Improving the risk adjustment formula

• Abstract
• Purpose of health policy or idea
• Characteristics of this policy
• Political and economic background
• Purpose and process analysis
• Expected outcome
• References

Abstract

The Committee for Social Security and Health of the Council of States has put a bill of law to discussion which is intended to improve the formula on the base of which risk adjustment among health insurers is carried out. There are two main innovations: taking morbidity indicators, in addition to age and sex, into consideration and the passage from a retrospective calculation model to a prospective one.

Purpose of health policy or idea

At an international level, health insurance experts agree that, for a health insurance system based on community rating and managed by a large number of private companies to function well, it is fundamental that this system has a well-performing risk adjustment mechanism at its disposal. In 1993 by means of an urgent federal decree Switzerland also decided to adopt a mechanism through which it could redress the balance of risk distribution present in the portfolios of the various health insurers.

The majority of the Swiss health economists agree that the present formula is not sufficiently precise to effectively inhibit cream skimming strategies. The context in which health insurance is managed in Switzerland no doubt encourages the spread of this kind of strategy: faced with a unitary package of services and a community rating premium each insured person has to choose the deductible (the amount of risk which is not passed on to the insurer in exchange for a reduction in the amount of the premium) and is free to give up (partly or wholly) his/her independence in the choice of doctor, by joining a system of managed care, again in exchange for a discount on the basic premium (see references [1] and [2]). Both these decisions represent signals sent to the insurer regarding the perception the individual has of his/her own risk of falling sick. In addition to this, besides compulsory insurance, the sickness funds offer integrative insurance products governed by private law, in whose ambit they have the power to apply individual premiums and gather information about the health of their clients.

Since the reform procedure indicated by the Federal Council provides for a strengthening of the role attributed to the market and more precisely to competition among health insurers (see the bill for deregulation of contracting relations with the service providers), it is right to expect the chance for the insurers to be able to operate cream skimming of the risks to increase (see references [3] and [4]). For this reason, by means of a draft bill, the Council of States suggests improving the risk adjustment mechanism with the aim of encouraging the creation of basic conditions able to promote fair competition among the sickness funds.

In order to develop effective cream skimming strategies a sickness fund must sustain higher administrative costs. Faced with these costs, however, the insurer can guarantee an economic advantage due to the fact that the health services reimbursed to a group of good risks are decidedly lower than the average and therefore less than the amount cashed from premiums too. In recent years the techniques used to skim risks have become more and more subtle and difficult to control. Since it is difficult to prevent cream skimming by means of decree laws or regulations the Council of States' proposal (see reference [5]) is to discourage these practices by reducing the return of investment, i.e. by obliging the sickness funds which have secured good risks to pay a suitable solidarity contribution to the risk adjustment fund and by offering those left with the bad risks the right to withdraw a solidarity contribution from the fund.

Main points

Main objectives

Further improvement of the risk adjustment formula

Type of incentives

Financial incentives

Groups affected

Health insurers, Insured, Service providers

Innovationsgrad	traditionell		innovativ
Kontroversität	unumstritten		kontrovers
Strukturelle Wirkung	marginal		fundamental
Medienpräsenz	sehr gering		sehr hoch
Übertragbarkeit	sehr systemabhängig		systemneutral

Political and economic background

Purpose and process analysis

Current Process Stages

	Idee	Pilotprojekt	Strategiepapier	Gesetzgebung	Umsetzung	Evaluation	Veränderung/Richtungswechsel
Implemented in this survey?

Origins of health policy idea

The "old" legislation on health and accident insurance (LAMI)

Within the scope of old law on health and accident insurance, passed in 1911, health insurance premiums were differentiated by classes of age and sex. Since at federal level health insurance was not compulsory, in order to encourage good risks (especially the young) to join early  the person's legal age was not considered in the calculation but his/her age at the time of joining a certain sickness fund. This practice granted young people a bonus for early membership, in exchange for participation in the health insurance, which they would enjoy for the rest of their lives. In short by joining a fund at the age of 20 the person could acquire the right to pay the premium reserved for 20-year-olds for all his/her life but on one condition: never to change the health insurer!

At the end of the 80s, almost 70 years after the law's introduction, 99% of the Swiss population was covered against the risk of falling sick (some cantons had meantime sanctioned compulsory insurance at cantonal level). As a result, since most citizens had joined the health insurance scheme at a young age, the differentiation of premiums based on age degenerated into a sort of community rating, thus implying for the young insured the payment of large contributions of solidarity with the older generations.

