The burden of smoking-related ill health in the United Kingdom

Transcription

1 The burden of smoking-related ill health in the United Kingdom S Allender, R Balakrishnan, P Sarborough, P Webster, M Rayner Researh paper Department of Publi Health, University of Oxford, Oxford, UK Correspondene to: Dr Steven Allender, British Heart Foundation Health Promotion Researh Group, Old Road Campus, Headington, Oxford OX3 7LF, UK; steven.allender@ dphp.ox.a.uk Reeived 23 May 2008 Aepted 27 Marh 2009 ABSTRACT Bakground: Smoking is one of the biggest avoidable auses of morbidity and mortality in the United Kingdom. This paper quantifies the urrent health and eonomi burden of smoking in the UK. It provides omparisons with previous studies of the burden of smoking in the UK and with the osts for other hroni disease risk fators. Methods: A systemati literature review to identify previous estimates of National Health Servie osts attributable was undertaken. Information from the World Health Organization s Global Burden of Disease Projet and routinely olleted mortality data were used to alulate mortality due in the UK. Population-attributable frations for smoking-related diseases from the Global Burden of Disease Projet were applied to NHS ost data to estimate diret finanial osts. Results: Previous studies estimated that smoking osts the NHS about 1.4 billion to 1.7 billion in 1991 and has been responsible for about deaths per annum over the past 10 years. We estimate that the number of deaths attributable in 2005 was (19% of all deaths, 27% deaths in men and 11% of deaths in women). Smoking was diretly responsible for 12% of disability adjusted life years lost in 2002 (15.4% in men; 8.5% in women) and the diret ost to the NHS was 5.2 billion in Conlusion: Smoking is still a onsiderable publi health burden in the UK. Aurately establishing the burden in terms of death, disability and finanial osts is important for informing national publi health poliy. Smoking is a major preventable ause of morbidity and mortality in the United Kingdom. 1 It is an important ause of ardiovasular disease, hroni obstrutive pulmonary disease (COPD), lung aner and a range of other aners, pepti uler and various other medial onditions. 1 In 2005, selfreported igarette smoking among British adults was 25% among men and 23% among women. 2 Despite reent redutions in the prevalene of smoking, it is still a serious publi health problem in the UK. 3 Beause smoking imposes a huge burden on population health and NHS resoures, quantifying the burden is important. Understanding the ontribution of smoking to the overall burden of ill health and to National Health Servie osts, ompared with that of other risk fators for hroni disease, helps inform the prioritisation of limited NHS resoures. In this way poliymakers an understand the relative osts of different risk fators and assess whether efforts to alter population levels of risk ould redue overall osts to the NHS. There have been various attempts to quantify the health and eonomi burden of smokingrelated ill health in the UK. The diret healthare osts of smoking were estimated to be between 1.4 billion and 1.7 billion in More reent estimates of the numbers of deaths attributable to smoking inlude in and per annum between 1998 and The omplexities of ost of illness studies mean they are not onduted regularly. Providing estimates of the health and eonomi burden of smoking that are up to date is neessary as the prevalene of smoking is dereasing (though less so in reent years), total healthare osts are inreasing and new interventions aimed at reduing the prevalene of smoking have reently been implemented notably the ban on smoking in publi plaes all of whih are likely to have had an effet on the burden of smoking-related ill health. These hanges are subjet to time lags in the hanging disease outomes and in the prevalene figures as shown by health surveys. A method is needed, whih is relatively fast and easy, to estimate the osts of smoking to the NHS. We have proposed a relatively simple and effiient method to generate estimates of the burden of ill health related to disease risk fators of publi health onern and have applied this method to generate estimates for diet, physial inativity and obesity The potential for suh a standard method that is simple and reliable is worthy of attention. The estimates this method generates give similar results to more omplex methods. 