Chapter 13 - Gosfield · Chapter 13 Making Quality Happen: In Search of Legal Weightlessness Alice G. Gos”eld1 KeyCiteL: Cases and other legal materials listed in KeyCite Scope

Chapter 13

Making Quality Happen: In Searchof Legal WeightlessnessAlice G. Gos�eld1

KeyCiteL: Cases and other legal materials listed in KeyCite Scopecan be researched through West Group’s KeyCite service onWestlawL. Use KeyCite to check citations for form, parallel ref-erences, prior and later history, and comprehensive citator in-formation, including citations to other decisions and secondarymaterials.

§ 13:1 Introduction§ 13:2 Reimbursement and Compensation§ 13:3 —Medicare Physician Payment§ 13:4 —A New Approach§ 13:5 —Physician Compensation§ 13:6 Anti-Kickback and Anti-Referral Laws§ 13:7 —Background§ 13:8 —Physician Incentive Plan Regulations§ 13:9 —Gainsharing§ 13:10 —Possibilities Explored§ 13:11 HIPAA§ 13:12 —Basic Regulatory Philosophy and Contents§ 13:13 —Quality vs. Privacy§ 13:14 —Joining Forces§ 13:15 Antitrust and Clinical Integration§ 13:16 Conclusion

§ 13:1 IntroductionThe contemporary moment in health policy is nothing

1The author is the Chairman of the Board of the National Com-mittee for Quality Assurance (NCQA), but the opinions expressedhere are exclusively her own and should not be imputed to NCQAin any way whatsoever.

1

@MAGNETO/NEPTUNE/AUTOMATE01/V�JUR/HLHB/CH13 SESS: 1 COMP: 04/08/02 PG. POS: 25

short of a Dionysian rhapsody of regulation, the inhospi-tality tradition gone riot, the formal and �nal enshrine-ment of the doctrine that everything not mandatory isprohibited.1

Although the hysteria of free-marketeers about mostregulation in health care often merits discounting asmere hyperbole, nowhere is the paralyzing weight ofhealth care regulation more evident than in the discon-nect between what is necessary to make real improve-ments in health care quality and the crushing burdenof the regulatory constraints on doing so. The clarity ofthis paradox is even more dramatic when we considerthat much of the di�culty stems from regulation whichclaims to be aimed at bolstering quality or at least ap-propriateness of care.

Imagine the following health care market:Physicians are paid based on how well they hew to

evidence-based clinical practice guidelines (CPGs) andresearch protocols. For services which conform to theselected CPGs, there is no requirement for prior autho-rization, concurrent review, or post-payment utilizationreview. Centers of excellence perform extremely risky,complex procedures, so not all physicians and facilitiesprovide all services. At the same time, research proto-cols are used to develop expertise in other settings whenclinical trial safeguards are strictly applied. Apart fromquality controls for market entry (i.e., credentialing,licensure, and threshold quality standards to providecertain services), no one worries about the �nancialrelationships between physicians and the entities whichserve their patients.

Once a guideline is agreed to, payors, includingMedicare, occasionally render post-payment audits onreasonable samples, simply to determine that guidelinesare being followed. Payment systems include (1) caserates where a single payment is made for treatment of

[Section 13:1]1Robinson, ‘‘Theory and Practice in The Design of Physician

Payment Incentives,’’ 79 The Milbank Quarterly 149, 174 (2001).

§ 13:1 Health Law Handbook

2


a speci�c condition for a de�ned period of time, and (2)payment based on quanti�cation of expected fee for ser-vice payments, but all systems re�ect CPGs. Other op-tions are available if proposed in a collaborativeconstruct among the providers and payers. Financialrisk is taken, if at all, at large group and system levelswhich involve more than one physician.

In some instances, otherwise competing physiciansjoin in clinically driven virtual organizations whichstandardize care in accordance with guidelines, mea-sure performance in accordance with those standards,and actively seek to improve care and take actionagainst participants who cannot e�ectively change theirbehavior. Within physician groups and organizations(including where hospitals employ physicians), thephysicians’ compensation re�ects guideline conformityand the strength of their doctor-patient relationshipsmeasured in terms of customer satisfaction surveys andpatient perception of care. Payors and health plansreward systems and physicians that excel in their qual-ity performance and patient relationships.

Hospitals work with the physicians to provide careconsistent with those guidelines and pay them for workthey do which bene�ts the hospital directly, whether (a)through traditional ‘‘gainsharing’’ concepts to motivateworkers, or (b) for consultant services necessary tomaintain the hospital’s programs, whether provided byphysicians or others. At the delivery system level, thefacilities and physicians collaborate to safeguard theclinical validity and dynamism of the application of theguidelines and research protocols which are the bedrockof the system.

Hospitals design their manpower recruitment plan-ning around the CPGs. Care teams are the core modelfor delivering care and are taken into account in thevehicles (e.g., medical sta� bylaws) by which cliniciansobtain the opportunity to render services within a facil-ity setting. There is relatively little emphasis on whichtype of licensed clinician performs speci�c tasks. Hu-man resources are assigned so the best quali�ed indi-

§ 13:1Making Quality Happen

3


vidual in the moment works to meet the patient’s needscoupled with a philosophy to use personnel at theirhighest and best use.

Hospitals develop and support clinical leaders forquality, oriented around implementing CPGs through-out the enterprise. Work�ow, information technologyand sharing, capital budgets and organizational struc-tures are driven by CPGs and the clinical needs of thepatients rather than by tax laws, cost accounting andsiloed information needs. Hospitals freely share infor-mation with and among physicians in the interests ofquality improvement. In addition to patient-speci�ctreatment data, to which clinicians have access for qual-ity improvement purposes, clinicians and others at alllevels of the system report untoward incidents and er-rors to a closed database which is used to study andimprove clinical and system performance.

Ancillary providers—durable medical equipment, re-habilitation, home health and pharmacy providers—areincluded in the treatment continuum and have accessto patient data as relevant to their needs to providetheir services, in accordance with the applicable CPG.Neither they nor the hospitals themselves are subjectto prior authorization or concurrent review once havingagreed to provide care in accordance with explicit clini-cal pathways.

Health plans not only manage clinician and providercredentialing and pay for services, they participate inlocal e�orts to improve care, by facilitating implementa-tion of CPGs and research protocols. Their vetting ofproviders (hospitals, physicians and others) turns onexplicit, agreed upon standards. They report perfor-mance of the systems and physicians to the public ac-cording to the activities for which each controls; in otherwords, physicians are measured for their knowledge oftheir clinical craft and their doctor-patientrelationships. Clinical outcomes and system guidelinesconformity is reported at the system level. Plans man-age early detection, disease management and publicreporting programs, in collaboration with providers.


4


They develop speci�c tools to permit bench markingand best practices models to evolve and facilitate theirimplementation within their networks. All the partiesbenchmark their performance against those of theirpeers. The results are used to revise CPGs and tochange processes to conform with what the data showsare best practices thereby renewing and maintainingthe vitality of the processes and standards usedthroughout.

While not available for all conditions or in all set-tings, for those physician groups, hospitals, systemsand plans which (1) explicitly agree to be evaluated byindependent auditors/accreditors to validate that theyhave a quality-driven infrastructure, (2) participate inbroader information sharing to a closed clearinghouseon errors so more participants can bene�t from lessonslearned, and (3) commit to public performance report-ing, an alternative compensation system for tort isavailable.2

A utopian view of millennial health care? Notnecessarily. Likely the market described is not evenideal and the description certainly is not a completepicture of what would be necessary to perfect a modelof clinical quality and payment. But much of what isdescribed in the scenario above is achievable and inmany ways re�ects the principles set forth in thelandmark study Crossing the Quality Chasm.3 There isno question, though, that the relationships and func-tions it describes are so far from today’s reality in mostsettings, that even partial movement toward the worlddescribed would create a far more clinically relevant,quality focused health care system. It would, in fact,begin to create an actual system with consistentprinciples of action which drive the major players andwhich emanate from certain core values.

2Gos�eld, ‘‘Liability for Quality: A Modest Proposal,’’ ManagedCare & Cancer 11-12 (Jan./Feb. 2000).

3Corrigan, et al., National Academy Press (Washington, D.C.,2001).


5


To the extent that signi�cantly improved quality ofhealth care has proven elusive over the last twenty-�veyears, it has not been mere lack of will that hasprevented real change. It must be acknowledged thatthere are many disparate factors which undermine andthwart transformation or even incremental transitionto simpler, more standardized, more clinically relevantsystems and programs which are more responsive tothe real concerns of patients, policy-makers, payors andphysicians. To tackle the practical implications ofchange though, there should be clear recognition thatthe doctor-patient relationship is the essential portal tothe rest of the system.

The critical role of physicians cannot be overstated.Patients enter the system through a doctor-patientencounter, and physicians drive the delivery of all otherservices. The Institute of Medicine has said that‘‘Transferring knowledge is care.’’3 The essential role ofphysicians is to ‘‘transform information into meaningfulexplanations of the present, predictions of the futureand changed futures, mainly for individual patientsand sometimes for whole populations.’’4 Whateverimpedes the application of standardized science to careand limits the strength of the primary relationshipthrough which knowledge can be transferred and heal-ing can occur will sti�e real quality improvement. Tomove markets more toward the scenario above, lawyers,policymakers, and those in the system with the will tomake a di�erence will have to take into account theregulatory environment which stands in the way ofmeaningful leaps forward, with particular emphasis onbarriers to deeper, more e�ective doctor-patientrelationships.

This article scans and elucidates some of the moreobvious legal and regulatory issues which are contained

3Corrigan, et al., National Academy Press, 72 (Washington,D.C. 2001).

4Reinertsen, ‘‘Health Care: Past, Present and Future,’’ GroupPractice Journal 37-43 (March/April 1997).


6


in the scenario, focusing on their implications primarilyas they relate to physicians. The operational barriers tobroad scale change in any community are manifold andpresent their own challenges; but they are beyond thisconsideration. To create a truly quality-relevantenvironment, many changes at organizational, struc-tural, dynamic and other levels will have to be adopted.Here, I consider only salient legal issues with a focuson physician reimbursement and compensation, anti-kickback and anti-referral laws, HIPAA, and antitrustand clinical integration, to enhance an understandingof how the major existing legal barriers can be man-aged (and where wholesale revision is necessary), soquality can advance more in the direction of our hypo-thetical market.

§ 13:2 Reimbursement and CompensationMany commentators have decried the fact that typi-

cal existing payment systems neither reward nor evenmotivate quality.1 Capitation has been touted as asystem which at least drives toward prevention andearly detection;2 but other observers have posited thatcapitation re�ects nothing about clinical necessity orquality and its incentives are so blunt and broad as toundermine medically necessary care.3 Percent ofpremium risk contracts pay to the provider just that—

[Section 13:2]1Corrigan, et al., National Academy Press Ch. 2 (Washington,

D.C. 2001), and Becher and Chassin, ‘‘Improving Quality, Minimiz-ing Error: Making It Happen,’’ Health A�airs 68-81 (May/June2001).

2California Association of Health Plans, Issue Briefs, ‘‘Capita-tion and Quality in Health Care,’’ 4/97, http://www.calhealthplans.com/ib�capitation�quality.htm; for an excellent analysis ofmultiple forms of physician compensation and their pitfalls, seeRobinson, ‘‘Theory and Practice in The Design of Physician Pay-ment Incentives,’’ 79 The Milbank Quarterly (2001).

3Court, ‘‘HMOs’ Gestures Won't Solve Patient Care Problems,’’San Diego Union-Tribune (July 12, 2001), http://www.consumerwatchdog.org/healthcare/nw/nw001779.php3; a settle-


7


percent of the premium, so these payment systemsincorporate inherently all of the disadvantages ofactuarial judgments, that is, an analysis of pastbehavior projected forward based on certain demo-graphic and epidemiological assumptions which may ormay not re�ect clinical appropriateness, let alonequality. Episode of care (EOC) reimbursement4 and itsvariants such as contact capitation5 contain essentiallyboth the same incentives and pitfalls as pure capitationmodels.

As capitation has come under increased skepticism inthe sweep of managed care backlash, fee for servicepayment systems have returned in some quarters6 andnever left in Medicare. To shift any market in signi�-cant ways, the in�uence of Medicare reimbursementmust be confronted, too.

§ 13:3 —Medicare Physician PaymentAs of 1992, Congress provided that Medicare pay-

ment to physicians must be made on the basis of a feeschedule which takes into account relative values of

ment between the Texas Attorney General and Aetna was seen asa major blow to capitation and its analogues—withholds andlimited specialist referrals. Klein, ‘‘Aetna consents to physiciancontracting principles in Texas,’’ AMNews (April 24, 2000), http://www.ama-assn.org/sci-pubs/amnews/pick�00/gv110424.htm.

4See Krane and McNair, ‘‘Episodes of Care Overview of Legaland Regulatory Issues,’’ Health Law Handbook 423 (Gos�eld, ed.2001).

5Robinson, ‘‘Theory and Practice in The Design of PhysicianPayment Incentives’’ 79 The Milbank Quarterly (2001), and Frank,‘‘Contact Capitation for Specialists: Can the Pie Be Divided MoreFairly?’’ 5 Cost & Quality (June 1999), http://www.cost-quality.com/5,2art3.html.

6Carlson, ‘‘Physician Financial Incentives: Another UM ToolBites the Dust,’’ Managed Care (June 2000), http://www.managedcaremag.com/archives/0006/0006incentives.html; andJacob, ‘‘Paci�Care Hopes to Regroup Financially in Colorado,’’AMNews 20-23 (April 2001), http://www.ama-assn.org/sci-pubs/amnews/pick�01/bisc0423.htm.


8


services measured against each other, multiplied by aconversion factor.1 The statute further requires auniform coding system which takes into account allphysician services.2 By regulation, the Secretary ofDepartment of Health and Human Services (‘‘DDHS’’)has adopted the American Medical Association’s(‘‘AMA’’) nomenclature, the Common Procedural Termi-nology (CPT).3 Unfortunately, one of the strongest bar-riers to the ability of physicians to focus on quality,spend more time, and deepen their relationships withpatients is the required documentation of both the scopeof services rendered along with some manifestation oftheir medical necessity. Some of the requirements arelocal carrier policies which set forth their interpreta-tions of national policy including, for example, whichdiagnostic code (ICD-9) must be used, thereby forbid-ding physicians to code the patient’s signs and symp-toms when a diagnostic service ordered to evaluate thepatient’s needs demonstrates an abnormal result.4 Bystatute, failure to put the diagnostic code on a claim issubject to a $2000 civil money penalty.5 With the adventof evaluation and management codes (E&M codes) cre-ated to quantify the cognitive e�ort associated withnon-procedural encounters between clinicians andpatients, new documentation guidelines were estab-

[Section 13:3]142 U.S.C.A. § 1395w-4.242 U.S.C.A. § 1395w-4(c)(5).356 Fed. Reg. 59,506 (Nov. 25, 1991); see also 42 C.F.R. § 415.40

(2002) in which HCFA reserved to itself the right to establishuniform national de�nitions of codes and modi�ers even as itadopted the AMA’s CPT designations for the bulk of physician ser-vices paid for on the then newly established fee schedule.