In those years new, rather aggressive insurance companies appeared on the Swiss market; they began to offer very favorable terms of cover for the young and rather expensive for the elderly. The price strategy of the new companies was made possible by the fact that the elderly, by changing fund, would have lost the bonus for joining early and would have ended up paying a premium on a par with their actual age, while the young (good risks) by changing fund would have ended up by getting out of the intergenerational solidarity requested in the "historical" sickness funds, thus characterized  by an elderly clientele which was redeeming its "early joining bonus".

Thus began a progressive ebb of young people from the historical sickness funds towards the new insurance companies; these movements of the insured altered the composition of the portfolios of the various sickness funds, threatening to make the whole system collapse. To prevent the de-solidarization in progress from breaking the pact between generations, in 1993 the Federal Council decided to institute a risk adjustment fund based on the age and sex of the insured. The aim of this fund, introduced by an urgent federal decree, was essentially to put a brake on the process of separation of the population into groups of mainly young people insured with the new sickness funds and mainly elderly people with the historical sickness funds.

The solution of the Federal Health Insurance Act (FHIA)

In 1996, when the Health Insurance Act came into force, the differentiation of premiums based on age and sex was abolished and substituted by a community rating premium, i.e. the same for all adult insured with a sickness fund, resident in a particular region. The package of services insured was made uniform by law and so the free transfer of the insured was set up (sanctioned also by the obligation for the funds to insure anyone at all). At first the legislator thought that it would be necessary to maintain risk adjustment for a limited period of time in order to allow the composition of the portfolios to re-balance autonomously thanks to competition and the new rules of the game; this composition was extremely disproportionate due to the above-mentioned mechanism. For this reason the institution of the risk adjustment fund was regulated in the provisional dispositions of the FHIA, in article 105, and limited in time to 10 years.

In Switzerland in the last 10 years there has been a progressive concentration of health insurance on the market, characterized by processes of merger and acquisition. In 1995 the sickness funds active in Switzerland were still 166, 10 years later in 2004 there were only 92 (-45 percent).

At the same time a decrease in the number of companies which withdraw a contribution from the risk adjustment fund can be observed. These went from 100 in 1995 to 23 in 2003, while the companies which pay contributions to the fund remained more or less at the same level (67 in 1995, 70 in 2003). These figures would seem to justify the hypothesis that the companies destined to disappear or be absorbed are those characterized by an unfavorable composition of their portfolio of insured persons. Since 1996 the volume of risk adjustment is in constant growth, with an average annual progression of approximately 10 percent. In 2004 1.1 billion francs were redistributed among the sickness funds, a sum which is equal to 6.75 percent of the net services borne by the health insurers. The volume compensated for corresponds to 64 percent of the cost differences which exist among the single health insurers. Although many experts had emphasized not only the need to make risk adjustment definite but also to refine its operating mechanisms, in 2004 the Federal Council proposed that parliament merely prolong the risk adjustment fund in its present form for a further five years; this decision was ratified by parliament at the end of 2004 (see reference [6]).

This is how the risk equalization fund mechanism works at present in Switzerland (see reference [7]). The insured are subdivided into 30 groups (15 classes of age, excluding children, for both sexes). An independent institution calculates the total average cost (of all the inhabitants) in each of the 26 cantons and the average cost recorded in the 30 groups. Theoretically the adjustment mechanism allows for each insurer to pay into the risk equalization fund, respectively withdraw from the fund, the difference between the costs of the group to which each of its clients belongs and the average costs of the population. In fact the balance between amounts due and credits is actually transferred. For an insurer with a portfolio of customers which is well distributed regarding age and sex, in exactly the same way as the total population in each canton, the balance between the amounts due and credits would equal zero. The qualities of this formula are its simplicity and the correspondence between payments and withdrawals.

Most experts active in the academic field recognize that the present formula is not good enough to prevent cream skimming strategies (see reference [8]). In fact within each class of age the difference between the average cost and the individual cost of each subject is very high (in other words a strong variance is recorded in the individual costs of the members of the various classes). As a result the fact that an individual belongs to one of the 30 groups does not constitute a good estimator of his/her individual costs. Let's try to illustrate the situation with a numeric example. In 2003 the insured belonging to the male sex aged between 36 and 40 caused an average monthly cost of 103 francs, the male insured between 86 and 90 an average monthly cost of 798 francs, while the male population on the whole causes an average cost of 170 francs (see reference [9]).

