Abstract
COPD is a major cause of emergency admissions worldwide. In many countries the majority of COPD care is provided by primary care teams. This study examined variation between primary care teams in rates of COPD admission and assessed the role of prevalence, deprivation, practice performance, and general practitioner (GP) and nurse supply on risk of COPD admission. Retrospective observational study of National Health Service admissions for COPD of patients registered with London general practices over four years (2006–2009). We sought associations using negative binomial regression between COPD admissions and population factors, practice organization and practice performance. Trends in rates of COPD admissions across London were stable between 2006 and 2009. COPD admission rates varied substantially between practices (2006/7: median 13.68/10,000 population (IQR 7.83–22.70)), with almost a 3-fold difference across the interquartile range each year. Practice characteristics varied to a similar extent. Variation in practice COPD admission rates was associated with diagnosed prevalence of COPD (Rate Ratio 2.06, 95% CI 1.84–2.3) and increasing levels of deprivation (RR 1.01, 1.006–1.01). Other practice characteristics, including GP and nurse supply, and practice performance scores were not predictive of practice level COPD admission rates, when controlling for COPD prevalence and socio-economic status. Main predictors of variation in rates of COPD admissions were prevalence of diagnosed COPD and socioeconomic status. The absence of evidence that variation in primary care services for COPD was associated with rates of COPD admission emphasizes the importance of primary prevention of COPD if COPD admission rates are to fall.
Introduction
Chronic obstructive pulmonary disease (COPD) is one of the largest causes of emergency admissions worldwide and the second largest cause in the United Kingdom (Citation1). Emergency admissions for COPD account for more than 60% of the management costs of the disease, and their prevention has been a high priority in health service strategic planning (Citation2, Citation3). Frequent exacerbation-related hospital admissions for COPD are associated with worsening lung function, quality of life, and mortality (Citation4, Citation5). Various strategies at regional, hospital, general practice and individual patient level have been promoted to reduce the risk of admission including weather forecasting, risk stratification, pulmonary rehabilitation, early intervention, and self-management (Citation6–Citation10). In the United Kingdom rates of emergency admission to hospital for COPD have been notably stable between 2002 and 2010 at a time of intensive efforts to improve the quality of its management including large increases in drug spending on the disease, and widespread take-up of pulmonary rehabilitation (Citation11, Citation12).
COPD is most prevalent in areas of socio-economic deprivation (Citation13). The greatest concentration of people with COPD in the UK is to be found in cities. In London, where COPD is the largest single cause of emergency admissions, there is evidence of considerable over-treatment of COPD, driven perhaps by recognition of the continuing heavy burden on patients and health services (Citation14). There is also considerable variation in London in the quality of COPD management in lung-function testing, diagnosis, prescribing, and influenza immunization (Citation14–Citation16). Two recent publications have provided opposing estimates of the impact of general practice level factors on risk of admission for COPD (Citation17, Citation18). In this study of admissions in the largest conurbation in the UK with a population of 7.5 million, served by 1600 general practices, we have examined the variation in primary care in rates of admission for COPD. We have assessed the role of prevalence, deprivation, general practice characteristics and general practice performance on the risk of admission for COPD.
Method
This study was a retrospective observational study of National Health Service (NHS) admissions for COPD of patients registered with London general practices (practices) during the four years between January 2006 and December 2009. All patient data were anonymised at the point of collection. Ethical approval was not required.
Hospital admissions data
All NHS hospital admissions must be notified to the NHS Information Centre Hospital Episode Statistics (HES) database (Citation19). Admissions are recorded as one or more finished consultant episodes (FCE) which together form a hospital spell (Citation20). We sought data from HES on all COPD hospital admissions during the period 01/01/2006 to 31/12/2009, of adult patients (age on admission ≥ 18 years) in which the principal diagnosis for the first FCE (the reason for admission) was COPD (International Classification of Diseases, 9th and 10th revision codes J40X–J44.9)(21). Admissions could have been to hospital anywhere in England. Patients had to be registered with practices in London Primary Care Trusts. Primary care trusts (PCTs) were freestanding NHS bodies responsible for delivering health care and health improvements in their locality. Data were provided by HES with unique identifiers for each patient so that analysis could be conducted using the number of patients admitted per practice in a given period in addition to the total number of admissions per practice.