4See, for example, ‘‘ICD-9-CM Coding for Diagnostic Tests,’’ Illi-nois Michigan Medicare Part B Bulletin (Nov. 2001).

542 U.S.C.A. § 1395u(p).


9


lished and then re�ned further.6 Local carrier applica-tions vary around the country.7

The legal perils in the combined weight of thesemultiple requirements lies in their link to post-paymentutilization review and ultimately to false claimsliability. In its Semi-Annual Report for 2001, the O�ceof the Inspector General found inappropriate or errone-ous fee for service payments in the amount of $11.9 bil-lion during �scal year 2000. Unsupported and medi-cally unnecessary services continued to be the pervasiveproblems.8 Although these numbers regarding failure todocument services e�ectively are noteworthy, physiciancomplaints about the time it takes to document just theE&M services have escalated:

The current guidelines emphasize the clerical skill ofphysicians and do not re�ect the intensity of the physi-cian service. They clutter the medical record with un-necessary verbiage, interfere with physician-to-physiciancommunication and waste precious physician time. Inaddition, while HCFA and the AMA maintain that theseare only guidelines, it is clear that the government willuse them as a basis in fraud investigations.9

The AMA reportedly has survey data showing thatphysicians spend one hour completing Medicare formsand meeting administrative requirements for every fourhours of patient care.10 The administrative burdenshave become so great, that fewer physicians in fee for

6‘‘Documentation and Evaluation Guidelines’’ (1995) and (1997),http://www.hcfa.gov/medlearn/emdoc.htm.

7See http://www.lrmp.net for a listing of carrier policies for pay-ment, including documentation of services.

8OIG SemiAnnual Report 2 (Oct.-Mar. 2001), http://oig.hhs.gov/semann/01ssemi.pdf.

9American College of Cardiology, Evaluation and ManagementDocumentation Guidelines, http://www.acc.org/advocacy/advoc�issues/emwebsite.htm.

10Qian, Bureaucratic Malpractice, The Business of Health, AnNYU School of Journalism Special Report (May 1, 2001), http://journalism.fas.nyu.edu/mag/boh/qian.htm.


10


service medicine are accepting new Medicare patients,with some communities �nding that only 34% of physi-cians will do so.11 Studies now show that clinical oncolo-gists are spending, on average, more than four timesthe time in documenting care and completing paperwork than twenty-�ve years ago.12

Those dilemmas pertain only to whether the data re-corded in the medical record substantiates the scope ofthe service performed. Another problem for physiciansis that there is no generalized source for standards ofmedical necessity. They are not listed. They are notpublished. There are not even normative parametersmade available to physicians on an ongoing basis.

The Peer Review Organization (‘‘PRO’’) program wasoriginally supposed to make medical necessity judg-ments for inpatient, outpatient and physician care,based on norms, criteria and standards; but its orienta-tion has been regulatorily reoriented so much and sooften over the last two decades, that the statute andthe implementing operational requirements seem to befrom two di�erent worlds. Although the statute callsthese entities ‘‘Utilization and Quality Control PeerReview Organizations,’’ the Secretary of DHHS in thelatest pronouncements on the subject has now dubbedthem QIOs, ‘‘Quality Improvement Organizations,’’which will have a major focus on nursing home care.13Today, the program has little impact on physicians, un-less they engage in grossly �agrant violations of stan-

11Qian, Bureaucratic Malpractice, The Business of Health, AnNYU School of Journalism Special Report (May 1, 2001), http://journalism.fas.nyu.edu/mag/boh/qian.htm.

12American Society for Clinical Oncology, Excessive PaperworkTaking Away from Cancer Patient Care, Professional Mentoringand Research (May 12, 2001), Annual Meeting News Room, http://www.asco.org/people/nr/html/01/ex.htm.

13Lovern, ‘‘QIOs: Eye Mission: Better Reporting,’’ ModernHealthcare 12 (Dec. 10, 2001); Lovern, ‘‘Gotcha Gives Way to Guid-ance,’’ Modern Healthcare 20 (Dec. 17, 2001).


11


dards or fail in a substantial number of cases substan-tially to conform with a PRO’s standards.14

In Medicare generally, the determination of medicalnecessity of the physicians’ services is made only inpost-payment audits for physicians who are identi�edon the basis of data aberrancies or bene�ciary com-plaints and thereafter on prepayment review for aphysician targeted as problematic.15 Therefore, whetherany physician will be subjected to audit is unpredict-able; so all must expend the same energies, lest theyend up in the snares of the enforcers. None of thesepayment mechanisms re�ects any explicit quality stan-dards, parameters, protocols or guidelines.

As for direct recognition of quality, during 2001, somemanaged care payors have publicized their e�orts toreward quality and pay physician groups more moneybased on their quality performance. Some de�ne qual-ity as achieving speci�ed targets.16 Some referencepatient satisfaction.17 Still others target those HEDIS18

measures which focus on services provided byphysicians.19 Medicare fee for service is still onlycontemplating a move in that direction with the

1442 U.S.C.A. § 1320c-5; 42 C.F.R. §§ 1004.30 et seq. (2002). Fora discussion of the evolution of the PRO program see Gos�eld,‘‘Utilization Management, Quality Assurance and PracticeGuidelines,’’ AHLA, Health Law Practice Guide 25-1 to 25-78(2001).

15See 4.0 et seq., and Medicare Program Integrity Manual Ch. 2,http://www.hcfa.gov/pubforms/83�pim/pim83toc.htm.

16Aquilina, ‘‘Buyers Reward Physicians for Quality’’ (Feb. 2001),http://www.qiphysician.com/cgi-bin/article.cgi?article�id+885.

17Jackson, ‘‘California HMO: Doctor Bonuses Based on PatientSatisfaction,’’ AMNews (July 30, 2001), http://www.ama-assn.org/sci-pubs/amnews/pick�01/bil10730.htm.

18Health Plan and Employer Data and Information Set, pub-lished by NCQA.

19Eaton, ‘‘Health Plans Rewarding Doctors for Meeting QualityStandards Rather Than Controlling Costs,’’ Managed Care Report,Physicians Financial News (Sept. 30, 2001), http://www.pfnpublishing.com/quality.htm.


12


expectation that eventually Medicare payment willreward higher quality.

To change such a cumbersome and clinically irrele-vant payment system, whether it is applied only inMedicare or, as is often the case, in private, commercialcontracts too, is potentially the most critical step toradically reorient to an explicit quality-driven modelhow care is provided in this country. The physicians’stakes in the current context are far more appreciatedtoday than ever before. The furor that met the govern-ment’s e�orts to revise the E&M documentation stan-dards in 1998 was so intense, that those standards werepulled and have not yet been reissued.20 The inability ofthe regulators to craft something which speaks mean-ingfully to physicians, even after all the complaintsabout the 1997 standards, underscores the need for anew conceptual framework. Crossing the Quality Chasmmade it clear that new payment methods would need tobe tested and piloted if quality is to advance. TheMedicare statute already provides that money is avail-able to develop and engage in experiments and demon-strations to determine whether other methods and ap-proaches, could enhance e�ciency and utilization‘‘including a change to methods based on negotiatedrates, [and] . . . would have the e�ect of increasing thee�ciency and economy of health services under suchprograms through the creation of additional incentivesto these ends without adversely a�ecting the quality ofsuch services.’’21

20See, for example, ‘‘New E&M Guidelines Postponed Inde�nitelyby HCFA,’’ San Bernardino County Medical Society, The Bulletin(June 1998), http://www.sbcms.org/southcalphysician/articles/art6c-98.htm for an understanding of the absurdity of thoseguidelines see ‘‘How to Use the New E/M Documentation Guide-lines,’’ ACP-ASIM Observer (Nov. 1997), http://www.acponline.org/journals/news/nov97/newem.htm.

2142 U.S.C.A. § 1395b-1.


13


§ 13:4 —A New ApproachImagining further our hypothetical market, consider

what would happen to the amount of time physicianscould spend with patients if we were to dramaticallysimplify the requirements for documentation of o�cevisits by making payment based on delivering servicesexplicitly in accordance with an agreed upon CPG.CPGs o�er a new and potentially far simpler approachto documentation for payment. For example, in thetreatment of congestive heart failure (‘‘CHF’’), a goodCPG, such as the one published by the Institute forClinical Systems Improvement,1 sets forth the clinicalsymptoms which lead the physician to suspect CHFand then enumerates the components of a comprehen-sive initial evaluation. Whether the treating physicianis the attending physician or a consultant, the visitcode is easily determined. The guideline states theextent of the history to be reviewed, the scope of thephysical examination to be performed, and the diagnos-tic testing to initiate—all the required elements forE&M coding. The guideline itself states the medical ne-cessity for all the tests, based on clinical evidence. Thecomplexity of the clinical decision-making is apparentfrom the application of the guideline’s terms, and neednot entail documentation of irrelevant negative �ndingswhich make no di�erence in the patient’s treatment atall. Where the patient has other co-morbid conditionssuch as hypertension or diabetes, those clinical guide-lines are referenced and also brought into play.

Evaluation of the test results with the patient, deter-mination of an appropriate pharmacologic regimen andpatient education are all included in the guideline andjustify additional o�ce visits which a physician practicecan template with expected, applicable procedural anddiagnostic codes. Steps to on-going evaluation of the ef-fectiveness of outpatient therapies (using diagnostic

[Section 13:4]1ICSI ‘‘Health Care Guideline: Congestive Heart Failure in

Adults’’ (Nov. 2000), http://www.icsi.org/guide/CHF.pdf.


14


testing and doctor-patient interaction) with appropriateadjustments in response, when to hospitalize, when torefer to subspecialists, and when to consider revascular-ization are all set forth. All of these translate into ap-plicable CPT and ICD-9 codes.

A physician practice willing to commit explicitly totreat CHF patients in accordance with such a guidelinecould be paid, even by Medicare, for the agreed uponrange of CPT codes, depending on the course of thepatient’s condition. Once agreed to, this techniquewould obviate the need for the physician to articulatein lengthy documentation the negative �ndings in thephysical exam or the medical necessity of the serviceprovided, merely to justify payment. A payer adoptingthis system might legitimately seek to conduct limitedpost-payment audits to assure that the guideline, infact, was the basis for treatment, but this methodologywould free signi�cant physician time spent in writingdata which has no bearing on treatment but is createdpurely to satisfy anticipated potential post-paymentfraud audits.

If Medicare were to adopt or even endorse such anapproach, physicians using it would see their liabilityfor false claims based on underutilization, over-utilization or misutilization virtually disappear.2 Therisk of exclusion based on quality failures would alsoevaporate. PROs would focus their energies on otherthan those physicians because the carrier acceptance ofthe CPG payment system would be binding on the PRO.3Although a totally new approach to payment anddocumentation would be preferable, until policymakersand regulators can see their way clear to sweepingreform, at least the approach in our hypothetical mar-ket can be crafted to accommodate the existing legal

2For a discussion of existing fraud and abuse liabilities whichsound in quality issues see Gos�eld, ‘‘Legal Mandates for PhysicianQuality: Beyond Risk Management,’’ Health Law Handbook 285-322 (Gos�eld, ed. 2001).

342 C.F.R. § 476.85 (2002).


15


mandates without perpetuating their administrativelunacies.

In commercial insurance, as distinct from Medicare,precisely the same technique could apply. Agreed uponCPGs, collaboratively selected by the plan or payor andthe physicians, provide a basis for the physicians toprovide care from a unitary and explicit clinicalplatform. Where the private context and Medicare areconsistent, the inevitable result would be to not onlymake the physicians’ lives simpler by not having toengage in documentation which serves neither a clini-cal nor a healing purpose, but also to make gatheringdata about care simpler. In reviewing their conformityto a CPG, physicians would then �nd it easier toidentify opportunities to be more e�cient and improvetheir processes. The roles of others—both clinicians andproviders—indelivering care would be more clearlydelineated.4 Plans would have better, more explicit dataon which to evaluate and motivate their networks.

There are communities in which multiple payors arebeginning to agree to use explicit and common CPGs.5While science ought not be local, and Medicare is anational insurance program, where clinicians in a com-munity seek to apply collaboratively selected guidelines(which meet the attributes for good clinical practiceguidelines as enunciated by the Institute of Medicine inits earlier guidelines study)6 to provide the basis forpayment, Medicare could permit such a mechanism

4To the extent the Medicare hospital payment system werecoordinated, the more powerful the whole undertaking would be.DRGs are not based on CPGs, though, they are based on historicalresource utilization patterns, no matter whether good or bad interms of clinical quality. Even so, hospitals and physicians work-ing together can apply CPGs within DRGs and some already do so.

5See ‘‘Minnesota Health Plans and Physicians Create aNational’’ Minnesota Model ‘‘For Improving Health Care,’’ PressRelease, McHugh Communications Inc. (Mar. 13, 2001).

6Clinical Practice Guidelines: Directions for a New Program,National Academy Press 59 (Field and Lohr, ed., Washington D.C.,1990).


16


under a pilot methodology until proof of the concept isclear. Still further, this approach requires physicianself-discipline and internal practice mechanisms to sup-port its application. Not all physician groups can orneed adopt it. Although the Medicare statute adoptsthe fee schedule as the foundation for fee for servicepayment, it does not require uniform application of it.

Today, quality payment from plans and insurers tophysicians and groups only go so far. Moreover, whenthe payment increment for quality varies from 5% ofthe capitation payments to $.30 per member per monthfor excellent performance7 one inevitably asks howmuch changed behavior these bonuses can stimulatewhen they are layered on top of basic payment systems(e.g., fee for service and capitation) which have nospeci�c quality or even clinical relevance.

