Quality and legal aspects in public care procurement

Purpose – The purpose of this paper is to assess the applicability of care quality concepts as contract award criteria for public procurement of health and social care, using the case of Sweden. Design/methodology/approach – Based on a literature review, European and Swedish legal texts, government regulations as well as 26 Swedish court review cases concerning care procurement have been analysed. Findings –Methods used for assessing care quality are seldom useful for predicting the quality to be delivered by a potential contractor. Legal principles of transparency and equal treatment of tenderers make it necessary to apply strict requirements for verification. Research limitations/implications – Results refer primarily to a Swedish context but could be applicable throughout the EU. Further studies of relations between award criteria and public/private collaborative practices for improving care quality during contractual periods are desirable. Practical implications – Local and regional procurement officials should benefit from a better understanding of how quality criteria should be designed and applied to the award procedures for care contracts. Care providers in the private sector would also be able to develop their quality strategies and present their abilities more efficiently when tendering for public contracts. Social implications – Issues of quality of health and social care are of obvious importance for social sustainability. Public awareness of care quality problems is evident and often a cause of media concern. Originality/value – This investigation pinpoints the difference between traditional care quality thinking and the legal principles underlying contract award in public procurement of care services.


Introduction
Global demographic change implies rising numbers of elderly people needing care services.Sweden is one of the countries where the age pyramid changed its shape early on, and the expansion of care has been accompanied by outsourcing to private sector providers, with the public sector typically acting as buyer of care services.When public authorities procure services from private providers, it is often obvious that contracts should not be awarded with lowest price as the only criterion for provider selection.Frequently, public procurement, which is in the focus of the present investigation, involves also one or more quality criteria, as for construction projects (Waara and Bröchner, 2006), public transport (Camén, 2010;Rönnbäck, 2012) and other services, including care (Bergman and Lundberg, 2013).For health and social care procurement (or commissioning: Murray, 2009), we may expect the quality issues to be particularly difficult to manage, although a huge number of health quality studies have been produced over many decades.However, in the context of selecting a private provider, there are two major obstacles to using earlier research on care quality.
The first obstacle lies in translating methods and measures from retrospective studies of care quality into predictive tools for the context of public procurement: will a particular provider deliver the desired quality level after having gained the contract?Darby and Karni (1973) distinguish between search, experience and credence qualities of a good.Search qualities are those that can be ascertained prior to purchase, thus relevant to a context of service procurement, be it private or public; credence qualities are expensive to judge even after purchase.Most health care services are associated with credence qualities, and difficulties related to measurement of credence qualities make the monitoring of care contracts costly.
Second, will quality criteria derived from earlier care quality research stand up to judicial scrutiny in case of a conflict arising from a contract award decision?Few authors have noticed the growing importance of how the legal system treats health care quality.McHale (2002) discusses how English courts have dealt with review cases concerning issues of professional practice.Such review cases concern the illegality, irrationality and procedural impropriety of decisions.However, it is both costly and time-consuming to litigate.While these cases are post-event (scrutiny of alleged clinical negligence), "legislation may also facilitate this process/determining the appropriate standard of care/ in a proactive manner".In the USA, it is competition law, including antitrust and consumer protection, that can be said to raise issues of health care quality (Sage et al., 2003).The two Institute of Medicine reports To Err is Human (IOM, 1999) and Crossing the Quality Chasm (IOM, 2001) underscored how law is related to quality of care.Over time, courts started taking consumer preferences and not only professional standards into account when identifying quality.Predictive assessments of care quality are implicit when allowing health insurers to select providers and purchase care services based on both price and quality.Also pointed out by Sage et al. (2003) is the potential for public purchasing to develop and implement market-oriented solutions to quality problems, mentioning standardized consumer information as an example.Highly pertinent is the identification by Hyman (2004) of barriers (relevance, complexity, framing, scepticism, inadequate demand) to legal use of health care quality research in the context of US competition law.Within EU, the relation between health care quality and competition law, as interpreted by the courts, is less clear and less important (cf.Fornaciari, 2009).
Nevertheless, while unknown before 1989 in European procurement directives as well as in most national legislation concerning public services, "quality" as a legal term has now spread widely.The purpose of this investigation is to assess the applicability

649
Quality and legal aspects of care quality concepts as contract award criteria for public procurement of health and social care.We use the case of Sweden as our point of departure since the national emphasis on care quality measurement is obvious in international comparisons (Kollberg et al., 2005), in addition to the fact that the share of health and social care procurement is high, also seen in an international perspective (Anell et al., 2012).

