Garber S, Ridgely MS, Bradley M, Chin KW. Payment Under Public and Private Insurance and Access to Cochlear Implants. Arch Otolaryngol Head Neck Surg. 2002;128(10):1145-1152. doi:10.1001/archotol.128.10.1145
Cochlear implants are expensive, yet often cost-effective. However, among hundreds of thousands of potential US candidates, only about 3000 received implants in 1999. To analyze whether insurance reimbursement levels may contribute to low access rates.
Surveys were performed during 1999 and 2000 of physicians and audiologists at clinics providing cochlear implant services, selected hospitals where surgery is performed, and state Medicaid agencies. Secondary data were obtained on Medicare payment rates and hourly incomes of otolaryngologists and audiologists.
One hundred thirty-one physicians (response rate 67.9%), 111 audiologists (74.0%), 60 hospitals (73.2%), and 44 Medicaid agencies (86.3%).
Reimbursement rates for selected Current Procedural Terminology codes and for cochlear implant systems (devices); time required to perform services; additional time not reimbursed; and device purchase prices.
Medicare and Medicaid payment rates often fail to cover costs of aural rehabilitation. Medicare sometimes and Medicaid often fails to cover surgeon costs. Sometimes private insurance does not cover hospitals' device costs. Under Medicare, in 1999 hospitals lost more than $10 000 per device for inpatient surgery and about $5000 per device for each outpatient surgery. Device reimbursement in 2002 for outpatient surgery under Medicare is about $3773 higher than in 1999. Medicaid device payment policies vary greatly and fail to cover costs in at least 18 states, accounting for 44% of national Medicaid enrollment.
Efforts to expand access to cochlear implants may be impeded by financial incentives. Facilitating access for Medicare and Medicaid patients could require changes in payment policies.
COCHLEAR IMPLANTS are an established treatment for severe to profound sensorineural hearing loss.1 Treatment can cost more than $40 000,2 including approximately $20 000 for a device.3 Evidence indicates, however, that these costs are typically outweighed by resulting benefits, such as reduced costs of special education2 and improved quality of life.4
It has been estimated that there are 460 000 to 740 000 severely to profoundly hearing-impaired individuals in the United States.5 New data reported herein suggest that approximately 3000 people received cochlear implants in the United States during 1999. Factors limiting access to cochlear implant technology may include (1) lack of knowledge about the performance of the technology among primary care physicians, (2) lack of insurance or knowledge about or interest in cochlear implants among hearing-impaired Americans, and (3) objections to cochlear implants from representatives of the deaf community.6 We analyze another possible factor, namely, financial incentives for professional and institutional providers stemming from payment policies of public and private insurers.
Data were collected through national mail surveys of physicians and audiologists who work with cochlear implant recipients, a telephone survey of hospitals in which cochlear implant surgeries are performed, and a telephone and mail survey of state Medicaid agencies. Following procedures developed in consultation with RAND's Human Subjects Protection Committee (Santa Monica, Calif), we identified the sponsor in survey materials and respondents were promised confidentiality and anonymity.
Practices or clinics that treat candidates for cochlear implant surgery were the unit of analysis. A list of practices was developed from postings on the Internet sites of Advanced Bionics Corporation (Sylmar, Calif), and Cochlear Corporation (Englewood, Colo), the only companies selling cochlear implant prosthetic systems in the United States during the study period. Because each company wants to help candidates locate a clinic that uses its device, we expect that their postings are comprehensive, and, therefore, we believe our list of practices accounted for almost all cochlear implant surgeries in the United States during the study period.
Initially, 231 practices were classified as eligible. Of these, 25 were randomly chosen for pilot tests. For the main survey, a physician questionnaire was mailed to a physician at each of the remaining 206 practices, and an audiologist questionnaire was sent to 187 audiologists, one for each practice for which a potential audiologist respondent was identified through a telephone call to the practice. Both questionnaires requested information on characteristics of practices and patients, time required to perform services identified by Current Procedural Terminology (CPT) code, time spent on activities that are not reimbursed, and average fee-for-service payment rates from private insurance, Medicare, and Medicaid. Each questionnaire concluded with the open-ended question: "Is there anything else you would like to tell us about cochlear implants?" The second of 2 mailings was followed by intensive telephone efforts to encourage participation. Data collection ran from August through December 1999.
The hospital survey focused on numbers of surgeries performed, costs of purchasing devices, and reimbursement rates when separate payments were made. This information is rarely available from a single hospital department, and a telephone survey enabled an interviewer to identify a knowledgeable individual for each item and contact each directly. Our telephone calls to practices identified 213 hospitals at which we were told physicians performed cochlear implant surgery. Resources did not allow us to survey all of them. We chose 96, oversampling hospitals in states where our information from state Medicaid agencies was incomplete. Data collection ran from September 1999 through March 2000.