Simplifying a little (in fact in the example we don't consider the differences between the 26 cantons) the present risk adjustment mechanism means paying 67 francs (170-63) a month for each man between 36 and 40 years of age into the fund, while 628 francs (798-170) a month is withdrawn for each man between 86 and 90. Now let's suppose that a man of 38 with a chronic illness, obliged to consume health services costing 800 francs a month, is insured with the same sickness fund as a man of 88, still in excellent health, whose monthly costs amount to "only" 200 francs. Thanks to community rating both pay the same premium, let's say 300 francs a month. Without risk adjustment the balance of the two policies would be: (a) a 500-franc monthly deficit for the young man and (b) a 100-franc monthly surplus for the elderly. Thus in this example the risk adjustment mechanism ends up exasperating the balance of the two policies, since the insurer has to pay the equalization fund a further 67 francs for the young man (therefore the final deficit will amount to 567 francs), whereas for the elderly man the insurer would cash in the 798 francs adjustment in any case, thus bringing the positive balance of the policy to +898 francs.

This numerical example well illustrates the problems caused by an adjustment mechanism which is not able to operate a suitable matching between costs (or risks) and transfers / payments to the equalization fund.

In recent years the health insurers have refined their own cream skimming strategies. Since the law obliges the sickness funds to accept any insured within the ambit of compulsory insurance, the skimming has to be carried out subtly and in a hidden way; targeted marketing campaigns, sponsoring of sports events, policy sales via internet, combining compulsory insurance with supplementary cover which is interesting for those in good health (e.g. cover for fitness activities), worsening the information desk - which is used in particular by sick clients, delays in reimbursement of the services, even to the extent of the payment of proper provisions to the so-called terminators, (insurance brokers whose mandate is to persuade especially expensive insured of the (inexistent) advantages they could have if they transfer to a competitor sickness fund.

Initiators of idea/main actors

• Regierung

• Parlament

• Kostenträger

Approach of idea

The approach of the idea is described as:
new:

Stakeholder positions

In recent years some institutions and private research centers have begun to study the way to further refine the risk adjustment mechanism, especially with reference to the progress reached in this sector in other nations, above all in Holland. The most significant contributions brought to the discussion in these years are: (1) the works of Konstantin Beck (see references [10] and [11]), (2) a research project realized in the ambit of the national program of research No. 45 (Future Problems of the Welfare State) under the supervision of Prof. Alberto Holly (see reference [12]) and (3) the research of Stefan Spycher (see references [13], [14] and [15]), among which a report of evaluation commissioned by the Federal Office of Public Health stands out; this was in the ambit of the project of the third revision of the Federal Health Insurance Act. The principal merit of these studies is that of using data already available today on the morbidity and on the clinical history of the insured (e.g. the occurrence of hospital stays in the previous year, ICD-10 code of hospital diagnosis, the presence of a chronic disease inferred from the prescription of particular pharmaceuticals), in order to be able to foresee in a more accurate way the risk factors and therefore the expected costs for the various categories of inured. The most evolved models of compensation are based on the risks and no longer on the effective costs, by resorting to statistical forecasting models. In this way the prospective models do not relax the incentive of health insurers to strictly control the appropriateness and the level of expenditure: then where real cost instead of expected cost is compensated, it is right to assume a relaxed attention of the health insurer with regard to all possible savings.

A certain consensus exists among the political parties as to this being an opportune moment to improve the risk adjustment mechanism. In fact it is a reform which should increase the fairness of the system (avoiding the painful consequences for sick individuals due to risk selection) and at the same time contribute to strengthening the regulatory function played in the Swiss health system by competition among health insurers. The cantons (see reference [16]) would also like to see risk adjustment functioning better, although they recognize that the task of regulating health insurance constitutionally falls on the Confederation. Among the health insurers no consensus reigns, reflecting the present situation characterized by sickness funds specialized in cream skimming strategies  and therefore with a particularly favorable portfolio on the one hand, and by less aggressive sickness funds who are obliged to insure a higher than average percentage of bad risks. At the moment the real enemy of a reform of risk adjustment seems to be the Federal Council, which - through the Minister of Home Affairs - has induced parliament to prolong the present formula for five more years and on several occasions has sided against every proposal to strengthen risk adjustment further. The hostility which the head of the Department of Home Affairs, which the Federal Office of Public Health belongs to, has shown towards the bill may be interpreted in the following way: a functioning mechanism of risk adjustment could in fact stop the concentration process in progress on the health insurance market; in other words the federal government would like to see a further reduction in the number of sickness funds to ten large funds, better able to manage the tasks which will in future be given to the health insurers.