Methodology
After an initial literature review of the care quality literature, we investigate how the "quality care" concept is defined and understood in documents regulating the service of elderly care, including court cases with public procurement conflicts.This elderly care context was chosen for our content analysis of texts because of: first, its particular challenges when defining and assessing quality, as quality often is thought to be a largely subjective matter determined by users or customers; in this context the elderly themselves might suffer from cognitive impairment, so that relatives or friends are relevant stakeholders; and second, the important volume of care services to elderly delivered by private firms; and third, public concern with controversial media reports of mistreatment of elderly by procured care firms in Sweden.We relied on a straightforward application of systematic content analysis (Hall and Wright, 2008), where pre-selected terms were located in the texts.Additionally, simple tabulation of term frequencies in court verdicts was performed.
Our analysis should be applicable to most national sets of regulative frameworks.It departs from how care quality is conceptualized in national legislation.Next, we look at how requirements, explicit and implicit, on procedures and criteria have been formulated in regulations concerning public procurement.Ultimately, the objective here is to identify acceptable approaches and concepts from quality theories and practices in general.Our illustration is the case of public procurement of care services for elderly in Sweden.To address the issue identified and to get a deeper understanding of applicability of care quality concepts as contract award criteria for public procurement, our data consist of: first, all relevant European public procurement directives since 1971; second, Swedish statues concerning health care and social care for the 1970-2013 period;third, Swedish 2009third, Swedish -2014 verdicts in administrative appeal cases, based on keyword searches in the InfoTorg legal database; and fourth, an overview of care quality literature.We relied on the InfoTorg legal database with full text verdicts from administrative courts.Here, we searched the January 2009-June 2014 period for review decisions where the three keywords procurement (upphandling), care (vård) and quality (kvalitet) occurred jointly.The results were screened to ensure that only care services contracts were in focus, which was the case for a total of 26 verdicts.

Literature review
Our review of the relevant literature begins with the topic of quality of care, followed how quality is related to public procurement and how quality is treated in the European procurement directives and in European case law.Finally, earlier studies of care quality in the context of public procurement are brought up.

Quality of care
The literature on quality of care is huge and also predates the stream of generic service quality studies that began in the 1980s (Martínez and Martínez, 2010).Currently, the US Institute of Medicine (IOM, 2006) recognizes six quality-related goals: safety, effectiveness, patient-centeredness, timeliness, efficiency and equity.

TQM 28,4
Already Sheps (1955) distinguished between three purposes of evaluations of hospital quality: regulatory, stimulus for the improvement of quality, and studying the effects of specific programmes or procedures on the quality of care.It is the third purpose of evaluation which comes closest to the use of quality criteria in procurement/ commissioning of care services, but then it takes the form of a predictive evaluation.Donabedian (1966) outlined three approaches (outcome, process, structure) to assessing quality of medical care; it is structure and process that are more relevant for procurement/commissioning of care services.Outcome would refer to documented outcomes from earlier contracts with other clients.
In principle, satisfaction survey results from tenderers' prior contracts could be required by a procuring entity as part of the documentation for assessing the degree of "quality" when awarding a contract.The entity could require that tenderers submit proof of a certified quality assurance system that includes the use of satisfaction surveys.Furthermore, it could be specified that the contracting authority is to receive results from satisfaction surveys of a particular type and at stated intervals.Measurement of patient and personnel satisfaction with care is obviously related to the care quality issue.
A pioneer study as that of Abdellah and Levine (1957) has been followed by numerous later investigations; Abusalem et al. (2013) have reviewed 23 studies of patient satisfaction in home health care, listing the dimensions and scale types used in these studies, which show great variety of approaches, unlike what has emerged as standard practices for customer satisfaction measurement in consumer services.Obviously, the applicability to (health) care of the SERVQUAL scale (Parasuraman et al., 1985(Parasuraman et al., , 1988) ) depends on the validity of the original US executive interviews and customer focus groups, considering that the four services chosen were retail banking, credit card, securities brokerage, and product repair and maintenance.Three of these services have been transformed fundamentally in the last 30 years by ICT applications; only the fourth has had and retains a slight resemblance to health care services.Thus the five "scale dimensions" of tangibles, reliability, responsiveness, assurance and empathy may be partly irrelevant to a care context.
The five SERVQUAL dimensions can be compared with five "quality characteristics" or "quality attributes" of home services for the elderly (continuity, availability/time, influence, personal relation and "the suitability of home helpers in terms of competence and personal manners") chosen by Edebalk et al. (1995).Lee et al. (2000) proceeded by adding two dimensions (core medical service, professionalism/skills) to the five SERVQUAL dimensions when assessing health care service quality.More ambitiously, Koerner (2000) retraced much of the method used for developing SERVQUAL and did so in an inpatient nursing context; factor analysis led to five mostly different service "quality dimensions": uncertainty reduction, reliability, compassion, relationships and individualized care.Furthermore a USA-UK comparison of health care quality in nursing homes has relied on a version of the SERVQUAL scale itself, although eliminating "assurance"; the statistical results were slightly weak (Kilbourne et al., 2004).Several types of criticism have been levied against the original SERVQUAL measures (Ladhari, 2008(Ladhari, , 2009)); among critical comments that are of interest when analysing care satisfaction, validity of items has been questioned as well as the fact that SERVQUAL focuses on the process of service delivery rather than outcomes.
If involvement of those who receive care is a policy objective, a quality focus on outcomes may imply that "outcomes for people" are more important than the achievement of a predetermined level of service quality (Willis and Bovaird, 2012).Wreder et al. (2009)

