8 - Choosing providers

Summary

Choice of health care provider has become an increasingly important feature of health care policy in many countries (Thomson & Dixon, 2006; Bevan & Helderman, 2010; Or et al., 2010), particularly in countries where choice had been previously unavailable. It was introduced as a means to generate competition among providers, thereby improving quality, efficiency and responsiveness, while in some cases choice was also meant to improve equity of access (Reid, 2003; Ringard et al., 2013). Moreover, giving patients and users choice of who, when and what forms of care will be available to them is in keeping with the political declarations and policy commitments towards more person-centred health services (Cacace & Nolte, 2011). Increasingly, individual patients and users are being thought of as consumers and expected to play a key role in their own care, while helping to shape the health system that serves them.

Introduction

Choice of health care provider has become an increasingly important feature of health care policy in many countries (Thomson & Dixon, 2006; Bevan & Helderman, 2010; Or et al., 2010), particularly in countries where choice had been previously unavailable. It was introduced as a means to generate competition among providers, thereby improving quality, efficiency and responsiveness, while in some cases choice was also meant to improve equity of access (Reid, 2003; Ringard et al., 2013). Moreover, giving patients and users choice of who, when and what forms of care will be available to them is in keeping with the political declarations and policy commitments towards more person-centred health services (Cacace & Nolte, 2011). Increasingly, individual patients and users are being thought of as consumers and expected to play a key role in their own care, while helping to shape the health system that serves them. Yet the market-type patient choice has not worked as intended, producing little benefit under specific conditions, which limits its usefulness as a policy tool in public health systems.

The types of choice introduced in European health systems reflect their structural and institutional requirements and the wider policy environment in which they operate. Understanding what motivates the adoption of certain policies can help to predict and evaluate the effects of choice on the intended objectives (e.g. efficiency, quality, responsiveness, equity and personalization). This chapter seeks to systematically explore the different types of choice of provider in primary and specialist care implemented in selected European health systems. It considers policy drivers behind expanding (or in some cases restricting) provider choice in an attempt to better understand the potential benefits and limitations. In examining the rationale for introducing patient and user choice, it critically appraises evidence of how choosing primary and specialist care providers works in a small number of single-payer health systems, where such options were previously limited. It does not consider insurance-based systems, where choice and competition among insurers operate simultaneously, although these could increase or limit patient choice of provider; this is discussed in greater detail in Chapter 9 on choice of health insurer.

The chapter draws on the theories of choice and empirical evidence to analyse the impact by considering the following:

• To what values, besides ‘choice’, does the theory explicitly or implicitly appeal? In other words, is choice a means to achieve other goals such as equity or efficiency or is it a value of its own?

• At what point do conflicts arise (i.e. ‘tensions’, ‘trade-offs’, ‘contradictions’, ‘inconsistency’) between choice and these other values?

• What practical constraints on the exercise of choice are implied (e.g. limits on the amount of complex technical knowledge patients can be expected to process, financial and structural constraints, or asymmetries of power and unwillingness of professionals to enable choice)?

It reviews the evidence of the impact of choice on care outcomes, efficiency, equity and patient empowerment and it puts forward proposals for empirical evaluation of choice in different settings and for models of choice that are closer to the reality of patient care. The institutional conditions necessary to realize effective patient choice are also discussed. The final section argues for recognizing the value and importance of the variety of aspects involved in patient choice, to propose a more balanced framework of choice taking account of users’ diverse needs and the resources they can realistically draw on when making their care-related decisions. The chapter concludes by assessing whether choice in its present forms has contributed to person-centred health systems or can support this objective in the future.

The concept of choice: origin, logic and rationale

The literature on choice is wide-ranging and closely associated with the concepts of freedom, autonomy and democracy (Fotaki, 2006). Normative theories assert that people would exercise choice if they chose ‘rationally’. Policy-makers often rely on these theories to identify the possible outcomes resulting from policies promoting patient choice, the mechanisms which would produce such outcomes, and the required conditions for these mechanisms to work effectively (Fotaki et al., 2006). Additionally, descriptive theories have been used to explain how people actually exercise choice.