Practice characteristics
Data on practice characteristics were obtained from the NHS Information Centre (NHSIC) and included list size, the prevalence of COPD, the number of full-time equivalent GPs (excluding GPs in training), single-handed (sole practitioner) status, and training practice status (Citation15). Annual data on full-time equivalent nurses per 1000 population per PCT were applied to each practice, as practice-level data were not available. Practice socio-economic status was based on the 2007 Index of Multiple Deprivation (IMD) score derived from each practice's postcode (Citation22). IMD is based on national census data and local authority data, and reflects deprivation specific to a geographical area. It combines a range of social, economic and housing issues into a single deprivation score for each small area in England. IMD scores in 2007 ranged nationally from 0 (least deprived) to 86 (most deprived)(22). Proportions of patients from ethnic minority groups were recorded because of the relationship previously demonstrated between frequency of antibiotic use and the proportion of patients of white ethnicity (Citation23).
General Practice performance data
Practice performance data were obtained from the NHS Information Centre Quality and Outcomes Framework (QOF) database for each London practice for the years 2006 to 2010 (Citation15, Citation24). COPD management performance characteristics of interest were: proportion of COPD patients who had received the influenza vaccination in the preceding 12 months; confirmation of diagnosis using spirometry; recording of FEV1 in the previous 15 months; recording of good inhaler technique, and patients who had had a review by a health professional in the preceding 15 months.
Patient characteristics for each individual admission were not available in the admissions data, but assessing whether smoking might be a predictor variable for the rate of admissions was a crucial component. QOF smoking indicators were used as a broad view of a possible impact on admissions, as they detail the percentage of patients with any chronic disease who smoke and whose notes contain a record that smoking cessation advice or referral to a specialist service has been offered within the previous 15 months.
Two QOF performance indicator variables on ability to book an appointment with a GP within 2 working days (PE7) and the ability to book an appointment with a GP more than 2 days ahead (PE8) were previously reported to be influential on admissions (Citation17). They were introduced into the QOF dataset in 2008–09 and 2009–10 and were included in our analysis as a broad measure of access to GP care from a patient's perspective.
Exclusion criteria
Practices with a total list size of fewer than 1000 patients and practices serving mainly students and university staff where the percentage of patients aged over 45 years was < 10% were excluded. In these practices the low numbers of admissions would lead to their practice characteristics being over-represented and carrying disproportionate weight in the analysis. Practices whose characteristics and performance data (QOF) were not available were also excluded
Statistical analysis
Annual hospital admissions for COPD of patients in each practice was the primary outcome variable. A dataset was constructed containing the admissions data and population characteristics and performance data for all eligible practices in London. Negative binomial regression models were used to explore univariable associations between COPD admissions and practice characteristics. Variables whose association with COPD admissions reached the 10% significance level (p < 0.1) in univariable analysis were entered into a multiple regression analysis to ensure that variables that might contribute to the multivariable model even though not significant in univariable analysis would be included. Negative binomial regression was performed due to the wide dispersion of the data. Adjustment was made for practice list size including the log (list) size as an offset term in the model. Statistical analyses were performed using IBM SPSS statistics 20 (IBM, New York, USA).
Results
Admissions
A total of 1566 eligible London practices, serving a population of 8,456,186, were included in the analysis. 35 practices (2.1%) with lists smaller than 1000 and 5 practices (0.3%) with less than 10% of their patients over 45 years were excluded. Practice characteristics and performance data were not available for 90 practices (5.6%). 39,087 patients from 1566 (92%) eligible London practices, serving a population of 8,456,186, were admitted to hospital with COPD between January 2006 and December 2009. 52604 admissions were analysed. Annual practice admission rates for COPD together with practice characteristics and performance data are shown in Table . In each year studied, the rate of COPD admissions per 10,000 patients on the practice list varied almost threefold between practices across the interquartile range, e.g., 2006/7 median 13.7 (IQR 7.8-22.7), although there was little change over that period in the median rate of admissions (2006–13.7; IQR 7.8-22.7; 2009–13.4; IQR 7.8-22.0). Annual and monthly mean rates of admission over the four years of the study are shown in Figures and .
