Are English treatment centres treating less complex patients?
Introduction
Many countries are introducing activity-based funding arrangements to pay for healthcare services, replacing either retrospective cost-based remuneration or prospective budgets [1]. Cost-based reimbursement can encourage cost inflation and increase activity levels above those that are clinically necessary and efficient (‘supplier-induced demand’); this risk is heightened if physicians are paid by fee-for-service. Paying providers a prospective budget can incentivise efficiency, but, unless the budget is appropriately funded and allocated, this approach may compromise the quality of care or exacerbate health inequalities.
Activity-based funding entails remunerating healthcare providers a fixed price per patient treated. In principle, this can incentivise efficient provision, inducing providers to reduce costs, increase activity and improve the quality of care [2]. In practice, activity-based funding requires that the remuneration adequately takes account of the differences in patient care requirements. Countries that have adopted activity-based funding use a categorisation system that differentiates patients according to their expected care requirements.
In the United States, where activity-based funding was first introduced in 1983 for the Medicare Program, Diagnosis-Related Groups (DRGs) are used to indicate the expected level of resources necessary to treat a patient with a given diagnosis [3]. Since then, other OECD countries have introduced similar arrangements. For example England, France, Canada and Australia have respectively developed their own Healthcare Resource Groups (HRGs), Groups Homogenes de Malades (GHMs), Casemix Groups (CMGs) and Australian National DRGs (AN-DRGs) [4], [5].
However the categorisation system used to differentiate between patients can never account for all cost variation. This is because so much of health care is individually tailored to each specific patient according to the severity of their condition, their co-morbidities and their circumstances. This would not create funding problems if differences across providers were random, where it is a matter of chance whether any particular patient is more or less expensive than the average patient in the DRG (or variant thereof) to which they are categorised. With sufficiently large volumes, these differences cancel out. Problems arise if the differences across providers are systematic, with one type of provider more likely to treat low-cost patients and another treating more high-cost patients within any given DRG.
There is a substantial literature on why within-DRG casemix might differ according to the characteristics of hospitals, such as whether they have a teaching function, are specialist or general hospitals, and according to ownership type [6], [7], [8], [9], [10]. One of the key reasons why within-DRG casemix might differ, particularly under a system of fixed price reimbursement, is that it is profitable for providers to engage in patient selection. Providers may select patients with lower expected costs (a practice known as ‘cream-skimming’ or ‘cherry picking’) and decline treatment to patients with high expected costs (‘dumping’). There is substantial evidence that hospitals in the United States engaged in patient selection in response to the introduction of the prospective payment system for Medicare patients [11] and that the problem persists, despite adjustments to the funding mechanism [12], [13], [14], [15].
In most public healthcare systems, providers are usually required to accept ‘all comers’. However this does not rule out the possibility of indirect selection, either by providers or by patients themselves [16]. Nor is the requirement a certain guarantee against direct selection, particularly as the institutional structure of many public health systems is evolving rapidly, with private providers playing an increasing role.
In this paper, we consider two inter-related institutional changes to the English public healthcare system that might encourage selection. The first is the development of ‘treatment centres’ which are stand-alone facilities dedicated to the provision of a limited range of high-volume procedures. The second is the increasing reliance on the private sector to deliver care to public patients in these settings. Pricing policy needs to take this evolution into account to ensure an equitable system of reimbursement.
A classic view of the hospital is one in which departments compete for shared resources such as diagnostic facilities, operating theatres, wards and staff [17]. Recognising that some departments might lose out, recent attempts have been made to ring-fence resources for elective surgery to protect them from competition with others departments [18]. In such a vein, the Royal College of Surgeons of England has long advocated a separation of elective and emergency work to avoid staff having to be on call for emergencies when treating elective patients, which often results in cancellations of planned operations [19]. While there are a number of ways to separate elective and emergency care [20], the treatment centre model offers a physical separation of beds, theatres, diagnostic facilities and staff to achieve a more predictable workflow. Treatment centres are dedicated to elective care, with no requirement to provide emergency treatment, and are designed to specialise in one or two high-volume procedures, such as hip and knee replacements or cataract removals, and to avoid taking on complex operations.