Larkin & Mitchell (2016) have argued that choice supported by competition in care services can be seen to be intrinsically a ‘good thing’. It can help in achieving desirable policy goals such as improving efficiency and quality of care by allowing patients and users of services to decide which services best meet their needs, and it provides the means for individuals to acquire a greater sense of control (Iyengar & DeVoe, 2003) and intrinsic motivation. Research from the adjacent fields of social care and elderly care points to the potential positive psychological effects that choice has on increasing one’s sense of personal independence (Arksey & Glendinning, 2007; Sandman & Munthe, 2010), as well as individual physical and mental well-being (Morris, 2006; Rabiee & Glendinning, 2010), which are also relevant in the context of health and long-term care. Moreover, choice is frequently associated with principles of citizenship (Markus & Schwartz, 2010). In short, choice has intrinsic value to patients; it feels good to be able to choose as it enables each person to pursue precisely those objectives and activities that best satisfy their own preferences within the limits of their resources (Saltman, 1994). Choice of provider in primary care settings could promote continuity of care while fostering a trusting relationship between doctor and patient, and contribute to positive health outcomes (Starfield, 1994; Goold, 1999). However, choice can also lead to anxiety, stress and regret (Loomes & Sugden, 1982; Daly, 2012; Baxter & Glendinning, 2013). Patients and service users may avoid exercising choice due to fears of potential or anticipated negative consequences (Ryan, 1994; Goold, 1999; Fotaki, 2014). This raises the issue of the patient’s right to refuse treatment from providers that they did not choose, the option of a second opinion, and the option not to choose.

Choice in the policy context: motivation, drivers and expectations

In health policy, choice was influenced by the neoclassic economics and neoliberal ideology developed throughout the 1970s, now permeating all aspects of society (Chang, 2014). At the same time, choice can also be seen as a response to long-standing demands by patient and user groups for autonomy and for greater control over the health care resources available to them (Barnes, 1999; Fotaki, 2011), enabling them to better manage their own conditions.

The following sections provide a brief overview of choice policies in selected European countries. The focus here is on countries where patient choice of provider was introduced from the 1990s and include Denmark, England, Norway and Sweden.

Denmark

In 2007 Denmark embarked on the most radical reform of the political administrative system since the first democratic constitution in 1849 with the thirteen counties merging into five regions and the 271 municipalities being amalgamated into 98 (Andersen & Jensen, 2010). Denmark has a decentralized system with regions having relative freedom to choose the volume of hospital activity allocated to different specialisms but central government is responsible for legislating these regional initiatives and their financing (Vrangbæk et al., 2012).This semi-decentralized model of health care was expected to stimulate active participation by local people in their own health care and to ensure the responsiveness of the system to the specific needs in each local area (Stubbs, 2015).

In a move to improve the efficiency of the health care system, in 1993 the government in Denmark introduced user choice of hospital. In 2002 the government introduced a waiting time guarantee (‘extended free choice’) of two months from referral, subsequently reduced to one month (Larsen & Stone, 2015). The 2007 reform also gave patients the option of choosing a specialist provider from outside their region if they were unhappy with the treatment offered or if waiting times were too long. However, primary care doctors continue to be responsible for referrals, acting as gatekeepers to the system; for example, in 2011 GPs chose the hospital on behalf of 76% of their patients (Pedersen, Bech & Vrangbæk, 2011). At the same time, introducing patient choice of hospital has led to each region in Denmark offering hospital services according to demand (Vrangbæk et al., 2007). Early evaluations of choice reforms in Denmark (as well as in Sweden and Norway) found limited use of choice by patients due to their lack of knowledge regarding reforms, insufficient support from GPs and limited information, although there was an upward trend in the uptake of choice (Vrangbæk et al., 2007).

England

In England a series of legislative changes have introduced increased levels of choice in all aspects of patient care in the National Health Service (NHS) over the last decades. The first attempts to introduce elements of choice date to experiments with quasi-market reforms aiming to introduce competitive mechanisms in health care services in the early 1990s, the so-called internal market (Le Grand, 2003). The policy sought to make services more responsive to users’ needs by giving health authorities a budget to contract services from hospitals, which had to compete for contracts. At the same time, general practitioners’ practices were encouraged to take up a portion of the budget to purchase some services for patients on their own lists, again requiring providers to compete (GP Fundholding). This enabled the referral of patients to a hospital of their own or their GP’s choice. However, in practice this kind of choice was not vigorously pursued, resulting in half-hearted and isolated responses rather than a choice revolution (Tuohy, 1999). If anything, choice of provider is likely to have diminished during this time, because the internal market set up contracts with specific hospitals, so that GPs and patients could only choose from among these options (Robinson & Le Grand, 1994; Fotaki, 1999).