10 This paper uses this simple method to estimate the osts of ill health attributable in the UK. The results are ompared with previous estimates in partiular estimates generated by a systemati review of studies of the eonomi osts of smoking-related ill health in the UK. METHODS First, a systemati review of previous studies of the eonomi osts of smoking-related ill health in the UK was arried out. This review aimed to inlude all studies, published between 1997 and 2007, whih alulated the ost of smoking in the UK or any ountry of the UK. Seondly the burden of ill health due in all four ountries of the UK was alulated using routinely olleted mortality data and data on NHS osts, together with population attributable frations (PAFs) and other information from the World Health Organization s Global Burden of Disease Projet. The PAFs generated by this projet estimate the proportion of disease that an be Tobao Control 2009;0:1 7. doi: /t

2 Table 1 Summary of studies inluded in review with diret healthare ost of ill health attributable to smoking in the United Kingdom Referene Setting Diret ost (method) Disease inluded in estimate Parrott et al, England, 1991 Method 1: Caner 1.5 billion (attributable frations) Lung, upper respiratory sites, Method 2: 1.4 billion (ost omparison: smokers versus non-smokers) oesophagus, bladder kidney, stomah, panreas, myeloid leukaemia Respiratory Chroni obstrutive pulmonary disease, pneumonia Cirulatory Ishaemi heart disease, erebrovasular disease, aorti aneurysm, amyoardial degeneration, atheroslerosis Digestive Uler of stomah and duodenum Buk et al, UK, 1991 Method 1: Not speified 1.7 billion (attributable frations) Method 2: 1.4 billion (using 1994 General Household Survey data and ost to the NHS of treating illnesses aused by smoking) attributed to a partiular risk fator by sex and region. This study used the same methods that have been used to estimate the ost of ill health attributable to other risk fators for hroni disease in the UK Systemati review The systemati searh of the literature was onduted using the searh terms ost and smoking in Medline, CINAHL, Embase, Cohrane Library, National Health Servie Eonomi Evaluation Database, EonLit, Siene Citation Index, Soial Siene Citation Index, Index to Sientifi and Tehnial Proeedings and the Health Management and Poliy Database from the Healthare Management Information Consortium. An initial searh strategy was developed and tested on Medline and the searh terms were then used for the other databases. Studies were inluded if they related to the ost of smoking to the NHS in the UK and were published in English between August 1997 and September The time limits were applied to ensure that only reent ost estimates were onsidered. The referene lists of all papers were reviewed to identify other potentially relevant studies. Only studies reporting a new analysis of the ost to the NHS in the UK were inluded. Studies reporting previous analyses or estimates based on no analysis were exluded. Burden of disease study The method for alulating the health burden due and the diret ost of smoking to the NHS involved five steps. They were: Step 1: Diseases where smoking is a risk fator (as defined by WHO) were identified from the World Health Report for Step 2: Data on years of life lost (YLL), years of healthy life lost to disability (YLD) and disability adjusted life years () for the year 2002 were taken from WHO s Global Burden of Disease Projet for WHO Euro-A Region (United Kingdom and other European ountries with very low hild and very low adult mortality) Data on mortality for the year 2005 was obtained diretly from Offie for National Statistis for England and Wales, and the General Register Offies for Sotland and Northern Ireland Step 3: In 1996, the National Health Exeutive published a study whih asribed NHS osts to different diseases as defined by the International Classifiation of Disease (ICD 9). 16 NHS total osts were defined in this doument as the sum of NHS inpatient and outpatient osts, NHS primary are expenditure, NHS pharmaeutial expenditure and NHS net ommunity are servies expenditure. The proportion of total NHS expenditure (on NHS inpatient and outpatient osts, NHS primary are expenditure, et) by disease ode in was applied to the total ost of the NHS in for all four ountries in the UK to provide an estimate of the total NHS osts for diseases that were identified as being related The diseases related were onsidered to be those asribed within the Global Burden of Disease Projet inluding mouth and oral aner, trahea/bronhus and lung aner, hroni obstrutive pulmonary disease and ardiovasular disease. Step 4: Estimates of the burden of disease in the PAFs were alulated by the Global Burden of Disease Projet. 