In its World Health Report 2008, the World Health Organization (WHO) advocates in favor of a central role for primary care in health care systems. The WHO defines the specific features that should characterize primary care to ensure improved health and social outcomes: person-centeredness, continuity, comprehensiveness, and integration. Person-centeredness is about adapting medical advice to individual life circumstances. Continuity would allow the best use and sharing of information between individuals and primary care providers (PCPs). The concepts of comprehensiveness and integration stress the multiple roles of a PCP: health promotion and prevention, diagnosis and treatment or referral, and chronic or long-term care. As for the organization of health systems, the WHO promotes to switch the entry point to the health system from hospitals and specialists to PCPs.
Evidence has been reported about favorable medical outcomes in systems with an emphasis on primary health care. For instance, continuity of care contributes to lower all-cause mortality. Person-centeredness is responsible for an improved quality of life and increased treatment compliance. Comprehensiveness contributes to better health outcomes and to fewer patients admitted for preventable complications of chronic conditions. The effect of the supply of primary care is not so clear-cut. A superficial observation negatively relates the number of PCPs per capita to the mortality rates. However, more PCPs are expected to work in areas with a worse case-mix of patients. Accounting for this simultaneity effect, the negative relationship between the number of PCPs and the mortality rates disappears.
On top of the evaluation of the benefits of primary care, the contribution of health economics to the trend toward strengthening primary care has been taking the following forms: to analyze possible organizations to bring primary care upfront; and to think of ways to make individuals and health care providers adhere to the aim of strengthening primary care. A lot of attention has been devoted to the gatekeeping role of PCPs and to the incentives of PCPs and patients to adequately use primary care as an entry point to the health care system. The selection of patients and PCPs into different primary care organizations is also an interesting issue even though it has been less debated so far. The notions of primary care, gatekeeping, incentives, and selection are therefore the core of this article. The aim is to understand each of these aspects independently from each other as well as their interactions with each other.
The remaining of the article is structured as follows. The following section discusses gatekeeping versus direct access to specialists. The next section reports on patients’ incentives to use primary care as an entry point to the health system. It also analyzes the PCPs’ incentives to fulfill their roles. The next section tackles the issue of selection that appears when several organizations for providing primary care or accessing specialized care coexist. The penultimate section extends the discussion to the supply of specialized care. The final section concludes the article.
Gatekeeping
The most obvious way to bring primary care upfront is to forbid patients’ direct access to specialists. The PCP is thereby empowered with a gatekeeping role. Patients can access specialized care only after the PCP has issued a referral. The WHO has stressed the importance of the gatekeeping system as an organizational model to structure health care. Gatekeeping is typical of the health care systems in Denmark, Finland, Ireland, Italy, the Netherlands, Norway, Portugal, Spain, and the UK; whereas Austria, Belgium, France, Germany, Greece, Iceland, Luxembourg, Sweden, and Switzerland allow free access to most medical specialists.
Empirical comparisons between gatekeeping systems and systems with free access to specialists repeatedly report the following three effects. Gatekeeping decreases patients’ satisfaction, even though it earns a better acceptance in countries where specialists are in short supply as in the UK. Also, gatekeeping is significantly associated with a lower utilization of health services and lower expenditures.
To appreciate the influence of gatekeeping on the utilization of medical services and on the resulting expenditure, it is important to understand the possible relationships between gatekeeping, medical utilization, and medical expenses. Gatekeeping is primarily meant to limit the use of expensive specialist services to the necessary cases only and to avoid them for patients needing primary care only. Therefore, a decrease in utilization and expenses can reflect an efficient use of medical services only if it decreases unnecessary visits to specialists. Empirical evidence on unnecessary care under free access to specialists is therefore needed to support this relationship; otherwise it is admitted to think that gatekeeping can cause a decrease in necessary specialized care too.
Another aspect of the relationship between gatekeeping versus free access, utilization, and expenses is selection. Gatekeeping in the public system coexists with free access in the private sector in countries such as Spain and the UK whereas they coexist in the private sector in Switzerland and in the USA. When both gatekeeping and free access systems coexist, the authors expect gatekeeping to attract members who are healthier on average than the free access system does. This selection process would automatically result in lower medical utilization and expenses for the gatekeeping system, independently of a possible gain in efficiency. Limited evidence is available about the existing efficiency effect, once the selection bias is accounted for.