In 1999 a treatment centre (TC) was opened at Central Middlesex Hospital, supposedly the first such facility in Europe [21]. The English Department of Health decided to extend the TC model [22], implementing a programme in April 2003 to establish treatment centres throughout England, prioritising “areas that have traditionally had the longest waiting times, such as ophthalmology and orthopaedics” [23]. Later that same year it was decided to allow private – or ‘Independent Sector’ – providers to establish treatment centres catering for NHS patients [24]. By late summer of 2007, almost 100 treatment centres were open, approximately half of which were operated by the private sector [25].
Under activity-based funding in England, providers are paid the same fixed price per patient categorised to the same HRG (the English version of DRGs). This arrangement is predicated on the assumption that the same types of patient are being seen in treatment centres as in hospitals. However, there are several reasons why casemix might differ between public providers (NHS trusts) and treatment centres (private or public). First, if the treatment centres are ‘private’, they may wish to treat a less severe casemix to boost profits. In contrast, even if public providers are able to retain surpluses, these must be re-invested in the hospital. Public providers, therefore, are under less pressure to engage in selection of low-severity patients. Second, treatment centres tend to be less well equipped than NHS hospitals, making them less suited to provide complex care. Treatment centres might try to avoid high-severity patients to reduce the risk of emergency transfers to an NHS trust. Third, NHS trusts tend to be more ‘prestigious’ and attract highly specialised doctors with the skills to treat more complex cases. In contrast, treatment centres focus on more routine work which requires a lower degree of specialisation and can be dealt by less experienced staff. In this paper we analyse patient-level data to test whether the complexity of patients coded to the same HRG differs by provider type: public hospitals or treatment centres.
Section snippets
Data and methods
To assess the characteristics of patients treated in treatment centres and hospitals we analysed the hospital episode statistics (HES) for 2006/7. HES comprise individual patient records about every NHS patient treated as a day case or inpatient in England. Each patient record includes a number of data ‘fields’, containing demographic (e.g. age, gender) and clinical information (e.g. diagnosis, procedures performed). Information is also available about the socio-economic characteristics of
Results
A total of 3,334,535 patients are included in the analysis of the twenty-nine highest volume HRGs, of which 77,358 (2.3%) were treated in treatment centres. Table 2 details the numbers of patients according to the HRG to which they are allocated and Table 3 provides descriptive statistics about the casemix indicators for these HRGs. Table 2 confirms that treatment centres tend to concentrate on less complex procedures. This is evidenced, for instance, by the high proportion (27%) of low
Discussion
Although activity-based funding arrangements in England are predicated on the assumption that within-HRG casemix is the same across treatment settings, our analysis finds evidence that hospitals are treating patients of greater complexity than treatment centres. Specifically, patients treated in hospitals are more likely to come from more deprived areas; to have more diagnoses; and to undergo significantly more procedures than patients seen in treatment centres. That said, patients seen in
Acknowledgements
The project was funded by the Department of Health in England as part of a programme of policy research. The views expressed are those of the authors and may not reflect those of the funder.
References (42)
- et al.
What do not-for-profit hospitals maximize?
International Journal of Industrial Organization
(2002) - et al.
Medicare upcoding and hospital ownership
Journal of Health Economics
(2004) Creaming, skimping and dumping: provider competition on the intensive and extensive margins
Journal of Health Economics
(1998)- et al.
Meta-analysis in clinical trials
Controlled Clinical Trials
(1986) - et al.
Activity-based financing of health care—experiences from Sweden
International Journal of Health Planning & Management
(2007) - et al.
Contracting for health services with unmonitored quality
Economic Journal
(1998) Hospital reimbursement incentives: an empirical analysis
Journal of Economics and Management Strategy
(1997)- et al.
Hospital case payment systems in Europe
Health Care Management Science
(2006) - et al.
Behavioral differences between public and private not-for-profit hospitals in the Italian National Health Service
Health Economics
(2007)