11.1.3.4. Technical efficiency

 

There is considerable controversy surrounding the development of an appropriate measure of efficiency in a complex human service sector such as health. Economic theory presents one of the main measures of efficiency as technical efficiency, concerned with outputs and not with the distribution of these outputs. It can be argued that efficiency can be broadly ascertained through an examination of the resources committed to the health system e.g. costs and utilization rates (Figueras et al, 2004). Others have attempted to generate a single measure of efficiency, or productivity of the health system. Macro level studies of efficiency have been conducted with OECD data, however, relying on aggregate measures of expenditure and health status; thus they should be interpreted with caution (Retzlaff-Roberts et al, 2004; Starfield et al, 2005). Moreover, these studies do not indicate causation, but rather identify associations between health system features or expenditure and outcomes.

 

Health system productivity can be defined as the level of output generated by a given set of inputs. The UK has been innovative in developing a measure of health system productivity attempting to incorporate elements of quality of care into the measurement of the output of the system (Department of Health, 2005). The quality indicators that have been used include survival rates, waiting times, patient experiences and longer-term survival rates for myocardial infarction. These quality indicators are combined with more objective indicators of output such as activity levels at various levels of the system. This is still a premature attempt at analyzing performance, but further work should provide some insight into the costs and benefits of the NHS.

 

Efficiency and provider payment methods

 

The methods used to pay healthcare providers create powerful incentives that affect provider behaviour and the efficiency, equity and quality outcomes of health system financing. Specifically, these payment methods can be used to influence the price and quantity of healthcare. In the health service, there are three basic methods of physician payment: fee-for-service, salary and capitation. There are many variations of these payment systems, but the basic principles remain always the same. Payment can either be prospective (includes salary, capitation and line-item or global budgets) or retrospective (usually in the form of fee-for-service and case-based payment for hospitals).

 

Fee-for-service is an agreed upon value for a specific service which is to be provided. The incentive is to provide the best service at a reasonable cost to maintain the confidence of patients. Fee-for-service payment works well when there is an adequate supply of providers, minimal interference in the negotiation of prices and freedom to choose the provider. It fails when providers attempt to control supply and demand (cartels) or when insurers set prices at a level too low for providers to survive (rent controls). Regulations can be implemented to prevent these examples of potential failures.

 

Salary is the payment of a negotiated amount of money for a fixed period of time, within which providers are committed to providing services. The number of patients seen, services provided, and the cost of services do not affect the payment. Legislation can cover overtime pay and holiday pay based on the amount of time worked. However, salaries remain linked to services as they must come from the payment for the service to the third party or through taxation in a public system.

 

Capitation is the payment of a set amount of money to the provider to insure that services are provided to the user in a given time period. The provider agrees to provide all agreed services and bares the risk that the negotiated amount will cover costs and leave a profit. The user agrees to obtain the agreed services only from the designated provider unless additional money is paid out.

 

Each of these have different inherent incentives. Fee-for-service systems have an incentive to increase activity, and could have an incentive to target the poor depending on the structure of the fee schedule, while salary and capitation methods control costs but provide an incentive to decrease activity, shift patients’ costs onto others and do not encourage the targeting of the poor. These theoretical incentives of different remuneration strategies have been supported by empirical studies. For example a study in Copenhagen of a selection of GPs changing from a full capitation model to a partial fee-for-service/partial capitation method found that: a) activity increased by adding services to existing patients, not by adding new patients (the latter being less at the doctor’s discretion); and b) referrals to specialists decreased as GPs were taking on more services previously provided by specialists (Krasnik et al. 1990).

 

Across Europe, these are the main approaches for paying providers. Primary care providers are most commonly paid through a combination of capitation and fee-for-service. Specialists are usually paid by salary in tax-funded systems, while in social insurance systems fee-for-service is the most common payment method. Fee-for-service is also the norm for privately delivered primary and outpatient care. Several studies have found supportive evidence for the actual effects of the payment method on the physician’s behaviour (Chaix-Couturier et al, 2000; Gosden et al, 2006). Therefore, more countries are experimenting with blended or mixed payment schemes, which include elements of two or more methods to moderate these negative incentives. Table 11.3 shows the physician payment methods for physicians in Europe.