In 2003, under the New Labour government, patients in England (Scotland, Wales and Northern Ireland have followed different non-market approaches since devolution) were offered a choice of five providers for elective treatments, such as hip or cataract surgery; this was expanded to about 150 approved providers from public, private or not-for-profit sectors in 2006 and took the form of an ‘extended choice network’ in 2008. Although there was no evidence of strong public demand for choice of hospital as such, there was considerable public concern about waiting times, and the newly introduced choice options particularly benefited patients in areas where existing services were poor and had long waiting times (Coulter, Le Maistre & Henderson, 2005; Dawson et al., 2004). In addition to improving quality and efficiency, the policy of offering choice to all was intended to extend the opportunity to choose different providers beyond the articulate and those who could afford to access private health care (Department of Health, 2003). This second attempt at creating a market within a single payer system was justified on the basis of having to keep up with the presumed demands of patients who were increasingly thought of as consumers and who were expected to ‘reveal their preferences’ through choice (Le Grand, 2007). There was, however, little evidence of whether market-based choice could work in publicly funded health services.

The 2012 Health and Social Care Act further expanded the commitment to provider choice. From 2015 patients may choose to register with a GP practice outside the GP-practice’s catchment area, although this scheme is voluntary for GP practices. In addition, patients, along with their GPs, have been given the possibility to choose the best services for their needs from an NHS, third sector or independent private sector provider as long as these are approved by the commissioners.

Norway

In Norway the structural reform of 2001 divided the responsibility for health care between the state, the four regions and the 428 municipalities (Lian, 2003). The reform also introduced choice of GP to improve accessibility, continuity and quality of primary health care (Ringard et al., 2013), especially for older people and those with chronic health problems, while addressing the problem of low recruitment of GPs (Holte et al., 2015). It introduced free registration with any physician licensed by municipalities, including those in private practice, to work in primary care (capacity permitting) as it was expected that this would strengthen the physician’s personal responsibility for continuity of care and availability (Grasdal & Monstad, 2011). In parallel with the reform, the capacity of primary care doctors increased by about 10% (Iversen & Lurås, 2011). Evaluations found that the reform has been popular among doctors and that the population has become more satisfied with access to care. There are also indications of improved equity in access to specialist services (Grasdal & Monstad, 2011). However, while mostly beneficial for improving equity of access, continuity of care and patient satisfaction, it is difficult to disentangle these outcomes from the increases in capacity in primary care. Also, challenges remain regarding the integration between independent private primary care doctors’ services, in particular for those working mostly in small practices, and other primary and specialist care activities. Finally, there is recognition of a need for more patient orientation, and more decentralized services close to where patients live to reduce costs (Rørtveit, 2015).

Sweden

The Swedish health system is similar to those in other Nordic countries in that it shares the same commitment to universal access and equality, and it is characterized by a decentralized structure. It differs, however, with regard to primary care, with an overall lower investment and fewer providers; also, GPs do not act as gatekeepers (Anell, 2015). Choice of health care provider has been introduced gradually since the 1990s (Fotaki & Boyd, 2005). The 1994 national family physician reform introduced an element of choice in primary care but this was discontinued the following year when the government changed from conservative to social-democrat over a dispute about extending choice to private providers in primary care.

Beginning in January 2010 the government made it compulsory for county councils to provide patients with a choice of primary care provider and freedom of establishment for those private units that did accept requirements and payment principles determined by county councils (Anell, 2015). Patients had to be given the option of a public or private provider, with county council funding allocated according to the individual patient’s choice. In keeping with the decentralized model of the Swedish health system, ten out of the 21 counties already had some arrangements in place at that time (Anell, 2011). In 2015 the government introduced unrestricted choice of provider in primary care and outpatient specialist care. These latest reforms led to the establishment of over 270 new private primary care practices operating for profit throughout the country, with some researchers foreseeing potentially negative impact on equity (Burström, 2002), which will be discussed below.

In specialist care, most county councils have adopted some form of public competition since the mid-1990s, particularly in Stockholm and other urban areas (Winblad, 2008), offering increased choice of hospital. However, only a minority of patients (who in many cases could self-refer themselves) and physicians exercised their right to choose a hospital at this point. Research has found that referrals are mostly based on medical grounds while the patient’s wish to choose a specific provider is considered less important (Burström et al., 2017). The 2015 health reform supports choice in outpatient specialist services and choice related to second opinion for treatment of life-threatening diseases nationally; there is no national policy related to choice of inpatient care in general.

However, choice in primary care remains a controversial issue in the Swedish debate, although much of the criticism does not revolve around choice as such but rather concerns the free establishment of private providers, which has found increasing support over time.

In summary, policies introducing patient choice of provider in various countries differed in content and context, and reform agendas have changed over time. For instance, in Norway and Sweden the emphasis was on improving access to primary care, while choice policies in Denmark and England were driven, at least initially, by policy concerns about waiting times, although in England the focus progressively shifted towards introducing competition in specialist services. A common feature for all systems described was the growing importance of individualized market-based forms of choice, although this may now be changing with the shift towards person-centred care as will be discussed in the concluding section of this chapter.

