The Economics of Oncology: Health System Reform and Clinical Research

February 1, 1995

First in a series of articles on how economic factors, managed care, and other aspects of health care reform affect physicians in the various oncology specialties, and ultimately their patients. The series was compiled for ONCOLOGY

ABSTRACT: First in a series of articles on how economic factors, managedcare, and other aspects of health care reform affect physiciansin the various oncology specialties, and ultimately their patients.The series was compiled for ONCOLOGY by Joseph S. Bailes, MD,who is Chairman of the Clinical Practice Committee of the AmericanSociety of Clinical Oncology. Dr. Bailes is a partner in TexasOncology, P.A.


In the last several years, there has been considerable legislativeactivity at the state and federal levels regarding reform of thehealth insurance system, especially after President Clinton madefederal legislation a top priority 2 years ago. A matter thatis relatively minor in the overall constellation of issues, butwhich is of particular importance to oncology, is insurance coverageof, and patient access to, clinical trials.

This article reviews current coverage of clinical trials, discusseshow health care reform may affect insurance coverage of clinicaltrials, and summarizes how the issue was treated in the intensivedebate on health care reform during the past 2 years. While thecongressional debate failed to produce federal reform legislationin 1994, the topic is certain to be revisited in the future.

Insurance coverage of clinical trials is an aspect of broaderpolicy on coverage of new technologies. Both private health insuranceplans and such public programs as Medicare and Medicaid typicallyexclude coverage of "experimental" or "investigational"procedures [1].

A drug or device is usually considered investigational for insurancepurposes if it has not been approved for marketing by the FDA.Thus, drugs and devices subject to investigational new drug ordevice exemptions (INDs and IDEs) are typically not covered byinsurance. A procedure is usually considered investigational ifit is not accepted as safe and effective by the medical community.The use of an informed consent form is often seen by insurersas evidence that a particular procedure is investigational. Theresult of these policies is that clinical trials are usually excludedfrom insurance coverage.

Although insurance plans exclude experimental or investigationaltreatment, they cannot necessarily enforce their views regardingwhether particular procedures lack coverage. Insurance plans canbe sued by beneficiaries on the ground that the plan administratorsare misinterpreting the coverage rules. In the last few years,the major issue litigated in this area has been coverage of autologousbone marrow transplants for breast cancer and other conditionsthat many insurers regard as unproven.

The courts have often been sympathetic to plaintiffs seeking accessto coverage of bone marrow transplants. Courts have often overruledcoverage denials by concluding that the plan's exclusionary languagewas unclear, or by directly holding that the bone marrow transplantprocedures were not experimental. The decided cases have splitbetween those that have upheld insurers excluding coverage ofbone marrow transplants for certain conditions and those thathave sided with patients.

The tide seems generally to be turning in favor of upholding insurerdecisions, as insurers have revised contract- ual language inresponse to early decisions holding that particular language wasunclear [2]. There is also anecdotal evidence, however, that insurersare increasingly extending coverage to what they regard as experimentalprocedures out of fear of punitive judicial judgments, such asthe recent $89 million verdict against a health maintenance organizationfor denial of coverage [3].

Partial Coverage of Costs

Although clinical trials are ostensibly not covered by insurance,insurance plans have nevertheless unknowingly often paid manyof the costs. Insurance claim forms frequently do not requireinformation that would disclose whether the patient was enrolledin a clinical trial, and as a result, insurers often cannot tellfrom the face of a claim that a patient received an investigationalservice.

The Inspector General of the Department of Health and Human Servicesrecently initiated an investigation that may, especially dependingon its outcome, change the way health care providers bill forpatient care costs associated with investigational procedures[4 ]. As part of an investigation under the False Claims Act,the Inspector General issued subpoenas to over 100 hospitals regardingclaims submitted to Medicare for certain investigational cardiovascularprocedures. The unstated premise of the investigation is thatthe submission of a claim to Medicare for costs associated withan investigational procedure may not be simply a claim that Medicaremay deny, but may also constitute a false claim. The theory ofthe investigation seems to be that health care providers may commita criminal offense when they submit a claim for costs associatedwith clinical trials, since they know Medicare does not pay forclinical trials.

Even in programs that deny coverage of clinical trials, however,there may be some intentional limited coverage. For example, inthe case of in-patients, Medicare, which pays a fixed amount basedon the patient's DRG, attempts, to some extent, to separate thecosts related to the investigational service from costs that wouldhave been incurred in any event. Thus, if a patient is admittedto a hospital without enrollment in a clinical trial having beendecided in advance, and once admitted is enrolled in a surgicaltrial, Medicare will pay for the admission as if the patient hadbeen admitted for nonsurgical treatment. On the other hand, ifthe admission was for the sole purpose of enrolling the patientin a clinical trial, Medicare will deny payment of the admissionin its entirety.