Table 1. Practice admission rates, practice characteristics, and practice performance data for 1566 (92%) eligible London practices, population 8,456,186. 2006–2009
Figure 1. COPD admissions per 10,000 patients on the list (mean and 95% confidence intervals) in London, UK from 2006 to 2009.
Figure 2. COPD admissions per 10,000 patients on the list (mean and 95% confidence intervals) for each calendar month in the four years from 2006 to 2009.
Practice characteristics and performance
There was variation in practice characteristics such as the median practice list size (4635.5; IQR 3003.0–7173.5), the diagnosed COPD prevalence (0.81%; IQR 0.5–1.1), and the deprivation score (24.97; IQR 14.16–36.35). There was relative uniformity in the performance quality of the practices. The QOF total points achieved/total points available (0.9699; IQR 0.9277–0.9897), median QOF points for smoking cessation offered to patients with chronic diseases (0.9362; IQR 0.9095–0.9649), and median QOF score for COPD diagnosis confirmed by spirometry (0.8858; IQR 0.8014–0.9412) had a narrow spread. Practice characteristics were relatively stable between 2006 and 2009 although the prevalence of diagnosed COPD rose from 0.81% (IQR 0.5385–1.1408) in 2006 to 0.92% (0.6321–1.2641) in 2009, and median list size rose from 4635.5 (IQR 3003.0–7173.5) in 2006 to 4928.0 (IQR 3194.0–7620.0) in 2009. The practice performance data also remained stable over the 4 year period studied, apart from the prevalence of COPD diagnosed by spirometry which rose briefly in 2008/9.
Associations (univariable and multivariable negative binomial regression) between practice characteristics and performance indicators and practice admission rates are shown in Table . Deprivation, diagnosed COPD prevalence, percentage of White patients, percentage of patients aged over 45 years, percentage aged over 85 years, and COPD patients who received influenza immunisation in the- preceding 12 months, were each predictors of admission rates for COPD in the univariable analysis. In multivariable analysis IMD score and COPD prevalence were independent predictors of practice level admission rates when adjusting for ethnicity, practice list age structure, number of full-time equivalent GPs, practice rates of influenza immunisations per patient with COPD, and number of practice nurses per 1000 patients at PCT level. Other practice characteristics and practice performance indicators had no influence on these associations.
Table 2. Association of practice performance and population characteristics with COPD admission rates: univariable and multivariable analysis using negative binomial regression
Discussion
Trends in practice rates of admission for COPD were stable for the four years 2006–2009 across the metropolitan area of London. Rates varied substantially between practices with almost a threefold difference across the interquartile range in all four years. Variation in practice admission rates for COPD was associated with diagnosed prevalence of COPD and geographically-based socio-economic status (IMD score). Other practice characteristics, including GP and nurse supply, and practice performance scores (QOF scores) were not predictive of practice level admission rates for COPD when controlling for COPD prevalence and socio-economic status.
These findings suggest that population factors are the key predictors of hospital admission for COPD. The variation in COPD admission rates between primary care teams is not explained by the considerable variation in the characteristics and healthcare performance of those teams. Changes in admission rates for COPD will depend on improvement in the primary and secondary prevention of the disease. It is unlikely that changes at primary care level in the organisation and delivery of services for COPD will result in a reduction in rates of admission for COPD. That evidence was strengthened by the recent observation that high rates of practice prescribing of combined inhaled long-acting beta-agonists and corticosteroids and of long-acting muscarinic antagonists for COPD were not associated with reduced practice rates of admission for COPD when controlling for practice list size, COPD prevalence, and geographically based socio-economic status (IMD score)(11).