Implementation and the evidence of impact: how far has provider choice delivered on its promises?

Do patients want choice and feel empowered by it?

Choice has been used not only as a policy instrument for achieving the policy goals of efficiency, quality and equity, but also to promote service user empowerment and autonomy (Fotaki, 2011). The development of the active, critical consumer is considered an important end in itself, even if people cannot always act as a perfectly informed agent.

A 2012 survey of patients’ involvement in health across the EU found that some expressed a desire for a more balanced relationship with their doctors, which would allow patients to participate more actively in their care. This finding was particularly strong for younger and well-educated people, those with chronic conditions, and those living in western Europe (Eurobarometer, 2012). However, these observations do not necessarily imply a demand for more choice. At the same time, a review of how people use choice in public services in England found that ‘having choice’ was seen to be important by the vast majority of respondents (Boyle, 2013). Those with lower education were more likely than those with at least a degree to respond positively to having this opportunity, although people from disadvantaged backgrounds may be less able to exercise choice and are therefore less likely to benefit from it.

The individual characteristics and circumstances of patients and users of health services are likely not only to influence their choices, but also to determine whether they exercise choice at all. For example, an evaluation of the London Patient Choice Project, which was established to offer NHS patients in England more choice over where and when they receive treatment, found that old age, low educational attainment, family commitments or low income all had an impact on patients’ choice of a non-local hospital, meaning that they were less likely to travel to a non-local hospital if they were offered the choice (Burge et al., 2004). Distance remains an issue for many people and a lack of public transport can make choice difficult for people who are unable to afford a car (Dixon et al., 2010).

The London Patient Choice Project also found that differential access to information for people with low educational attainment and those for whom English is not their first language could lead to variations in uptake of choice (Dixon et al., 2010). A related empirical study concluded that patients in England who are not highly numerate and health-literate were less able to use the available information to make complex decisions about hospital choice without some expert support (Boyce et al., 2010). Comprehending the options and making trade-offs between quality, safety, patient experience and location posed difficulties, and the way information was presented made a difference to how patients used it. Similar difficulties were observed in other health systems such as in Sweden regarding accessibility (Anell, 2015; Victoor et al., 2012) and the role of information when choosing hospitals in Denmark (Birk & Henriksen, 2012; Pedersen, Bech & Vrangbæk, 2011). This highlights the need for adequate support structures to be put in place if choice is meant to work for all. Evidence from shared decision-making suggests that structured support may help reduce health inequalities when the intervention is adapted to disadvantaged groups’ needs (Durand et al., 2014) (see also Chapter 11).

The type and degree of choice patients want and value is not self-evident either. Research by the UK-based consumers’ association Which? in 2005 found that choice was seen to be of relatively low priority for many people compared with other aspects of service delivery in the NHS. The majority of respondents were more concerned with having safe, good quality services provided locally, and not so much about having diverse providers to choose from (Which?, 2005). Patients tend to favour a provider they know and trust and opt for choice only when no such provider is available (Taylor-Gooby & Wallace, 2009). The evidence from across Europe further suggests that in addition to satisfaction with the health system (Eurobarometer, 2012), perceptions of choice are influenced by an individual’s personal health situation, age and gender. For instance, early studies from Sweden found that older patients appeared to be both interested in choice of primary care doctor and happy about the amount of choice offered, while highly educated young people, and women in particular, were found to both exercise and favour choice more when compared to other population groups (Rosén, Anell & Hjortsberg, 2001; Anell, Rosén & Hjortsberg, 1997). These age and gender factors were also confirmed for England. Overall, evidence suggests that patients appear to be more interested in choosing treatments especially when they are chronically ill and have knowledge about their disease (Coulter, 2010). The willingness to engage in treatment decisions is, however, often influenced by the severity of the medical condition and the complexity of the procedure involved: the more life-threatening the disease and technologically advanced the treatment, the lower tends to be the patient’s desire for choice (Fotaki et al., 2008). Patients’ preference for choice might also be different in primary care as opposed to specialist services but there is little comparative research on these issues.

It is important to note that, for example, in England retaining the public and universal aspects of the health system has tended to be of greater concern than demands for choice, and the marketization of public services has been considered as a threat to universal and free provision of health services provided by the NHS. When ranked on a scale of one to five in a 2010 MORI survey, fairness in public services came first, while choice and the personalization of services was last for the majority (63%) of the British population (2020 Public Services Trust, 2010). Similar public concerns about the impact of recent privatization have been noted for Sweden, with evidence suggesting that while the public may be in favour of provider choice, they were sceptical about profit incentives in tax-funded markets and about the payment of dividends by health care providers to their owners (Anell, 2015).