Actors and positions

Regierung
Canton	sehr unterstützend		stark dagegen
Federal Council	sehr unterstützend		stark dagegen
Parlament
Committee for Social Security and Health of the Council of States	sehr unterstützend		stark dagegen
Kostenträger
Health insurers with a particularly favorable portfolio	sehr unterstützend		stark dagegen
health insurers with a lot of bad risks	sehr unterstützend		stark dagegen

Influences in policy making and legislation

The Committee for Social Security and Health of the Council of States suggests inserting two new variables, referring to health status, into the formula for risk adjustment: (1) the occurrence of a hospital stay in the previous year, since there is evidence that on average people show higher medical expenditure even in the year following the hospitalization and (2) the belonging to a group of chronically ill, on the base of a prescription of a certain active principle (ATC Class). The committee proposes passing from the retrospective model - today the compensation takes place on the base of the average effective costs of the current year- to a prospective model. Substantially, instead of the present 30 classes 60 ordinary classes would be created (15 classes of age, 2 sexes, people with and without hospital stays the previous year) for whom the average costs would be forecast on the base of the previous year's data, and 13 extraordinary classes, where an attempt would be made to group together some kinds of chronically ill. The draft bill would be particularly detailed and technical and this is a rather unusual fact for Switzerland. As a rule the laws contain only general principles, relegating the technical details to an application ordinance whose formulation falls on the relevant federal office (in this case the Federal Office of Public Health). There are some who interpret the decision to draw up an unusually detailed text of law as a tangible sign of the committee's lack of trust in the Federal Office, which till now has seemed reticent in the face of the proposal to further refine the risk adjustment mechanism.

Legislative outcome

pending

Actors and influence

Regierung
Canton	sehr groß		kein
Federal Council	sehr groß		kein
Parlament
Committee for Social Security and Health of the Council of States	sehr groß		kein
Kostenträger
Health insurers with a particularly favorable portfolio	sehr groß		kein
health insurers with a lot of bad risks	sehr groß		kein

Positions and Influences at a glance

Adoption and implementation

The consultation process should finish by the end of October 2005, as the committee intends submitting the final text of law to the Council of States during the winter session of the federal chambers. If the parliamentary procedure should go ahead without delays, part of the reform could come into force on 1 January 2008. Since in order to identify chronically ill patients some changes are necessary to the present survey methods of pharmaceutical prescriptions, the 13 extraordinary classes might not even come into force before 2010. The only unknown factor on the decision times is linked to the possibility that the reform of risk adjustment may be voted on together with the much more controversial reform of hospital financing. In this case the decisional times might turn out to be much longer.

Monitoring and evaluation

The cantons recommend realizing a simulation of the risk adjustment mechanism according to the new formula in order to evaluate its impact and consequences better.

Expected outcome

The reform of risk adjustment represents a first important step towards a fairer competition in the field of health insurance and constitutes a premise to realize the other reforms on the agenda.

However, in order to make the market function in this sector two extra measures are necessary:

• the clear separation of the activities of basic insurance, governed by social law, from those of the complementary insurance, governed by private law. In other words each insurer should decide in which of the two sectors he intends to operate and abandon the activity in the other sector. This separation might make the choice of basic insurance more price elastic than it is today.
• the introduction of free transferability for the mandatory reserves. At present the legal reserves constituted by a health insurer are not transferred in case of loss of members. Therefore they may be absorbed by the insurer who loses part of his clientele (higher administrative costs, increased managers' salaries, profits transferred from the non profit sector of the basic insurance to the lucrative sector of complementary insurance) and at the same time must be reconstituted by the insurer who acquires new clients.

Impact of this policy

Qualität	kaum Einfluss		starker Einfluss
Gerechtigkeit	System weniger gerecht		System gerechter
Kosteneffizienz	sehr gering		sehr hoch