The results are consistent with the findings of Purdy et al in 2011 from a study of UK national data over one year from 2005–06 (Citation18). They are at odds with the findings of Calderón-Larrańaga in 2010 examining similar UK national data over 3 years between 2006 and 2009, and which have been disputed (Citation17, Citation25). The findings will be of interest to health care commissioners and managers concerned with the unchanging rates and escalating costs of COPD admissions.
In a separate study in 44 practices in London we have shown high levels of overtreatment of COPD with high dose inhaled corticosteroids (Citation13). High rates of practice prescribing of these drugs for COPD have been motivated by the hope of avoiding exacerbations and in turn preventing admissions. The lack of relationship between high practice COPD prescribing and reduced COPD hospital admission rates for COPD is in keeping with our findings here.
Strengths and limitations
The strength of this study lies in the large number of general practices whose data has contributed to the analysis. Primary care is provided by general practices in many countries. In the UK more than 97% of the population is registered with general practices and this enables a reliable estimate of the variation in population prevalence of diagnosed COPD to be obtained. Through this sampling frame we have been able to show that the wide range of practice rates of COPD admission was not explained by practice characteristics and performance scores which were also distributed widely. The findings mirror our recent national study of primary care COPD prescribing and primary care admission rates, and extend the examination of the relationship between primary care factors and risk of COPD admission (Citation10).
The main weakness of this study lies in the absence of patient level data, both in terms of disease severity and in treatment received. Smoking prevalence data for COPD patients were also not available. We have attempted to account for the lack of patient specific data by including summary data on smoking prevalence. It is possible that individual interventions may have reduced the risk of admission in individual patients, but admission rates over the four years of the study at a time of rising prescribing and rising uptake in pulmonary rehabilitation have been notably stable (Citation11,Citation13). This suggests that any individual experience of reduced risk of COPD admission was at least matched by an absence of change in the overall population.
Routinely collected and coded data, such as the admissions data from HES, can raise concerns about accuracy and reliability. However, there has been evidence of improving quality and consistency in HES data prompted by the use of these data in determining hospital admission income (Citation20). QOF data may be similarly criticised but they remain the best available data source in the UK and have also been used universally in the UK in the determining of GPs’ practice income (Citation26).
Conclusions
Variation in rates of admission for COPD between general practices was most strongly associated with population factors including the prevalence of COPD and socio-economic deprivation. General practice characteristics including numbers of clinical staff, organisation of care, and performance scores were not associated with COPD admissions rates. This suggests that there is little evidence to encourage healthcare commissioners or primary care teams in urban areas to seek treatment interventions at primary care level that might decrease hospital admission rates of COPD. It is likely that COPD admission rates will only fall with improvements in primary and secondary prevention.
Funding
Funding: Guy's and St Thomas’ Charity, Reference No. G060703. Dr Harries was supported by an NIHR In-practice fellowship.
This research was supported by the National Institute for Health Research (NIHR) Biomedical Research Centre at Guy's and St Thomas’ NHS Foundation Trust and King's College London. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health.
Declaration of Interest Statement
The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the article.
References
- Damiani M, Dixon J. COPD Medical Admissions in the UK. Newmarket, UK: Hayward Medical Publications, 2004.
- Britton M. The burden of COPD in the U.K.: Results from the Confronting COPD survey. RespirMed. 2003; 97 Suppl C:S71–S9.
- Department of Health. An Outcomes Strategy for Chronic Obstructive Pulmonary Disease (COPD) and Asthma. London: 2011 7/1/2011.
- Suissa S, Dell'Aniello S, Ernst P. Long-term natural history of chronic obstructive pulmonary disease: severe exacerbations and mortality. Thorax 2012; 67(11):957–963.
- Seemungal TA, Hurst JR, Wedzicha JA. Exacerbation rate, health status and mortality in COPD–a review of potential interventions. Int J Chron Obstruct Pulmon Dis 2009; 4:203–223.