A similar distinction is drawn with respect to treatment of adverseeffects from a clinical trial. Even though Medicare will not payfor treatment costs in a clinical trial itself, if the patientsubsequently develops a complication requiring treatment, Medicarewill fully cover the costs of the subsequent treatment, even thoughthe need for treatment can be traced directly to treatment ina noncovered clinical trial.

Types of Costs Involved

There are several types of costs that are incurred in clinicaltrials:

The cost of the investigational drug, device, or procedure.

The cost of collection and analysis of trial data.

The costs of patient care, including hospital, physician, laboratory,and other services.

Each of these categories is treated somewhat differently underthe current system.

FDA regulations generally prohibit manufacturers from chargingfor investigational drugs, and manufacturers therefore supplyfree drugs to investigators conducting clinical trials. FDA rulesare more liberal in allowing charges for investigational devices,however, with the result that patients (or some other party) mayincur the cost of the device used in a clinical trial.

Data collection and analysis are ordinarily paid for by the sponsorof the clinical trial. For example, funds may come from a pharmaceuticalor medical device manufacturer or from the National Institutesof Health.

Patient care costs are generally the responsibility of the patient,or are borne by the institution at which the clinical trial isconducted. Because these costs are the outgrowth of an investigationalservice, insurers typically will not cover them. This is the caseeven though the patient might have incurred substantially similarcosts if the patient had undergone standard therapy instead ofbeing enrolled in the clinical trial.

Although public debate about coverage of clinical trials sometimessuggests that the focus of interest is coverage of the investigationalprocedure itself, in reality the principal source of costs isfrequently patient care, most or all of which might have beennecessary even if the patient had undergone standard therapy.High-cost experimental procedures, such as bone marrow transplantsdominate the discussion, but almost certainly do not accuratelyreflect the added costs of clinical trials in general.

Proposed Coverage of Trials

Many of the major bills considered by Congress during the healthcare reform debate of 1993 and 1994 contained provisions extendinginsurance coverage to patient care costs associated with clinicaltrials. These provisions were part of the bills' descriptionsof standard benefit packages. The primary function of a standardbenefits package is to describe the benefits that must be providedby employers, public subsidy programs, or whatever other mechanismwas used in the bill to expand health insurance coverage to thosewho cannot now afford it.

The principal impetus for inclusion of the provisions appearedto be the view that covering certain clinical trials is not onlybeneficial in itself, but also has the virtue of explicitly denyingcoverage of investigational procedures conducted outside the qualifyingclinical trials. For example, investigational bone marrow transplants-theprocedure frequently used as an example-would be covered onlyin certain trials, and not when used in nonresearch situations.This approach is seen by its advocates as hastening the resolutionof questions as to whether new technologies are safe and effective,while at the same time providing an outlet for at least some patientsto have insured access to new treatments.

The provisions covering clinical trials differed somewhat frombill to bill, with the following differences being the key issues:

Qualifying Trials--The bills varied in their criteria forthe trials that would be eligible for coverage. Generally speaking,all the bills covered trials approved by the Department of Healthand Human Services, which includes the National Institutes ofHealth and the FDA. Some of the bills explicitly included drugtrials conducted pursuant to investigational new drug applicationsapproved by FDA, while others, by simply referring to trials approvedby Health and Human Services, left somewhat unclear whether theexistence of an approved investigational new drug applica- tionwould, by itself, be sufficient to afford coverage. Trials approvedby nongovernmental entities in accordance with NIH guidelineswere also usually covered. A significant area of variation amongthe bills was coverage of trials approved by the Departments ofDefense and Veterans Affairs, which some bills, including theAdministration bill, would cover, while other bills would not.

Services Covered--Perhaps the greatest difference amongthe bills was the specific services that would be covered by insurance.There was consensus that the administrative costs of a study shouldnot be covered by insurance and, similarly, that costs of drugsthat are now borne by pharmaceutical manufacturers should notbe transferred to insurance. Beyond that, however, there was adivision of opinion.