651
Quality and legal aspects identified several groups of stakeholders in public eldercare, starting with consensus among nurses who all mentioned the patient's relatives, doctors and their colleagues.Examples of studies of caregiver's work satisfaction in care for older people are given by Suhonen et al. (2013).Service provision is an interactive process (Grönroos and Voima, 2013), which is brought out clearly in the study by Gill et al. (2011) on service co-creation in community-based aged health care.
The problems of comparing results from care satisfaction surveys have several sources: many: satisfaction measures may primarily reflect interpersonal care experiences, including the effect of language barriers, patient experiences of their health status regardless of care they have received, feelings of fulfilment of patients' a priori desires; furthermore, the timing of satisfaction measurement (Manary et al., 2013).Ambiguities of satisfaction survey interpretation increase when patients may suffer from dementia (Zank and Leipold, 2001).A recent Swedish study has compared care satisfaction among old people receiving public care at home and in other accommodation (Karlsson et al., 2013), finding that functional impairment and health complaints overshadowed differences in where they lived.
In order to assess and predict the quality of service in advance, the quality level or quality maturity of the organization can be measured.The existence of certified management systems against the ISO 9001:2008 standard might be a potential indicator.Fulfilling requirements according to this standard does show that there is a well-documented methodology within the organization, rather than proving an ability to meet customer needs, requirements and expectations (Poksinska, 2010).A more direct way to assess the quality maturity of an organization is to use quality award models such as the Malcolm Baldrige National Quality Award, the European Excellence Award and the Swedish Quality Award.Although this goes beyond the immediate needs for contracting, it is recognized that relying on award models can provide benefits such as increased awareness of overall issues and customer orientation, and furthermore that quality improvement initiatives are supported (Hendricks and Singhal, 1997;Corredor and Goñi, 2011).A drawback of these award schemes is that they require considerable resources for organizational assessment, especially in the phase of description of activities when employees engage in self-assessment, and that evaluations require specific skills (Eriksson, 2003).Looking back at the preceding decade of market-related change in the provision of social services in England, Knapp et al. (2001) identified the issues of high transaction costs and of to what extent price competition would damage quality of care.