§ 13:5 —Physician Compensation

Moving down to the next level of incentive, withingroups and systems, there have been few physiciancompensation systems which directly reward qualityperformance. This is also changing and represents yetanother focus to attend to in order to propel the qualityagenda. Compensation mechanisms within groups,particularly in markets that are not totally dominatedby managed care, have created virtually bi-polar physi-cian incentives for patient throughput and physicianproductivity measured in volumes of revenue, while atthe same time proclaiming the need for e�ciencies andutilization constraints to safeguard capitation dollars.No wonder physicians feel bu�eted.

Today, more physician groups are beginning to pegindividual physician compensation to qualityperformance. Some groups look to patient satisfactionscores, sta� and peer reviews, and utilization of preven-

7Maguire, ‘‘As They Struggle to Improve Quality, HMOs Try aNew Incentive: Bonuses,’’ ACP-ASIM Observer (June 2001), http://www.acponline.org/journals/news/jun01/bonuses.htm.


17


tive services to determine one-quarter of a physician’sincome.1 Others link total compensation to speci�cquanti�able measures of multi-variate factors such asnet medical revenue, reduced practice expenses, costand utilization, quality of care and patient satisfaction.These types of programs, which typically involve pri-mary care network compensation of individual physi-cians, limit base salaries to 65% to 75% of grosscompensation with the opportunity to earn bonuses thatmay reach 35% to 40% or more of base salaries.2 Over-all, though, the shift to quality oriented compensationis new and the literature on its application, as withmost physician compensation data, is surprisinglyscant.3 Although in the recent days of managed carereform and backlash there has been considerable heaton the subject, a closer investigation reveals very littleassociated light. The basic implications for quality infee for service payment versus capitation has itselfbarely begun to be studied.4 While the methods of pay-ment from health plans to physicians and systems hasbeen relatively little analyzed, the internal compensa-tion techniques and their impacts on quality withinphysician groups, or even from employers to physician-

[Section 13:5]1Maguire, ‘‘As They Struggle to Improve Quality, HMOs Try a

New Incentive: Bonuses,’’ ACP-ASIM Observer (June 2001), http://www.acponline.org/journals/news/jun01/bonuses.htm.

2Holm and Lipsky, ‘‘Using Compensation to Improve PrimaryCare,’’ MGM Journal 46 (July/Aug. 1999).

3See Andreae and Freed ‘‘A New Paradigm in Academic HealthCenters: Productivity-Based Physician Compensation,’’ 48 MGMJournal 44 (May/June 2001), and Forland et al., ‘‘Capitation, Sal-ary, Fee-For-Service and Mixed Systems of Payment: E�ects onthe Behaviour of Primary Care Physicians’’ (Cochrane Review),from The Cochrane Library, Issue 4 (2001), http://www.cochrane.org/cochrane/revabstr/ab002215.htm.

4Miller and Luft, ‘‘Does Managed Care Lead to Better or WorseQuality of Care?’’ Health A�airs 7-25 (Sept./Oct. 1997), and Ameri-can Association of Health Plans, Research Brief, ‘‘Quality of Careand Health Plans’’ (April 2000).


18


employees, has been virtually ignored in the researchfocused around health care �nancial incentives.5

As a matter of law, the physician employmentcontract is the primary legal determinant of physiciancompensation, as in�uenced by the common law ofemployment and ethical guidelines. In addition, how-ever, state law issues including corporate practice ofmedicine doctrines can get in the way of approachesthat could positively in�uence quality. These doctrinestake as axiomatic that employment by de�nition entailsdirection of the employee, stemming from ancientprinciples of master-servant. It is, therefore, againstpublic policy, so the doctrine goes, to permit non-licensed individuals or entities to corrupt the indepen-dent judgment of learned professionals—physicians.

In today’s world, the validity of this principle whenwe are trying to establish collaborative, aligned mecha-nisms to create quality-driven environments is not onlyquestionable, it leads to absurd organizational con-structs designed to accomplish by economic indirectionthat which cannot be done directly. Captive professionalcorporations whose sole shareholders are ‘friendly’ withthe business entity shareholders, independent contrac-tor relationships between physicians and business enti-ties, not-for-pro�t taxable membership companies,physician practice management companies or medicalservices organizations (MSOs) holding ‘‘sweep’’ accountsthat gather all the money physicians receive for theirservices and then pay them back a salary equivalent,are all techniques which lawyers have developed to copewith these restrictions. Each creates its own outlandishorganizational and operational realities to make thebusiness work within the common law framework. Thisonly adds to the complexity of health care delivery to-ward no quality end at all. Similar legal eccentricitiesresult from certain statutes aimed at eliminating‘‘improper’’ physician referrals.

5Robinson, ‘‘Theory and Practice in The Design of PhysicianPayment Incentives,’’ The Milbank Quarterly 79 (2001).


19


§ 13:6 Anti-Kickback and Anti-Referral Laws

If the general regulation of health care is a stellardemonstration of the inhospitality tradition noted byRobinson—‘‘That which cannot be understood withoute�ort is deemed ipso facto to be designed for fraud,monopolization, or some other antisocial purpose’’1—then the anti-kickback and anti-referral laws aresuper novas of this thinking. These statutes date backto the Mesozoic-era equivalent in health care, with theanti-kickback statute from the Triassic period and themore recent Stark statute from the Cretaceous era. Wecan only hope that what we see today is the delayedlight from their long ago �ameout re�ected in today’sregulatory environment, soon to extinguish completely.Until that happens, though, we must take into accountthe barriers their interpretation presents to our hypo-thetical market and the chances for improving quality.

§ 13:7 —Background

Their current obsolescence does not mean that whenthey were adopted these laws had no reasonableimpetus. It is their current applications which highlighttheir irrelevance to quality. The general propositionthat fraudulent acts under Medicare would be subjectto penalty was enacted in 1972. The legislative historyspoke of ‘‘penalties for certain practices which havelong been regarded by professional organizations asunethical, as well as unlawful in some jurisdictions,and which contribute appreciably to the cost of theMedicare and Medicaid programs.’’1 At that time it wasstated that the criminal penalty provision would includesuch practices as the soliciting, o�ering, or accepting of

[Section 13:6]1Robinson, ‘‘Theory and Practice in The Design of Physician

Payment Incentives,’’ The Milbank Quarterly 173 (2001).[Section 13:7]

1Sen. Fin. Comm. Rpt. to H.R. 1, Social Security Amendmentsof 1972, 208.


20


kickbacks or bribes, including the rebating of a portionof a fee or charge for a patient referral involvingindividuals, providers of healthcare services, and busi-ness entities such as corporations, companies, associa-tions, �rms, partnerships, societies, and joint stockcompanies. There is no further elucidation on theprohibited practices in 1972. In 1972, the violationswere to be treated as misdemeanors.

In 1977, the penalties were upgraded to felony status,and rewritten to clarify and restructure those provi-sions in then current law which de�ned the types of�nancial arrangements and conduct to be classi�ed asillegal under Medicare and Medicaid.2 ‘‘The bill wouldde�ne the term ‘any remuneration’ broadly to encom-pass kickbacks, bribes or rebates which may be madedirectly or indirectly, overtly or covertly, in cash or inkind.’’3

Ten years later, in 1987, the Secretary of DDHS wasgiven additional authority, requiring exclusion fromMedicare and Medicaid of any individual or entity whohas engaged in certain proscribed behaviors and permit-ting it in other situations. These authorities, along withconsiderably increased power to assess civil monetarypenalties in the amount of $50,000 for each violationwere further extended in the Health Insurance Port-ability and Accountability Act of 1996 (‘‘HIPAA’’), andthen even more so a year later in the Balanced BudgetAct.4

So, the anti-kickback statute was designed to preventwhat in the ordinary course of business would beunderstood to be kickbacks and bribes for referrals. Itwas clearly focused on cost, as stated in the legislativehistory. In its interpretation though, and likely toescalate since civil penalties do not require the govern-ment to meet the criminal burden of proof, the enforc-

2Sen. Fin. Comm. Rpt. to S.143, 10 (Sept. 26, 1977).3Sen. Fin. Comm. Rpt. to S.143, 12 (Sept. 26, 1977).4Subtitle A—Fraud and Abuse Control Program; Title XI of the

Social Security Act. 42 U.S.C.A. §§ 1320b et seq.


21


ers have come to sweep into their ambit virtually anytransaction in which there is an economic bene�t of anykind that redounds to a physician coincident with hisfederal patient referrals. Although safe harbor regula-tions were intended to quiet the considerable anxietiesof those in the industry regarding exactly how far thegovernment intended to go, one need only peruse theanti-kickback issues alone in the omnibus summariescontributed by Sandy Teplitzky and Craig Holden overthe fourteen editions of this Handbook, to learn justhow canonical the government has become both in thebreadth of its views of the statute’s scope and there�ned details of what relationships are protected ei-ther in safe harbor regulations or Advisory Opinions.We are well far a�eld from kickbacks and payments forreferrals.

That statutory scheme notwithstanding, others feltthat the law did not go far enough. In a crystallineexample of the inhospitality tradition, in 1989 ( e�ec-tive January 1, 1992), a law was enacted prohibitingphysicians from referring Medicare patients for the pro-vision of clinical laboratory services to entities withwhich they had a �nancial relationship, unless the rela-tionship met one of a number of exceptions. Introducedas the ‘‘Ethics in Patient Referral Act,’’ that law becameknown as ‘‘Stark I’’ after its sponsor. The statutory ver-sion became an issue because he subsequently intro-duced a major extension of the types of restriction whichhad initially applied only to clinical laboratory services.The second iteration of Stark became known as ‘‘StarkII,’’ and extended the services implicated by the anti-referral provisions to a range of other ‘‘designatedhealth services’’ including physical therapy, occupa-tional therapy, outpatient drugs, radiation therapy,imaging services including MRI, CT and ultrasound,durable medical equipment, home health services andall inpatient and outpatient hospital services.


22


Under the Stark II law,5 e�ective January 1, 1995, itis prohibited for a physician (or an immediate familymember) to refer a Medicare or Medicaid patient forany designated health service to an entity with whichthe physician or family member has a �nancial rela-tionship, unless the relationship meets one of seventeenstatutory exceptions. ‘‘Physicians’’ include podiatrists,optometrists, chiropractors (for whom Medicare other-wise recognizes exactly one service for payment) anddentists. It is prohibited to refer and the entity receiv-ing the tainted referral is prohibited from submitting aclaim. The statute de�nes a referral; and it is not thedirection of a patient to a source of business. Rather itis the mere ordering of a service, item or good, payableby Medicare—actually the statute says Part B, but thisis what happens with midnight drafting and is not tobe taken seriously—and the establishment of a treat-ment plan under which such a designated health ser-vice is to be provided.

Remarkably, then, a referral can entail no intentionalorientation of the patient to the entity where the�nancial relationship exists. Violations are punishableby $15,000 civil money penalty for each improper refer-ral and for each improper claim submission and$100,000 for a circumvention scheme. Payments madepursuant to a tainted referral are overpayments. Thereis also potential liability for false claims which areseparately punishable by triple the charges and up to$11,500 per improper claim.

Stark II de�nes �nancial relationships to include bothdirect and indirect compensation relationships. Stark IIde�nes a compensation relationship as any arrange-ment involving any remuneration between a physician(or an immediate family member of such physician) andan entity . . . . The term remuneration includes any re-muneration, directly, indirectly, overtly or covertly, in

542 U.S.C.A. §§ 1395nn et seq.


23


cash or in kind.’’6 Final regulations under Stark II haveonly been issued7 for part of the statute (‘‘Phase 1’’).Not all exceptions have been addressed to date inregulation. This has no e�ect on whether the statute isviolated or not.

The complexity created by the interrelationship be-tween two sets of laws addressing similar issues, in dif-ferent ways, in totally separate titles of the Social Secu-rity Act, not to mention the relatively scarce interpretiveguidance (and what there is arrives stunningly late)fosters its own inordinate compliance burdens. Considerthat the anti-kickback statute applies to all Medicare,Medicaid and, as of January 1, 1997, federally �nancedprograms, including CHAMPUS, Title V, etc. It appliesto whomever is involved in a relevant transaction.Violations can be criminal or civil, and an exclusionmay be imposed without a conviction or civil penalty. Ifyou do not conform with a safe harbor, then the situa-tion will be analyzed on a facts and circumstances basis.

Stark II, on the other hand, applies only to physicianreferrals and then only for the designated services inMedicare and Medicaid and where there is a �nancialrelationship between the referring physician and thereferred-to entity. There are two prohibitions underStark: (1) the prohibition against the referral; and (2)the separate prohibition against a claim submitted pur-suant to an improper referral. Stark does not o�ersafety zones. You either comply or you do not. For thesereasons, transactions are usually analyzed �rst underStark and then under the anti-kickback statute.

The anti-kickback statute has evolved over twenty-�ve years with increasingly broader and deeperapplication. The Stark statute was introduced in 1989based on then contemporaneous data about physicianreferrals to joint venture imaging centers in Florida.Both statutes were designed primarily for settings inwhich fee for service payment was dominant and

642 U.S.C.A. § 1395nn(h)(1)(A), (B).766 Fed. Reg. 855 (Jan. 4, 2001).


24


volumes of patients would increase revenue to the bill-ing entity. The continuing vitality of such concern isquestionable in a more managed environment andparticularly within managed care arrangements whereover-utilization is not the real risk; under-use is.8

Although there is no question that over-utilization tomaximize income is a quality issue, the complexity ofthese statutes goes well beyond that evil and views anyeconomic bene�t to physicians coincident with patientreferral as potentially improper. In many instances, theexact same actions in clinical terms when exercised in amanaged care setting would be ignored. This disjunc-ture demonstrates that these laws have little to do withquality. They barely focus on cost. They trigger intenseorganizational and structural analyses and restructur-ing based on where economic bene�t exists in relation-ship to patient referral. They distract enormous ener-gies in healthcare from its real problems and they missthe real point.9

The legitimate question to be answered in the

8The discount safe harbor regulations, managed care safeharbors and shared risk safe harbors are intended to control for is-sues associated with managed care discounting and other �nancialpractices of providers in exchange for increased and channeledmanaged care business, but they have little relevance to the qual-ity issues here.