Some of the bills attempted to exclude services related to thetrial that were in excess of the services that would have otherwisebeen provided to the patient. This has the appeal of seeming toguarantee that the system will not incur additional costs as aresult of covering clinical trials. It has the serious disadvantage,however, of being impractical. While some services might be easilyidentifiable as performed solely because the patient was enrolledin a clinical trial, others (eg, length of hospital stay) couldbe much more difficult to assess. Moreover, there may be offsettingsavings, such as where an investigational outpatient procedureis substituted for standard in-patient care. A controlled studyof costs accompanying the clinical trial might be required tosort out the cost effects, and this would be a totally infeasibleway to pay insurance claims.

In addition to the practical problems it presents, this approachfails to provide adequate coverage for procedures, such as bonemarrow transplants, where the investigational service may differsignificantly from standard therapy.

The alternative formulation in some of the bills was to coverservices furnished in accordance with the design of the trial(excluding the administrative and drug costs, which are excludedunder all bills). This language would provide broad coverage ofessentially all patient care costs. The one potentially significantexception is that some of the bills would (by analogy to exclusionof drug costs) have excluded costs of the investigational procedureitself.

Medicare--The bills took different approaches to the issueof whether Medicare should be modified to include coverage ofclinical trials. This issue has proved to be more contentiousthan coverage of trials under private insurance. Additional benefitsincluded in the standard benefits package for private insurancemay increase the premiums but do not significantly influence thefederal budget (except to the extent that federal employees areaffected). Therefore, Congress was relatively receptive to enhancementsof the private benefit package.

Changes to the Medicare program, however, can significantly impactthe federal budget. Under the applicable budget rules, increasedMedicare expenditures must be offset by savings or increased revenues.Although most of the health care reform bills included large futurereductions in the rate of Medicare spending, these savings wereused in the bills to finance other increased Medicare benefits,such as a new benefit for outpatient prescription drugs, or forhealth insurance subsidies for the general population. Thus, therewas reluctance on the part of some bill sponsors to include newMedicare coverage of clinical trials. The Congressional BudgetOffice, however, never developed an estimate of the costs associatedwith Medicare coverage of clinical trials.

Managed Care and Clinical Trials

Health maintenance organizations and other forms of managed careare growing rapidly in the United States. Managed care organizationsemploy utilization management, capitated fee arrangements, requirementsfor patients to use particular hospitals with which the managedcare plans have contractual relationships, and similar techniquesto control costs.

The growth of managed care has led to concerns that academic medicalcenters will suffer, as managed care organizations, for cost reasons,prefer to use community hospitals [5]. Similarly, the concernhas been raised that managed care organizations will enroll fewerpatients in clinical trials, which frequently take place at academicmedical centers, than will occur under a fee-for-service system.These concerns are often opposed, however, by the view that competitivemanaged care plans are the best hope for controlling health carecosts.

A number of the health care reform bills included provisions toprotect academic health centers and patient access to those centers.One approach, such as that in the President's bill, was to createan independent source of funds for academic health centers bytaxing all insurance premiums to support direct grants to thecenters. This approach would presumably allow such centers tocompete more easily on the basis of price with community hospitals,since their charges to insurers would be supplemented by the directgrants.

Patient access to academic health centers (and, by extension,to their clinical trials) was also addressed in some of the bills.Most of the proposed legislation included broadly worded requirementsthat managed care organizations provide appropriate patient accessto specialty facilities. Other bills, however, included much morespecific requirements and procedures, entitling patients to insuredcoverage at facilities outside the plan's regular network of providers.


The various health care reform bills evidenced a large degreeof agreement that there should be insurance coverage of clinicaltrials. The principal unresolved question was whether there shouldbe essentially full coverage or whether coverage should be somehowlimited to the costs incurred in standard treatment.

Potentially diminished patient access to clinical trials and toacademic medical centers in general is a more difficult issue.The widespread support for academic medical centers is at leastpartially offset by similar strong support for managed care, whichis growing for reasons unrelated to health care reform legislation.Thus, even if significant health care reform requiring coverageof clinical trials is enacted in the future, the current flowof patients into those trials may be altered by nonlegislativechanges in the health care system.


1. Physician Payment Review Commission: Coverage decisions andtechnology assessment. Annual Rep Congress 219-236, 1994.

2. Herrington E: Experimental healthcare reimbursement: A casesurvey. The Health Lawyer. 6(4):1, 3-9, 1993.

3. Anders G: More insurers pay for care that's in trials. WallStreet Journal, p B-1, Feb. 15, 1994.

4. IG nationwide hospital investigation seeks data on non-FDAapproved devices. BNA's Medicare Rep 5:732-734, June 24, 1994.

5. Blumenthal D, Meyer G: The future of the academic medical centerunder health care reform. N Eng J Med 329:1812-1814, 1993.