The Medicaid survey focused on policies for fee-for-service enrollees. We requested information, for children and adults separately, about coverage policies and payment rates for physician and audiologist services and for cochlear implant prosthetic systems (Health Care Financing Administration Common Procedural Coding System [HCPCS] code L8614). We attempted to collect information from all 51 Medicaid agencies through telephone calls to identify appropriate respondents, mailings, faxes, and intensive telephone follow-up. Data were collected from April 1999 through February 2000.
Cochlear implantation is covered nationally for Medicare patients who meet selection guidelines.7 We focused on traditional (fee-for-service) Medicare. Data for the period covered by the surveys were extracted from 3 files downloaded from the Health Care Financing Administration Internet site. The data for professional services are physician fees by CPT code.8 For inpatient surgery, hospital payment for the device is included in the payment for diagnosis related group 49, and we used Medicare Provider Analysis and Review data for fiscal year 1997.9 For outpatient surgery, separate payments were made for the device (HCPCS code L8614) according to a fee schedule.10 On August 1, 2000, the Health Care Financing Administration, since renamed the Centers for Medicare & Medicaid Services, implemented the outpatient prospective payment system (OPPS), a new method for reimbursing hospitals for outpatient services.11 We reviewed documents describing OPPS, listing payment rates, and discussing future payment policies.
Data were obtained from 131 physicians and 111 audiologists. Ignoring practices reclassified as ineligible because respondents reported that their practices had not performed an implant in a new patient in the past 12 months or no longer worked with patients receiving cochlear implants, the response rates were 67.9% and 74.0% for physicians and audiologists, respectively. Table 1 provides information about the practices and their patients receiving cochlear implants during the past 12 months. Almost 40% of the patients were covered by private fee-for-service insurance, and about 35% by fee-for-service Medicare or Medicaid.
Among 82 surveyed hospitals in which cochlear implant surgery had been performed during the past 12 months, 60 (73.2%) provided information. Of these 60 hospitals, 56 reported a total of 727 cochlear implant surgeries during the past 12 months (median, 8.0). Of the 51 state Medicaid agencies, 44 agencies (86.3%), accounting for 81% of national Medicaid enrollees,12 provided data.
Table 2 reports information about 7 candidacy testing and evaluation procedures and 2 postoperative procedures audiologists reported performing relatively often. One preoperative procedure (CPT 92506) and one follow-up procedure (CPT 92510) stand out as requiring relatively large amounts of time. Table 3 reports median fee-for-service reimbursement levels for the 9 procedures in Table 2, and Table 4 expresses reimbursement in dollar-per-hour terms. The medians give equal weight to each practice for which an hourly reimbursement rate could be calculated, and the weighted medians weight each hourly reimbursement rate by the number of the practice's patients in the payer category during the past 12 months.
Regarding additional time spent by audiologists that is not reimbursed, 90 and 89 respondents reported time spent before and after surgery, respectively. None reported zero unreimbursed time. Median and mean unreimbursed times were 3.0 and 4.6 hours, respectively, before surgery (interquartile range, 2.0-5.4) and 5.0 and 7.1 hours, respectively, after surgery (interquartile range, 2.0-10.0).
Of 6 procedures included in the physician survey, only cochlear implant implantation (CPT 69930) was reported as being used frequently enough to analyze reimbursement levels. Table 5 reports information about time required to perform the implant surgery, unreimbursed time of physicians spent before and after surgery, and distributions of reimbursement expressed alternatively per procedure, per hour in surgery, and per total hour.
Table 6 summarizes prices paid by hospitals during the past 12 months to purchase cochlear implant systems and reimbursement amounts from private insurers when hospitals were reimbursed separately for devices. Among the 27 hospitals that reported both average purchase prices and average private reimbursement rates, 15 reported purchase prices exceeding reimbursement and 6 reported the opposite. The median and mean differences between purchase price and reimbursement were $2000 and almost $3700 per device, respectively.
Medicare payment rates for calendar year 1999 ranged from $14 086.90 to $14 791.63 per device for outpatient surgery. National average reimbursement per short-stay discharge under diagnosis related group 49 was $8982.58 during fiscal year 1997.
Under Medicare's OPPS, outpatient services related to cochlear implant surgery are assigned to ambulatory payment classification group 259 ("Level VI ENT [Ear, Nose, and Throat] Procedures") with a payment rate for 2002 of $19 279.89, of which $16 243.65 is intended to cover or "offset" part of hospitals' costs of purchasing cochlear im plant systems.13 In addition, cochlear implant systems (and many other innovative devices, drugs, and biologicals) are eligible for "transitional pass-through" payments. These payments will cease for cochlear implants after December 31, 2002.14 The potential transitional pass-through payment for a cochlear implant system is calculated as a hospital's charge for a system multiplied by its cost-to-charge ratio minus the portion of the ambulatory payment classification rate intended to cover (some of) the costs of acquiring systems.11,15Actual transitional pass-through payments for 2002 are substantially less than potential payments, however, because of a legislative requirement that in 2002 total transitional pass-through payments may not exceed 2.5% of total Medicare payments under OPPS.16 For 2002, potential payments were reduced by 63.6% to conform to this requirement.13 For 2002, ambulatory payment classification group payment rates were constructed to incorporate 75% of estimated pass-through device costs.13 Thus, actual transitional pass-through payments in 2002 will average roughly $1970 per cochlear implant system, and total reimbursement for the device will be roughly $18 213, about $3773 more than the midpoint of the range of fee schedule payment rates for 1999 ($14 440).