9If the concern is whether Medicare pays too much for a ser-vice, fee schedules were created to establish in explicit terms thedollar value the public programs are willing to place on certainservices. If providers �nd a way to deliver services within thoseboundaries so as to make more money and can do so withoutcompromising patient quality, where is the problem? It appearsthat there is sub rosa, a ‘‘most favored nations’’ philosophy at workhere: if any customer can get the service at a lower price, theMedicare program should get that bene�t, too. There is, however,no such explicit statutory or regulatory provision in Medicare, al-though there is a provision which was adopted back in the pre-feeschedule, usual customary and reasonable era of Medicare physi-cian payment, which states that a provider can be excluded fromMedicare for charging the program substantially in excess of hisusual fees (42 U.S.C.A. § 1320a-7(b)(6)). There is no speci�edmethodology for determining those fees but the continued exis-


25


physician’s referral for any service is whether theprovider rendering the service meets quality thresholdsand whether that speci�c patient needs that speci�cservice on that speci�c day. Someone will make moneyfrom the referral service in any event. As long as thequality predicates are present, why does it matter whothat is? These statutes are completely incapable of ad-dressing the fundamental issue which can be solvedthrough the application of CPGs.

§ 13:8 —Physician Incentive Plan Regulations

The only purportedly directly quality-relevant pieceof this Byzantine regulatory scheme can be found in thephysician incentive plan regulations. These are an at-tempt to de�ne for two pieces of law: (1) the Stark stat-ute and (2) the separate statutory provision whichprohibits hospitals and prepaid health care organiza-tions from knowingly making incentive payments to aphysician (and prohibits physicians from accepting suchpayments) as an inducement to reduce or limit servicesto Medicare bene�ciaries or Medicaid recipients, allunder the same statutory provision. The physicianincentive plan statutory provisions were so contentiousthat from 1986 through 1990 they were enacted, re-pealed, and reinserted and then in 1993, as part of theStark II law were reintroduced as part of the physiciancompensation exceptions. 1 Thereafter, theMedicare+Choice program was enacted, taking theplace of the prepaid health plans that had been part ofthe original concerns around managed care paymentsto physicians. Congress then amended theMedicare+Choice provisions to allow physician incen-

tence of this provision raises at least questions if not eyebrows inan era of widespread managed care discounts. Again, however, thequality issues get lost in these cost control provisions.[Section 13:8]

142 U.S.C.A. § 1395nn(e)(3)(B).


26


tive plans that do not restrict medically necessary care.2If your map through this thicket now seems likesomething out of Lord of the Rings, we are not yet done.Regulations interpreting the delayed provision (whichstemmed from 1986) were originally published in 19963

and then were amended and revised in a new publica-tion at the end of the same year.4 When new regula-tions under Medicare+Choice were published in June1998, they announced an intention to adopt what hadbeen published in 1996 with some changes.5 Then, fur-ther regulations were published in June, 2000, makingminor clari�cations.6

Unlike the safe harbor regulations which de�ne onlywhat is safe, so anything not explicitly safe is analyzedon a facts and circumstances basis, these regulationscontain mandates. They confront circumstances whereMedicare managed care plans put individual physiciansand their ‘‘downstream’’ organizations at �nancial riskfor more than their direct services to patients. The goalis to prohibit any speci�c payment of any kind, directlyor indirectly, under an incentive plan, to a physician orphysician group as an inducement to reduce or limitcovered medically necessary services furnished underthe managed care plan’s contract furnished to an indi-vidual enrollee.7 The basic intent is unassailable, but aswe consider the regulatory approach, keep in mind thattwo years after their publication we still have almostno data about the actual impact of speci�c forms ofphysician incentives in either economic or clinicalterms. Still further, two years post-publication, econo-mists tell us:

The complexity of physician behavior, the emergence of

242 U.S.C.A. § 1395w-22(j)(4).361 Fed. Reg. 13,431 (Mar. 27, 1996).461 Fed. Reg. 69,034 (Dec. 31, 1996).563 Fed. Reg. 34,968 (June 26, 1998).665 Fed. Reg. 40,170 (June 29, 2000).742 C.F.R. § 417.479(d) (2002); 42 C.F.R. §§ 422.208, 422.210

(2002).


27


payment methods that blend fee for service and capita-tion, the interdependence of price mechanisms withnonprice mechanisms, the salience of organization as asupport for compensation systems, and the remarkablevariety and continual change in all arenas suggest thatpublic policymakers should adopt a stance of intellectualhumility and a tone of cautious optimism.8

Neither is evident in the physician incentive planregulations, which, in all fairness, were required byCongressional prescription.

The primary technique to safeguard enrollees was tode�ne in what circumstances physicians were placed at‘‘substantial �nancial risk’’ by a Medicare health planfor services more than their own. ‘‘Substantial �nancialrisk’’ was the topic garnering the most comments onthe March 1996 publication. Where the patient riskpool is less than 25,000 the plan must calculate theswing from lowest to highest potential amounts forwhich the physicians are at risk. If the result represents25% of the physician group’s possible compensationthen there is substantial �nancial risk. Bonuses, with-holds, capitation, combinations of them or any otherform of incentive must be taken into account. Astonish-ingly, given the pace of health care change, the analysison which even the 2000 regulations were based reliedon data in a 1990 study, which by de�nition has tore�ect an even earlier time. Where such �nancial riskexists, the protections are (1) required disclosure of howthe plan incentivizes its physicians, both to the govern-ment and upon request to bene�ciaries; and (2) stop-loss coverage and bene�ciary surveys if such risk ispresent. Where the risk devolves to intermediate enti-ties, they become part of the reporting chain. On theother hand, for all its breadth of application on the riskcontinuum, there is nothing in these regulations thatprohibits any transactions or sets parameters for riskassumption in any way. These are purely reporting and

8Robinson, ‘‘Theory and Practice in The Design of PhysicianPayment Incentives,’’ 79 The Milbank Quarterly 174 (2001).


28


informational requirements combined with a stop-lossinsurance mandate.

In considering their e�ect, in 1996 one commenter onthe regulations noted that quality and service bonusesseemed not to be the type of risk the regulators wereconcerned about. The commenter argued that bonusesfor high quality, full service capacity, long o�ce hours,accepting all new patients and cost-e�ectivenessrepresented no downside risk.9 The regulators’ response:

While we are supportive of a quality bonus payment,there is very limited experience with its use, andwhether a physician will actually receive it isspeculative. We will revisit the issue when more infor-mation is available on the nature, extent and experiencewith quality bonuses.10

The one real change in the �nal regulations is thatthe regulators woke up to allow that ‘‘Bonuses and othercompensation that are not based on use of referrals,such as quality of care furnished, patient satisfaction orcommittee participation, are not considered paymentsin the determination of substantial �nancial risk.’’11

As a practical matter the disclosures often require in-dividual physicians to �ll out forms they then send totheir organizational entities which then send them tothe plans. The disclosures purportedly were to be usedby the government ‘‘to monitor compliance, evaluatethe impact of the regulation, and ensure the delivery ofhigh quality care.’’12 In the overall maelstrom regardingphysician �nancial incentives, it should be rememberedthat states weighed in too, enacting laws to address

961 Fed. Reg. 69,041 (Dec. 31, 1996).1061 Fed. Reg. 69,041 (Dec. 31, 1996).1142 C.F.R. § 422.208 (2002) in the de�nition of ‘‘potential

payments.’’1261 Fed. Reg. 69,037 (Dec. 31, 1996).


29


physician incentives,13 thereby simply accreting moreand more layers of administrative burden on all theplayers. In considering the value of all this administra-tive weight, it should be noted as well that the OIG hastargeted physician payment incentives in Medicaremanaged care in its Work Plan in every year from 1996to 2001. Medicare managed care is reviewed by PROs.14There are still separate and complex quality programsimposed in Medicare + Choice. It would be hard not towonder if there isn't a real issue of regulatory overkillhere, for something about which we do not even havedata de�ning the problem.15

Now consider our hypothetical market. Case ratesbased on CPGs, de�ned fee for service payment withina range as determined by a CPG—these do not entail�nancial risk under the regulatory de�nition. Paymentis predetermined and re�ects the clinical needs of thepatient. Variability turns only on the patient’s condi-tion, which, in a well designed system would incorporate‘‘case breakers’’ or co-morbid conditions that shift thecare outside the de�ned CPG, thereby triggering analternative payment methodology. The easiest defaultin that instance would be straight CPT based fee forservice payment. How relatively simple by comparisonwith all of these questions about �nancial incentives!The proposed approach gives to physicians medicalmanagement risk while retaining to the insurance

13See Anderlik, ‘‘E�orts to Regulate Physician Financial Incen-tives,’’ (Jan. 20, 1999), http://www.law.uh.edu/healthlawperspectives/Managed/990120E�orts.html.

1442 U.S.C.A. § 1320c(3)(a)(4)(B). Medicare + Choice has its ownquality managed and reporting requirements in the QIMSC (Qual-ity Improvement System for Managed Care) and QAP1(QualityAssessment and Performance Improvement) provisions of theregulations and operational policy letters. http://www.hcfa.gov/medicare/mgdqual.htm.

15For a discussion of the essentially unmanageably complexregulatory context for Medicare + Choice, see Krasner, ‘‘Medicare +Choice: Swamped by Regulation,’’ Health Law Handbook 479-500(Gos�eld, ed. 2000).


30


entity, actuarial or incident risk, for which the insureris licensed but the physician group is not. Still, thephysician incentive plan rules apply to Medicare andMedicaid managed care entities which receive paymentsfrom the government and then manage �nancial riskbased on that government compensation. Another re-lated technique to motivate physician behavior wasbrie�y far more popular at the hospital-physician leveland even without formal regulations garnered anotherless than hospitable government response.

§ 13:9 —GainsharingLike much in healthcare, the concept of gainsharing

had a long history before its application by hospitals inrelationships with physicians. Beginning in the 1930sas a form of compensation referred to as the ScanlonPlan, through a variety of other iterations, techniquesto motivate workers by giving them a share in the suc-cess of the enterprise have been present up to and intothe current era, primarily, however, in manufacturingindustries:

Simply put, gainsharing is a group bonus plan—as op-posed to more traditional individual plans—designed toencourage workers to surpass previous company perfor-mance standards such as productivity, quality, or anyother objectives corporate management can adequatelymeasure. When the program is e�ective, the extra pro�tsor savings are paid out in cash above and beyond salary.Generally, the workforce receives about half of the gainsand the company garnishes an equal share.1

Where gain sharing has been successful it has beenimplemented as part of overall team involvement, withrealistic expectations among the workers, and carefulcommunication of the goals and measures selected.

[Section 13:9]1Taylor, ‘‘Gainsharing the Wealth More Companies Turn to an

Old-Fashioned Incentive,’’ Reengineering Resource Center (Aug.1996), http://www.reengineering.com/articles/aug96/gainsharing.htm.


31


‘‘Gainsharing is really designed to support a culture ofhigh involvement, teamwork and empowerment, and itdoesn't work as well in other kinds of environments.’’2

Medicare DRG payments to hospitals are intended topressure them to manage their own costs; yet the physi-cian ‘‘workers’’ on whom part of their productivitydepends are paid in a way which incentivizes them inthe opposite direction. It is not surprising, then, thatgainsharing would surface as a physician motivationtechnique, especially in settings where dominant physi-cian groups are responsible for large portions of hospitalexpenses based on their ordering, prescribing and treat-ment regimens. Hospitals do not practice medicine andcannot provide any of their services without a physicianorder to do so. Physician behavior, then, can have directand major impacts on hospital revenues which is whywe have so much rhetoric about aligning incentives inthe hospital sector long after that chant has faded frommanaged care.3

Against this background, and heavily marketed byconsulting �rms, hospital-physician gainsharing focusedmost often on speci�c high cost DRGs, often cardiac innature. Hospitals proposed to enter into �nancial ar-rangements of various types with their physicians toreward them for helping to cut hospital costs. Whetherthese programs were e�ective aside, the Inspector Gen-eral was not positively impressed.

2Taylor, ‘‘Gainsharing the Wealth More Companies Turn to anOld-Fashioned Incentive,’’ Reengineering Resource Center (Aug.1996), http://www.reengineering.com/articles/aug96/gainsharing.htm. The law at issue for our consideration is the 1986 provisionwhich originally included prepaid health plans in the incentiveplan prohibitions.

3Some who are serious worshippers at the shrine of economicsuspicion might argue far more strongly for checks and balancesmodel with intentionally diverging incentives. Whichever orthodoxyone follows though, it is fair to say neither model is purely presentin today’s world.


32


In a ‘‘Special Advisory Bulletin,’’4 the OIG character-ized the law on payments to physicians to reduce orlimit services to bene�ciaries as follows:

The statutory proscription is very broad. The paymentneed not be tied to an actual diminution in care, so longas the hospital knows that the payment may in�uencethe physician to reduce or limit services to his or herpatients. There is no requirement that the prohibitedpayment be tied to a speci�c patient or to a reduction inmedically necessary care. In short, any hospital incen-tive plan that encourages physicians through paymentsto reduce or limit clinical services directly or indirectlyviolates the statute.5

This is not mere inhospitality. This is battening downthe hatches and heading for the air-raid shelter. Stillfurther, reduction alone does not account for the factthat we now know that much of poor quality in healthcare involves over-utilization or mis-utilization.6 Reduc-ing and limiting inappropriate services, in fact, is oftenthe pro-quality policy. The OIG attempts to acknowl-edge that some good might result from gainsharing.

The OIG recognizes that hospitals have a legitimateinterest in enlisting physicians in their e�orts to elimi-nate unnecessary costs. Savings that do not a�ect thequality of patient care may be generated in many waysincluding substituting lower cost but equally e�ectivemedical supplies, item or devices; re-engineering hospitalsurgical and medical procedures; reducing utilization ofmedically unnecessary ancillary services; and reducing

4O�ce of the Inspector General, ‘‘Gainsharing Arrangementsand CMPs for Hospital Payments to Physicians to Reduce or LimitServices to Bene�ciaries,’’ (July 1999), http://oig.hhs.gov/frdalrt/gainsh.htm.

5O�ce of the Inspector General, ‘‘Gainsharing Arrangementsand CMPs for Hospital Payments to Physicians to Reduce or LimitServices to Bene�ciaries,’’ (July 1999), http://oig.hhs.gov/frdalrt/gainsh.htm.

6Chassin, Galvin, et al., ‘‘The Urgent Need to Improve HealthCare Quality,’’ 280 JAMA 1000 (Sept. 16, 1998), and Corrigan, etal., National Academy Press (Washington, D.C. 2001).