21 The PAFs for smoking, by sex and ondition and relevant to the UK, were extrated from the World Health Report The smoking PAFs were alulated using a base theoretial population in whih all individuals had no tobao use. Step 5: The burden (in terms of eonomi ost, mortality and morbidity) of smoking was alulated by applying the PAFs for diseases related to diseasespeifi data. Cost data for eah of the four ountries in the UK were derived using the method desribed in step 3. Mortality data four eah of the four ountries in the UK were taken from routinely olleted soures RESULTS Systemati review The searh strategy generated 4338 papers of possible relevane. After applying the inlusion riteria 4207 were rejeted on the 2 Tobao Control 2009;0:1 7. doi: /t

3 Table 2 Proportion of deaths (mortality), years of life lost in early deaths (YLLs), years of life lost in disability (YLDs) and disability adjusted life years () due -related diseases in WHO-EUR-A region, 2002 % of mortality % of YLLs % of YLDs % of (n = ) (n = ) (n = ) (n = ) Conditions related Mouth and oral aner Trahea/bronhus/lung aner Other aners Chroni obstrutive pulmonary disease Cardiovasular diseases Other medial onditions* Total related Communiable diseases, maternal/ perinatal onditions and nutritional defiienies Digestive diseases{ Diabetes mellitus Diseases of the genitourinary system Neuropsyhiatri disorders Musuloskeletal diseases Injuries Other Total *Inludes pepti uler and all respiratory onditions exept hroni obstrutive pulmonary disease. {Exluding pepti uler. basis of title or abstrat and a further 135 were rejeted after reviewing the full paper. The majority of papers were exluded beause they reported on the ost of smoking in ountries other than the UK mostly the United States. Some studies were exluded beause they only reported the osts of a partiular intervention, suh as smoking essation or niotine replaement pathes. There were two studies that met the inlusion riteria and reported on the eonomi ost of smoking to the NHS in the UK. 4 5 Both studies provide estimates for 1991 ranging from 1.4. billion to 1.7 billion and are summarised in table 1. Parrott et al 4 produed similar ost estimates using two different approahes; the appliation of attributable frations to 1991 ost data for diseases inluding a number of aners, respiratory and irulatory diseases and by omparing the osts of smokers and non-smokers. Buk et al 5 produed a slightly higher estimate also using a PAF although omparisons between these studies are diffiult beause the diseases inluded in the seond study were not speified. Burden of disease study Morbidity and mortality Table 2 shows the proportion of deaths, YLLs, YLDs and in WHO EUR-A region by disease for It shows that those diseases assoiated with smoking were responsible for 74% of all mortality, 67% of YLLs, 18% of YLDs and 41% of. Table 3 shows the total lost and the PAFs for smoking-related ill health in WHO EUR-A region for It shows that 12.1% (male: 15.4%; female: 8.5%) of all lost were diretly attributable. Cardiovasular disease was the largest ontributor to the burden of disease attributable, aounting for 3.8% of all lost followed by trahea/bronhus/lung aner (2.7%) and hroni obstrutive pulmonary disease (2.3%). Using PAFs derived from the WHO s Global Burden of Disease Projet, we estimated that of the deaths in the UK in 2005, deaths (18.6%) were diretly attributable (27.2% of all male deaths (77 154) and 10.5% of all female deaths (32 010)). The proportion of deaths attributable ranged from 19.7% in Sotland to 12.2% in Northern Ireland (table 4). Diret ost to the NHS Table 5 shows that the diret ost to the NHS for smoking attributable onditions was estimated to be 5.17 billion (5.5% of total healthare osts) in Of the 5.17 billion diretly attributable, the largest proportion was due to ardiovasular diseases ( million). Table 5 also shows the diret ost to the NHS for smoking attributable onditions by ountries in the UK. DISCUSSION We estimate that deaths (18.6% of all deaths) in the UK in 2005 an be attributed (27.2% of male deaths and 10.5% of female deaths). Our estimate is similar to that produed by Twigg et al that on average, eah year deaths (17.2% of all deaths) were from smoking-attributable auses aross the UK over the period (23% of male deaths and 12% of female deaths). 