 

Table 11.3. Physician payment methods in Europe

 

Countries	Primary care physicians	Ambulatory care specialists	Physicians in public hospital	Physicians in private hospital
Austria	60% by fee-for-service and 40% by fee-for-service and capitation.	90% by fee-for-service, 10% by capitation and fee-for-service.	90% by salary and 10% by fee-for-service.	90% by fee-for-service and 10% by salary.
Belgium	Fee-for-service.	Fee-for-service.	Fee-for-service.	Fee-for-service.
Bulgaria	Capitation (with some fee-for-service for preventive services).	Fee-for-service.	Salary, with bonus payments for performance.	Negotiable between employer and personnel groups.
Croatia	Capitation (with fee-for-service for preventive services, not exceeding 7% of annual capitation).	Fee-for-service	Fee-for-service	Fee-for-service
Cyprus	Salary (public sector); fee-for-service (private sector)	Salary (public sector); fee-for-service (private sector)	Salary.	Fee-for-service.
Czech Republic	Capitation and fee-for-service	Fee-for-service (with limit)	Salary.	Salary*.
Denmark	Blended payment (63% of income from fee-for-service, 28% from capitation).	Not relevant.	Salary.
England	86% by blended payment (capitation, practice allowance, fee-for-service for selected services, target payments for immunization), 14% by fee-for-service for private work.	100% by salary for public patients; fee-for-service for private patients.	100% by salary for public patients, fee-for-service for private patients.	100% by fee-for-service.
Estonia	Capitation, fee-for-service (to a maximum of 18% of capitation payment), monthly allowance, and additional payments for training and distance from hospital.	Fee-for-service with maximum (depending on contract).	Salary.	Salary.
France	Fee-for-service.	Fee-for-service.	Salary.	Fee-for-service.
Germany	100% by fee-for-service.	100% by fee-for-service.	Salary. Fee-for-service for private patients.	100% by salary.
Greece	Salary in public sector, fee-for-service in private sector.	Salary in public sector, fee for service in private sector.	Mainly by salary.	Blended payment (fee-for-service and salary).
Hungary	Capitation plus fixed amount based on practice size and location. Fee-for-service in private sector.	Salary.	Salary.	Fee-for-service.
Iceland	Salary (plus fee-for-service for after hours up to 10% of the total)	Fee-for-service	Salary
Italy	Capitation (ranging according to years of experience) plus fees for specific services (surgery and prevention) (fee-for-service for private work)	Pediatricians – same as GPs	Salary	Fee-for-service.
Ireland	Fee-for-service if higher income, patient capitation if lower patient income.		Salary. Fee-for-service for treating privately insured patients in public hospital.
Latvia	“Mixed capitation”	Fee-for-service	Salary	Fee-for-service
Lithuania	Capitation and fee-for-service.	Salary and fee-for-service	Salary and fee-for-service
Luxembourg	Fee-for-service	Fee-for-service (except for few neuron-psychiatric doctors paid salary).	Fee-for-service.	Fee-for-service.
Malta	Salary.	Salary (public); fee-for-service (private).	Salary.	Fee-for-service.
Netherlands	Fee-for-service if higher patient income, capitation if lower patient income.			Blended payment (salary and fee-for-service).
Norway	Blended payment (70% of income from fee-for-service and 30% from capitation).	Salary and fee-for-service in public sector, fee-for-service in private sector.	Salary.
Poland	Capitation.			Fee-for-service.
Portugal	Salary in public sector, fee-for-service in private sector.		Salary.	Fee-for-service.
Romania	Blended payment: capitation (85%) and fee-for-service (15%).	Fee-for-service.	Salary	N/A
Slovak Republic	Blended payment (capitation and target payments for preventive care).	100% by fee-for-service.	100% by salary.	Fee-for-service.
Slovenia	Salary and bonus payments.	Salary (with public contract); fee-for-service (without).	Salary.	Salary (with public contract); fee-for-service (without)
Spain	Blended payment (85% of income from salary and 15% from capitation).	100% by salary.	100% by salary.	Mainly by fee-for-service.
Sweden	Salary	Salary	100% by salary	100% by salary
Switzerland	96% by fee-for-service and 4% by salary.	90% by fee-for-service, 10% by salary.	Fee-for-service, salary and blended payment (fee-for-service and salary).	Fee-for-service, salary and blended payment (fee-for-service and salary).
Turkey	Salary	Private specialists paid fee-for-service.	Salary plus bonuses	Salary plus bonuses (fee-for-service for additional private work)

*Information as communicated by the Czech partner.

Sources: Siciliani and Hurst – OECD 2006; Bulgaria HiT 2007; Croatia HiT 2007; Cyprus HiT 2004; Czech HiT 2004; Estonia HiT 2004; Hungary HiT 2004; Iceland HiT 2003; Italy HiT 2001; Latvia HiT 2001; Luxembourg HiT 1999; Malta Hit 1999; Romania HiT forthcoming; Turkey HiT 2002; Slovenia HiT 2002.

