1. Background
1.1: History and Burden of COPD Chronic Obstructive Pulmonary Disease (COPD) is believed to have been first described by Theophile Bonnet in 1679 in the form of cadaveric 'voluminous lungs'(7). Despite being a major cause of disability and death in the UK(1), COPD still remains a relatively poor relation to many of the other current leading causes of death in terms of health resource allocation, patient and public understanding and the interest and motivation of health professionals to diagnose and manage the disease(8). Over the last three hundred years, COPD has been known to health professionals and the public in a variety of guises including emphysema, chronic bronchitis, smoker's cough and chronic airflow obstruction. The term, Chronic Obstructive Pulmonary Disease is believed to have been used for the first time in 1965 by William Briscoe at the 9th Aspen Emphysema Conference(7). Although the name has now been successfully unified, the negative associations, variable diagnosis rates and management strategies amongst patients with the disease continue. Sadly, a nihilistic attitude towards the value of diagnosis and treatment of the disease is still found in some clinicians, with some feeling the value of accurate and early diagnosis is small as there is no evidence for any disease course altering interventions(8). Whilst there are a lack of disease-modifying interventions other than smoking cessation, there are effective pharmacological and non-pharmacological interventions that have been shown to improve disease symptoms(4). Accurate, early diagnosis and disease education are the first steps to achieving high quality care throughout the whole patient journey up to and including palliative care.
COPD patients may be perceived as either being on a slow emphysematous downward decline or, to spiral rapidly through the cycle of recurrent exacerbations and hospital admissions to lung function decline and death (8-10). However, with the range of effective interventions now available (including pharmacological strategies, patient education, diet and exercise strategies, vaccination and targeted smoking cessation interventions) early diagnosis and ongoing proactive care should be the 'gold standard'(4).
The relatively recent evidence supporting pulmonary rehabilitation(11), lung volume reduction techniques(12), and modern developments in pharmacotherapy and preventative care have provided health care professionals with access to therapies which have proven benefits in terms of morbidity. The wealth of inhaled therapies flooding the market and the enormous cost burden of COPD to the NHS has led to more interest and research in the disease in the last decade. This research has generated an increased understanding of disease pathology, recognition of multiple disease phenotypes and the growing acknowledgement of the need to direct therapies to individuals rather than the disease population as a whole.
Increasing bed pressures in the NHS has shifted focus towards identification of those high-risk patients who are recurrent hospital attendees, and resources have been directed towards strategies to reduce hospitalisation and towards community based care. Currently, services vary enormously between hospital trusts and commissioning groups, ranging from Community Matron support, hospital admission avoidance and early discharge schemes, to full time, multidisciplinary, integrated COPD teams all of which are attempting to bridge the gap in service provision and communication between Primary and Secondary Care(4).
1.2: Risk Stratification within COPD In the last ten years, much research has been directed towards risk stratification within COPD, and whereas previous risk stratification has centred on lung function, i.e. FEV1, it is now recognised that the disease is far more complex(4). A variety of factors such as symptom burden, historic exacerbation rate, exercise tolerance, smoking status and co-morbidities have been shown to be significant markers of health status and risk in COPD (4, 13-15)and so, may be useful to predict deterioration and, consequent health and social care utilisation. Thus far, attention has been mainly focused on highlighting factors indicating deterioration into end stage disease with the aim of identifying how to most efficiently deploy health services to manage these patients with both high symptom burden and high levels of health care utilisation. Whilst this remains an important point of research there is a limit to what can be achieved at this, relatively late point in the disease process, where management is directed at achieving what is effectively, high quality, multi-disciplinary, palliative care.
We hypothesise that to make a bigger difference to patients and the health service in the longer term, attention needs to be focussed on the earlier stages of COPD before the bulk of irreversible lung damage has already occurred.