7 Other previous estimates of smoking attributable deaths inlude in 1995 by Callum et al, in 1997 by the Royal College of Physiians 23 and in 2000 by Peto et al. 6 This suggests that overall numbers of deaths attributable in the UK have not hanged muh in the past 10 years. It is diffiult to ompare estimates of the number of smoking attributable deaths beause of methodologial differenes, the hanging prevalene of smoking, hanges in the resident population (57 million in 1991 to 61 million in 2006 in the UK), et. For example Callum and White 24 used age-speifi attributable frations for London to derive estimates aross all boroughs and parliamentary onstituenies an approah that underestimates the attributable burden of smoking in areas Tobao Control 2009;0:1 7. doi: /t

4 Table Smoking-related onditions Total disability adjusted life years () lost and population attributable frations for smoking-related ill, by gender, in WHO EUR-A region, Male Female Total lost (n) PAF lost (%) lost %of all male lost (n) PAF lost (%) lost % of all female lost (n) PAF lost (%) lost Mouth and oral aner Trahea/bronhus/ lung aner Other aners Chroni obstrutive pulmonary disease Cardiovasular diseases Other medial onditions* Total related to smoking *Inludes pepti uler disease and all respiratory onditions exept COPD. {Exluding pepti uler disease. COPD, hroni obstrutive pulmonary disease; PAF, population attributable frations. % of all where smoking prevalene is higher than in London and overestimates the effet where smoking rates are lower than in London. Twigg et al 7 used the same published relative risk fators as Callum and White to alulate smoking attributable mortality for primary are trusts and strategi health authorities in England. These authors extended the previous approah by applying multi-level syntheti estimation tehniques to produe estimates at small area level and ombining primary are trusts estimates to reate figures for the four ountries within the UK. The deline in overall prevalene of smoking together with the inrease in smoking among younger age groups inreases the omplexity of the issue. In this study we used the WHO s (2002) population attributable frations for smoking-related onditions and the latest (2005) mortality data for the UK. This study therefore provides the most up-to-date estimates of the burden of smoking-related illness. Our study estimates that smoking ost the NHS over 5 billion in This equates to 5.5% of total NHS expenditure in The systemati review identified two studies that estimated the ost of smoking to the NHS. The first study estimated the ost to be between 1.4 billion and 1.5 billion in and the seond study estimated the ost between 1.4 billion and 1.7 billion in the same year. 5 However, the total NHS healthare ost has gone up from 31 billion in to 80 billion in , so although our estimate is higher than these previous estimates, it does not differ greatly as a proportion of total healthare osts (4.9% to 5.5% in 1991 ompared to 5.5% in ). We an assume that muh of the differene in ost was due to the inrease in overall healthare osts sine Another potential soure of differene between our estimate of the ost of smoking to the NHS and previous estimates is the PAFs used. Both the previous studies used smoking attributable frations alulated for 1991 taken from Callum (1995), 22 while our study used PAFs for 2002 from WHO s World Health Report. 11 Our estimate that smoking was responsible for 5.5% of total healthare osts is omparable with estimates of the eonomi burden in other developed ountries. These estimates range from 11.8% of total healthare osts (in the United States in ) to 3.3% (in Germany in ). In this paper we have underestimated the burden of smokingrelated ill health for three reasons. First, we have not inluded indiret osts in our eonomi analysis; seond, we have not inluded the burden due to passive smoking; and, third, we have not onsidered all onditions related. Indiret osts inlude those aused by prodution losses due related premature mortality, morbidity and informal are of people with smoking-related illnesses, but also other osts suh as sikness absene payments and the osts of smoking-related fires. For Sotland, Parott et al estimated that the annual ost of smoking-related produtivity losses, absene from work and fire assoiated with smoking were, in 2000, 40 million, 450 Table 4 Total deaths and deaths attributable by gender and ountries in the UK, in 2005 All ages England Wales Northern Ireland Sotland UK total All deaths Male Female Total Deaths attributable Male Female Total % deaths attributable Tobao Control 2009;0:1 7. doi: /t