 

 

 

In consideration of the incentives for each method, some alternative models have been devised. These include performance-based reimbursement in Sweden and the GP contract in the UK. In the US, experimentation with physician payment methods that stimulate physicians to expand the provision of preventive services, improve clinical outcomes and enhance patient safety and satisfaction are collectively called “pay-for-performance” programmes. These are based on the assumption that the structure of payment methods may not facilitate (or even prevent) the actions needed to systematically improve the quality of care (Institute of Medicine, 2001).

 

The analysis of payment mechanisms generally focus on the balancing of risk aversion with moral hazard (Eisenhardt, 1989; Sappington, 1991). The combination of retrospective (i.e. fee-for-service) and prospective (i.e. capitation, salaries) payment methods under the need to balance conflicting incentives falls within the scope of the economic literature on principal-agent relationships as well as the literature on optimal contracts in the context of multi-task agency relationships (Robinson et al, 2004).

 

Most European health systems have in place a hospital payment system based on global budgets, though increasingly case-based payments (often referred to as diagnosis related groups (DRGs)) – a fixed fee for service that is risk adjusted by case mix complexity – are being introduced to define the budgets or as a form of payment (see Table 11.3). While DRGs are a retrospective payment, budgets are prospective and may be ‘hard’ (i.e. penalties are incurred for overspending) or ‘soft’ (i.e. overspending is not penalized). The DRG payment system was first introduced in US Medicare in the 1980s, leading to initial cost savings and efficiency improvements. Currently, most European countries have moved from a full retrospective hospital payment system towards one linked to activity or within a fixed budget constraint aimed at encouraging efficiency in production.

 

Case-based payment are designed so that providers are paid an inclusive flat sum for a patient’s treatment according to their specific diagnosis or service need. They serve both to measure the costs of treating a given patient and to reimburse providers fairly and with the incentive to improve efficiency. Both purposes require an unbiased and accurate risk adjustment process, and a methodologically sound system (Busse et al, 2006). Technical challenges arise in designing a DRG system to ensure patients within one diagnostic group have equivalent costs and equivalent clinical diagnoses, severity and treatments. There is concern over how broadly or tightly DRGs should be defined: broad diagnosis groups may increase efficiency and reduce data manipulation but may increase quality skimping, whereas tightly defined groups may lead to over-treatment in order to reclassify patients into a higher paying DRG (Busse et al, 2006).

 

DRGs are used to set part or the whole of hospital budgets in Austria, Belgium, Denmark, England, France, Finland, Germany, Hungary, Ireland, Italy, the Netherlands, Poland, Portugal, Spain, Slovenia, Sweden, Switzerland (Carrin and Hanvoravongchai, 2003; HOPE - European Hospital and Healthcare Federation, 2006; Schreyogg et al, 2006). Although the principles of the initial US Medicare DRG system have been translated into the European context, each country has adapted the design to the suit their own system. Moreover, while significant variations are seen in the role DRGs play in the health system, and in some cases in their different roles within countries, what is common across European countries is that even when implemented for financing purposes, their objectives reach beyond financing to other aims such as increased transparency (HOPE - European Hospital and Healthcare Federation 2006). With many countries having introduced DRG systems, and with different systems of patient classification, there is increasing concern for maintaining and improving quality. Thus additional quality incentives have been developed alongside the DRG system (HOPE - European Hospital and Healthcare Federation, 2006).

 

The success of a DRG system depends on the incentives that are created, and whether these are in line with policy objectives. A DRG system is associated to several possible incentives, many of which have been identified in practice. For example, it may encourage providers to treat patients with lower expected costs than the reimbursement, which may or may not be beneficial. Also, on the contrary, it may lead providers not to treat patients with higher expected costs than the reimbursement, i.e. it may encourage the ‘dumping’ of complex, high cost patients (Busse et al, 2006). A common incentive created by DRGs is ‘up-coding’, which refers to the incentive to upgrade the severity of the DRG to increase reimbursement. Finally, cost shifting and quality skimping may arise because of the incentive to minimize costs within a treatment group (Busse et al, 2006).