In summary, patients’ willingness to exercise, and demand for, choice differs by age, gender, social characteristics and personal circumstances. Although service users might be attracted to the idea of having a choice in general, research shows that not all populations are equally able to exercise choice, as will be discussed next.

What are the impacts of choice on equity?

As noted earlier, one aim of individual patient choice of provider is to improve equity by removing barriers to access, although there are other important equity considerations such as improving health outcomes for those in greater need (Fotaki, 2010). Thus, the introduction of choice of primary care provider in Norway and Sweden was, at least initially, intended to improve access to primary care, while elsewhere relevant policies served to provide greater choice of specialist care, such as in Denmark and England. In England the introduction of choice of elective treatment in 2003 described earlier was specifically intended to enhance equity of access by permitting those unable to afford private health care a choice of provider already enjoyed by those who could afford to pay for it (Reid, 2003).

In Sweden some population groups in urban areas enjoyed improved access to primary care because of the increased number of private providers entering the market following the choice reforms (Anell, 2015; Dietrichson, Ellegård & Kjellson, 2016). However, the higher number of new primary care providers in densely populated urban areas might have negatively affected equality of access for patients outside urban areas (Burström et al., 2017). Also, evidence from some county councils from the 1990s suggested that relatively healthy people benefited more following the choice reforms than did others in terms of access to primary care (Saltman, 1994). This risk of inequality might be higher for specialist services (Devaux, 2015) as they generally tend to favour the better off while primary care is more pro-poor (Grasdal & Monstad, 2011). However, in the case of Norway, introducing choice and contracting of a higher number of primary care doctors operating in the private sector improved patient access to specialist services as well as decreasing the marginal effect of income on utilization (Grasdal & Monstad, 2011).

Evaluations of pilots introducing choice of hospital in various regions in England in 2002–2003 found that age, class, income and family obligations affected patients’ ability to travel to a non-local provider, and therefore their choices (Burge et al., 2004). Other studies reported no evidence of inequalities of access for patients participating in the same projects but these studies did not consider patients who were not offered choice (Coulter, Le Maistre & Henderson, 2005; Dawson et al., 2004). In many cases choice was only offered to a minority of patients, for example excluding older and sicker patients (Appleby, Harrison & Devlin, 2005).

Empirical research on the effects of reforms introduced into the English NHS during the 1990s suggest that socioeconomic differences that lead to variations in health care utilization are deeply ingrained, and that in the context of universal and comprehensive health systems small doses of ‘quasi market’ competition (with a few providers competing) modifies providers’ behaviour while having little or no effect on socioeconomic inequalities in health care (Cookson et al., 2010). Nevertheless, there is a risk of creating new inequalities over and above those that already exist and this might differ by the type of service and setting (primary or specialist care). This is because some patients receive preferential access and treatment under certain schemes, as was the case with the patients of GP fundholders in the UK (Manion, 2005).

There is also evidence that physicians are likely to change their behaviour to fit the market, which could benefit some patients more than others. For example, following the introduction of competition and choice in Sweden, GPs and specialists reported that these changes had enhanced their autonomy, income and employment prospects, while at the same time they could reduce their commitment to the normative foundation of the system, that is ensuring equal access according to clinical need (Bergmark, 2008). Thus, although choice and privatization might have improved access to primary care in Sweden in general, the reforms have also raised serious questions regarding their impact on equity, leading to calls for future regulation of providers (Anell, 2016). The full impact of choice on equity cannot be assessed without suitable data on quality or outcomes of care, which is currently lacking in the Swedish context. A study from the Netherlands found that surgeons felt they had to ‘sell themselves’ by advertising or marketing their performance when patients had the option to choose between them (Dwarswaard, Hilhorst & Trappenburg, 2011). It was noted that better performance would be easier to demonstrate for relatively minor routine conditions, such as varicose veins and hernia, which represented a significant source of income for hospitals; therefore, surgeons began to pay more attention to patients with such conditions, following patients’ preferences rather than medical need. Recent work by Visser et al. (2018), also in the Netherlands, noted that the introduction of consumerist communication technology in health care would assume a ‘universal individual’, creating tensions for health care professionals who aim for equal treatment of all patients, and which has paradoxically led to new inequalities among patients with differential abilities to access technology.

