Management of commissioning

German commissioning practice shows how powerful and sophisticated the monitoring aspect of managerial performance can become. Data were collected in real time. SHIs audited, confirmed and made payments continuously, collecting from hospital bills and medical records data about what activity was being paid for. Such data enabled one SHI, reputedly the most developed in this respect, to make routinely such analyses as volumes of hip-replacement revisions per provider, evaluations of disease-management programme costs and outcomes for diabetes, and analyses of case mix distributions and trends for any geographical level, any provider(s) and any DRG. Each SHI knew its own data for hospital case mix and compared them with the publicly available national figures, interrogating apparent inconsistencies between the two. The Medizinische Dienst der Krankenkassen (MDK), run jointly by the SHIs, routinely reviewed patient case notes in order to verify if the coding and therefore payment were appropriate given the clinical facts, but did not review the effectiveness of care.

Hospital activity was nevertheless far from transparent to German SHIs or patients.²⁶⁵ German hospital managers told us that when negotiating with commissioners they aggregated data and income data into large blocks:

Interviewer:

Why did you decide that?

Controllingschef, hospital 4:

To make in all one negotiation about the DRGs and one for nursing care for children. And it’s less transparent to the SHIs. . . . We have an orthopaedic department in [hospital 1] and one at [hospital 2], and the SHIs could see from our data, our Excel tables, we have done this here but more there . . . if we handle matters at a large scale we don’t have to discuss these things with the SHI.

Hospital staff also tended to dislike the Medical Review Board and its members.

Lessons for NHS commissioners concern the availability and analysis of performance activity data. DRG adoption alone is insufficient. NHS commissioners’ performance management, cost control and planning roles might be strengthened by having rights of access to provider data but German practice also suggested that gaining the modelling and monitoring benefits of a tariff system requires data warehouses and specialised staff, centralised at the level of some millions of patients, and an extensive IT infrastructure. These specialised functions may be more effectively and cheaply undertaken at regional or subregional level (e.g. by commissioning support units) than at CCG level. However, greater transparency and the resulting commissioner interventions appear, on German evidence, at times to strain relationships between commissioners and providers.

Negotiated order: the Rhineland model

The German health system contains nested negotiated orders at national, Land and provider levels. Annual negotiations involving all main national interest groups (federal associations of SHIs, doctors and dentists, hospitals and patient organisations), co-ordinated by the Gemeinsame Bundesausschuss and with the state as arbiter, agree the broad framework of health service planning and guidelines for quality of care. This ‘Rhine’ or ‘Ordoliberal’ approach was deeply rooted in German political culture.²⁶⁸ Although binding, these decisions were consensual. Consequently, doctors and SHIs could block changes. For instance, the Association of Ambulatory Physicians vetoed other doctors doing out-of-hospital surgical procedures. Similar negotiations establish a Land bed plan based on predicted needs for hospital services, make regional adjustments to DRG payments, and agree the points tariffs by which ambulatory care doctors are paid.

At provider level, the main currency of commissioning negotiations between SHIs and German hospitals was the number and case mix of episodes and, for ambulatory care doctors, medical acts. To launch the annual contract negotiations, German SHIs sent each hospital spreadsheets of DRG targets. The proposed case mix implied an overall number of DRG points, hence an implied budget. It was possible to reduce, even remove, groups of cases by reallocation within the total number of points, but the system did not so readily allow an overall reduction in case load, case mix or budget. National SHI federations, SHI national offices and local health-manager networks advised and updated SHI negotiators about the commissioning climate and local issues, but the hospitals had better data about their own case mix and internal costs than the SHIs did. The Land plan framed the ensuing commissioner–provider negotiations by defining each hospital’s bed numbers, overall case load, case mix and, in effect, ceiling for SHI-funded activity. Negotiations focused on the hospital’s DRG points allocation, its case mix and the nationally defined growth margin rather than clinical quality. Payment for new treatments not yet in the DRG system and for discretionary services were also negotiated. Assuming that some degree of planning of the overall profile of health-care provision is desirable, the German DRG system instantiated a concrete, detailed way of modelling and managing hospital activity, case mix and revenue costs, although it might be argued that deciding DRG volumes within the constraints of a Land bed plan puts the infrastructural cart before the epidemiological horse of health-care needs.

