Family Practice Advance Access originally published online on July 8, 2005
Family Practice 2005 22(5):529-531; doi:10.1093/fampra/cmi057
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Which general practices have higher list inflation? An exploratory study
a STaRNet, GKT Department of General Practice, Kings' College London, 5 Lambeth Walk, London SE11 6SP, b South East London Shared Services Partnership, Lower Marsh, London SE1 7EH, c South East London Strategic Health Authority, Lower Marsh, London SE1 7EH, d Prescribing Support Unit, Leeds LS2 7RJ and e Department of Primary Care & Social Medicine, Imperial College Faculty of Medicine, London W6 8RP, UK
Correspondence to Dr Ashworth; Email: mark.ashworth{at}gp-G85053.nhs.uk
Received 21 October 2004; Accepted 27 May 2005.
Ashworth M, Jenkins M, Burgess K, Keynes H, Wallace M, Roberts D and Majeed A. Which general practices have higher list inflation? An exploratory study. Family Practice 2005; 22: 529–531.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
Background. There is a 7% discrepancy between English population estimates based on census figures and those based on the registered lists of GPs. GP income under the 2004 new contract will be based on registered patient populations but a national ‘de-ghosting’ exercise will ensue, aiming to eliminate list inflation.
Objective. To derive an estimate of the variation in list inflation between individual general practices.
Methods. Letters were sent to all men aged 20–29 years and registered at practices based in three inner city Primary Care Trusts (n = 42 712). Non respondents received one further reminder. Cards were issued to each GP listing non responders. Patients were deducted from the GP list after six months if the GP did not verify the address.
Results. 42 712 letters were sent. 33.5% of registered patients were eventually deducted from the GP list (deduction figures only available for 20–24 year old group). Practice level deduction rates ranged from 7–76%. Practices with higher deduction rates achieved lower vaccination rates for 2 year olds (Pearson's r = 20.25; P = 0.005) and cervical smear rates (Pearson's r = 20.18; P = 0.04); they also had cheaper prescribing costs per ASTRO-PU (Pearson's r = 20.20; P = 0.03).
Conclusions. If these findings apply to the whole registered population, the national de-ghosting exercise is likely to result in large changes to the list size of some practices. Without correcting for list inflation, primary care research based on patient list size as the denominator may underestimate various measures of GP activity, particularly in deprived inner city areas. Resource allocation is also likely to be distorted by differences in list inflation.
Keywords. Census population, list inflation, registered population.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
List inflation is an inevitable consequence of a capitation based system of funding primary care. Some patients on a GP's list probably no longer exist because of death, emigration or moving house. These non-existent patients are often termed ‘ghosts’. Overall, in 2001 population estimates based on GP lists were 7.1% higher than official estimates derived from the 2001 census (Bulwinder Singh, Department of Health, personal communication). Not all of this difference is accounted for by ‘ghosts’; some is attributable to patients not completing the national census (including refugees and the homeless). National capitation based GP remuneration is reduced correspondingly so that public funding for GPs is not falsely inflated. Nevertheless, individual practices with relatively high list inflation figures will gain financially at the expense of colleagues in practices with lower list inflation.
Coupled with the new contract for GPs, the UK government is conducting a ‘de-ghosting’ exercise, aiming to eliminate list inflation.1 Levels of list inflation at Primary Care Trust (PCT) level have been estimated but values at practice level are largely unknown.2
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
The study was conducted in three inner city PCTs in 2002/3 consisting of 151 practices (24 practices were excluded because they had been the subject of a list inflation exercise in the previous year). After reaching agreement with the Local Medical Committee (LMC), all men aged 20–24 years and registered at practices within the three PCTs were sent a letter asking them to confirm that they were still registered with their local GP. One further reminder was sent to non-responders. Cards were then sent to the GPs of those patients who had not responded or whose original letter had been returned by the post office because the patient was not known at that address. If, after six months, the GP was unable to provide a current address for these patients then they were deducted from the GP's list. Six months later, the exercise was repeated for males aged 25– 29 years although patient deductions were not pursued for this second age cohort because the national list inflation exercise had been proposed by then. Patient deductions as a proportion of list size were compared with known practice performance variables.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
Letters were sent to 17 450 men aged 20–24 years and to 25 262 aged 25–29 years. Returns by the post office were 2582 (14.8%) and 5175 (20.5%), respectively. Non responses were 7232 (41.4%) and 10 362 (41.0%), respectively.
In total, 5838 (33.5%) of the non-responding patients were deducted (data only available for 20–24 year old males). The proportion of patients deducted in each practice ranged from 7.3% to 76%. The relationship between list deduction figures and various practice and performance indicators is shown in Table 1.
|
In each practice, the proportion of post office returns was closely correlated between the two age cohorts (Pearson's r = 0.74; P < 0.001) as was the proportion of non-responses (Pearson's r = 0.52; P < 0.001).