In summary, countries differed in their objectives by which introducing choice of provider should improve equity. There is evidence of choice leading to improved access to certain services, for some populations, and in some settings. Yet there might also be different and potentially negative consequences for equity where there is little additional support offered to those who are less able to exercise the option of choice. Indeed, patient choice of provider might exacerbate inequities in access due to pre-existing inequalities in income, class and individual circumstances, with the additional risk of individual choice leading to new inequalities.

Does choice improve the quality of care?

Quality is an intrinsically difficult concept to define, with definitions including a wide range of dimensions and indicators of process, such as waiting times, as well as the outcome of care, such as patient experience (Berwick, 2002). The economic assumptions driving choice policies in public systems where prices are fixed, such as the National Health Service in the UK, is that providers will strive to attract patients by improving quality if the market contains a sufficient number of competitors: hospitals in these instances will compete in terms of quality and not price (Gaynor, Morreno-Serra & Propper, 2012). Empirical studies measuring the relationship between competition and quality of care suggest that there are positive as well as negative consequences, and sometimes neither. For example, in the Netherlands there were reports of perceived decreases in quality of care after the introduction of regulated competition (Dwarswaard, Hilhorst & Trappenburg, 2011; Victoor et al., 2012; see also Chapter 9). The estimated impact of competition on quality of care has been considered to be small in other health systems such as in England (Dixon et al., 2010) and Sweden (Anell, 2015).

For example, empirical evidence from England found an association between the introduction of choice policies and improvements in the quality of care. For example, Cooper et al. (2011) demonstrated that death rates from acute myocardial infarction were slightly lower in geographical areas where there was greater potential competition between hospitals facing fixed prices. These competitive pressures were attributed to the effects of patient choice initiatives, although patients exercised choice mainly in relation to elective treatment, which was not the subject of the evaluation in this specific study (Pollock et al., 2011). Conversely, an evaluation of the impact of the internal market in England in the 1990s using negotiable prices found that greater competition was associated with higher mortality among patients with acute myocardial infarction (Propper & Burgess, 2004). However, as many other factors besides competition influence the quality of hospital services, including price structure, payment methods, internal organization and pre-existing culture, in addition to quality regulation systems and protocols, it remains difficult to clearly attribute observed outcomes to choice policies per se (Sutton et al., 2012; Ferlie et al., 2004).

There is also evidence from market-based systems such as that in the USA of providers tending to compete on quality by introducing expensive technology (particularly when they do not face hard budget constraints) that can lead to higher costs and squeeze out cost-effective care (Pauly, 2005). This appeals mainly to doctors but it also aims to attract patients by offering novel and usually more expensive treatments and diagnostic procedures.

Overall, most research on the impact of choice and competition in relation to quality is conducted in the context of specialist care. The available evidence of the impact of choice in terms of improved outcomes remains inconclusive. Alongside methodological weaknesses, reported improvements tend to be small or were derived from a very narrowly defined set of clinical indicators. Moreover, studies are often conducted under specific conditions that may not be universally applicable.

There is a lack of comparable studies in primary care, along with a lack of suitable data on quality of care besides patient satisfaction surveys. There is some, albeit limited, evidence for Sweden, and studies have failed to find a substantial impact on the quality of care following the introduction of patient choice in primary care (Dietrichson, Ellegård & Kjellsson, 2016; Fogelberg, 2014) and few patients compare providers before making their choice (Glenngård, Anell & Beckman, 2011; Swedish Agency for Health and Care Services Analysis, 2013; Wahlstedt & Ekman, 2016). There is some indication of improved patient satisfaction (in areas with alternative providers) but there is no general trend suggesting that satisfaction or quality of care has improved overall and as a consequence of the choice reforms alone (Gaynor, Morreno-Sella & Propper, 2013; Gaynor, Propper & Seiler, 2016; Gravelle et al., 2014; Moscelli, Gravelle & Siciliani, 2016) as these will often depend on precise institutional arrangements (Cellini, Pignataro & Rizzo, 2000). Population surveys show that trust in primary care has increased between 2009 and 2012 (Anell, Glenngård & Merkur, 2012) but this trend reversed after 2013. In Norway there have been improvements in patient satisfaction in most aspects since the 2001 reform, but it is not clear to what extent this can be explained by the parallel capacity increase in the number of primary care doctors (Iversen & Lurås, 2011).

On the whole, there is little robust empirical evidence that choice of provider leads to substantial quality improvements. Studies on increased patient choice of hospitals have shown mixed effects on health outcomes.

Does choice improve the efficiency of health care?