Public procurement
Procuring entities can influence quality by prequalification requirements (Eadie et al., 2012), by specifying quality standards (Kuypers and Gruppen, 2008;Enquist et al., 2011) in the tendering documentation and by applying quality criteria when awarding contracts.How a contract is formulated influences the ability and willingness of the service provider to manage quality (Camén, 2010).Contractors can be provided with a variety of contractual incentives for both cost reducing activities and quality (Albano et al., 2006).The tendering documents also provide the fundament for client monitoring of services actually delivered, which can be essential for quality development (Dean and Kiu, 2002).However, detailed and static contracts reduce the scope for flexibility in developing the contractual relationship (Camén, 2010) and can thus impact quality management.This lack of flexibility is stronger under public procurement legislation than in long-term private industry relationships (Camén et al., 2012).A contract awarded under the Act has a 652 TQM 28,4 fixed expiry date, which means that the relationship between the contracting authority and the provider ends when the contract expires.It is virtually impossible to award a well-known incumbent additional contract periods, referring to favourable experiences of working together during the current contractual period; instead a new procurement process must be initiated in order to award a new contract to a provider.Legally, the relationships between contracting authorities and their service providers have a clear beginning and end.
Hitherto, two award principles have been possible: first, choosing the economically most advantageous tender; or second, lowest price.The procuring entity has to declare in advance how tenders will be evaluated.If the economically most advantageous tender is chosen, the criteria are usually a predefined combination of price and quality.In principle, several alternative award mechanisms can be identified, and it can be shown, relying on microeconomic theory, that price-to-quality scoring has a number of shortcomings (Bergman and Lundberg, 2013).A fundamental approach to the relevant theory of scoring auctions has been developed by Asker and Cantillon (2008;2010; see also Dini et al., 2006).A Portuguese study reveals alternatives that are consistent with the European 2004 procurement directive (Mateus et al., 2010).

Quality in European procurement directives
Quality as a legal term has successively gained prominence in the European procurement directives.When the first Council Directive (71/305/EEC, Article 29) gave five examples of criteria when the award of contracts was made to the most economically advantageous tender, quality was not explicitly among them.Instead, the examples were "price, period for completion, running costs, profitability, technical merit", where at least the last example appears to have a strong relation to quality.Later, in Directive 89/440/EEC, Annex III mentioned "quality" and "quality assurance" in the definition of technical specifications, and in Directive 92/50/EEC (Article 36), the examples of criteria under most economically advantageous tender had been reshaped: "quality, technical merit, aesthetic and functional characteristics, technical assistance and after-sales service, delivery date, delivery period or period of completion, price".As already in the 1971 directive, the contracting authority was required to state the award criteria "where possible in descending order of importance" in the contract documents or in the contract notice.Again, in directive 2004/18/EC (Article 53), the list of examples of criteria was reordered and a few additions were inserted: "quality, price, technical merit, aesthetic and functional characteristics, environmental characteristics, running costs, cost-effectiveness, after-sale service and technical assistance, delivery date, delivery period or period of completion, price".Furthermore, the contracting authority was now required to specify in advance also the relative weighting which it gives to each of the criteria.
With the new 2014/24/EU directive (Article 67), the identification of the most economically advantageous tender may include "the best price-quality ratio", to be assessed on the basis of criteria "including qualitative, environmental and/or social aspects".Next, the directive mentions that such criteria may comprise (examples are then given): first, "quality, including technical merit, aesthetic and functional characteristics"; and under second, "organisation, qualification and experience of staff […] where the quality of the staff […] can have a significant impact on the level of performance […]".As to the legal semantics, this means that the quality concept has been moved to a higher hierarchical level instead of being listed in parallel with other concepts, which was the case beginning with the 1992 directive.In addition, the new 653 Quality and legal aspects directive allows for weightings to be expressed by "providing for a range with an appropriate maximum spread" and it also recognizes the situation where a weighting is not possible "for objective reasons", and then (similar to the 1992 directive) "indicate the criteria in descending order of importance".
However, it is certainly not only in the award criteria that a contracting authority may take quality into account.As an extreme, unlikely to occur in practice, if all requirements for quality could be entered in the specifications for a contract, then identifying the best tender and the award might be done on the basis of price alone.

European court rulings on award criteria
National interpretation of European procurement directives is influenced strongly by court rulings at the European level.Four important cases are relevant here.
The European Court of Justice (ECJ) Case C-87/94 Commission v. Belgium (25 April 1996) highlighted that directive text requiring contracting entities to state "in the contract documents or in the tender notice all the criteria they intend to apply to the award, where possible in descending order of importance" is intended to inform tenderers of the features to be taken into account in identifying the economically most advantageous offer."All the tenderers are thus aware of the award criteria to be satisfied by their tenders and the relative importance of those criteria.Moreover, that requirement ensures the observance of the principles of equal treatment of tenderers and of transparency".
Again, according to the ECJ SIAC Construction Case C-19/00 (18 October 2001), the principle of equal treatment implies an obligation of transparency in order to enable compliance with it to be verified.More specifically, this means that "the award criteria must be formulated, in the contract documents or the contract notice, in such a way as to allow all reasonably well-informed and normally diligent tenderers to interpret them in the same way".Finally, when tenders are being assessed, the award criteria must be applied objectively and uniformly to all tenderers.
The Wienstrom Case C-448/01 (4 December 2003) showed that criteria must be accompanied by requirements which permit the accuracy of the information contained in tenders to be effectively verified.Award criteria must be sufficiently clearly formulated to satisfy the requirements of equal treatment and transparency of award procedures.
Or, as stated by the Court in the "Max Havelaar" Netherlands Case C-368/10 (10 May 2012), "compliance with the principles of equality, non-discrimination and transparency requires that the award criteria are objective, ensuring that tenders are compared and assessed objectively and thus in conditions of effective competition".In the same case, the court points to "both the principle of the equal treatment of potential tenderers and the principle of transparency of the award criteria, the formulation of the award criteria being such as to allow all reasonably well-informed tenderers exercising ordinary care to know the exact scope thereof and thus to interpret them in the same way", which is wording only slightly different from the 2001 verdict.