33


unnecessary lengths of stay. Achieving these savingsmay require substantial e�ort on the part of the partici-pating physicians. Obviously, a reduction in health carecosts that does not adversely a�ect the quality of thehealth care provided to patients is in the best interest ofthe nation’s health care system. Nonetheless, the plainlanguage of section 1128A(b)(1) of the Act prohibits ty-ing the physicians’ compensation for such services toreductions or limitations in items or services provided topatients under the physicians’ clinical care.7

So, for physicians to expend what the OIG acknowledgeswould be ‘‘substantial e�ort’’ to assist their hospitalpartners to cut costs cannot be recognized economicallyif there is a reduction in services merely coincident withthe payment. It is di�cult to know what ‘‘tying’’ ofcompensation entails. Earlier in the Bulletin, the OIGsays that the reduction itself need not even occur, it isthe mere encouragement of reduction which is at issue.Citing a high risk of abuse and the di�culty in deter-mining if there were adequate and accurate measuresof quality of care that would provide assurance thatthere is no adverse impact on patient care, the OIGrefused to issue a safe harbor to address gainsharing,and indicated it would not even issue Advisory Opinionson the practice. The e�ect went beyond chilling toglacial. Most gainsharing programs were abandonedimmediately.

It was with some surprise, then, that eighteenmonths later the industry read an OIG Advisory Opin-ion8 not only precisely on point, but far more positive inits views. Unlike many of the gainsharing proposalswhich had been circulating at the time of the SpecialBulletin, here a group of cardiac surgeons would bepaid 50% of the �rst year cost savings appreciated bythe hospital as a result of the surgeons using speci�c

7O�ce of the Inspector General, ‘‘Gainsharing Arrangementsand CMPs for Hospital Payments to Physicians to Reduce or LimitServices to Bene�ciaries’’ (July 1999).

8OIG Advisory Opinion No. 01-1 (Jan. 11, 2001).


34


surgical supplies only as needed, agreeing to thesubstitution of less costly items and supplies in thespeci�ed surgical procedures and limiting the use of apre-operative medication to prevent hemorrhaging. It isnoteworthy that the physicians were not exclusivelycontracted to the hospital, but accounted for more than85% of the cardiac surgical procedures performed in thefacility. The compensation program was not even madeavailable to other physicians with privileges to performthe same procedures.

Savings were to be measured against a historical baseand reductions below the �oors for the then current us-age rates would not be included in the compensation tothe surgeons. The recommendations from outsideconsultants as to how to reduce costs, on speci�c DRGsfor speci�c procedures, constituted 19 separate measur-able actions. Cost savings at the end of the year wouldbe calculated for each recommendation separately.

A number of other safeguards were present in theproposal. No payment would be made for cost savingsgenerated by increases in the numbers of procedures,so that increased referrals would not increase thecompensation to the physicians. The payment programwould be managed and data maintained by an outsideProgram Administrator. If the data showed that morecostly patients were steered to other facilities duringthe year, the physician(s) making such referral deci-sions would be terminated from the program. Thesurgeons and the hospital would disclose the arrange-ment to patients prior to admission and if the decisionto operate was made after admission, prior to patientconsent.

The OIG indicated its concern with four aspects ofgainsharing generally: ‘‘(i) stinting on patient care; (ii)‘cherry picking’ healthy patients and steering sicker(and more costly) patients to hospitals that do not o�ersuch arrangements; (iii) payments in exchange forpatient referrals; and (iv) unfair competition (a ‘race tothe bottom’) among hospitals o�ering cost sharingprograms to foster physician loyalty and to attract more


35


referrals.’’9 The OIG analyzed the arrangement underthe hospital-physician incentive civil money penaltylaw and under the anti-kickback statute, ignoring theStark law as outside its purview.

Addressing the civil money penalty issue the OIGacknowledged that reduction of care is sometimes ap-propriate but said ‘‘whether current medical practicere�ects necessity or prudence is irrelevant for purposesof the CMP.’’10 Reduction in improper services wouldviolate the statute too, according to this analysis, inany instance in which the reduction was given economicrecognition in payment to the physicians. Surprisingly,then, announcing that they would not take actionagainst the participants, the OIG distinguished thesafeguards in this narrow and well de�ned programfrom the tenets of most gainsharing they had seenwhich lacked veri�able measures, did not link individ-ual action to speci�c payment amounts, did not haveany mechanisms to take into account other factorsbesides the physician actions as having produced thesavings, and, among other things, claimed qualityindicators which were of questionable validity andstatistical signi�cance.

The tandem piece was the anti-kickback analysiswhere the OIG looked to the personal services andmanagement services safe harbors. The touchstone ofboth of these, is, of course, fair market value, allowingpayments that do not re�ect the volume or value ofreferrals. Although the payment of 50% of savings doesnot meet the safety factor of aggregate compensationstated in advance, the baseline measure against histori-cal data acted as a cap on the amount the physicianswould get. The formula also speci�cally excludedincreased referrals as contributors to the cost savings.Taken together the OIG found a way to position itsanalysis as consistent with its prior sweeping hostility.Given that the Stark statute also has a personal ser-

9OIG Advisory Opinion No. 01-1, 6 (Jan. 11, 2001).10OIG Advisory Opinion No. 01-1, 8 (Jan. 11, 2001).


36


vices exception, and that fair market value is also thegravamen of its protections, it is unlikely that againsharing program found to meet OIG requirementswould be seen to violate the Stark statute on thesepoints.

Still, in the overall analysis of gainsharing andhospital-physician collaboration, it is interesting thatthe requestor of the Advisory Opinion asserted—albeitin the form of a ‘‘certi�cation’’ —that ‘‘this paymentmethodology will generate payments to the SurgeonGroup that will be consistent with fair market value forservices rendered to the hospital in arms-lengthtransactions.’’11 The OIG does not dwell on, but musthave accepted this assertion, since there can be nosafety in something that does not re�ect fair marketvalue, although the OIG generally refuses to opine onthe subject because it is statutorily precluded from do-ing so. Here it stepped out to say ‘‘While we areprecluded from opining on whether a payment is fairmarket value, the payments under the Proposed Ar-rangement do not appear unreasonable, given amongother things, the nature of the nineteen recommendedactions, the speci�city of the payment formula and thecap on total remuneration to the Surgeon Group.’’12Reasonableness is not, despite the OIG’s good inten-tions, synonymous with fair market value, at least ac-cording to their own regulations, which merely statethat fair market value is that which pertains ‘‘in arms-length transactions.’’ Where does this lead us? To a farbetter land than the one we have traversed.

§ 13:10 —Possibilities Explored

The physician incentive plan regulations pose moreof an administrative burden than an operational one,since they neither prohibit nor mandate any form ofincentive. Payment based on clinical guidelines cannot

11OIG Advisory Opinion No. 01-1, 6 (Jan. 11, 2001).12OIG Advisory Opinion No. 01-1, 13 (Jan. 11, 2001).


37


be seen as falling within the ‘‘substantial �nancial risk’’de�nition of those regulations. Good news for our hypo-thetical market.

In contrast, the views of the OIG on hospital-physician �nancial alignments is not a welcoming one.Still, two relevant and positive conclusions can bedrawn—one directly on point, and one positive in itsimplications: (1) When properly designed, narrowlyfocused, and �xed in amount, duration and scope,�nancial relationships between hospitals and physi-cians that are intended to improve the quality-safetyenvironment can be crafted to e�ectively navigate thehills and valleys of these still barely charted terrains;and (2) There is no evidence cited or yet available on abroad basis that paying the physicians a part of thehospital’s savings for doing the hospital’s work is eithere�ective or is a sustainable business model for this crit-ical relationship. Once the initial savings accrue, if theydo, how much more cutting can the physicians gener-ate? Since the compensation for their economic valuecomes out of lowered expenses and the least expensiveresources are none at all, there is a very �nite bound-ary to the longevity of these arrangements. This mayreassure the OIG but it does not create a decent modelfor an on-going relationship. If the physicians are notmotivated to make these changes without the reward ofa piece of the savings, what will make them do it afterthe initial payment? Why would they not revert to priorbehavior? Moreover, while physicians are responding tothat incentive, what are they doing about their morebasic challenges in terms of solidy�ng their doctor-patient relationships by, for example, increasing theirtime with those patients, lowering expenses in thepractice so less empathetic patient throughput is notwhere they end up? Still further, it is hard to imaginethat disclosing to the patient the type of �nancial rela-tionship in the proposed arrangement will strengthenthe bonds of trust between doctor and patient.

How can the payment in a model like the proposedarrangement be considered fair market value? This


38


could only be the case if other consultants who do notrefer patients and have the ability to a�ect the actualimplementation of programs (and one would ask howthat could happen?) charge and are paid on the samebasis. This is also speculative at best. How can all ofthis be positive? Taken together, the best message maybe that gainsharing as a business model betweenhospitals and physicians does not merit anywhere nearthe attention it has garnered to date, thereby liberatingboth parties to focus on what each does best.

As to how they can work more e�ectively together, letus consider the many activities within a hospital aroundwhich physician input is critical to de�ne, create andsustain a clinical culture of quality.1 The long traditionof organized medical sta�s has had as its core responsi-bility, delegated from the hospital board to the physi-cians, the selection of the pool of physicians who wouldjointly utilize the facilities of the institution (privileg-ing, credentialing, quality monitoring of the physicians).This same tradition has depended on the volunteerspirit of the sta� members to contribute considerabletime and energy in the exercise of this function. Al-though viewed as heresy in some physician quarters,some hospital advocates have long held that this activ-ity should be assigned to paid consultants or employees.2The economic value of these judgments to the hospitalin terms of risk management and quality performanceis obvious. There is no reason physicians cannot be paidfor these activities on a fair market value basis. Suchan approach would assure that many of the failures of

[Section 13:10]1For a consideration of why the role of physicians is so

important in hospitals and a consideration of the speci�c valuesphysicians bring to bear in what they do and where hospitals oughtseek to bring them to the table as true collaborators, see Gos�eld,Quality and Clinical Culture: The Critical Role of Physicians inAccountable Healthcare Organizations, American Medical Associa-tion (1998), http://www.ama-assn.org/ama/pub/article/371-477.html.

2Horty, ActionKit for Hospital Law (Mar. 1989).


39


physicians to exercise these roles e�ectively3 (inconsis-tent attendance at meetings, failure to evaluate datacompletely, giving short shrift to di�cult judgmentsbecause of time pressures, among others) would beremedied.

In our hypothetical community, the most criticaldriver within the hospital and throughout the contextis CPGs. The model will not work if the evidence basedfoundations for it—the CPGS—are not selected care-fully and wholeheartedly endorsed by the physicianswho will live by them for their own direct payment, andfor most other activities in which they engage. Conse-quently, to select, implement, monitor and safeguard astimely and valid the CPGs at work in the system arefundamental activities for which physicians can be paidfair market value on an on-going basis. Analysis of thedata which emerges from tracking conformity toguidelines or deviation from them and its implicationsfor improved care also must involve the physicians. Re-lated activities pertaining to elimination of clinicalmistakes by the physicians as part of the care teamcannot be conducted without physician involvement.Financial and non-�nancial incentives to be used withinthe facility or system to sustain physician engagementcannot work without physician input in their establish-ment and operation. All of these activities can be seenas fairly consistent with, although something of anexpansion in, traditional medical sta� roles.

A broader view of the value of physicians to the modelwould take into account a variety of hospital relatedfunctions in which physicians, historically, have beenless involved. Payor contract negotiations where the is-sue is whether the payor’s rate can support the qualityat issue in the applicable CPGs is rarely an activity to

3See Gos�eld, Quality and Clinical Culture: The Critical Role ofPhysicians in Accountable Healthcare Organizations, AmericanMedical Association (1998), http://www.ama-assn.org/ama/pub/article/371-477.html, for a discussion of the ways in which physi-cians undermine their own e�ectiveness as collaborators.


40


which physician leadership has been invited. Mosthospital �nancial o�cers believe that physicians cannotunderstand the �nancial issues at work in thesecontracts. As a result, not surprisingly, physicians oftendistrust the resulting payment arrangements that thefacility administration negotiates, even where thephysician’s own payment is not on the table. Still fur-ther they often believe, and usually are right, that thehospital’s �nance driven negotiators have little under-standing of the fundamental clinical issues these pay-ment terms create. Basing these negotiations on CPGscan highlight in far more obvious terms where thepayor’s proposed rate will compromise quality.4 Finally,where CPGs drive payment, negotiations can turn onthe clinical implications of payment, which, as we haveseen, is virtually impossible in traditional paymentmodels, whether fee for service, capitation, or itshospital equivalent—DRGs—which are based on histori-cal utilization of resources and not patient clinicalneeds.

One of the most fundamental challenges for everyonein health care today is how to maximize the use of in-formation systems without creating undue administra-tive burden. Physicians are rarely included in thediscussions within hospitals about the choice of theirsystems until the selection and capital investment havealready been made. Typically, the hospital has had dif-ferent information technologies for di�erent activities(billing and collection, medical records, �nancial report-ing) and none of them has taken into account their clini-cal implications nor whether the reports they generatecan be understood or used by the physicians whom thefacility would seek to engage. In our hypothetical sce-nario, the information systems and their use are criti-cal to the operation of the model. The involvement of

4The success of such an approach, however, depends on a solidunderstanding of what it costs to treat a patient for a condition.This understanding has to start from a basic statement of whatthe patient needs to have done by de�nition a CPG.


41


the physicians in the selection and design of thesemechanisms is, therefore, essential if the model is towork e�ectively and not smack of a cram down ap-proach which is neither useful nor trustworthy as faras the physicians are concerned. Their involvement herecan also be recognized with fair market value paymentsfor their input.

Three other basic functions in our model sustain itsclinical consistency: strategic planning (what is thebusiness of the system and the brand we seek to proj-ect); budgeting (do we have the appropriate resourcesin place to make our strategic plan happen) andmanpower planning (do we have the appropriateconstellation of personnel to ful�ll these goals). Mosthospitals have typically not only not involved physi-cians in these processes, they have not taken into ac-count in these critical management issues the clinicalmandates that our posited scenario entails. Unlessthese activities are built on a bedrock of evidence basedclinical purpose, the resulting system will simplyreplicate the failures identi�ed in Crossing the QualityChasm. To make quality happen, a di�erent approachto these basic business judgments will be necessary.The more consistent they are with quality needs themore the entire approach inherently makes the busi-ness case for quality at all levels of the enterprise.