- Roberts JA, Maslin TK, Bakerly ND. Development of an integrated chronic obstructive pulmonary disease service model in an inner-city region in the UK: initial findings and 12-month results. Prim Care Respir J 2010; 19(4):390–397.
- Bryden C, Bird W, Titley HA et al. Stratification of COPD patients by previous admission for targeting of preventative care. Respir Med 2009; 103(4):558–565.
- Kruis AL, Smidt N, Assendelft WJ et al. Integrated disease management interventions for patients with chronic obstructive pulmonary disease. Cochrane Database Syst Rev 2013; 10:Cd009437.
- Seymour JM, Moore L, Jolley CJ et al. Outpatient pulmonary rehabilitation following acute exacerbations of COPD. Thorax 2010; 65(5):423428.
- Effing T, Monninkhof EM, van der Valk PD et al. Self-management education for patients with chronic obstructive pulmonary disease. Cochrane Database Syst Rev 2007(4):CD002990.
- Harries TH, Seed PT, Jones S et al. Chronic obstructive pulmonary disease hospital admissions and drugs–unexpected positive associations: a retrospective general practice cohort study. NPJ Prim Care Respir Med 2014; 24:14006.
- Hogg L, Garrod R, Thornton H et al. Effectiveness, attendance, and completion of an integrated, system-wide pulmonary rehabilitation service for COPD: prospective observational study. COPD 2012; 9(5):546–554.
- Eisner MD, Blanc PD, Omachi TA et al. Socioeconomic status, race and COPD health outcomes. J Epidemiol Commun Health 2011; 65(1):26–34.
- White P, Thornton H, Pinnock H et al. Overtreatment of COPD with inhaled corticosteroids–implications for safety and costs: cross-sectional observational study. PloS One 2013; 8(10):e75221.
- NHS Information Centre. National Quality and Outcomes Framework Statistics for England. Leeds: The NHS Information Centre; 2014 [cited 2014 30 07 2014]. Available from: http://www.hscic.gov.uk/qof.
- Martin A, Badrick E, Mathur R, Hull S. Effect of ethnicity on the prevalence, severity, and management of COPD in general practice. Br J Gen Pract 2012; 62(595):e76–e81.
- Calderon-Larranaga A, Carney L, Soljak M et al. Association of population and primary healthcare factors with hospital admission rates for chronic obstructive pulmonary disease in England: national cross-sectional study. Thorax 2011; 66(3):191–196.
- Purdy S, Griffin T, Salisbury C, Sharp D. Emergency respiratory admissions: influence of practice, population and hospital factors. J Health Serv Res Pol 2011; 16(3):133–140.
- NHS Information Centre. Hospital Episode Statistics Leeds: NHS Information Centre; 2014 [cited 2014 30 07 2014]. Available from: http://www.hscic.gov.uk/hes.
- Holt PJ, Sinha S, Ozdemir BA et al. Variations and inter-relationship in outcome from emergency admissions in England: a retrospective analysis of Hospital Episode Statistics from 2005-2010. BMC Health Serv Res 2014; 14:270.
- World Health Organisation. ICD-10 Version:2010 Geneva: World Health Organisation; 2010 [cited 2014 05 08 2014]. Available from: http://apps.who.int/classifications/icd10/browse/2010/en.
- Noble M, McLennon D, Wilkinson K et al. The English Indices of Deprivation 2007. London, UK: Department for Communities and Local Government; 2008.
- Wang KY, Seed P, Schofield P et al. Which practices are high antibiotic prescribers? A cross-sectional analysis. Br J Gen Pract 2009; 59(567):e315–320.
- Ashworth M, Kordowicz M. Quality and outcomes framework: Time to take stock. Br J Gen Pract 2010; 60(578):637–638.
- White P, Jamieson AL. Primary healthcare factors and hospital admission rates for COPD: no association. Thorax. 2013.
- Ashworth M. The quality and outcomes framework in the United kingdom: indicators in transition. J Ambul Care Mgmt 2012; 35(3):200–205.