by GPs and hospitals. <50% took action when concordance reports showed variance (under or over use) from prescribed treatment. 2. Lack of ability to detect hypoxic patients with >20% community healthcare sites not having access to an oximeter. 3. Fire safety officers are rarely advised about the persistent smoker (only 16%) despite the potential risk to patients, their families and the general public. Local guidance on appropriate steps to take is rare (35%). 3 HOS units denied LTOT for smokers and one assessed this by exhaled carbon monoxide measurement. 4. A variety of methods for protecting patients from excessive oxygen are favoured but use appears limited. When asked what policy respondents favoured, universal precaution (as promoted by ambulance guidelines) was most popular (60%) while 20% favoured oxygen cards and 20% patient specific protocols (PSP). 5. A specific local policy for removing oxygen when no longer indicated or used is rare (<25%). This, coupled with inadequate follow-up of patients started on oxygen during hospital admission, suggests significant waste with the current oxygen provision. 6. Respondents indicated guidance on oxygen removal, contract monitoring, assessment for ambulatory oxygen and training in arterial or capillary blood gases as being required.Conclusions Problems in healthcare coordination, public and patient safety and in removing oxygen once ordered were common. There is a need to integrate hospital and community teams and to prepare for safe mobilisation and contract management so that a quality home oxygen service can be provided in the future.

Methods We have compared the Inspiratory Flow rate (I) with the flow metre against Maximum Inspiratory flows taken from a flow volume loop (FVI) taken as part of routine lung function testing in 100 sequential subjects attending the Cardio Respiratory department for lung function testing. Results We have found major variability in the FVI on flow volume traces despite attempts to obtain traces with maximum volume and effort. Only 36% of subjects had variability between attempts of <1 l/ s, with 64% showing variability between attempts of >1 l/s, 24% of >2 l/s, and 3% >3 l/s. For measurements using the Inspiratory Flow metre If I>2 l/s all of the 38% of subjects showed FVI of 2 l/s or more. With I of <2 l/s there was agreement between the two methods 60.3 l/s in 26%, and a further 14% with FVI of <2 l/s. 40% of subjects with Inspiratory Flow (I) of <2 l/s had FVI of <2 l/s. But in 22% of subjects I <2 l/s but FVI >2 l/s. FVI-I showed mean difference for these subjects of 2.4 l/s (range 0.9e4 l/s). In total 78% of subjects showed concordance of Maximum Inspiratory Flow to >2 l/s or >2 l/ s between the two measurements and for 22% the inspiratory flow metre reading of <2 l/s did not reflect maximum Inspiratory Flow. Conclusion There are major variations in the Maximum Inspiratory Flow measured with a flow volume loop but for a simpler measurement with an Inspiratory Flow metre if Maximum flow is >2 l/ min then it is unlikely that Inspiratory flow is compromised. A simple clinic based measurement can be useful to exclude limitation of Inspiratory Flow but if abnormal further investigation is needed.

No abstract