For NHS commissioning, German experience seems to confirm that a consequence of making DRGs the commissioning currency is to focus commissioners’ negotiations on service volume, case mix and cost rather than on service quality, which commissioners therefore need to manage through another mechanism. Giving hospitals the right to reimbursement for whatever patients they could attract tied the German commissioners’ hands when negotiating with providers. SHIs strengthened their hand by negotiating jointly with providers, and negotiating with each provider separately.

Discursive control

German informants mentioned emic discourses used in commissioning management and negotiations. They tended to refer to the different parties’ rights and obligations under the nationally negotiated agreements and regulations. These arguments cut both ways. When SHIs claimed to represent patients’ interests, the hospitals replied that they – and SHIs – were equally obliged to ensure that patients could get the services that they (patients) chose. Apart from having to work within the Land bed plan, considerations of public accountability did not appear to figure much. Neither did EBM, and still less in primary than secondary care.²⁶⁹ Provided they stuck to treatments authorised under the Land plan, law and regulations, German hospitals’ treatment methods were beyond SHI scrutiny. The place of EBM was more at national level. When the GB-A decided which new therapies, devices, pharmaceuticals or models of care to include in the DRG tariff, it used above all evidence about effectiveness from the Institut für Qualität und Wirtschaftlichkeit in Gesundheitswesen.

An implication for NHS commissioners appears to be that a shared framework of norms is what makes discursive control possible. The norms do not necessarily have to be technical ones. Neither does the norm of ‘complying with health policy’ have to be defined any single way (e.g. compliance with a negotiated settlement vs. compliance with the latest government pronouncement). What matters is that commissioners and providers share the same norms.

Incentives

Diagnosis-related group-based payments gave hospitals a financial incentive to increase activity,²⁷⁰ whereas the SHIs wished to avoid patient numbers, hence costs, spiralling out of control. Only a few treatments (e.g. short-term nursing care at home) required SHI consent to pay. Otherwise, the DRG tariff system left German commissioners little discretion for using financial incentives to renegotiate provider behaviour locally. As a workaround, SHIs negotiated with hospitals a ‘corridor’ (Flur) for the main groups of DRGs, agreeing an expected level of activity for each, what rebates the SHI would receive should the volume or case mix fall below that range and the payment for justified additional work above it. Since 1998, German SHIs have offered selective contracts, restricting subscribers’ choice of providers in return for lower subscriptions, but many patients assumed that only providers with difficulty attracting patients accepted such contracts.

Neither did the tariff system directly reward providers for improving the clinical quality of care (as opposed to attracting more patients). SHIs had discretion to pay providers for new and off-tariff treatments, but otherwise were obliged to pay for all acute hospital treatment irrespective of quality. Complex and long-term care presented different problems. For chronic care, the SHIs paid a per diem Pflegekost (care cost) and will from 2013 have the option to do the same for some psychiatric services, instead of making cost-plus payments as hitherto. Some experimental integrated care projects constructed interorganisational care pathways linking primary and secondary providers for certain patient groups, but these required specially negotiated contracts because DRGs were available not for network-based care provision, but only for paying single providers. For ambulatory care, the points tariff also had to be modified to accommodate disease-management programmes, that is preventative case management and continuous care for certain chronic conditions (e.g. diabetes, COPD). Some 14,000 such schemes existed but only about 5.5% of people were enrolled in them. Integrated care and disease-management programmes represented only 1% of health-care spending. The benefits appeared to be improved care, at least for diabetics,²⁶⁹ rather than cost savings.

These problems and workarounds have parallels in England. German SHIs independently invented what NHS commissioners call ‘cap-and-collar’ agreements, and independently discovered that the construction of cross-organisational (‘integrated’) care pathways requires substantial modifications to tariff payment systems. A hidden limitation of a tariff system is the requirement for a separate, parallel system for undertaking the commissioning function of managing clinical quality in care providers, raising transaction costs in consequence. It is possible to define DRGs in terms of clinical procedures but this negates a theoretical advantage of DRGs: that providers are paid per episode of care (ideally, by outcome), remaining free to introduce innovative clinical procedures.

