a number of process deficiencies.3–5 There is also the risk, in the absence of justifying trial evidence, that the preferred use of NIV in AECOPD might be extended to all hypercapnic patients, irrespective of circumstance or underlying disease process. That this is a real risk might be inferred from the BTS audits where the indication for NIV was not COPD in over 30% of cases.3 ,4NIV development in the UK has been largely outside the organisational ‘umbrella’ of critical care. This may have adversely affected resource allocation and contributed to a lack of integration in NIV and IMV patient pathways. Other unintended consequences might be a restriction on access to invasive ventilation and delay in the development of extended applications of NIV, such as accelerating extubation and its use in the management of …

Introduction Non-invasive ventilation (NIV) in motor neurone disease (MND) is an evidence-based therapy, recommended by NICE. A single centre randomised trial of 41 patients underpins much of current practice,1 it was suggested our patient cohort may differ from those in the original trial work. Methods Retrospective review of all patients offered NIV from 01.01.2013 to 30.06.2015. Data was taken from the initial neurology referral, and NIV set-up. Demographics were compared with the Newcastle study1 (Table 1). Twelve month survival, and/or death post NIV initiation were assessed. Results Sixty-three patients were offered trial of NIV; 5 declined admission, and 7 declined NIV. Fifty-one patients were discharged with NIV, of whom 4 rapidly discontinued ventilation. Forty-seven patients were followed as NIV users, 35 for at least a year or to death. Fifty-seven percent were documented as having bulbar symptoms, the severity of which were not formally assessed. Twenty-nine percent received formal carer support at NIV initiation. Of the 35, 24 (68.6%) died within one year of NIV commencement, and median survival for all deaths was 177 days (range 4–630 days). Patients who died were significantly more likely to have bulbar dysfunction (18/24, p = 0.003) with a trend to reduced survival, median 149 vs. 239.5 days non-bulbar (p = 0.09). Twenty patients are alive at data collection, current median survival 292 days (range 7–793 days) and this data will affect results. Those with carers in place had a significantly lower ALSFRS-R (26.3 vs 32.4, p = 0.008) and shorter median survival (135 days). Of those dying or surviving at least a year, 22/35 (63%) were issued with cough-assist support (18/22 mechanical in/exsufflation). Conclusions Our cohort and outcomes are similar to those in the Bourke trial. Patients with bulbar disease, and/or pre-existing care input may have worse survival. Current users will be followed up to complete the dataset for survival. The impact of bulbar disease, cough augmentation2 and carer need remain uncertain. Ways to better assess and support these groups should be sought, and adequately powered randomised trials in these areas developed. References 1 Bourke SC, Tomlinson M, Williams TL. et al. Effects of non-invasive ventilation on survival and quality of life in patients with amyotrophic lateral sclerosis: a randomised controlled trial. Lancet Neurol. 2006;5:140–7 2 Rafiq MK, Bradburn M, Proctor AR. et al. A preliminary randomized trial of the mechanical insufflator-exsufflator versus breath-stacking technique in patients with amyotrophic lateral sclerosis. Amyotroph Lateral Scler Frontotemporal Degener. Published Online First: 3 Jul 2015. doi:10.3109/21678421.2015.1051992

A 53-year-old woman with spinal muscular atrophy and a 7-year history of nocturnal non-invasive ventilation (NIV) use via nasal mask and chinstrap was admitted electively. Outpatient review suggested symptomatic hypercapnia and hypoxaemia. Use of her usual NIV resulted in early morning respiratory acidosis due to excess mouth leak, and continuous face mask NIV was instigated while in hospital. Once stabilised, she elected to return to nasal ventilation. At outpatient review, respiratory acidosis reoccurred despite diurnal use of NIV. Using the patient's routine ventilator and a novel mouthpiece and trigger algorithm, intermittent daytime mouthpiece ventilation (MPV) was introduced alongside overnight NIV. Control of respiratory failure was achieved and, vitally, independent living maintained. Intermittent MPV was practicable and effective where the limits of ventilator tolerance had otherwise been reached. MPV may reduce the need for tracheostomy ventilation and this case serves as a reminder of the increasing options routinely available to NIV clinicians.

Introduction HMV can be initiated and monitored as either inpatient or outpatient. There is little evidence for best practice in this field and inpatient ventilation beds are a scarce resource. We evaluated patients, with sub-optimal HMV, admitted to our tertiary unit for adjustments to consider whether these admissions were successful, and hence an effective use of resources. Methods Patients were identified from our ventilation unit’s database. Notes, oximetry and ventilator download from pre-admission, pre-discharge and post-discharge were retrospectively analysed. Results In a 6-month period (June–December 2013) 30 patients were admitted to our unit for adjustments of HMV. 43% were female. Obesity related sleep disorder formed the majority of underlying conditions (53%), with musculoskeletal deformities (20%) and neuromuscular conditions (10%) also frequently seen. Median length of stay was 2 days. HMV was discontinued during admission in 2 cases in line with patient wishes. 19 (63%) were deemed to have had successful admissions, defined as normalisation of at least one abnormal ventilation parameter (pCO2 >6.0, desaturations >14/hr, time below 90% of >30 min, mean saturations of <88%, usage >6 hrs, leak <50 L/min). Of the 19 successful admissions, 6 showed sustained improvement post-discharge. 11 (37%) admissions were deemed unsuccessful, poor baseline usage and missed outpatient appointments were observed in this group. Noteworthy improvements were made to oximetry parameters during admission, although not all of these were maintained post-discharge (Table 1). Ventilator leak and usage information was available for 22 (73%) patients. Excess leak (50 >L/min) was seen in 10 patients pre-admission, only 1 patient had excess leak post-discharge. Pre-admission usage of <2 h/night was seen in 6 patients, only 1 showed sustained improvement in usage. 8 patients were admitted with usage of 2– 4 h, 4 improved post discharge usage to >6 h and only 1 showed deterioration in usage. Conclusion Admitting patients for adjustments to HMV can improve ventilation parameters yet only some of these improvements are maintained after discharge. There appears to be a subset of patients who do not benefit from inpatient admissions, particularly patients with poor baseline usage. We suggest careful selection of patients to ensure effective use of limited resources.