Quality in public procurement of care
A global overview on quality-based purchasing in health care has been made by Waters et al. (2004).In the UK, quality criteria, premia, incentives are now common in local authority social care commissioning (Hardy and Wistow, 1998;Knapp et al., 2001;Rubery et al., 2013).The Swedish policy orientation and change towards an increased reliance on care outsourcing has been described in a European perspective by Pavolini There are few studies of quality and non-contractible quality under public care contracting.Eggleston and Zeckhauser (2002) applied transaction cost analysis to the quality problem.Empirical studies based on such approaches have failed to give clear patterns of quality reduction or increase upon outsourcing from the public sector.Stolt et al. (2011) concluded from Swedish statistics that private care providers seem to emphasize service aspects rather than structural quality factors; Brennan et al. (2012) reviewed a number of studies that indicate lower quality associated with marketisation of care.Mutiganda (2014) takes a pessimistic view based on a case study in Finland of procured public care of the elderly, whereas Bergman et al. (2014), comparing mortality rates, have found Swedish evidence for good effects of competition and private provision of care under contract.There is little agreement among these authors on how care quality outcomes should be measured, and there are no simple answers to the question of how to take into account the prior condition of older individuals when a private provider begins to operate under a particular contract.

The Swedish context
In this study we have chosen to focus on care services in Sweden and procurement of these services.Especially, we are interested in "quality of care" as formulated concept in documents regulating such services.The 290 municipalities are legally responsible for care for the elderly; each municipality can chose to deliver care services in-house (by municipal employees) or to outsource them.Local authorities in Sweden increasingly buy care and the scope has gradually increased throughout the 1990s (Nutek, 2007).In 2005 the purchase of care was estimated to SEK33 billion, corresponding to about EUR3.5 billion (Nutek, 2007).According to data from 2004, all counties put out any part of the care to private providers and only 12 municipalities out of 290 handled care of the elderly entirely in-house (Nutek, 2007).The number of employees in commercial nursing homes increased from less than 1 per cent of the workforce in 1990 to 16 per cent in 2010 (Stolt et al., 2011).

National legislation on care provider selection
If the care service is outsourced, the supplier is selected by means of the public tendering process regulated by the Swedish Act on Public Procurement (LOU), revised in 2007 and implementing the European procurement directives.More recently, the 2008 Act on System of Choice in the Public Sector (LOV) has been introduced, regulating what applies when authorities allow individuals themselves to choose a supplier from among approved suppliers in a system of choice.In systems of choice, prices are established in advance.LOV applies to municipalities and counties when they introduce systems of choice for health and medical care and for social services.Counties must have systems of choice in primary care.For municipalities, and for the counties' other activities, introducing systems of choice is voluntary.Since quality criteria are applied in LOV systems mostly in local, retrospective evaluations of service deliveries, and the Act itself fails to mention quality, our analysis here is confined to LOU contracting, where quality is treated explicitly during the selection process.4.2 "Quality" in Swedish care legislation While it is not within current competence of the EU to regulate health care quality (Hervey and McHale, 2005), Swedish care legislation has developed on its own,