We noted in the introduction to this chapter that competition between health care providers has been considered central to improving the efficiency of publicly funded health systems. Efficiency in this context can be defined as the optimal allocation of scarce resources and providing the best value for money (Palmer, 1999). It is seen by some as a solution to rising costs and demand (Le Grand, 2007). Choice, then, can – at least in theory – enhance efficiency by favouring providers who offer better services at lower cost (Bartlett, Roberts & Le Grand, 1998). However, the principles of the commercial sector do not readily apply in health care. This is because service users often have to base their choices on insufficient information (Arrow, 1963) or they may be induced to make choices that suit providers, especially when there is a financial incentive to do so (Rice, 2002).

The evidence of whether patient choice of provider does positively impact efficiency remains mixed. For instance, one review of the impact of choice in England concluded that any increases in efficiency that were observed after the introduction of related policies (as measured by, for example, an increase in the number of elective surgery patients treated as day cases, a decrease in the length of inpatient stays, or reductions in avoidable admissions) could not be attributed to patient choice alone as there were also other policies and trends which could have encouraged such results (Civitas Institute, 2010). There is limited evidence from Sweden suggesting that implementing provider choice may be associated with an increase in costs (Bergmark, 2008).

In the context of specialist services, when these involve fee-for-service payments, providers may classify treatments as being more risky and expensive in order to generate additional revenues in these instances, a practice which can be found in the market-type based health systems (Kuttner, 2008). This gaming of the system, combined with an increased supply of specialist services that followed the introduction of competition, choice and per case payment in the 1990s in Sweden, made strategic priority setting and resource allocation by county councils more difficult, creating new threats to efficiency (Bergmark, 2008). Another form of gaming was observed under the internal market in the NHS in the 1990s where hospitals competing with each other became intentionally less productive shortly before obtaining trust status, so as to look more efficient under the new arrangements when compared with those that did not (Söderlund et al., 1997).

In summary, the evidence that patient choice of provider leads to greater efficiency is not persuasive because it is difficult to single out a specific policy initiative as the ‘cause’ of a specific ‘effect’. Any measured efficiency gain may also be achieved by gaming the system and compromising quality. Furthermore, introducing competition and choice between providers to improve efficiency relies on an implicit belief that existing public providers with restricted choice are intrinsically inefficient (and private providers with extended choice for patients are intrinsically efficient), which has little basis in evidence, although non-market systems may create their own inefficiencies due to the sub-optimal allocation of resources.

The limitations of provider choice: policy lessons

The key implication that policy-makers need to consider concerns the usefulness of provider choice for promoting the goals of public health systems and for supporting person-centred care. The introduction of provider choice in single-payer systems such as Denmark, England, Norway and Sweden were shown to produce some benefits for some population groups in some settings, in particular those who are most likely to benefit from a higher supply of providers and those who are willing and able to use the information available. At the same time, there are a number of undesirable effects, especially in specialist and/or hospital services. Evaluations of choice policies in health care find that they rarely lead to more social efficiency or increases in welfare (Schwartz & Cheek, 2017) at a population level. This is because of the complexity of the choices involved and patients’ unequal ability to navigate these. Choice policies in health care may also negatively impact on equity, and may fail to meet patients’ interest in improving the quality of services provided locally (which patients prefer), once policy assumes their willingness to travel afar to find the best provider since patients with caring commitments (Burge et al., 2004) and those who do not own a car are less likely to travel any distance (Dixon et al., 2010).

The theory of market imperfections in health care considers how choices are actually made, and demonstrates the problems of replicating simplistic economic choice models in health care. People’s ability and willingness to make choices is influenced by their beliefs, cultural values and expectations as well as by their life circumstances, personal characteristics and experiences of health care services (Fotaki et al., 2006; Visser et al., 2018). People are seldom rational choosers, least of all in relation to health or care services, a reality that psychologists and economists both acknowledge (Kahneman & Tversky, 1979; Thaler & Sunstein, 2008; Hansen et al., 2015).

Although these limitations may apply more to some types of care than others, choice is often impaired in health care and cannot on its own promote person-centred care for all. Patients often lack the information needed to make meaningful choices about providers and their care, and there is therefore a need to better understand the information needs of people to help their choices or indeed, where people are unwilling to exercise choice, to provide appropriate support. At the same time, as it was noted earlier, there are population subgroups that are more motivated and better able to make informed choices about their own care, such as people with long-term conditions. This can be turned into a strong argument for choice in primary care, where the role of the service is to support them in their choice but with a default option available for those who do not want, or who are unable, to choose.

Although it is possible to treat people who seek support from the health service as customers, this may not be compatible with ways of thinking and acting that are crucial to good quality health care. Good care grows out of collaborative and continuing attempts to attune professional knowledge and technologies to diseased bodies and complex lives (Mol, 2008). When making complex health decisions, patients often rely on their intuition and emotions, which also involves the avoidance of regret (Ryan, 1994; Loomes & Sugden, 1982) as well as trusted networks (Pescosolido, 1992), rather than the impersonal data.