Finally, a projected figure for patient deductions in the second age cohort was derived by extrapolating the proportion for the first age cohort. Based on this projection, deductions for 25–29 year old males would have ranged from 9.8%–56.3%.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
List deductions varied tenfold between general practices. Our study was conducted in three inner city PCTs known to have 18% overall list inflation (national range: –1% to 27%) (Pers. comm. Bulwinder Singh). Although we found close correlations between our two age cohorts, it is unlikely that list inflation figures would be consistent across all age and sex cohorts, and likely that data from this cohort might overestimate total list inflation. Young men living in inner cities are probably the group least likely to respond to official letters (and practices may have added to their apparent list inflation figure by failing to respond to notification cards). Nevertheless, practices with higher list inflation figures for young males were less likely to have achieved cervical smear and vaccine targets, suggesting that list reported inflation affecting other age groups might have reduced practice performance by increasing the size of the denominator population. Poor organisation within a practice may also have contributed to both list inflation and failure to achieve targets.
Studies comparing GP performance and costs have rarely corrected for practice level list inflation. We are aware of only one such survey 3 which estimated list inflation using manual case record searches of small samples of registered patients in each of 16 participating practices. Non-attenders over a two year period were identified. List sizes were ‘corrected’ accordingly, resulting in improved levels of recorded health promotion activity in each practice.
Failure to correct for list inflation, particularly in inner city areas, is likely to result in under-estimates of several important primary care performance and quality indicators. Disease incidence and prevalence levels will also be falsely depressed in practices with high list inflation. Estimates of GP workload based on consultation rates per 1000 registered patients will tend to underestimate the true consultation frequency unless the list has been ‘de-ghosted’. Similarly, practice prescribing costs are usually interpreted according to the weighted list size; high levels of list inflation may give the false impression of economical prescribing. The current national list inflation exercise is likely to reduce one source of bias in primary care research.
Under a capitation based payment system, the more patients that are registered with the GP, the higher the earnings of that GP. This contrasts with a fee-based system in which GP income is based on specified items of clinical activity. Several European countries currently operate a capitation based payment system,5 much of Eastern Europe has adopted this system6 and New Zealand is considering shifting to a capitation system.7 Distortions of health service measures dependent on list size denominators are likely to arise in all countries where primary care funding is capitation based.
|
Declaration |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
Funding: this work was part funded through the South Thames Research Network (STaRNet), London. STaRNet London is funded by the NHS Executive. MA is one of the STaRNet lead GPs.
Ethical approval: St Thomas's Hospital LREC.
Conflicts of interest: none.
|
Acknowledgments |
|---|
We would like to thank David Lloyd (Prescribing Support Unit, Leeds) for helpful advice about interpreting national figures for list variation. The International Unit of the Royal College of General Practitioners helped obtain information on international funding systems for GPs.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
1 GP Bulletin, Issue 31. London: Department of Health; June 2004. http://www.publications.doh.gov.uk/gpbulletin/Issue31.htm#tn10
2 Majeed FA, Cook DG, Given-Wilson R, Vecchi P, Poloniecki J. Do general practitioners influence the uptake of breast cancer screening? J Med Screen 1995; 2: 119–124.[Medline]
3 Robson J, Falshaw M. Audit of preventive activities in 16 inner London practices using a validated measure of patient population, the ‘active patient denominator’. Health Eastenders Project. Br J Gen Pract 1995; 45: 463–468.[Web of Science][Medline]
4 Majeed FA, Cook DG, Anderson HR, Hilton S, Bunn S, Stones C. Using patient and general practice characteristics to explain variations in cervical smear uptake rates. Br Med J 1994; 308: 1272–1276.
5 Gervas J, Perez Fernandez M, Starfield BH. Primary care, financing and gatekeeping in western Europe. Fam Pract 1994; 11: 307–317.
6 Lember M. A policy of introducing a new contract and funding system of general practice in Estonia. Int J Health Plann Mgmt 2002; 17: 41–53.
7 Cumming J, Mays N. Shifting to capitation in primary care: what might the impact be in New Zealand? Aust Health Rev 1999; 22: 8–24.[Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  N S Crowcroft Action on immunisation: no data, no action Arch. Dis. Child., November 1, 2009; 94(11): 829 - 830. [Full Text] [PDF]
|
|
|
|
 |
|
 S. Connolly and D. O'Reilly Variation in care home admission across areas of Northern Ireland Age Ageing, July 1, 2009; 38(4): 461 - 465. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Moser, J. Patnick, and V. Beral Inequalities in reported use of breast and cervical screening in Great Britain: analysis of cross sectional survey data BMJ, June 16, 2009; 338(jun16_2): b2025 - b2025. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||