References

Sources of Information

1. Gardiol, L., P-Y Geoffard and Chantal Grandchamp (2003), Separating Selection and Incentive Effects: an Econometric Study of Swiss Health Insurance Claims Data," DELTA Working Papers 2003-27.
   www.delta.ens.fr/abstracts/wp200327.pdf
2. Lehmann H-J. e P.  Zweifel (2004), Innovation and risk selection in deregulated social health insurance, in: Journal of Health Economics (23), 997-1012.
3. Oggier, W. (2004), Risikoausgleich oder Risikoselektion? Einige gesundheitsökonomische Gedanken zu den aktuellen Reformvorhaben auf Bundesebene, in: Schweizerische Ärztezeitung, 85/31, 1626-1629.
   www.saez.ch/pdf/2004/2004-31/2004-31-639.PDF
4. Spycher, S. (2004), Die Reform des Risikoausgleichs als Vorbedingung für die Aufhebung des Kontrahierungszwangs?, in: Schweizerische Ärztezeitung, 85/31, 1630-1635.
   www.saez.ch/pdf/2004/2004-31/2004-31-711.PDF
5. Bundesgesetz über die Krankenversicherung. Teilrevision (Vorlage B, Risikoausgleich). Vernehmlassungsentwurf der ständerätlichen Kommission für soziale Sicherheit und Gesundheit (SGK-S) betreffend den Risikoausgleich vom 30. August 2005.
6. Amtliches Bulletin (2004), Ständerat - Herbstsession 2004 - Zwölfte Sitzung - 08.10.04-08h00. Bundesgesetz über die Krankenversicherung. Teilrevision. Gesamtstrategie, Risikoausgleich, Pflegetarife, Spitalfinanzierung.
   www.parlament.ch/ab/frameset/d/s/4705/112877/d_s_4705_112877_112953.htm
7. Beck, K. (2000), Growing importance of capitation in Switzerland, in: Health Care Management Science 3, 111-119.
8. KVG: Die Schweiz muss das Risikoausgleichssystem unter den Versicherern ändern. Forscher fordern die politischen Behörden zur Revision des KVG auf. Medienmitteilung vom 15. September 2004.
   www.snf.ch/de/com/prr/prr_arh_04sep15.asp
9. Bundesamt für Gesundheit (2005), Statistik der obligatorischen Krankenversicherung 2003 und 2004.  www.bag.admin.ch/kv/statistik/d/2005/KV_2003_DE_v030205.pdf
10. Beck, K. (2001), Capitationberechnung in der Schweiz: The State of the Art, in: Managed Care 1, 12-16.
    www.forummanagedcare.ch/archiv/2001/1/mc1-01-s12_s16.pdf
11. Beck, Konstantin (ed.) (2004): RAN Eine Publikation des European Risk Adjustment Network. Mit Beitragen von K. Beck, A. Holly, S. Spycher, C. van de Voorde, W. van de Van, D. Chernichovsky, J. Wasen, Luzern. www.css.ch/d_ran_230804-3.pdf
12. Holly, A. et al (2004), Health-based risk adjustment in Switzerland. An exploration using medical information from prior hospitalization. National Research Program 45 "Future Problems of the Welfare State".www.snf.ch/downloads/com_prr_arh_04sep15_3.pdf
13. Spycher, S. et al. (2004),Risikoausgleich Krankenversicherungsgesetz. Experten-/Forschungsberichte zur Kranken- und Unfallversicherung. Grundlagen zur 3. KVG-Revision, Teilprojekt, Bern: Bundesamt für Gesundheit.
    www.bag.admin.ch/kv/forschung/f/doc/Ber05_Risikoausgleich_d.pdf
14. Sycher, S. (2004), Risikoausgleich im KVG - wie weiter?, in: Soziale Sicherheit, 2/2004, 109-112.
    www.bag.admin.ch/kv/analysen/d/2005/CHSS0204_d_4p.pdf
15. Stefan Spycher (2002): Risikoausgleich in der Schweiz - Notwendigkeit, Ausgestaltung und Wirkungen. Bern: Paul Haupt Verlag.www.buerobass.ch/pdf/2002/Dissertation%20Stefan.pdf
16. Council of the Swiss cantonal health ministries (2005), Verfeinerter Risikoausgleich. Hauptanliegen der GDK zur neuen Vorlage der KVG-Revision
    www.gdk-cds.ch/fileadmin/pdf/Gesundheitsoekonomie/KVG-Revision/Positionspapiere_200507/Risikoausgleich-Hauptanliegen-d.pdf

Author/s and/or contributors to this survey

Luca Crivelli (proof reading by Iva Bolgiani and Mary Ries)

Empfohlene Zitierweise für diesen Online-Artikel:

Luca Crivelli (proof reading by Iva Bolgiani and Mary Ries). "Improving the risk adjustment formula". Health Policy Monitor, October 2005. Available at http://www.hpm.org/survey/ch/a6/3