The effects of gatekeeping versus free access are also dependent on the financial incentives they are associated with. For example, gatekeeping is often associated with PCPs’ financial incentives to limit referrals to specialists, whereas system with free access provides generally little incentives of this kind. Therefore, the lower medical utilization and costs observed in gatekeeping systems might be due to the financial incentives rather than to the gatekeeping barrier itself. The empirical literature on gatekeeping versus direct access to specialized care so far has not disentangled the effect of both patients’ and PCPs’ financial incentives from the effects of constrained access to specialists.
Some theoretical arguments help comparing gatekeeping with direct access considering the optimal provision of incentives to PCPs. The next section discusses how to provide adequate incentives to PCPs. At this stage and for the sake of comparison between gatekeeping and free access, it can be mentioned that incentives to PCPs are meant to minimize two types of possible errors: the use of specialized services when unnecessary (type-I error) and the lack of specialized treatment when necessary (type-II error). Without adequate incentives to PCPs, gatekeeping is expected to generate more type-II errors than type-I errors. Conversely, free access may result in more type-I than type-II errors. With adequate (though costly) incentives to PCPs, one can minimize both types of errors in a gatekeeping system because decisions are in the hands of the PCP. However, type-I errors would remain in a free access system because those are independent of the PCPs’ decisions. Therefore, when optimal incentives are provided, gatekeeping performs better, in theory. Conversely, free access might perform better when the patients’ pressure to refer anyway is high or when the quality of the patients’ self-health information is either highly accurate (in which case the patients’ self-referral is very efficient) or weakly accurate (in which case the PCPs’ financial incentives are very costly).
Incentives
Incentive mechanisms are increasingly popular in the health care sector to deal with the inefficiencies caused by asymmetric information between physicians, patients, and third-party payers. Incentive mechanisms exist for both patients and physicians to encourage the adequate utilization of health care services. The incentives for the efficient utilization of primary care versus specialists is discussed here, starting with the incentives directed to patients and following on with those directed to PCPs.
In a free access system, patients can be motivated to access PCPs first by making them bear an additional out-of-pocket payment when they choose to directly visit a specialist. This system is in use in France since 2005 as well as in some health maintenance organizations in the USA. It is a soft version of the gatekeeping system. However, financial incentives directed to patients bring the issue of equity in the access to health services, which was absent from the pure gatekeeping system. The financial incentives directed to patients may also limit the scope for quality (non-price) competition between specialists. Indeed, if the patients’ co-payment is proportional to a specialist’s fee, the best reply of the specialist to an increase in patients’ co-payment is to decrease his fee to sustain demand. The specialist earns thereby less revenue per patient, which results in lower incentives to invest in costly quality. However, for these financial incentives to prove efficient in terms of utilization of primary versus specialized care, empirical evidence is needed about the actual behavior of patients with and without out-of-pocket payments. The French experience with out-of-pocket payments has proven disappointing because direct access to a specialist was very limited even when no such payments were due.
As for the PCPs, incentives are generally provided through their remuneration scheme and they are expected to influence their referring behavior. The aim is to limit the discretionary and unnecessary referrals to specialists. Many authors have written about the incentive properties of the most traditional payment schemes. If one ignores the issue of referrals, the classical analysis of traditional payment systems yields the following conclusions. Fee-for-service (FFS) payments may encourage physicians to provide too many medical services to maximize their revenue. Capitation may lead physicians to limit either the amount or the quantity of the medical services they provide. FFS can thus be responsible for excessive health care costs and utilization, whereas capitation can be responsible for low quality/amount of care. Salaried doctors have an incentive to minimize their effort during the consultation because they receive the same income irrespective of this effort.
However, some of the aforementioned incentives can be reversed in the case of PCPs who can refer patients to specialized care. PCPs paid by capitation can save on personal costs by simply referring their patients to expensive specialized care. PCPs’ altruism reinforces this effect because referrals allow for own cost minimization without prejudice to the patient.
Salary may bring the same incentives as capitation regarding referrals. To minimize their effort during consultation, salaried doctors have an incentive to refer more often than needed. This is even more so if one considers that PCPs derive utility from the well-being of their patients, because referrals do not harm the patients.