Provider competition

Except for selective and integrated care contracts, German commissioners could not choose their providers. Only the Land government could select or deselect hospitals as providers, by including them in the bed plan or not. For ambulatory care doctors and dentists, not even this mechanism was available. The German equivalent of an AQP policy prevented commissioners from using selection (competition) of providers as a means of controlling them, despite – indeed, because of – patient choice of provider.

Hospital capacity was 5.66 acute beds per 1000 population (2010), reflecting German health care’s greater orientation towards secondary rather than primary medical care and community health services, and ‘a hospital on every hill’. In the west, Germany also had a tradition of religious foundations providing hospital care, especially small local hospitals. Corporate providers were also entering the hospital market. That appeared, on German experience, to promote market concentration and centralisation on the provider side, with firms such as Aesculapias developing a strong national organisation to counter those of the SHIs. [Competition between SHIs also produced concentration on the commissioner side of the quasi-market (TK data, 2010), as in the Netherlands.²⁷¹] Nevertheless, public- and third-sector hospitals were no less – perhaps more – efficient than private ones.^272,273 Attempts to control the health system by regulation and tariffs pushed provider competition into the marginal ‘windows’²⁷⁴ not foreclosed by regulation, tariff or policy fiat. When competing providers are entitled to payment once patients have chosen them, commissioners’ control over provider costs is weakened from a budgetary cash-limited system into one that, at most, contains care costs within ‘corridors’.

If NHS commissioners are to harness provider competition as a medium of governance, German SHIs’ experience suggests that commissioners need scope to deselect providers, for instance from providing services for specific care groups. Long-term provider competition may become self-weakening as providers merge or fail in the face of competition, which suggests that commissioners would need to be continually searching for possible new providers and (insofar as they can influence such events) default towards taking a critical view of proposed provider mergers. Comparing German and UK experience suggested that the possibility of provider redundancy, hence excess provider capacity, may be required to sharpen provider competition. That would appear to confront commissioners with having to decide whether to control costs through provider competition or by reducing the numbers of providers.

Juridical controls

German commissioning relied on nationally standardised regulations, contracts and legal entitlements, clearly specified decision-making processes and participants in them, and allocations of decision-making powers among particular institutions. The range and number of services offered, and remuneration rates, were stipulated at national level for all SHIs.²⁶³ German SHIs’ obligations to patients were legally prescribed. In disputed cases a first step was to seek an independent opinion from MDK about the medical necessity of the treatment in question, followed by appeal to the Schiedstelle (administrative court) and then to the civil courts, but even Schiedstelle cases were infrequent (maybe one or two a year for the largest hospitals) and expensive (€7000 or more per case). The hospitals won perhaps 80% of these cases. A lesson for NHS commissioners was that, even in the juridically oriented German system, judicial remedies were for commissioners and providers alike costly, unpredictable instruments of last resort, used only exceptionally. Most important, regulations intended to maximise provider diversity and competition for patients removed provider competition as a medium of commissioner power.²⁶⁷

Lombardy health policy context

Like its English counterpart, the Lombardy region of the INHS had a quasi-market structure, a mixture of public, third-sector and corporate secondary-care providers, between which commissioners were legally required to ensure fair competition. Public hospitals were semi-autonomous ‘private firms’ (AOs). All organisations wishing to provide publicly funded health care had to pass a four-stage accreditation process, whose last stage was the award of a contract. Local HAs (ASLs) commissioned services for populations of similar size to those of English PCTs, although unlike English CCGs they controlled only 2% of the budget for ambulatory and diagnostic care. The regional authority, a branch of the regional government, commissioned both social care and the rest of health care. It had considerable discretion over what kind of quasi-market structures to establish. It could, for instance, vary the national DRG tariffs and take ‘make-or-buy’ decisions. All region authorities directly managed some hospitals, although Lombardy retained only one (Table 10).