 

In the light of the numerous perverse incentives, some countries make use of a blended payment system. Countries in Western Europe provide an interesting example of convergence towards a mixed payment system for funding hospital services. Many countries currently use some form of case-mix system, based on an adaptation of the DRG system combined with a global budget cap. Each system has adapted the specifics of the case-mix measure and/or the application to fit within the local funding framework, or to address the objectives prioritized within the local hospital environment. For example, in Norway, there is a combination of payment by actual and expected activity: DRG payments are combined with risk-adjusted capitation (Smith, 2004).

 

The development, implementation and modification of the DRG systems vary across Europe. A review of the methodology used to develop DRG systems in 9 European countries (Denmark, England, France, Germany, Hungary, Italy, the Netherlands, Poland and Spain) reveals two general strategies: adopting or modifying an existing system (with different modifications across countries), or developing a new system (e.g. in the Netherlands and Austria) (Schreyogg et al, 2006). Reimbursement rates are calculated on the basis of different data sources across the countries. For instance, while England includes all hospitals to determine average costs (300 hospitals), the Netherlands chooses hospitals that are representative according to some criteria (including 23 or 22% of all hospitals using DRGs), Italy, Germany and Spain include those hospitals with predefined cost accounting standards (including 8 (1%), 214 (12%), and 18 hospitals respectively). There is a trade-off between the quality of data and the degree to which the data are representative of the country’s hospitals (Schreyogg et al, 2006). Calculating reimbursement rates also requires decisions to be made about outlier cases (i.e. treatment episodes with much higher or lower resource use than average) and about defining prices. For the former, different methods can be used to detect outliers using mathematical trimming. To set prices, products first have to be defined, which can be either via medical procedures, groups of medical procedures, or most commonly a combination of clinical, demographic and resource consumption data. In England and France prices are established directly by the average cost per DRG, while in other countries, DRG cost weights are used. Whether with average costs or weights, further adjustments are needed to account for differences in hospital structure, region and other factors affecting the cost of service delivery. These adjustments are made differently in each country. Finally, many continue to make modifications to improve accuracy and also to encourage the uptake of new technologies.

 

The impact of the DRG hospital financing systems introduced in Europe have been measured to differing degrees. In Austria, hospitals became more cost conscious, hospital activities more transparent, information for insurance funds improved, and hospital costs increased more slowly, but there were no differences for patients. In Belgium the length of stay declined, significant efforts were directed towards detecting fraud (therefore potential waste and administration costs), and the hospital budget was better controlled. In Denmark, hospital activity increased, and the private sector increased its market share. In Finland, transparency and productivity appeared to improve, and there were estimated financial savings. In Italy, costs of services have fallen along with the average length of stay and number of beds, patients experienced lower waiting times and increased accessibility, insurance plans expanded their benefits packages because of increased financial transparency, and hospital budgets were better controlled (HOPE - European Hospital and Healthcare Federation, 2006). In England the DRG-style Health Related Group (HRG) system introduced in 2003/2004 to improve performance has been characterized by slower and more difficult than anticipated implementation, higher than expected costs, and to date no evidence of improved efficiency or increased activity (Boyle, 2007).

 

Administrative costs

 

Administrative costs reflect part of healthcare funding that is not directly related to improving health. Included in these costs are the planning, management, regulation, collection of funds and handling of insurance claims. Often expenditure in this area can be used to infer the level of efficiency in which the healthcare system operates. OECD country data show that social health insurance countries generally have higher administrative costs (i.e. Germany, Luxembourg, the Netherlands), with costs in Austria being relatively lower and closer to those of tax-funded systems. The lower level of administrative costs in Austria and France could be due to the lack of choice permitted between sickness funds, with the exception of Luxembourg (Figueras et al, 2004). In spite of this, the Netherlands in 2006 generalized the provision of insurance through competing private insurance funds though with heavy state regulation (see Section 11.6 Financing Health Care). In the Central and Eastern European countries that have embarked on healthcare financing reforms, the separation of the collection of health insurance premiums from the collection of general taxes has likely increased the administrative costs within the country (Thomson et al, 2004).

 

Private health insurance (PHI) is associated with much higher levels of administrative costs than statutory public health insurance systems. This results from the extensive bureaucracy involved in private insurance markets related to assessing risk, setting premiums, designing benefits packages and reviewing claims. For example, administrative costs of PHI in Germany, Luxembourg, the Netherlands and France are about 10% of total costs, while in Austria, Belgium, Italy and Portugal they may be up to 25% of costs (Mossialos and Thomson 2004). The costs of administration are much lower for the statutory health insurance systems, around 3-5% in most countries. It can be argued, therefore, that the additional costs of administration associated with PHI could largely be avoided and thus can be considered as inefficient.