PCPs paid on an FFS basis earn more revenue when treating the patients on their own rather than referring them to a specialist. In that case, FFS would lead to lower total costs and quality compared to capitation, if it is supposed that the costs and the quality of specialized care are higher than those of PCPs’ care and it can be abstracted from professional duty considerations.
In theory, fundholding shares the same incentives as FFS regarding referrals. Fundholding enables PCPs to receive a fixed budget from which to pay for primary care, drugs, and non-urgent hospital treatment for patients. It has been used in the UK between 1991 and 1999 and reintroduced in 2005. Again, if non-urgent hospital treatment is more expensive than primary care, PCPs have an incentive to limit referrals.
Empirical studies clearly confirm the positive theoretical relationship between capitation and the number of (unnecessary) referrals. Empirical support also exists for associating fundholding with a lower rate of referrals.
As many professionals, physicians might actually be heterogeneous in the way they respond to financial incentives because they are actually heterogeneous in both the ability and sense of professional duty. Under FFS or fundholding, it can be expected that very altruistic yet not very able PCPs refer all patients to specialists. PCPs who are relatively altruistic and very able might decide to either treat or refer according to their diagnostic. The very selfish yet able PCPs might treat all their patients to either maximize their earnings under FFS or minimize their expected expenditure under fundholding. Empirical evidence is needed to eventually confirm these theoretical predictions.
Selection
Incentives need not be uniform for a given population within a health care system. For instance, physicians in the US can work either in a traditional FFS setting or in a managed-care organization with a capitation arrangement. Primary care practices in the UK in the beginning of the 1990s had the choice to adopt the fundholding scheme or not. PCPs in France can voluntarily participate in the Contract for Improving Individual Practice (CAPI) scheme, which pays PCPs a performance payment for satisfying guidelines on prescription and prevention behavior. On the demand side, patients in the US can enroll into either gatekeeping health plans or plans allowing direct access to specialists.
In theory, the ability of selecting one or another type of organization may result in a pooling of individuals with the same profile in each organization type, which potentially results in increased inequalities. It can also increase efficiency, thanks to a better match between individuals and organizations.
Allowing PCPs to select between FFS or capitation can be optimal if savings on specialists’ costs are not the main concern of a regulator. Otherwise, all PCPs should be paid on an FFS basis to avoid the incentive that associates capitation to excessive expensive referrals. Limited empirical evidence exists about either the existence or the lack of PCP selection into one or another plan. The French experience with the CAPI system shows no PCP selection according to their profile. Conversely, there is some evidence of British selection concerning groups of PCPs enrolling in the fundholding system in 1991. Evidence about the effects of selection is not available so far.
On the patients’ side, patients enrolling into gatekeeping health plans are expected to be less likely to see a specialist than are others in plans with unrestricted access to specialists. There is significant evidence of selection into plans with gatekeeper and/or network selection in the US. Self-selection occurs because individuals, possessing knowledge of their own health attributes and economic constraints, select plans accordingly. These attributes that partly determine the individual’s choice of health plans also affect their expected utilization of services. Individuals in plans that require signups with a PCP have more visits to nonphysician providers of care, more surgeries, and hospital stays but substantially fewer emergency room visits.
To sum up, there are three channels through which the choice of remuneration scheme may affect PCPs’ output or productivity: First, certain kinds of behavior may be encouraged by the scheme itself (the incentive effect); second, certain kinds of physicians may be attracted to certain types of physician practices, which, in turn, are influenced by the remuneration scheme (the physician selection effect); third, certain kinds of patients may be attracted to certain types of physician practices, which, in turn, are influenced by the remuneration scheme (the patient selection effect).
The Supply Of Specialized Care
Gatekeeping systems have developed in countries with a limited supply of specialists, as in the UK. There is also empirical evidence that the supply of specialists is an important system determinant of referrals. Therefore, controlling the market for specialists might help improving the organization of primary care. For instance, in health systems with lots of PCPs and few specialists per medical discipline, specialists enjoy in theory a high level of monopoly power eventually leading to high fees. Therefore, increasing PCPs’ qualification may decrease the monopoly power of specialists.
Increasing the number of PCPs makes the primary-care market more competitive too. Together with a capitation payment for a patient-list system, more PCPs may experience a patient shortage from the more intense competition. There exist empirical evidence that this may lead to more referrals. Intuitively, against more competition, PCPs refer patients more often, responding positively to patient requests. Therefore, the cost-saving effect of the substitution of specialists by PCPs may be weakened by the PCPs’ reactions.