Framing the issue of choice in the context of market competition roots it in assumptions originating in neoclassical economics about humans as disembodied and socially disembedded individuals pursuing their self-interest. This leads to a significant narrowing of the concept of choice, and of the users of health services as rational ‘choosers’ exercising their preferences. Choice and independence are indeed powerful concepts, but interdependency is an essential part of social life and never more so than in relationships involving care (Fotaki, 2015).

Innovations and future developments: implications for person-centred care

The desire of service users for more autonomy and greater control over the health care they receive should not be discarded along with the consumerist market model but rather should be seriously addressed on its own terms (Beresford, 2008). In many ways patients are obliged and increasingly willing to make health-related decisions as co-producers of their health together with health care professionals, and as citizens and community members they participate in co-designing health services. Often these choices are governed by social values and the need for cooperation and recognition, not by mere self-interest (Taylor-Gooby, 1999); patients’ involvement is most effective when used as part of a broader ethos of care (Health Foundation, 2012).

Various practical ways of strengthening elements of ‘voice’ in the system should be considered. Enabling people to use voice, beyond the option to exit and choose a different provider, would allow patients and service users to assume responsibility for their health in ways that are different from the individualistic personalization agenda. The example of co-production of public services, with the users of services as active asset-holders of resources rather than passive consumers, demonstrates the benefits of promoting collaborative rather than paternalistic relationships between staff and service users, where the focus is on the delivery of outcomes rather than the services (Needham & Carr, 2009). The degree to which patients and professionals each hold agency for these co-produced outcomes varies widely, but the concept has profound implications for improving health care quality, safety and value.

Overall, the expansion of choice can empower patients, if it is appropriately linked to their direct participation in decision-making processes. This can occur, for instance, by involving them (individually or collectively) in managing their health resources as in existing co-production schemes (Batalden et al., 2016; Baker, 2010; Needham & Carr, 2009) and by assisting them in deciding what is best for them (Barry & Edgman-Levitan, 2012). Supporting patients in the process of choosing can help overcome the information disadvantage and some of the socioeconomic barriers associated with market-based choice. In all cases, patients and service users should be clear about what is involved in their choices, and the potential consequences, not just for their immediate care but for the future provision of care for them and their families and community.

Choice is also a key value embedded in contemporary approaches to framing the delivery of health care services as can be seen in the emphasis placed upon it and its integration in the movement towards person-centred care. The idea of person-centredness implies that an individual’s decisions and preferences are at the heart of all their interactions with health care practitioners, who are expected to support these despite the degree of confusion over what is meant by ‘person-centredness’ and the types of changes that are needed to promote it (see Chapter 2).

Conclusions

Promoting market-based individual patient choice, first introduced in the planned health systems in the UK and Sweden in the 1990s, has become a standard health policy objective in health and social care in many other countries. Two rationales typically make the case for consumer choice in health systems. First, it is as a method to stimulate providers to improve the quality of services offered; and second, as a benefit in its own right that is valued and desired by patients. Moreover, in many western societies choice is increasingly seen as an expression of an individual’s unique identity (Schwartz & Cheek, 2017).

However, the idea of offering patients choice and making them act as consumers in a market-place has serious limitations when applied to health and social care. Overall, policies based on these assumptions have been found wanting, for both theoretical and empirical reasons. People have various needs, which are further augmented in times of dislocation, vulnerability and stress, and many cannot or do not want to make such complex choices themselves. Choice works best in instances where it is supported by trusted people and with the help of decision aids.

Furthermore, any impact on quality and costs will depend on the precise institutional setting in which choice of provider is implemented. Reliance on competition to promote choice in health carries the risk of reproducing existing inequalities while simultaneously introducing new ones related to health literacy and access to information linked to users’ educational status and ability to pay.

Choice is more likely to work if policy design reconsiders what it means and what types of choice are important to patients. Policy design should be informed by the social and psychological factors affecting individuals’ health-related decisions, such as their previous experience and social bonds as family and community members. To achieve this, policy-makers might consider interdisciplinary frameworks and alternatives to market mechanisms which could offer a more balanced view of how choice works and what choices matter to patients if they are to promote person-centred care.

Note

I would like to thank Professor Colin Leys, Dr Sally Ruane and David Roland for commenting on an earlier version of the policy report on this topic produced for the Centre for Health and the Public Interest. I would also like to thank Professor Anders Anell for his valuable comments and insights offered, in particular on Sweden and Norway. Any errors are mine alone.