TABLE 10

Italian regions: proportion of beds in directly managed INHS hospitals

Of the media of commissioner power over providers, discursive (ideological) control predominated, followed by negotiated order and the managerial (above all, regional) performance of commissioning. Incentives and provider competition were less important, juridical controls least of all (Table 11).

TABLE 11

INHS (Lombardy) mode of commissioning

Elements of population-based commissioning coexisted with client-based commissioning (DRG tariffs), but the former, undertaken mainly at regional level, predominated. More like the UK than Germany, a range of community health services and intermediate care providers existed.

Managerial performance

Commissioners planned their local health economy. Every January the regional authority set overall inpatient case mix and numbers for each ASL, and the range, level and volume of non-tariff payments (FNTs; see below). Each ASL then made a territorial plan and negotiated contracts with its accredited ‘preferred providers’, chosen on a value-for-money basis. The ASL agreed admission levels with each provider and the contract duration. Providers had two kinds of contract. One, with juridical status, was awarded as the last stage of the accreditation process. It stated the main rules and quality standards with which the provider must comply. An annual operational contract fixed the provider’s budget and activity level for that year, but had no legal status. Both contracts were managed at ASL level, although the regional DGs oversaw the whole process, allocated resources and actually licensed the providers. Comparing the roles of local commissioners (ASL in Lombardy, PCT or CCG in England) and regional organisations made the Lombardy INHS appear more centralised than the English NHS; but it was the reverse at national level.²⁷⁷ Compared with the English NHS, the strength of this highly vertical system was its control over providers and its homogeneous way of gathering information and data, which has been described (p. 209)²⁷⁶ as a ‘quasi-administered’ system.

Nevertheless we found at regional level a considerable fragmentation of management, a ‘silo’ approach coupled with a lack of information sharing and of transparency in decision-making. In Lombardy, health services commissioning remained divided from that of social services, at both regional and ASL levels. Indeed, the so-called integrated care for frail people with health and social care needs was commissioned by the family DG, even though the regional health fund financed these services. Within each ASL, the social care department (attività socio sanitarie integrate) and its director were responsible to the ASL general director for the key strategic policies, but still maintained a good deal of autonomy, networking with different institutional stakeholders such as the municipalities, social services, schools and other local actors. Other local (ASL) commissioning staff had limited discretion but when they tried to exercise it they were often admonished by the regional DGs. Primary care was an exception due to the social capital on which it called and its different territorial characteristics, which let commissioners collaborate more actively with providers.

Providers’ contracts stipulated external audits of services and peer reviews of service quality. (Each Emilia Romagna hospital also had an advisory committee, which included patient representatives and often conducted user surveys.) Commissioners appeared to consider monitoring by means of administrative data management too laborious and inefficacious as a means of quality control in Lombardy, although inspection of medical records was used in Emilia Romagna. However, monitoring data were not necessarily converted into information through which commissioners could exercise governance over providers:

of course we have all the data and whenever we do something we perfectly know who will be influenced by that intervention and the expected results. . . . We conduct a lot of survey and data analysis, . . . surveys and tables for the national accounting court . . . but no documents and scientific studies have been made. . . . Research of course could be done. But there is also a point of the political willingness and interest in doing that: to open the data and information as well as a clear interest in confronting with other realities.

Director-General, regional government

Health technology assessments and evidence-based protocols were applied, but not systematically and very little by commissioners at ASL level.

For English NHS commissioners, one lesson appears to be that interdisciplinary rather than uniprofessional organisational structures are more conducive to an integrated approach to service commissioning. More striking, though, is the proof of concept of the feasibility of regionally managed commissioning by organisations with the latitude to select providers and set contract duration. To overcome the information asymmetry between providers and commissioners, and to make the health system more transparent to the public, required the political will on the part of the commissioners and the higher-level bodies which oversee them.