5 Table 5 Perentage of total NHS osts attributable for different diseases in and and ost (% NHS total osts) 5 PAF (%) England Wales Sotland Northern Ireland Total NHS osts ( m) Cost attributable to smoking ( m) Total NHS osts ( m) Cost attributable ( m) Total NHS osts ( m) Cost attributable to smoking ( m) Total NHS osts ( m) Cost attributable ( m) Total UK ost attributable ( m) Smoking-related onditions Mouth and oral aner Trahea/bronhus/lung aner Other aners Cardiovasular diseases , Chroni obstrutive pulmonary disease Other medial onditions* Total related Infetious diseases Neuropsyhiatri disorders Musuloskeletal diseases Injuries and aidents Other Total *Inludes pepti uler disease and all respiratory onditions exept COPD; {Exluding pepti uler disease. COPD, hroni obstrutive pulmonary disease; DALY, disability adjusted life years; PAF, population attributable fration. Tobao Control 2009;0:1 7. doi: /t

6 What this paper adds The ost of smoking-related ill health to the National Health Servie in the UK is 5.2 billion per year, nearly 5.5% of the total NHS budget. The ost of smoking to the NHS, as a proportion of the total NHS budget, has not hanged substantially sine the early 1990s. The ost of smoking to the NHS is omparable to the ost of diet-related ill health, estimated to be approximately 6.0 billion per year. million and 4 million, respetively. 27 Inluding indiret osts would inrease our estimate of the finanial burden onsiderably. The British Medial Assoiation estimates that eah year in the UK at least 1000 people die as a result of seondhand smoke 28 and more than hildren under the age of 5 are admitted to hospital beause of the effets of passive smoking. 29 Parrott and Godfrey have estimated that eah year in the UK 410 million is spent in treating hildhood illnesses related to passive smoking. 30 If these estimates are orret then inluding the effets of passive smoking would substantially inrease our estimates of the health and finanial burden due. The analysis presented here only onsidered onditions for whih the WHO Burden of Disease Study has alulated PAFs. Our report did not inlude other onditions reported to be assoiated with smoking suh as infertility, ompliations during pregnany, stillbirths, neonatal deaths, post-menopausal osteoporosis, eye diseases, et The WHO Global Burden of Disease data are also limited by the lak of age-speifi information. The appliation of a single estimate aross the total adult population assumes that disease patterns are onsistent aross all age groups. A further limitation to this and similar studies is the absene of urrent NHS ost data by disease ategory. In this study the data had to be extrapolated from 1992 data (beause there is no more reent study) and learly methods of treating diseases assoiated with smoking (and the osts assoiated with those methods) have hanged signifiantly sine then. In 2002, the Department of Health initiated the National Programme Budget Projet (NPBP), whih provides detailed expenditure information for primary are servie under 23 programmes of are based on medial onditions suh as mental health, ardiovasular disease and aner. While these data are only available for England and only for some speifi disease ategories they provide a hane to understand the hange in relative proportion of osts for these diseases between 1992 and In the NPBP reported problems of irulation as osting 6.1 billion, obstrutive airway disease 534 million and infetious diseases at 1.2 billion. For England our method returned estimates 9.7 billion for ardiovasular disease, 1.7 billion for COPD and 834 million for infetious disease. The methods used for this study only provides a piture of the past burden of smoking beause of its retrospetive nature, and does not reflet the suess or otherwise of reent interventions. This is beause of the lag time between any intervention and the subsequent influene on disease outomes. As one example, Peto 6 found that people who stop smoking before middle age will avoid 90% of the risk attributable to tobao but the effet of essation would not show itself on hospital osts until later life when the symptoms of lung aner would manifest. These limitations notwithstanding, omparative information about the osts of smoking and other risk fators for hroni disease is important in making resoure alloation deisions regarding smoking essation and other preventive ations. Previous work using the methods used here has