655
Quality and legal aspects successively increasing its reliance on quality terminology.However, the development is not fully consistent, as will be seen: is "patient safety" included in "care quality" or not?
As a legal term, kvalitet has a long history in Swedish statutes, but then referring to the quality only of tangible goods.The first instance of kvalitet of public services was introduced in the late 1970s as a change to the Local Government Act, then limiting local authority powers to devolve responsibility to bodies with representatives of labour market organizations for the quality of local government services (Bill 1978/ 79:188).Sweden has separate legislative frameworks for health and social care, currently and primarily the 1982 Health and Medical Services Act and the 2001 Social Services Act.The county councils are responsible for health-related services, while after a 1992 reform, municipalities are responsible for social care.In the 1982 Act, Section 2a centres on "good care", which means that services must "(1) be of good quality and cater to the patient's need of security in care and treatment; (2) be readily available; (3) be founded on respect for the self-determination and privacy of the patient; (4) promote good contacts between the patient and health and medical personnel".Given this enumeration, it is obvious that "good quality" must be interpreted as subordinated to "good care".
Kvalitet as a term entered the Social Services Act (as Section 7a) in the late 1990s (Bill 1996/97:124, pp. 52-54).The government then stated that quality could not be viewed only in a client/user perspective but should also be assessed in a staff, management and citizen perspective.Furthermore, attaining good quality in social services was seen as requiring a number of ingredients, such as rule of law, individual influence and easily available care and services.The encounter between social workers and clients was seen as core to social work; trustworthy collaboration between the individual and social services staff as well as respect for individual integrity were said to be of great importance for quality.Moreover, it was said in the Bill to be essential that the social services show responsiveness and empathy.Monitoring and evaluating of services, including a voice for those who take part of or use the service efforts, were emphasized; it was also said that the attempts to define good quality in social services led to the conclusion that quality development should focus on all parts of the activities: the organizational structure, the work process and the result obtained.
In general, it emerges that the government distinguished between antecedents or preconditions of quality and quality itself.In the same 1996/1997 Bill, the government developed further an analysis of quality in care for older people, underlining as a quality aspect that relatives should be brought into the planning of care and be supported in various ways.In addition, it was stated that well-functioning collaboration within social services and together with other affected municipal entities was of great importance for quality in the care of older people.However, in the revised and current 2001 Social Services Act (Chapter 3, Section 3), it is only stated that "Measures within social services shall be of good quality.[…] The quality of activities shall be systematically and continuously developed and assured", thus a much shorter text than in the corresponding article of the Health and Medical Services Act.

Review cases involving "quality" in care services procurement
There is no thorough overview available, but a sampling of quality criteria used in local and regional public procurement of care services indicates great variety.Certain municipalities tend to specify quality and associated processes as numerous detailed requirements, relying on assigning points to just a few quality subcriteria when identifying the economically most advantageous tender.In general, between two and 656 TQM 28,4 five quality subcriteria is the usual range, and these often include staff numbers according to educational levels and also processes for quality assurance.There are examples of tenderers having to describe how they are to achieve "meaningful everyday experience" or similar vague concepts, a text that will then be graded by officials or expert readers.
An analysis of contested care procurement decisions in Swedish local and regional government during January 2009-June 2014, where the cases have involved "quality", reveals the principles used by administrative courts.The time period chosen reflects that the revised Swedish national legislation, implementing the 2004/18/EC procurement directive, came into force for new procurement processes on 1 January 2008.
Most of the contested award decisions concern procurement of institutional care services for older people.Some have concerned youth care homes.Degree of detail in the criteria as formulated in the tendering documentation can be contentious, given the consequent risk of arbitrary decisions by the procuring entity.How scale points for quality are assigned during evaluation is often contentious.
We find "transparency" and "equal treatment" to be prominent as principles."Transparency" is interpreted more than once as including "openness and predictability", or as being "clearly formulated", or as a combination of "clarity and predictability".Occasionally, issues of "proportionality" and "non-discrimination" arise, as well as "objective requirements".Reliance by procuring entities on the outcome of earlier treatment might impair "openness and predictability" of an evaluation.Proportionality is discussed typically when compulsory requirements, e.g. for a specified type or item of documentation of bidders' quality systems, have not been fulfilled.
Proportionality is often questioned by tendering firms, but the courts then find mostly that their tenders had not contained information that the tendering documents had stated as compulsory.A frequent context for referring to transparency is when the evaluation criteria have been imprecise; sometimes, the procuring entity is found to have failed to apply its own predefined criteria during tender evaluation, which in its turn could be seen as in conflict with the principle of equal treatment.
The total number of cases is given in Table I.There were 21 cases in administrative courts of first instance, and five cases went further to appeal courts during the period in question.During the second half of the period, the volume of cases in lower courts increased.The recent European Directive 2014/24/EU on public procurement derives (Article 1) the five principles of equal treatment, non-discrimination, mutual recognition, proportionality and transparency from the free movement of goods, freedom of establishment and the freedom to provide services.However, analysis of the EU principles is complicated by the uncertain or inconsistent relation between them, in particular how transparency is related to the other principles (Arrowsmith et al., 2000, p. 73ff.;Prechal and De Leeuw, 2007;Bovis, 2009).