1.3: The DOSE score A variety of scoring systems exist to try to risk stratify COPD. The DOSE score (Dyspnoea Obstruction, Smoking and Exacerbation) is an index of severity in COPD developed by Jones et al in 2009 from an audit data set of patients from Devon in the UK(16). The DOSE Index points associated with each category of all four variables (mMRC Dyspnoea Scale Score, FEV1% predicted, smoking status, exacerbations per year) are added to build the DOSE Index score. It was derived from multiple regression analyses of possible markers of health status in COPD and their relationship to quality of life scores. A scoring system was developed according to clinical and statistical strength and this was validated against three other data sets from London, Holland and Japan. It correlates well with other COPD severity scoring systems such as the BODE (Body mass index, degree of airflow Obstruction, Dyspnoea and Exercise capacity) index(17) and ADO (Age, Dyspnoea, airflow Obstruction) scores, but carries the advantage of using some modifiable variables all of which are available from existing Primary Care datasets.
Jones et al demonstrated a 'high' score (\>4) to show the greatest risk of hospital admission, increased hospital bed days and risk of respiratory failure(16). A high DOSE score has since been demonstrated to predict an increased five year respiratory mortality of 42.4% vs 11.0% in those with a low score(18) and to predict an increase in respiratory symptom burden over the subsequent two years from scoring(19). A high score has also been shown to predict an increased risk of subsequent exacerbation frequency, hypercapnoea and poorer exercise capacity (20). In these patients we aim to optimise symptoms and risk and to provide high quality palliative care, which may span many years. Despite this, as demonstrated by Sundh et al, there will still be a mortality of over 40% at five years in the high DOSE score group(18), associated with a considerable symptom and care burden.
Evidence exists to support poorer outcomes in COPD patients with certain co-morbidities, particularly those related to heart disease, osteoporosis, Body Mass Index and mental health (14). COPD prevalence is also strongly related to socioeconomic status. Other parameters such as living alone, the presence of other respiratory co-morbidities, a persistent inflammatory state reflected by raised inflammatory markers, the presence of dementia and gastro-oesophageal reflux disease all seem likely to be associated with poorer outcome and have less data or conflicting data associated with their link to COPD(20).
The original paper by Jones et al noted a mean increase in DOSE score of 0.18 DOSE points/year over nine years in the London cohort(16) but one would expect considerable variation between patients in their rate of decline. The group is currently undertaking 'CASCADE', a five year, retrospective, database search of anonymised primary care records (The Hampshire Health Record) in which, DOSE scores are recalculated over five years and an increase in score is used as a surrogate marker for health status deterioration. In those identified at baseline to have a low DOSE score (\<4), the subsequent rate of change in DOSE scores and use of hospital emergency services will be used to divide the cohort into 'high risk' and 'low risk' groups. 'CASCADE' also identifies, at baseline, the co-morbid and social demographics mentioned in the above paragraph in these patients. Using a regression analysis, we hope to identify which co-morbid and social demographics can risk-stratify the patients at baseline, ultimately creating a model, applicable in Primary Care.
In CASCADE II we will collect data regarding co-morbidities and social demographics from a real world COPD population. As one of the secondary analyses this will provide us with pilot data to further develop our risk prediction model in a real world clinical setting.
Evidence for treatments that affect disease progression rates in COPD is scarce (with the exception of smoking cessation) perhaps partly as they are often a complex patient group both clinically and socially who are highlighted late in their disease process. Within the CASCADE II study we aim to go some way to remedy this deficit by further exploring the potential relationship between interventions already recommended (e.g. formal diagnosis using spirometry, optimal inhaler prescription, participation in pulmonary rehabilitation and optimisation of comorbidities) and respiratory health status when adequate time and expertise is prospectively allocated to recognise and optimise all aspects of these patients' health.
2. Hypothesis We hypothesise that a complex intervention to optimise clinical and supportive care in early stage COPD patients will improve health outcomes and will be cost effective.
3. Study Aim Within a population of COPD patients deemed 'low risk' by DOSE score but with respiratory symptoms, this study looks to establish whether intervention in the form of a prospective, in depth, medical assessment and individualised optimisation of care will alter their respiratory health status and clinical outcomes at twelve months in comparison with standard care.
4. Study Design A cluster randomised controlled interventional study of the impact on health status of prospective assessment and individualised medical optimisation using GP practices as clusters in a cohort based analysis.