Negotiated order

Unlike their NHS equivalents, most senior commissioning managers in Lombardy (and Emilia Romagna) had occupied the same role for 10 years or more. Long-term stability, investment in training commissioners and long-standing relationships with providers gave commissioners a deep knowledge of the regional health system, its development and commissioning mechanisms. These relationships and the trust which had accumulated were an important medium of commissioner influence over providers, especially the AOs but also private providers. Commissioners’ relationships with providers have been described as more like a ‘compact’ than a ‘contract’ (p. 3).²⁷⁸ In both Lombardy and Emilia Romagna, disagreements between commissioners and providers were usually resolved negotiatively and, in Lombardy, tended to concern quite focused activities and services. It was a way in which clinicians participated in commissioning. (In Emilia Romagna, few people other than clinicians participated.) Most commissioning negotiations were about adjusting recent historical patterns of provision. Nevertheless, there was also an informal vertical hierarchy in decision-making about commissioning, with the regional DGs and directors-general for health and for social care services at its apex. The other important interest group (in both regions) was the GPs’ trade union, which negotiated GPs’ contracts at both national and regional levels.

For NHS commissioners, the Lombardy and Emilia Romagna cases suggest that frequent changes in commissioning personnel may be counterproductive for commissioners’ exercise of ‘relational’ governance over providers. The Italian examples also indicate the practicality of giving clinicians (medical managers) a substantial role on either side in commissioning negotiations.

Discursive control

This negotiated order partly rested, too, on shared ideologies. Because programme planning, resource allocation and therefore commissioning were mainly driven by political decisions, they had an ideological substrate and a disciplinary force over the professionals and managers involved. A regional school for public managers, and training programmes for health-care managers and professionals, helped foster common disciplinary and ideological norms across the regional health system. A key, distinctive shared belief was the idea of subsidiarity, expressed by many theories (e.g. the New Public Management, the so-called Third Way, the new public governance).²⁷⁹ In Lombardy it shaped a unique view of government, focused not on the division of powers among different layers of government (‘vertical subsidiarity’) but on ‘horizontal’ subsidiarity, conceived as a sort of division of labour between the public sector and civil society. It reinforced the ‘knightly’ public service ethos, (p. 195)²⁸⁰ which interviewees expressed more strongly than provider or professional interests. In Emilia Romagna, in contrast, the Alma-Ata declaration²⁸¹ was a chief reference point. These (emic) principles appeared more pervasive and influential than evidence-based (etic) discourse.

As with Germany, a lesson for NHS commissioners appears to be that what makes governance by discursive means possible is not so much how the norms are formulated – whether as policy or technical norms – but the fact that commissioners and providers share the same norms.

Financial incentives

Lombardy commissioners’ only lever of control through DRG tariffs was by negotiating a production ceiling in the providers’ contracts. For ambulatory and diagnostic services, a provider was guaranteed 95% of the previous year’s expenditure and case load. For activity from 97% to 103%, the tariff was cut by 30%; from 103% to 106%, it was cut by 60%; and above that the tariff was zero. This arrangement incentivised providers to achieve the 106% level, to grow the next year’s starting budget level. As cost-control pressures increased, ‘ceiling budgets’ were introduced into provider contracts, enforced by tariff caps should service use exceed the planned budget. Neither would the regional authority reimburse providers for treatments not included in its annual plan. In these ways the Lombardy commissioners contained costs by (in effect) constructing a hybrid of cost-and-volume (below the 106% level) and block contracts (above 106%).

In response to other limitations in the tariff system, commissioners kept some services outside it, financing them as ‘functions with no tariff’ (funzioni non tariffate, FNTs). This was partly because tariffs could not be introduced for all services at once. Outpatient, diagnostic and some highly specialised clinical services were still paid for through cost-plus fees for service. Rehabilitation, residential services, domiciliary services and long-term care were still paid for per diem. For other services (A&E, dialysis, foetal and neonatal pathology), commissioners considered payment by DRG undesirable because they believed it would restrict access. Neither was it practicable to finance teaching and research through tariffs. FNTs had also been used to support private and non-profit hospitals that could not access other public funds for their activities, and indeed public hospitals that had exceeded their budgets. However, as cost pressures increased, extra-budgetary payments were gradually reduced. Other exceptions to the tariff system included integrated care management; ASLs would assess each patient’s needs, decide the necessary treatment plan and give the patient a payment voucher to cover the cost. From 2012 the regional DGs paid a variable premium (adjustment) of ± 2% of the budget to providers according to the provider’s performance against certain quality standards. The local ASLs also allocated 2% of each provider’s previous year’s income for activities of its choice.