Quality and legal aspects
Table II shows how the administrative courts of first instance have invoked transparency, equal treatment and other fundamental principles in their 21 review cases.Of the five principles invoked by these courts, two occur more frequently: equal treatment and transparency.It is difficult to identify trends during the period.The administrative court of appeal is the court of second instance.There are four such courts, and during the period, they have reviewed five cases involving quality, care and procurement.The administrative courts of appeal invoked equal treatment and transparency in a few out of five review cases, but they often avoided stating the principles explicitly in their decisions, althoughor perhaps becausethe court of first instance had done so.

Conclusion
The analysis presented here shows that the use of quality specifications and quality criteria for the award of care service contracts, as well as the scope for quality improvement, is limited by requirements for transparency and verification in the context of public procurement.Thus it is paradoxical that "quality" as a term has spread widely through legislation during the last decades, while there is little room for exploiting the findings of quality research when public authorities choose to outsource care services.
There are many obstacles to inclusion of a wider range of quality concepts in the procurement of care.Two major concerns identified in this paper are transfer of methods and measures from retrospective studies of care quality into predictive tools suitable for the procurement context, and whether quality criteria derived from earlier care quality research will stand up to judicial scrutiny.Moreover, the legally required expiry dates for awarded contracts reduces the scope for long-term perspectives on quality improvement.Quality concepts and tools appropriate for public procurement criteria are few.Procurement officials may hesitate to include innovative criteria from the care quality literature because of perceived risk of judicial review, preferring to rely on tried and tested simpler practices.
When awarding a contract for care services, the contracting authority has to prescribe and also to predict the quality of the services to be delivered by a successful tenderer.Although much work has been carried out on retrospective measurement of care quality concerning health outcomes, many important issues remain unresolved.When service providers submit tenders, it is not always possible to rely on surveys of outcomes and patient satisfaction during earlier contracts, since lack of standardization and sources of potential bias give rise to uncertainty that makes objective verification difficult.It is a matter of future research to investigate whether the development of  strictly national standards for use in public care procurement creates unacceptable barriers to cross-border service trade, given that there is little international consensus on how to measure care quality.The concept of "improvement science" has recently emerged to provide a comprehensive framework for research focused on health care quality (Marshall et al., 2013).Improvement science is dominated by concepts and models from quality management and is guided by a patient-oriented perspective.This field of research coincides with an increased emphasis on evidence-based medicine and practice as well as a need to develop and spread knowledge more effectively, thereby contributing to quality improvement in health care.The role of commissioning for care quality improvement has been identified in an English NHS setting by Gillam and Siriwardena (2013), but the ability of procuring entities and providers jointly to improve quality during a contractual period is limited because of the restricted scope for modifying a public contract, once awarded to an external provider.Our results refer primarily to the Swedish context but are applicable throughout the European Union.Further studies of the relation between award criteria and public/private collaborative practices for improving care quality during contractual periods are desirable.Although the state-of-the-art of quality management and of public procurement on the other hand is well advanced in each of their fields, a number of questions remain on how quality criteria can be understood in contexts where services are provided by means of a contractual relationship between two parties, while it is a third party, here the care recipient, who is the primary beneficiary.In line with Balbastre-Benavent and Canet-Giner (2011), there emerges a clear need to examine how quality is operationalized, measured, evaluated and acted upon in such contexts, in particular regarding the use of various quality models.
There are at least three important implications for policy makers from this study.First, it appears fruitful to promote contracts of longer duration, also making it possible to reward care providers for a history of quality improvements, or to penalize them for absence of improvement practices.Second, it is urgent that quality concepts and tools are developed further to fit the context of public procurement of care services.Although there are many different concepts and tools available, few seem to be adequate in this context.Third, one way forward may be to benchmark procurement procedures in which a wider range of quality concepts and tools have been used successfully.