estimated 2002 osts due to poor diets in the UK to be 6 billion, overweight and obesity to be 3 billion and physial inativity to be 1 billion This study enables a omparison to be made between the diret ost of smoking to the NHS and other disease risk fators. This omparison should help poliymakers prioritise publi health interventions and make effetive use of limited NHS resoures. Funding: SA, PS and MR are supported by the British Heart Foundation. The sponsor had no role in the study design, data olletion, interpretation, writing or submission of this manusript. Competing interests: None. REFERENCES 1. Peto R, Lopez AD, Boreham J, et al. Mortality from smoking in developed ountries, : indiret estimation from national vital statistis. Oxford: Oxford University Press, Davy M. Time and generational trends in smoking among men and women in Great Britain, /05. Health Stat Q 2006;32: Petersen S, Peto V. Smoking statistis. London: British Heart Foundation, Parrott S, Godfrey C, Raw M, et al. Guidane for ommissioners on the ost effetiveness of smoking essation interventions. Thorax 1998;53(suppl 5):S Buk D, Godfrey C, Parrott S, et al. Cost effetiveness of smoking essation interventions. London: Health Eduation Authority, Peto R, Lopez A, Boreham J, et al. Mortality from smoking in developed ountries: 2006 update. Aessed 22 May Twigg L, Moon G, Walker S. The smoking epidemi in England. London: Health Development Ageny, Rayner M, Sarborough P. The burden of food related ill health in the UK. J Epidemiol Comm Health 2005;59: Allender S, Foster C, Sarborough P, et al. The burden of physial inativity-related ill health in the UK. J Epidemiol Comm Health 2007;61: Allender S, Rayner M. The burden of overweight and obesity-related ill health in the UK. Obesity reviews 2007;8: World Health Organization. World Health Report Geneva: WHO, World Health Organization. Comparative quantifiation of health risks: global and regional burden of disease attributable to seleted major risk fators. Geneva: WHO, England and Wales, Offie of National Statistis. Deaths registered by ause and area of residene. mortality. Aessed 22 May Sotland General Register Offie. Deaths registered by ause and area of residene. Aessed 22 May Northern Ireland General Register Offie. Deaths registered by ause and area of residene. annual_reports/2005/table6.4_2005.xls. Aessed 22 May NHS Exeutive. Burdens of disease: a disussion doument. Wetherby: Department of Health, Department of Health. Resoure aounts HC 1680: November London: Stationery Offie, Publiations/PubliationsPoliyAndGuidane/DH_ Aessed 22 May Information Servies Division. Sottish Health Servie Costs (Costs Book 2006). Edinburgh: ISD, Aessed 22 May Department for Health and Soial Servies. Wales Programme Budget results Welsh Health Cirular (WHC 2007,15). Cardiff: Welsh Assembly Government, Department of Health, Soial Servies and Publi Safety. Summary of HPSS Expenditure in Northern Ireland (April 2005 Marh 2006). A summary report. Belfast: DHSSPS, Murray CJL, Lopez AD. Global mortality, disability, and the ontribution of risk fators. Global Burden of Disease Study. Lanet 1997;349: Callum C. The UK smoking epidemi: deaths in London: Health Eduation Authority, Royal College of Physiians. Niotine addition in Britain: a report of the Tobao Advisory Group. London: Royal College of Physiians, Callum C, White P. Tobao in London: the preventable burden. London: Smokefree London and the London Health Observatory, Miller LS, Zhang X, Rie DP, et al. State estimates of total medial expenditures attributable to igarette smoking, Publi Health Reports 1998;I43: Tobao Control 2009;0:1 7. doi: /t

7 26. Neubauer S, Welte R, Beihe A, et al. Mortality, morbidity and osts attributable to smoking in Germany: update and a 10-year omparison. Tob Control 2006;15: Parrott S, Godfrey C, Raw M. Costs of employee smoking in the workplae in Sotland. Tob Control 2000;9: British Medial Assoiation. The human ost of tobao. Passive smoking: dotors speak out on behalf of patients. Edinburgh: British Medial Assoiation (Sotland), Royal College of Physiians of London. Smoking and the young: a report of a working party of the Royal College of Physiians. London: RCP, Parrott S, Godfrey C. Eonomis of smoking essation. BMJ 2004;328: Bartehi CE, MaKenzie TD, Shrier RW. The human osts of tobao use first of two parts. N Engl J Med 1994;330: Solberg Y, Rosner M, Belkin M. The assoiation between igarette smoking and oular diseases. Surv Ophthalmol 1998;42: Tobao Control 2009;0:1 7. doi: /t