Background The apnoea–hypopnoea index (AHI) is used to define Obstructive Sleep Apnoea Syndrome (OSAS). Some subjects however, present primarily with excessive daytime sleepiness (EDS) and loud snoring, but investigation may reveal an elevated Respiratory Disturbance Index (RDI) with most events comprising Flow limitations. Little UK based data exists regarding treatment outcomes in this group. Methodology/results 118 subjects (mean age 52 years; Epworth sleepiness scale score (ESS) 13.58 (5.30); 80% male) presented between November 2011–October 2013 to the Sleep Service with EDS as a primary symptom, loud snoring, RDI >15 with AHI≤11 (Mean RDI 21.77 (9.43)); AHI 8.03(2.74); ODI 6.72 (4.49) and were treated with CPAP. At 30 day compliance review, 60% (71/118) had benefited from CPAP with mean ESS pre-CPAP 14.13 (5.12) falling to 7.70 (4.82) following CPAP. The mean BMI was found to be significantly higher in those 71 subjects benefiting from CPAP (33.20 (SD 8.13) v 30.26 (SD 7.40); p = 0.04) but no significant differences were noted in baseline Epworth score, age, gender, AHI, RDI, ODI and Pulse Transit Time (PTT). This “Flow Limitation” cohort was compared with 261 subjects (mean age 56 years; ESS 12.47(5.61); 82%Male) diagnosed with OSAS during the same time period (Mean AHI 37.11 (19.94); mean ODI 31.15 (19.74) and treated with CPAP. 76% (199/261) of the OSAS group reported benefit from CPAP; ESS fell from 13.24 (5.35) to 6.60 (4.74) following CPAP therapy. Comparing the “Flow Limitation” group with the “OSA” group, the mean BMI (32.03(7.94) v 34.70(8.65); p = 0.004) and age (51.75(12.34) v 56.20(12.18); p = 0.001) were significantly lower in the “Flow Limitation” subjects but no significant difference was noted in baseline ESS. Those deriving benefit from CPAP in the OSA group demonstrated significantly higher CPAP usage (4.45(2.24) v 3.83(2.15) hours/night; p = 0.04). Conclusion Basing treatment decisions on AHI rather than RDI may miss a proportion of patients exhibiting similar levels of EDS as those with OSAS who would otherwise have gained benefit from CPAP. Despite the observed benefit, CPAP usage appeared lower in this “Flow Limitation” cohort who appeared overall to be a younger group with a lower BMI compared to those with OSA.

Background Non-invasive ventilation (NIV) is an established treatment for patients with acute ventilatory failure. It can be successfully provided on a specialist ward, rather than intensive care (ICU) when certain criteria are met. It is frequently delivered outside ICU when a patient is deemed not suitable for invasive ventilation. Methods Deaths in 2012 on our dedicated ventilation unit were analysed as part of ongoing clinical governance. Information on demographics, admission diagnosis, respiratory and metabolic acidosis, consolidation or pulmonary oedema on chest radiograph reports, Glasgow Coma Score (GCS), serum creatinine and hospital length of stay prior to NIV were recorded. Escalation of care and resuscitation decisions were noted. Results There were 228 admissions for acute NIV, with 31 recorded deaths (13.6%), 22 case notes were available for review. Mean age was 79 years, 77.3% had known COPD, admission median MRC score of 4, and 18.2% had been in hospital for >7 days before NIV. All had acute hypercapnic respiratory failure. Not for resuscitation decisions had been made for 95.5% prior to NIV, and 100% had NIV as a ‘ceiling of care’. Mean pH was 7.25 (SD 0.06), similar to previous reports of admissions to our unit1, 22% had mixed acidosis (BE <-2.0 mmol/l). GCS was <8 in 9% and 36.4% had serum creatinine >100 µmol/l, all triggering alerts for acute kidney injury. Admission diagnoses are shown in figure 1. Radiographic consolidation was reported in 59.1% and pulmonary oedema in 18.2%. Conclusion The mortality of patients receiving acute NIV is low2. Most deaths had an underlying diagnosis of COPD, they were an elderly frail group, deemed inappropriate for escalation to critical care. There were multiple risk factors for NIV failure on initiation of therapy. Whilst a trial of NIV may have been appropriate based purely on blood gases, it was at high risk of failure and discussion about end of life care may have offered an alternative approach. References Chakrabarti et al. Thorax 2009;857–62. Roberts et al. Thorax 2010;43–8.