This discussion has unfolded in the context of theimplications of the anti-kickback and anti-referral lawswhich are fundamentally about �nancial relationships.At the same time, cost pressures in the system haveincreased; so, where will the money come from to paythe physicians for the economic value they bring to thehospital in its quality-driven e�orts? Facilities andsystems adopting these principles will, obviously, haveto make judgments regarding which of these functionsmerit payment. But, it can be expected that where theoperational principles coincide for the physicians andthe facility—where the same CPGs are used in the of-�ce, in the hospital, and in relationship to payors—where consistency is present throughout, physicians


42


will see their own self interest in fostering a settingwhere there is less administrative burden on them anda greater opportunity to do the right thing for theirpatients. Where they have the ability to maximize theirown e�ciencies by using a consistent logic, their longtradition of volunteerism can be called on to sustaintheir interactions with the hospital. The ultimate bal-ance between fair market value recognition of physi-cian economic value as quality collaborators andvolunteerism to make a better clinical culture can onlybe de�ned within the speci�c setting where such asystem is adopted. Along the way yet another complexregulatory landscape awaits.

§ 13:11 HIPAAPerhaps the most daunting, time consuming to

understand, complicated, labor and systems intensiveregulations to hit health care in the last thirty yearsare those published under HIPAA—not under the fraudand abuse provisions which were hair raising them-selves, but under the oxymoronically named ‘‘adminis-trative simpli�cation’’ provisions.1 The initial intenthad been to simplify claims processing by creatinguniform standards for electronic claims reporting. Therewas signi�cant concern for the security of thesemechanisms.2 There was also attention paid to patientcon�dentiality and privacy so, in a bizarre reversedefault provision, the statute provided that if Congressfailed to enact privacy standards within three years ofthe law’s enactment, the Secretary of HHS was directedto do so. The privacy standards were contemplated asonly one piece of a complex set of requirements for a

[Section 13:11]1Part C, Title XI, Social Security Act; 45 U.S.C.A. §§ 1320d et

seq.2For a brief discussion of legal authorities dealing only with se-

curity measures, which will not be discussed further here, seeShay, ‘‘A Survey of Security in Electronic Health Information,’’Health Law Handbook 27-53 (Gos�eld, ed. 2001).


43


more electronic, uniform, claims submission system inhealth care. These components included transactionsand data elements, unique health identi�ers, securitystandards, electronic signatures, code sets and transferof information among health plans. The focus was onhealth plans and electronic interchange of data.

Despite considerable e�orts to stimulate congressio-nal action, Congress defaulted. So, in a parting shot atthe end of the troubled Clinton Administration, somefour years after the law’s enactment, privacy standardswere published on December 28, 2000.3 The prefacealone is 336 pages in the Federal Register to explain 31pages of regulations. This is no Dionysian rhapsody.This is a hyper-manic Bacchanalian frenzy. The actualcontents of the regulations are far worse if we are seek-ing the smooth interchange of data among critical qual-ity collaborators who must be engaged to facilitate qual-ity, not only in the aspirational terms of Crossing theQuality Chasm, but in today’s vineyards of qualityimprovement. In these regulations, truly, anything thatis not mandatory is prohibited and what is mandatoryis considerable. The hue and cry which resulted is sowell known that it is not necessary to catalogue or citeit here, although by DHHS’s own count they receivedmore than 11,000 letters and comments on the �nalregulations, let alone the proposed ones.

By July, 2001, the O�ce of Civil Rights in DHHS4

sought to quell the turmoil with an easily digestiblelisting of Frequently Asked Questions (FAQs) posted on

361 Fed. Reg. 82,462 (Dec. 28, 2000).4That OCR has administrative responsibility for the law says

something signi�cant about how rules intended to facilitate healthcare system data interchange became rules on patients’ rights. Themission of the O�ce in its own words follows: ‘‘The Department ofHealth and Human Services, through the O�ce for Civil Rights,promotes and ensures that people have equal access to and op-portunity to participate in and receive services in all HHSprograms without facing unlawful discrimination. Through preven-tion and elimination of unlawful discrimination, the O�ce for CivilRights helps HHS carry out its overall mission of improving the


44


their website in the form of Guidance.5 These simpli�ca-tions themselves print to 51 pages in length. Acknowl-edging the regulatory philosophy set forth in this docu-ment is critical to understand how the privacyregulations will a�ect quality for those who fall withinits domain: health plans, health care clearinghouses,and providers which transmit protected health informa-tion electronically.6 The compliance date is April 14,2003.

§ 13:12 —Basic Regulatory Philosophy andContents

That there is a need for con�dentiality and privacy ofpersonal health care information is not in question. Themedical record spawns interchange and commerce inhealth care information:

Insurers freely romp through it. Fragments zoom fromagency to agency. Computer �les are hacked, faxes aresent to the wrong number, messages are left on thewrong machine, wastepaper �ies through the streets—entropy scatters con�dential medical data everywhere.Malice and happenstance can also play a role: yourneighbor’s husband lands a clerical hospital job—asactually happened to one of my patients—and next thing

health and well-being of all people a�ected by its many programs.’’http://www.hhs.gov/ocr/mission.html.

5OCR Guidance, ‘‘Statements for Privacy of IndividuallyIdenti�able Health Information,’’ TA 164.000.00 (July 6, 2001),http://www.hhs.gov/ocr/hipaa/�nalmaster.wpd.

6Because of the profound complexity in the regulatory schemeand the fact that this consideration is limited to a basic understand-ing of regulatory philosophy and how it runs at cross purposeswith quality, references in this discussion will be made to thegreatest extent possible to the Guidance. Where the guidance doesnot address an issue, the citation will be to the regulation itself.


45


you know the intimate details of your venereal diseasesare all over town.1

These regulations are designed to prevent and punishthese occurrences.

The basic philosophy of the regulations is that patientconsent must be obtained for the use and disclosure ofpersonal health information (‘‘PHI’’) to be used in treat-ment, payment or healthcare operations (‘‘TPO’’). Verystrong patient control over health care data bothmaintained and disseminated about them is a hallmarkof these rules. The regulations apply only to ‘‘coveredentities’’—health care providers that transmit PHIelectronically, plans and health care clearinghouses.The rules are not the same for all players. Specialburdens are imposed on providers. For them to conductthe basic functions of TPO even if they do not share theinformation with any other party, a patient consentmust be obtained. The only exception among healthcare providers is those who have an indirect treatmentrelationship with the patient, such as a pathologist andin some instances a radiologist. Although the consentneed be obtained only once, it can be revoked andrestricted by the patient. The covered entity (physician,hospital, other provider) may refuse the patient treat-ment and need not accede to the restriction but is boundby any restriction to which it agrees. Health plans anddata clearinghouses do not need consent to use PHI inTPO.

The consent burden is individual; and the consentobtained by one covered entity may not be relied uponby another. The absurdity of this requirement is suchthat OCR admits that a pharmacist cannot use basicinformation about a patient when a physician calls in aprescription, unless the patient has provided a writtenconsent to the pharmacist; nor may a specialist or

[Section 13:12]1Zuger, ‘‘On Record: Con�icting Obligations,’’ New York Times

F5 (Jan. 1, 2002).


46


U51014

Author: Revise with 3/27 proposed amendments or not?

hospital to whom the patient is presenting for the �rsttime use PHI to set up appointments or schedule proce-dures!2 Although there is an exception for emergencysituations, there is no de�nition of these and the healthcare provider ‘‘must exercise their professional judg-ment to determine whether obtaining consent wouldinterfere with the timely delivery of necessary healthcare.’’3 The Secretary states that DHHS intends topropose modi�cations to these aspects of the rule.

Consent, however, is only the �rst part of the patientcontrol features. For any use or disclosure which is notwithin the de�nitions of TPO an ‘‘authorization’’ mustbe obtained. Unlike consents, all covered entities mustobtain ‘‘authorizations’’ for uses which are not for TPO.An authorization is a customized document whichstates the proposed uses and disclosures of the PHI; ithas an expiration date. There are even stricter stan-dards for certain categories of data. A provider wouldneed to obtain an authorization (the longer, more oner-ous form of patient approval) to disclose PHI maintainedin psychotherapy notes.

The third administrative component of the informa-tional aspects of these rules is the requirement thatentities provide patients with ‘‘Notice’’ of their TPOuses of data so the consent is meaningful.4 How mean-ingful is a question since, although the Notice muststate how the provider uses PHI, the rules do not

2OCR Guidance, ‘‘Statements for Privacy of IndividuallyIdenti�able Health Information,’’ TA 164.000.00, 9 (July 6, 2001),http://www.hhs.gov/ocr/hipaa/�nalmaster.wpd.


445 C.F.R. § 164.520 (2002).


47


require that the individual read the Notice nor that thecovered entity explain each item in the Notice.5

The Notice is similar to those required of �nancialorganizations but goes further. The Notice must de-scribe and include at least one example of the types ofuses and disclosures the covered entity is permitted tomake for each of the elements of TPO. Other speci�clanguage must be included; and the Notice must alsoinclude a description of each of the other purposes forwhich the entity is permitted or required to use or dis-close PHI without consent or authorization. Although amere list of the categories noted in the regulations canbe crafted without real di�culty, it is doubtful theregulators themselves could state all the implied ap-plications of data under the permitted uses. Still fur-ther for each of fund-raising, appointment reminders,or a group health plan disclosure to the sponsor of theplan, a separate speci�c statement must be included.Six enumerated patient’s rights must be set forth.Reportedly the American Hospital Association drafted amodel Notice which was twelve pages in length!

Where disclosure of PHI will be made and is permit-ted, the rule imposes boundaries on what may bedisclosed, even after consent has been obtained. Thedisclosure must be limited to the ‘‘minimum necessary’’PHI to accomplish the purpose. This limit does notpertain to disclosures to or requests by a health careprovider for treatment purposes, pursuant to an autho-rization when it is requested by the individual, uses ordisclosures required for compliance with the standard-ized transactions rules, to DHHS for enforcement or asotherwise required by law.

The ‘‘minimum necessary’’ standard has generatedprofound concern in the healthcare industry. From oneend of the implementation spectrum many are con-cerned over the labor required to determine what data



48


to select as ‘‘the minimum necessary’’ and whether thatjudgment will stand up to the heat of scrutiny fromenforcers. At the other end is concern as to whether itis ever permitted to turn over a complete medicalrecord. Obviously such a minimum necessary standardcannot incorporate de�nitions to cover all situationswhere PHI will be disclosed. ‘‘The rule requires coveredentities to make their own assessment of what PHI isreasonably necessary for a particular purpose, giventhe characteristics of their business and workforce, andto implement policies and procedures accordingly.’’6 TheGuidance attempts to emphasize the reasonablenessstandard which the government expects providers toapply to make this standard real. Although some com-menters raised concern over the need to restructureexisting work�ow systems, including redesign of o�cespace and upgrades of computer systems to meet thesestandards, the OCR has said such wholesale redesign isnot necessary, On the other hand ‘‘covered entities mayneed to make certain adjustments to their facilities tominimize access, such as isolating and locking �lecabinets or records rooms, or providing additional secu-rity, such as passwords on computers maintainingpersonal information.’’7

When the covered entity uses subcontractors toperform TPO, the rule goes further and requires themto inquire of and demand of their ‘‘business associates’’assurances that they have in place many of the samesafeguards. Here the rule graciously at least allowsproviders and plans to give PHI to business associateswithout additional consent, but still conditions suchdisclosures on the provider or plan obtaining throughcontract ‘satisfactory assurances’ both that the business




49


associate will use the information only for the purposesfor which they were engaged by the covered entity andalso will impose similar —although as characterized byOCR, far narrower—safeguards. Business associatesare de�ned as third parties to the covered entity, suchas billing companies, utilization review entities, or otheroutside suppliers who perform TPO functions on behalfof the covered entity. These requirements do not pertainto covered entities which disclose PHO to providers fortreatment purposes—‘‘for example, information ex-changes between a hospital and physicians with admit-ting privileges at the hospital’’8 —even if they wouldotherwise meet the de�nition of a business associate.

The regulations go further to take on yet anotheraspect of con�dentiality which the statute did notaddress. Many people today are aware of the horrorstories of computer hackers violating privacy throughintentional acts and other disclosures of electronicinformation. Most con�dentiality breaches, however,are made through mere human mistake or carelessness.Many observers have known of the longstandingprinciples of con�dentiality in doctor-patient relation-ships, and that as a practical matter, in hospital set-tings and physician o�ces inadvertent disclosure of in-formation occurs, including through conversation inelevators, for example. There is no question that theselapses can be signi�cant to the patient when they occur.No one would argue that all types of providers shouldwork hard to prevent such disclosures and that policiesand procedures to maximize con�dential handling ofdata are worthwhile.

This entire regulatory construct, though, has beendeliberately and explicitly extended to encompass anycommunication of individually identi�able PHI in allforms, electronic, written, oral or any other. Rememberthat the only providers who are covered entities are



50


those that transmit PHI in electronic form. So a threephysician practice without a computer or fax machinewould be exempt from all of these requirements, includ-ing the oral communication limitations. Reference tooral communication was not present in the statute, norin the proposed regulations. It stems from the regula-tory philosophy that ‘‘If oral communications were notcovered, any health information could be disclosed toany person, so long as the disclosure was spoken.’’9Covered entities must ‘‘reasonably safeguard’’ PHI fromany violation of the rule whether intentional orunintentional.

The implications of this requirement in the real worldalso have raised alarms. One of the FAQs addressed byOCR is ‘‘If health care providers engage in con�dentialconversations with other providers or with patients,have they violated the rule if there is a possibility theycould be overheard?’’10 The OCR ‘‘understands that over-heard communications are unavoidable,’’ which is atleast some manifestation that they have had momentsof being in the here and now. Since those disclosuresare not the primary concern in our consideration of howto reorient health care processes and systems withincommunities to improve quality infrastructures, theinstructions and directions in OCR’s observations willnot be further elaborated here. But the reader whoenjoys both Stephen King novels and Miss Manners’ et-iquette books will enjoy both the tone and contents ofthis government guidance.

The detail in these rules is extraordinary even in theparagons of regulatory excess which we have consideredso far; and the privacy rules, awe inspiring as they are,are only part of the overall package of issues to make




51


real the implementation of what used to be administra-tive simpli�cation. The privacy rules, alone, however,hold serious implications for our hypothetical healthcare market.

