Discussion

This study demonstrates that each year in England around 2.6 million episodes of inpatient care are delivered to approximately 1.1 million smokers. At an individual and population level, intervening to maximise the number of smokers who receive effective NHS support to stop smoking as part of their hospital care represents an opportunity for both health promotion and cost management through a reduction in demand on NHS services. This is particularly true of people treated in mental health settings, among whom current smoking rates are double those in other specialities.

THIN dataset is broadly representative of the UK population and estimates of the prevalence of current smoking from THIN are comparable with those derived from national survey data,5 adding credibility to our estimates of the burden of smoking and opportunities for intervention at a national level.

NHS Stop Smoking Services provide freely available behavioural support and pharmacotherapy, but these services are predominantly community-based and hence not immediately available to smokers admitted to or discharged from hospital. Separate funding streams for community and hospital services mitigate against service integration since both services can argue that the other is responsible for inpatient smoking cessation provision. Our own recent trial data demonstrate that treating smoking as a default at the point of admission substantially increases the proportion of smokers accepting support,6 and this approach is now recommended in NICE guidance3 and the British Thoracic Society's Recommendations for Hospital Smoking Cessation Services.7 The findings of the present study indicate that default provision of behavioural support and pharmacotherapy to smokers treated as inpatients in English hospitals has the potential to reach 1.1 million smokers each year. NICE data indicate that this approach is cost-effective for both hospitals and wider society.3 The routine provision of smoking cessation support to smokers seen as outpatients, not captured in this study of inpatient data, will further increase the number of smokers reached. Indeed, HES outpatient data indicate 61.5 million outpatient appointments were attended by patients aged 20+ in England in the 2010–2011 financial year, approximately four times the number of inpatient FCEs, substantially increasing the number of opportunities to intervene with smokers.4

Our finding that around 47% of people treated in hospital are current or ex-smokers, compared with about 35% of those not treated, is broadly consistent with the strong role that smoking plays in causing disease, particularly respiratory and cardiovascular disease. However, the proportion of hospital patients who had quit smoking was also generally high, at around two-thirds, and broadly similar between specialties, with the exception of mental illness. Over 60% of people treated for mental illness were ever-smokers and 50% were current smokers. People with mental health problems are more likely to be smokers and to be more heavily addicted, but the extraordinarily high prevalence of continued smoking in this group reflects the previously highlighted failure of existing approaches to engage with this patient group.

Our study thus identifies both the importance and the magnitude of the opportunity to intervene to treat smokers and reduce the prevalence of smoking in NHS secondary care users. Implementation of current NICE guidance to deliver smoking cessation interventions as a routine component of secondary care provision could have a major impact on the prevalence of smoking, and hence on the morbidity and mortality that smoking causes.