Modelling the prevalence, healthcare costs and number of deaths in chronic obstructive pulmonary disease in England and Scotland

Abstract

Introduction Chronic obstructive pulmonary disease (COPD) has emerged as a major policy focus for health systems throughout Western Europe. This reflects the increased prevalence, associated healthcare utilisation and costs of COPD, and the potential to substantially improve outcomes through achieving reductions in smoking. The aim of this PhD was to develop projections for the prevalence, healthcare costs and number of deaths in people with COPD in England and Scotland over a 20-year horizon (i.e. from 2011 to 2030). Methods I undertook a phased programme of work, which began with a systematic review of the published and unpublished literature to identify models that were suitable for estimating and/or projecting the prevalence and disease and economic burden from COPD. This involved searching Medline, Embase, CAB Abstracts, World Health Organization (WHO) Library and Information Services and WHO Regional Indexes, and Google over the time period 1980-2013. The models were then critically appraised for their quality of reporting. From these, I selected the Dutch Model developed by Erasmus University for generating projections. Suitable data sources from both England and Scotland were identified, sourced and carefully processed in order to run the modelling exercises. Rates of incidence and prevalence were calculated using English and Scottish healthcare datasets and population data were obtained from the Office for National Statistics (ONS) and the General Register Office for Scotland (GROS). Relative risks for all-cause mortality among people with COPD were calculated from the Clinical Practice Research Datalink and mortality data were obtained from the ONS and GROS. The Model was thus adjusted to apply to England and Scotland. I then travelled to the Netherlands to work with the developers of the Dutch Model and ran a baseline model and an array of sensitivity analyses with modified inputs to the Model. Finally, my Rotterdam colleagues calculated uncertainty intervals for some of the estimates using probabilistic analysis. Results Using the probabilistic means and uncertainty intervals, in England, the modelled prevalence of diagnosed COPD among males of all ages in 2011 was 1.8% (95% uncertainty interval 1.8-1.9) increasing to 2.0% (1.7-2.1) by 2030. In females, in England, the baseline estimate was 1.8% (1.7-1.8) in 2011 increasing to 2.4% (2.0-2.6) in 2030. In Scotland, the modelled prevalence among males was 1.9% (1.8-1.9) in 2011 and this was projected to stay the same at 1.9% (1.7-2.2) by 2030. In females in Scotland, the estimated prevalence was 2.2% (2.1- 2.3) in 2011 and was projected to increase to 2.5% (2.1-2.7) in 2030.Using the Model I estimated that overall in 2011 there were a total of 952,000 (941,000-966,000) people with diagnosed COPD in England and 106,000 (103,000-110,000) in Scotland and that these numbers would increase to 1,325,000 (1,117,000-1,408,000) in England in 2030 and 125,000 (113,000-136,000) in Scotland in 2030, respectively. The greatest increase in COPD was projected to be in females over 65 years of age in both countries. The total annual direct healthcare costs of COPD in England were projected to increase from £1.60 (95% uncertainty interval 1.18-2.5) billion in 2011 to £2.35 (1.85-3.08) billion in 2030. In Scotland, costs were projected to increase from £170 (128-268) million in 2011 to £210 (165-274) million in 2030. These costs were calculated in terms of 2011 costs without the application of any economic trends (i.e. no annual increase applied for inflation). The number of deaths among people with COPD in England was estimated to be 99,000 (93,000-129,000) in 2011, increasing to 129,000 (126,000-133,000) in 2030. In Scotland there were estimated to be 10,000 (9,000-12,000) deaths in 2011, increasing to 14,000 (13,000-15,000) in 2030. The Dutch Model demonstrated a 39% increase in the number of people with COPD in England and a 17% increase in Scotland between 2011 and 2030. It provided an estimate of a 30% increase in deaths among people with COPD in England and of a 43% increase in Scotland. Overall, there was a projected 46% increase in the direct healthcare costs required to care for people with COPD in England and a 23% increase in Scotland between 2011 and 2030. The reasons for these differences are largely due to higher COPD-related excess mortality in Scotland and to differences in the data used for populating the model in both countries. Conclusions There are likely to be substantial increases in the number of people with COPD, associated morbidity, direct healthcare costs and mortality in both England and Scotland over the next two decades. These increases in numbers will predominantly occur in females over 65 years of age and are likely to have substantial societal impact in terms of organising the health and social care for this frail population.