§ 13:13 —Quality vs. PrivacyThe exchange of information about care is one of the

basic tenets of quality improvement in that you canimprove only that which you can measure in the �rstplace. Since so much of health care emanates fromprovider-patient encounters and even more from doctor-patient interactions, the patient’s medical record anddata which derives from it, naturally creates PHI. The�rst challenge under these regulations turns on whowill be engaged in gathering and reporting data andwhat patients must consent to with respect to theinformation. ‘‘Health care operations’’—the ‘‘O’’ inTPO—is de�ned to include:

[Q]uality assessment and improvement activities includ-ing outcomes evaluation and development of clinicalguidelines, provided that the obtaining of generalizableknowledge is not the primary purposes of any studiesresulting from such activities; population-based activi-ties relating to improving health or reducing health carecosts, protocol development, case management and carecoordination, contacting of health care providers andpatients with information about treatment alternatives;and related functions that do not include treatment.1

Privileging and credentialing; evaluating practitionerand provider performance; conducting or arranging formedical review or auditing including for fraud andabuse compliance; business planning and developmentincluding development or improvement of methods ofpayment or coverage policies are all included in healthcare operations. In addition, ‘‘treatment’’ includes notonly the actual clinical care of the patient hands on,but also the coordination or management of care by a

[Section 13:13]145 C.F.R. § 164.501 (2002) De�nitions.


52


health care provider ‘‘with a third party’’ in terms ofrisk assessment and case management as well as thecoordination of health care or other services amonghealth care providers and third parties authorized bythe health plan or the individual.2

Before your relief so overtakes you that you skip thenext section, remember that a provider entity engagedin any of these activities must obtain a patient consentto do so where PHI is involved (as it is in all of these atsome point). So the Notice to the patient of any suchentity’s privacy policies must include references to theseactivities. Still further, even in our hypothetical com-munity, some of these functions will be performed bybusiness associates of the relevant players. We do notknow how the enforcers will interpret ‘‘with a thirdparty,’’ but, at a minimum it means a third party wholikely is a business associate of the provider, eventhough not technically subordinate to the providerentity as is implied by subcontracts. Each business as-sociate must have a contract which gives to the pri-mary covered entity, the requisite satisfactoryassurances.

Patients who receive the Notice may refuse to giveconsent. Those instances are relatively easily managed,in HIPAA terms at least, since the provider may refusetreatment in such a case. What, however, is the likeli-hood a provider will refuse to treat a patient who isoverwhelmed by the privacy Notice and deems it easiersimply to deny consent? The �rst problem, then, is thatpatients may restrict the entry of their data into healthcare operations so a full picture of quality may not exist.Once having entered treatment, however, the patientmay also later revoke or restrict the consent. Then,what happens to his data when the covered entitywants to conduct a quality review?

Remember further, that one covered entity may notrely upon another’s consent, even for health careoperations. So if a hospital obtains the patient’s consent

245 C.F.R. § 164.501 (2002) De�nitions.


53


for its quality improvement activities, it may not dis-close any PHI contained within its data to the healthplan for its activities, unless the hospital is consideredthe business associate of the plan in performing thosefunctions for itself, or crazier still, the plan is consideredthe ‘‘business associate’’ of the hospital and, therefore,is performing TPO on behalf of the hospital. This, ofcourse, is exactly backwards from the logic that oper-ates in the real world. And, without direct access to thePHI, how can the plan verify the validity of the data?3

Still further, for quality improvement purposes, whatcan we assume about the accuracy of the data main-tained as PHI, since patients have the right not only torestrict uses of data but also to ‘‘amend’’ the very datain the record?4 And how far does an amendment go?What if the amendment nulli�es basic facts stated inthe record? The circumstances in which this will hap-pen may be small, but consider the real life story of aman who as a transgendered person mandated that allreferences in the medical record be to ‘‘she’’ and ‘‘her’’and that no notation be made that he was a biologicalman, nor, among other things, of the hormones takenover more than twenty years.5 As described by ‘‘her’’treating physician, ‘‘Concealing a patient’s travels fromone gender to the other would be more than a littlewhite elision, though. It would be a deliberate act ofU.S.C.A. action that would spawn many more (like analibi for why I was inexplicably neglecting this patient’sroutine gynecologic care).’’6 If patients get to pick andchoose not only which data may be used in health careoperations, but also which data is maintained for treat-ment purposes, many of the fundamental premises ofquality improvement go out the window.

3See discussion of ‘‘Organized Health Care Arrangements,’’below.

445 C.F.R. § 164.526(a) (2002).5Zuger, ‘‘On Record: Con�icting Obligations,’’ New York Times

F5 (Jan. 1, 2002).6Zuger, ‘‘On Record: Con�icting Obligations,’’ New York Times

F5 (Jan. 1, 2002).


54


At a minimum, any quality improvement has to turnon an assumption that records of services rendered are,in fact, accurate. It is beyond the scope of this article toaddress the patient’s rights and associated administra-tive appeals these regulations set up between providersand patients,7 but the regulatory scheme is so focusedon patients’ rights that it may well have the e�ect, inthe end, of undermining their most critical right inhealth care—the right to the best health care possiblefor them, rendered in a setting which continually strivesfor improvement. Although the regulators here purportto have struck a balance between privacy and quality,others in the world of quality assurance have drawnthe lines di�erently in ways which place real importanceon quality improvement while still safeguardingprivacy.8 These rules manifest a fundamental valuesorientation that when in doubt, patient privacy takesprecedence over all other system needs.

Of course, some might argue that the regulations onlydeal with identi�able health information and that toavoid the concerns just discussed, the discloser needonly ‘‘de-identify’’ the PHI. And so, the regulations havecontemplated such an approach, and enunciate therequirements for such de-identi�cation,9 which issigni�cantly more easily said than done. De-identi�edinformation is information that does not identify an in-dividual and about which there is no reasonable basisto believe the information could be used to identify an

7In addition to revocation and refusal of consent, amendment,restriction and access to the data, they also include a right tocomplain; and they provide appeals rights wherever a right isdenied. See King, ‘‘HIPAA; Privacy Standards: Driving a WedgeBetween Patients and the Health Field,’’ Health Law Handbook3-26 (Gos�eld, ed. 2001). Still further there is an entire compliancepiece which is far more prescriptive than the OIG’s guidance forfraud and abuse compliance.

8See NCQA-JCAHO, Protecting Personal Health Information: AFramework for Meeting the Challenges in a Managed Care Environ-ment, http://www.ncqa.org/docs/reports/con�de.doc.

945 C.F.R. § 164.514 (2002).


55


individual. This is a di�erent standard from the PROdata disclosure regulation which forbids the release of‘‘con�dential information’’ which is de�ned, inter alia,as data that explicitly or implicitly identi�es a patient.10

The HIPAA privacy regulations recognize that in thenew electronic age, fragmented and dissociated infor-mation can be reconstructed to identify someone. So, acovered entity may determine that health informationis not individually identi�able only if—in the nowclarion spirit of inhospitality—nineteen speci�c formsof identi�cation have been removed with respect notonly to the individual himself, but also his relatives,employers or household members! One can only theorizeabout which degrees of a�nity and consanguinity theseregulations intend to incorporate in the term ‘‘rela-tives,’’ nor what standards of residence would pertainto ‘‘household members.’’ Likely the employer could bereadily identi�ed and all inappropriate references to itremoved.

Fear not though. There is another option. In a moretrustful government mode than that otherwise in playhere, instead of relying on the checklist, a covered entitycan rely on the judgment as to whether the data hasbeen su�ciently de-identi�ed by ‘‘a person with ap-propriate knowledge of and experience with generallyaccepted statistical and scienti�c principles and meth-ods for rendering information not identi�able,’’ thatthere is very little likelihood the data could identifysomeone.11 Statisticians everywhere are cheering, nodoubt. Still, the risks associated with whether thatexpert views the data in a way shared by enforcers iscompletely unknown. Fools rush in where angels arequaking in their seraphic boots. If there has not been astrong business case for quality improvement up untilnow, to have to de-identify data in order to conduct thisactivity that is so essential for a strong healthcaresystem is a virtually insurmountable barrier. Luckily

1042 C.F.R. § 480.101(b) (2002).1145 C.F.R. § 164.514(b)(1) (2002).


56


our fellowship of the ring may �nd some pathwaysunder this mountain.

§ 13:14 —Joining ForcesThere are two fundamental approaches to sharing

PHI for collaborative quality initiatives which can serveour hypothetical market: (1) aggregations of coveredentities into single entities;1 and (2) organized healthcare arrangements. For the former, the de�ning charac-teristic is that otherwise legally distinct entities sharecommon ownership and control, such as a chain ofhospitals across twenty states.2 Under this de�nition,in our scenario a system of hospitals, nursing homes,and a faculty practice plan within the same communitymight be able to qualify as an a�liated covered entityand use a single shared notice of privacy practices anda single consent form. Of course, the organizationalinterrelationships must be documented, and the li-abilities which otherwise would pertain to each of theelements still would be in e�ect.

Of greater signi�cance for our hypothetical market isthe ‘‘organized health care arrangement’’ (‘‘OHCA’’)de�ned as:

[C]linical or operational integration among legally sepa-rate covered entities in which it is often necessary toshare protected health information for the joint manage-ment and operations of the arrangement. They mayrange in legal structure, but a key component of thesearrangements is that individuals who obtain servicesfrom them have an expectation that these arrangementsare integrated and that they jointly manage theiroperations.’’3

The discussion recognizes the need for physicians,nurses, and hospital personnel to interact in commoncause for TPO in a hospital setting even when they are

[Section 13:14]145 C.F.R. § 164.504(d) (2002).265 Fed. Reg. 82,503 (Dec. 28, 2000).365 Fed. Reg. 82,494 (Dec. 28, 2000).


57


not otherwise legally under one rubric. This is �ne asfar as it goes, but it is the only example, and an easyshot at that, of what they consider a truly integratedhealth care setting. One can presume that where non-employed physicians of the medical sta� join with thehospital to evaluate care as part of their review of theirown care, or as part of gainsharing, or even as part ofactivities similar to those of a PHO, without actuallyforming a contracting entity, these could potentiallyqualify as clinically integrated. Although the preface tothe regulations states that ‘‘While protected health in-formation may be freely shared among providers fortreatment purposes under other provisions of this rule,some of these joint activities also support the healthcare operations of one or more participants in the jointarrangement.’’4

The OHCA de�nition goes on to recognize an ‘‘orga-nized system of health care’’ in which more than onecovered entity participates and in which the participat-ing covered entities hold themselves out to the publicas participating. Here the joint arrangements mustinclude at least one of the following: (1) utilizationreview among the covered entities or by a third partyfor them; (2) quality assessment and improvementactivities in which the covered entities evaluatethemselves or a third party does it for them; or (3) pay-ment activities where there is shared �nancial risk andthe involved covered entities share data which isreviewed by other covered entities in the arrangementto administer the shared risk.5 Such an organizationalstructure may use a joint notice and joint consent.6 Pre-sumably a health plan in relationship to its contractednetwork of providers together would qualify, but, amaz-ingly, this simple aspect of health care today is not ad-dressed in the relevant discussion.

Unfortunately, the regulations o�er little more by

465 Fed. Reg. 82,494 (Dec. 28, 2000).545 C.F.R. § 164.501 (2002).645 C.F.R. §§ 164.506(f)(1), 164.520(d) (2002).


58


way of explanation or application. The best we can �ndat this point in these prefatory discussions is that com-munities seeking to work along our model will have toconsider structuring relevant components as an orga-nized health care system. However, what quali�es as‘‘holding itself out’’ to the public as participants for thepublic’s ‘‘having an expectation’’ that an entity is clini-cally or operationally integrated or jointly managed isutterly unelucidated in the discussion. This is particu-larly true where the entity is clinically integrated foroperations but not necessarily for treatment.

The major virtue of the prefatory acknowledgementand de�nitional recognition seems to lie in the fact thatthe de�nition of a ‘‘business associate’’ which otherwisewould be applicable. That de�nition requires that thecontracted partner is performing a function on behalf ofthe covered entity. In our model, the participants maybe co-equal and the functions performed serve some-times their own needs but, often as part of the jointenterprise, bene�t all. They are not acting technicallyon behalf of any one of them. This de�nition o�ers tan-talizing but still mysterious possibilities. There is a def-inition of ‘‘data aggregation’’ which refers to a businessassociate combining PHI it receives from one entitywith PHI received as the business associate of anotherentity ‘‘to permit data analyses that relate to the healthcare operations of the respective covered entities.’’7 In amore hospitable environment, quality driven partici-pants might breathe a sigh of relief but the institution-alized inhospitality religion of this program o�ers littlecomfort. Data aggregation is not mentioned in thelimited provisions on organized health care arrange-ments and is mired in the restrictions of business as-sociate relationships. Still, with enough contractual as-surances throughout the shire, these functions could beperformed with everyone functioning as the businessassociate of everyone else to get the necessary data ag-

745 C.F.R. § 164.501 (2002) De�nitions.


59


gregation to perform real benchmarking and qualityimprovement.

Our scenario posits freely shared information aboutclinical quality and patient perceptions of care. Itanticipates credentialing and involvement of physiciansat the hospital or the healthcare system in a wide vari-ety of activities that would have to fall under theorganized health care arrangement if all the playersare to have access to the fundamental clinical and otherinformation which emanates from PHI. The wholeundertaking is cloaked in clinical integration of carebased on CPGs. This integration is also important toanother legal morasse.

§ 13:15 Antitrust and Clinical Integration

Emerging from the cobwebs and corners of HIPAA,we now �nd ourselves confronting another aspect of thelaw which has stymied both plans and providers, al-though the predominant concern has come from theprovider arena—namely the ways in which the antitrustlaws can thwart the equities of provider-payerinteractions. In our scenario we have proposed thatpayment schemes be jointly worked out between payorsand providers, but life rarely evolves so neatly. In thereal world, to negotiate rates, someone has to step outand put the �rst number on the table. In the world ofantitrust where the inhospitality tradition is onesided—anything which does not support competition isvenal, evil and for secret, nefarious purposes—negotiat-ing parties must assess their risks. Whether the riskinherent in any negotiation (will the number be accept-able? did we leave money on the table?) is all that theproviders have to fear if they take the laboring oar orwhether the Federal Trade Commission and Depart-ment of Justice may take an interest, are the choicesthat will be important to the ability to use the paymentprinciples espoused here to motivate quality-drivenbehavior.