Efficiency gains that are usually attributed to gatekeeping cannot be taken for granted. In the short run, better matches between patients and specialists may lead to efficiency gains. However, in the long run, specialists have an incentive to adjust their specialization so that differentiation between specialists increases. This would increase the monopoly power of specialists, which might counteract the positive short run effect.
Conclusion
This article has discussed issues related to the organization of the primary-care sector. The following important relationships have been reported. Gatekeeping arrangements result in lower expenditures and utilization of health care services although no significant effect has been proven on health care outcomes. Concerning PCPs’ incentives, capitation is associated with an increase in referrals to specialists, whereas the fundholding scheme seems to limit these expensive referrals. Concerning the choice of a health plan, patients opting for a gatekeeping plan are less likely to see a specialist than are others in plans that allow direct access to specialists. However, when regulating the primary care sector, it is important to anticipate its consequences on the behavior of specialists.
France is now witnessing a movement from solo PCP practice to group practice. This change may result in new incentives and behaviors. An opportunity for relevant research appears there, to follow on the recent interest of economists for group practice and norms.
References:
- Aakvik, A. and Holmas, T. H. (2006). Access to primary health care and health outcomes: the relationships between GP characteristics and mortality rates. Journal of Health Economics 25(6), 1139–1153.
- Allard M., Jelovac, I. and Leger, P. T. (2010). Physicians selection of a payment mechanism: capitation versus fee-for-service. GATE Lyon St Etienne, Working Paper 1024.
- Allard, M., Jelovac, I. and Leger, P. T. (2011). Treatment and referral decisions under different physician payment mechanisms. Journal of Health Economics 30(5), 880–893.
- Brekke, K. R., Nuscheler, R. and Straume, O. R. (2007). Gatekeeping in health care. Journal of Health Economics 26(1), 149–170.
- Deb, P. and Trivedi, P. K. (2009). Provider network and primary-care signups: do they restrict the use of medical services? Health Economics 18, 1361–1380.
- Devlin, R. A. and Sarma, S. (2008). Do physician remuneration schemes matter? The case of Canadian family physicians. Journal of Health Economics 27, 1168–1181.
- Forrest, C. B. (2003). Primary care gatekeeping and referrals: effective filter or failed experiment? British Medical Journal 326, 692–695.
- Forrest, C. B., Nutting, P. A., von Schrader, S., Rohde, C. and Starfield, B. (2006). Primary care physician specialty referral decision making: patient, physician, and health care system determinants. Medical Decision Making 26, 76–85.
- Garcia Marin˜oso, B. and Jelovac, I. (2003). GP’s payment contracts and their referral practice. Journal of Health Economics 22(4), 617–635.
- Gonzalez, P. (2010). Gatekeeping versus direct-access when patient information matters. Health Economics 19, 730–754.
- Iversen, T. and Ma, A. (2011). Market conditions and general practitioners’ referrals. International Journal of Health Care Finance and Economics 11, 245–265.
- Luras, H. (2004). General practice: four empirical essays on GP behavior and individuals preference for GPs. Thesis for doctoral dissertation, Department of Economics, University of Oslo, Working Paper 2004: 1.
- Malcomson, J. M. (2004). Health service gatekeepers. RAND Journal of Economics 35(2), 401–421.
- Naiditch, M. and Dourgnon, P. (2010). The preferred doctor scheme: a political reading of a French experiment of Gatekeeping. Health Policy 94, 129–134.
- Porteiro, N. (2005). Regulation of specialized medical care with public and private provision. European Journal of Political Economy 21, 221–246.
- Schwenkglenks, M., Preiswerk, G., Lehner, R., Weber, F. and Szucs, T. D. (2006). Economic efficiency of gatekeeping compared with fee for service plans: a Swiss example. Journal of Epidemiology & Community Health 60, 24–30.
- Velasco Garrido, M., Zentner, A. and Busse, R. (2011). The effects of gatekeeping; a systematic review of the literature. Scandinavian Journal of Primary Health Care 29(1), 28–38.
- WHO (2008). The World Health Report 2008 – Primary Health Care (Now More Than Ever).