Lombardy commissioners also adopted programmes to ‘nudge’ patient choice to reduce cross-boundary patient flows, especially for highly specialised services (e.g. neurosurgery), because the Lombardy commissioners would have to pay the regions receiving these patients. (Cross-boundary flow was also an issue in Emilia Romagna.) For example, Mantova ASL, on the border of the Emilia Romagna and Veneto regions, introduced training schemes to encourage GPs to meet their patients’ diagnostic and specialised ambulatory care needs within the region, and created an incentive for hospitals to hire specialists to treat such patients.

In the fact that commissioners in Lombardy independently invented something similar, NHS commissioners may find some endorsement of their own ‘cap-and-collar’ modifications to the HRG system. The experience of commissioning in Lombardy suggests further modifications: retention of pre-DRG payment systems and the application of non-DRG payments to stimulate provider compliance with clinical quality standards, the development of interorganisational care pathways, and provision of specialist treatments within the region.

Provider competition

Until the banking crisis, the system of paying providers for up to 106% of last year’s case load was largely used to foster provider competition. One consequence was that the proportion of small providers fell, partly through mergers. Private providers’ market share increased, partly because of public–private partnership schemes which involved contracting out the operational management, and temporarily the property, of nine AO hospitals, and a network of specialised services and departments, especially for diagnostics and rehabilitation. Private providers concentrated mainly on long-term care, less clinically risky and more lucrative than acute care,²⁷⁶ although provider competition impeded the construction of stroke networks.²⁸² Since 2002, competition has decreased. Because new provider entry had saturated the supply of hospital beds, both regionally and nationally, a regional decree closed the publicly funded health system to new providers. Fixed tariffs limited the scope for price competition, production ceilings the scope for competition on quality. Consequently the competitive incentives for hospitals, and after 2007 also for diagnostic and laboratory services, to improve services weakened. However, patients’ choice of provider could still slightly change a provider’s share of the INHS quasi-market, and still played a role in rewarding provider quality and the providers’ reputation. Domiciliary care provision was more competitive, but, since the reform of assistenza domiciliare integrata (integrated home care) services was still in a pilot phase, it remains to be seen if that situation will continue. In general, though, provider competition was sacrificed and tariffs ‘bundled’ in the interests of cost control, and to a lesser extent planning.^283,284 This experience and the consequent trade-off decisions appear relevant to NHS commissioning too.

Juridical controls

Through regional law, regulations and decrees, juridical control was pervasive, extending to defining the roles of different types of manager,²⁸⁵ the annually planned workloads and the overall lines of health system development. A major challenge for the Lombardy regional health system was the permanent conflict of competencies and roles between the regional health general directorates and the ASLs, which in practice had only a limited role in actively programming, purchasing and controlling the health services allotted to them. There was no regulatory authority for competition comparable with the UK Office of Fair Trading. Nevertheless, tight financial and juridical regulation left commissioners and providers little margin for autonomy and discretion on either side, defined a clear division of labour between them, and created a joint commitment to avoiding conflicts. Conflicts between providers and commissioners were usually settled informally, either by direct negotiation, or, if still unresolved, by the regional DG, but without any mediation. Doctors in management roles in some providers maintained direct relationships with regional DGs, going over the heads of their ASL. Exceptionally, disputes were taken to the ordinary courts, but the costs were high and procedures time-consuming and, from the commissioners’ standpoint, of little consequence because any remedy or sanction imposed on the provider was usually administrative. The courts seldom withdrew a provider’s accreditation. At most, decrees supplied part of the normative framework through which commissioners might exercise discursive influence over providers.

Although comprehensive juridical powers were available to commissioners, they were used sparingly, commissioners being costly, unpredictable instruments of last resort. The same lesson for NHS commissioners emerges as from the German case.