The antitrust rules, in general, are well beyond our


60


consideration here, but they typically prohibit both perse price �xing where otherwise competing providersjointly establish the prices they will accept and groupboycotts where competitors jointly refuse to do businesswith an unattractive partner.1 In the information andguidance o�ered by the Federal Trade Commission andthe Department of Justice regarding the activities ofprovider networks, ‘‘joint pricing agreements will not becondemned as per se illegal if the participants haveintegrated their activities through the network in away that is likely to produce signi�cant e�ciencies thatbene�t consumers and the price agreements are reason-ably necessary to realization of those e�ciencies.’’2

When the initial antitrust safety zones were pub-lished, the forms of risk that were permissible amongotherwise competing providers turned both on the typeof �nancial risk taken (e.g., capitation, percent ofpremium or withholds) and the percentage of providersin the networks. Another type of risk which the guid-ance allows is ‘‘global fees’’ or ‘‘all inclusive case rates.’’Where there are agreements by a network:

[T]o provide a complex or extended course of treatmentthat requires the substantial coordination of care byphysicians in di�erent specialties o�ering a complemen-tary mix of services, for a �xed, predetermined, pay-ment, where the costs of that course of treatment canvary greatly due to the individual patient’s condition,the choice, complexity of length of treatment or otherfactors.3

That statement describes an a�rmative safety zone

[Section 13:15]1For a brief but useful statement of many of these issues see

Leibenluft, ‘‘Antitrust Issues Raised by Rural Health Networks’’(Feb. 20, 1998), http://www.ftc.gov/bc/ruralsp.htm.

2‘‘Antitrust Issues Raised by Rural Health Networks’’ 4 (Feb.20, 1998), http://www.ftc.gov/bc/ruralsp.htm.

3‘‘Statements of Antitrust Enforcement Policy in Health CareIssued by the Justice Department and Federal Trade Commission(Aug. 28, 1996), BNA’s Health Law Reporter 1309 (Aug. 29, 1996).


61


meaning that the regulators would not move to enforceif those factors were present. The case rate approach,as stated there, though, requires physicians of multiplespecialties. There ought be no reason that physicians ofa single specialty could not be put similarly at risk, andnot even for complex care. Our CPG based payment inour scenario can work well for the congestive heart fail-ure CPG discussed in terms of Medicare reimbursement.Acknowledging that other forms of risk may be equallyas bene�cial, the enforcers allow that they do not fore-close consideration of other models that put the provid-ers at the dreaded ‘‘substantial �nancial risk.’’

Whether such risk exists or quali�es in our modelmay, happily, be irrelevant since the enforcers allowyet a totally di�erent, far more quality driven approach:

Physician network joint ventures that do not involve thesharing of substantial �nancial risk may also involvesu�cient integration to demonstrate that the venture islikely to produce signi�cant e�ciencies. Such integra-tion can be evidenced by the network implementing anactive and ongoing program to evaluate and modifypractice patterns by the network’s physician participantsand create a high degree of interdependence andcooperation among the physicians to control costs andensure quality.4

Where the agencies have o�ered guidance but not asafety zone, they do not enumerate a checklist ofqualifying characteristics. Rather they describe a sce-nario where they would not take enforcement action,absent other major problems. In addition, the physiciannetwork approach is also available for and they explic-itly describe a multi-provider mechanism in the form ofa PHO.5

Consider the following features of their elaborated




62


arrangement in comparison with our posited scenario:An IPA is established by a group of community physi-cians and it moves to implement systems to establishgoals relating to quality and appropriate utilization ofservices by the IPA physicians. It regularly evaluatesboth individual and network aggregate performance inrelationship to those goals and modi�es individualparticipants’ behaviors in relationship to the goals. TheIPA itself will engage in case management, preauthori-zation of some services, and concurrent and retrospec-tive review of inpatient stays. In addition the IPAdevelops and adopts practice standards and protocols togovern treatment and utilization of services, andactively evaluates the physicians’ performance in rela-tionship to those standards (i.e., CPGs).

The regulators posit a signi�cant capital investmentin the systems necessary to the applicable reportingfunctions around clinical performance, patient satisfac-tion and individual and aggregate network behavior.The IPA provides detailed reports to payors about theperformance measured. The IPA physicians are paid byhealth plans on a fee-for-service basis.

The IPA will retain an agent to develop a fee schedule,negotiate fees and contract with payers on behalf of theventure. Information about what participating doctorscharge non-network patients will not be disseminated toparticipants by the IPA and the doctors will not agreeon the prices they will charge patients not covered byIPA contracts.6

In terms of the participants and what they represent inthe community, the IPA includes three geographicallydispersed primary care group practices that togetheraccount for 25% of the primary care doctors in the hy-pothetical town. A number of specialists to whom theyrefer also participate. They constitute between 20% and35% of the specialists in each relevant market. All the



63


physicians participate in other managed care plans inthe market. Here the regulators have stated that theprice agreement is subordinate to and reasonably nec-essary to achieve the e�ciencies posed by the clinicalintegration. In the PHO model, the story is almostexactly the same.7 It is noteworthy that the types ofintegration stated here are also the types of clinicalintegration noted by the HIPAA regulators under theirdiscussion of organized health care arrangements. Infact an IPA acting in relationship to either a hospital ora health plan are the very few non-hospital basedexamples the HIPAA regulators o�er.

What, then, can we conclude about the 1996 notionsof the antitrust enforcers six years later with respect toforms of organizational functioning that would not bearundue scrutiny or enforcement? We know that they canand do take action against physicians who jointlybargain for fees without either clinical or �nancialintegration, even when they claim the protection of hav-ing unionized. In a high pro�le action in Tampa, an or-ganization of two dozen competing surgeons purportedto use the infamous ‘‘messenger model,’’ but the enforc-ers did not share their interpretation of the messengermodel rules, which explicitly prohibit joint negotiating.8We know that the AMA’s short lived physician union,Physicians for Responsible Negotiation, was disbandedafter eighteen months in light of the Supreme Court’sdecision in NLRB v. Kentucky River Community Care,Inc.9 which held that health care workers in privatehospitals could not join unions if their duties includedrawing upon their professional training and experi-


8See Klein, ‘‘Justice Dept. Slaps Florida Physicians withAntitrust Charges,’’ AMNews (Feb. 15, 1999), http://www.ama-assn.org/sci-pubs/amnews/pick�99/pick0215.htm.

9NLRB v. Kentucky River Community Care, Inc., 121 S. Ct1861 (2001).


64


ence in supervising others.10 Federal investigations havebeen pursued in Connecticut and Ohio.11 A Delawareinvestigation stopped a purported union in its trackswhen its ‘‘messenger’’ bargaining with a major payer inthe state, represented that he was speaking for 90% ofthe orthopedists in the state. The resulting settlement,trumpeted as an agreement allowing physicians to usethird party messengers,12 gave them nothing more thanwhat the safety zones had stated would be legitimatesome �ve years earlier. What’s a well motivated groupof physicians to do?

Our hypothetical scenario falls squarely within theantitrust agencies’ published parameters. In settle-ments stimulated by enforcement of otherwise indepen-dent physicians’ e�orts to bargain collectively for fees,the enforcers have reiterated their position that clinicalintegration is a viable alternative. In our scenario,CPGs drive payment, credentialing and a wide varietyof other behaviors. The regulators do not take any posi-tion regarding the breadth of implementation of CPGswithin a community. There may be issues of networksize, and scope of the application of the CPGs; but theseare not the essential concern. It is the linking of thoseCPGs to payment issues that is the biggest problem, inantitrust terms. Even there, it is the concerted action ofthe physicians alone, or too many of them joinedtogether against the payer, in an e�ort simply to raiseprices that incurs attention and enforcement. Wherethe arrangement is worked out among all the partici-pants, the requisite conspiracy of one side against theother does not exist. Moreover, if our scenario plays out

10‘‘Highlights of PRN Report to Delegates at AMA Annual Meet-ing June 15-21, 2001,’’ http://www.4prn.org/kriver.html.

11Peterson, ‘‘Update on Joint Negotiations with Insurers,’’ Amer-ican Academy of Orthopedic Surgeons, http://www.aaos.org/wordhtml/bulletin/oct01/�ine6.htm (Oct. 2001).

12Albert, ‘‘Don't Shoot the Messenger: Path Opens for ContractTalks,’’ AMNews (Nov. 12, 2001), http://www.ama-assn.org/sci-pubs/amnews/pcik�01/prl11112.htm.


65


as modeled, who would be the complainant to theantitrust enforcers?

Of course, there is always a chicken and egg problemin using clinical integration to cloak bargaining tacticsin safety: how much integration and functionality mustbe in place before the physician-network can approachthe payors? As we have seen so far, there is no brightline here, but, again, practicalities and the smell testtell us that ‘‘we're �xin’ do to it’’ may be a statementabout the per se nature of the antitrust risk and is notlikely to satisfy the payors that the physician networkis engaged in something more than naked price �xing.The bottom line, however, is the credibility of—andperhaps even more the motivation for—the joint action.Where physicians move to build a truly clinicallyintegrated environment for quality purposes, so as toavoid having to engage in unnecessary documentation,to permit them to improve care to patients in all theways that evidence-based medicine can, because theywill be able to strengthen their doctor patient relation-ships and return to what makes medicine meaningfulto them, truly the fee arrangements are subordinate tothe quality purposes. To embark on clinical integrationsolely to be able to bargain for higher fees is folly, if nothighly ine�cient at best. It is the collaborative,coordinated, integrated behavior, where all the partiesin the community are working together that makes theundertaking worth it. And in that case no matter whogoes �rst in negotiating terms, or whether the healthplan facilitates the development of the infrastructuredescribed by the antitrust regulators, which partymoves the physicians along the continuum to true clini-cal integration will not matter. Our hypothetical mar-ket is a win-win for all concerned under the antitrustrules and otherwise: The community gets better, moreevidence based care. The parties are not adversaries inantitrust terms; and the fee setting aspect of the ar-rangement whether case rates, global rates, fee for ser-vice or something else, can be structured to be subordi-nate to the true purpose which is to make better qualityhappen faster.


66


§ 13:16 ConclusionTo create truly quality-driven healthcare in a mar-

ket, or community, or even a single system, as in thescenario that has anchored this consideration, is anenterprise that raises a host of profound challenges,many of them operational. To make our story realrequires new forms of work�ow in hospitals and physi-cian o�ces including such changes as Drop In GroupMedical Appointments1 which are conducted in lessscheduled o�ce practices, and di�erently organized ex-amination rooms,2 and better systems for patient safetyin hospitals,3 and better collaboration and coordinationmerely to improve pure clinical processes of care suchas that in the national (and even international) qualityimprovement project known as ‘‘Pursuing Perfection.’’4None of these is a small task by itself.

But that would not be enough. The landscape de-scribed requires cultural, organizational and leadershipchanges to engender a clear and committed vision tiedto self renewing processes to make its aspirations real.At its core, above all, it requires good science whichre�ects the application of what we know to be right,which, by de�nition, changes over time. And, it shouldbe observed, it also requires skills and support for thepsycho-dynamics of strong, meaningful doctor-patient

[Section 13:16]1Walpert, ‘‘Patients Tired of Waiting? Try These Strategies,’’

ACP-ASIM Observer (Oct. 1999), http://www.acponline.org/journals/news/oct99/waiting.htm; Domin, ‘‘Doctors Find QualityTime with Patients in Groups’’ (Oct. 6, 2000), http://www8.cnn.com/2000/HEALTH/10/06/doctors.time; No�singer, ‘‘Enhance Sat-isfaction with Drop-in Group Visits,’’ http://www.hippocrates.com/February-March2001/02features/02feat�digman.html.

2Bagley, ‘‘Building for Tomorrow: The Idealized Design of Clini-cal O�ce Practices,’’ Fam. Practice Mgt. (May 2000), http://www.aafp.org/fpm/20000500/13buil.html; and http://www.ihi.org.

3See http://www.leapfrog.org.4http://www.ihi.org/pursuing perfection.


67


relationships that stand at the frontlines of all of healthcare delivery. The range of things necessary to makethe American health care system a better one character-ized by singularly improved quality is considerable.

To embark on this journey, from where we are todayto where we need to be, and can be, is not merely aquest for a mythological talisman. It is a responsibilitythat comes with what health care is about, its realpurpose. Those of us who come to the work of healthcare quality from a wide variety of perspectives areoften confronted with the plaint of ‘‘but where is thebusiness case for quality?’’ While in a world of �nitehealth care dollars this is understandable, and theexpense of quality processes has to be taken into ac-count, the question is almost beside the point—there isno other case in health care than the one for quality. Inthe scenario we have considered here, greater e�cien-cies and lower administrative costs cannot help butemerge from more streamlined, standardized and clini-cally relevant management systems by comparison withthe fragmented, error prone, unsatisfying, custom-crafted delivery system familiar to us today. The busi-ness case is inherent in the operation of the componentsthat make our scenario a uni�ed, truly clinicallyintegrated, understandable system, based on science.Yet there is nothing in the construct which assumes itsuniform application, in the same way, in every market,or even within markets or across the country. It merelyprovides principles of operation which guide the activi-ties down whatever speci�c path the travelers choose.

But, the thrust of this consideration has been on thelaws which, however well intentioned at their incep-tion, we now see have added unnecessary complexityand a crushing, burdensome and meaningless weight tothe task of making quality happen.5 The major regula-tory programs in reimbursement, anti-referral and

5And, there are other ancillary legal issues not addressed hererelating to, for example, how to implement within medical sta�organizations systems which may recognize di�erently the physi-


68


fraud and abuse, and privacy are excessive, intrusiveand directly noxious if not a�rmatively toxic to themanagement simplicity we need to deal with the inher-ent complexities of health care delivery and science.Only the antitrust laws hold out the promise of receptiv-ity to a more idealized quality orientation that willpermit those who are engaged to do well by doing good.

To make this journey we need to travel lighter. Weneed multiple pathways, even in the national insuranceprogram of Medicare. There is much that the lawimposes on this passage today which is outrightsuper�uous and restrictive. Still, as we have seen, mostof the existing terrain is navigable, but not without te-dious and careful baggage management and courseplotting. Wouldn't the trip be easier if those who are inthe business of setting the ground rules, namelyCongress and the regulators, could conceive of theirrole as one which requires a consistent platform of uni-�ed purpose and science, explicitly supporting anddepending upon the use of evidence based clinicalpractice guidelines as the sextant, as they in�uence thecourse of the health care system?

cians who are on board for various elements; how licensure lawsrestrict personnel assignments or how guidelines are developed,translated and implemented; or how performance measurementsystems are designed and their results reported. All will raiselegal concerns.


69



Documents

Chapter 13 - Gosfield · Chapter 13 Making Quality Happen: In Search of Legal Weightlessness Alice G. Gos”eld1 KeyCiteL: Cases and other legal materials listed in KeyCite Scope