Medicaid policies for reimbursing hospitals for the prosthetic system vary widely in form, specificity, and generosity (Table 7). Respondents in several states italicasized inpatient surgery, and the categories in the first column of the table reflect that italicasis. In cases where states reported different policies for outpatient surgery, these are noted as comments. Only 9 of the 60 hospitals that provided data reported levels of device reimbursement from Medicaid, with a median of $13 800. Moreover, 29 hospitals reported that in past 12 months they had not billed Medicaid for a cochlear implant device, had not received a separate payment from Medicaid for a cochlear implant device, or both.
If we assume that 14 annual surgeries per practice as reported by physicians (Table 1) applies to the 193 practices considered eligible for the main survey and to the 25 practices used for pilot testing, the total number of cochlear implant procedures performed in the United States in 1999 was approximately 3050. Are financial incentives a contributor to limited access?
Because we cannot observe responses to financial incentives, we analyze incentives quantitatively and draw inferences about potential behavioral responses. For audiologist and physician services, costs of serving patients receiving cochlear implants include professionals' time required and expenses for other required resources such as office staff, office space, insurance, and equipment. Substantially increasing access to cochlear implant services would almost certainly involve significant costs in addition to professionals' time.
Data on hourly incomes of audiologists from the 1999 Omnibus Survey of the American Speech-Language-Hearing Association (Rockville, Md) were provided in October 1999 by Jeannette Janota, PhD, of that organization. Median salaries in 1999 for audiologists with Certificates of Clinical Competence in Audiology were $20.00 (n = 142) and $23.18 (n = 160) per hour for audiologists working 26 or more and 25 or fewer hours per week, respectively. Hourly incomes of otolaryngologists are gauged from American Medical Association Socioeconomic Monitoring System data. Median figures for hours per week in professional activities, weeks worked, annual incomes, and annual professional expenses of self-employed otolaryngologists17 suggest net and gross hourly incomes of approximately $93 and $207, respectively. The ratio of median professional expenses to net income is 1.2, and excluding nonphysician payroll expenses, the ratio is about 0.85.17
Suppose that practice costs associated with audiology services add 100% to audiologist salaries (ie, about halfway between the ratios for otolaryngologists alternatively including and excluding nonphysician payroll expenses). Then, total cost per hour for audiology services would be $40 to $46. The median hourly reimbursement rates reported in Table 4, none of which factors in audiologist time that is not reimbursed, are typically more than $46 for private payers. However, they are near or less than $40 per hour for some procedures for Medicare and Medicaid. Notably, these include the 2 procedures—CPT 92506 and 92510—that require relatively large amounts of time per session and numbers of sessions (Table 2). Reimbursement for aural rehabilitation is likely to be particularly important, because several sessions are required.
In addition, reimbursement rates for aural rehabilitation reported by some Medicaid agencies are much lower than the medians in Table 3. For example, 7 agencies reported a rate of $20 or less per session for children. If financial incentives for audiologists to serve cochlear implant recipients impede access, a leading factor is likely to be payment levels for aural rehabilitation after surgery in public insurance programs, especially in some states under Medicaid.
The 25th percentile of private reimbursement for surgery per total hour of $250 (Table 5) exceeds our estimate of average gross hourly earnings ($207). For Medicare, median reimbursement per total hour is $192, and the 25th percentile value is $144 (Table 5). The former figure is not much less than that for average gross hourly income for self-employed otolaryngologists ($207), and the latter is closer to the figure for average net hourly income ($93). Thus, for surgeons who spend about the average time per Medicare patient and whose hourly gross incomes are near $200 per hour, Medicare reimbursement may cover their costs of surgery.
The median reimbursement for CPT 69930 per total hour for Medicaid is $156 (Table 5), about halfway between the figures for gross and net hourly incomes. The 25th percentile value for hourly reimbursement is $100, which is close to the figure for average net hourly incomes. Thus, for many surgeons, reimbursement may contribute to financial disincentives to serving Medicaid patients. In addition, one Medicaid agency reported a reimbursement rate for CPT 69930 of only about $600, and 11 agencies reported rates of less than $1000. Moreover, 3 Medicaid agencies reported that CPT 69930 is not covered for children, and 11 agencies reported that it is not covered for adults. We infer that in several states Medicaid payment rates fail to cover costs for many, but not all, surgeons, and that in some of these states Medicaid policies amount to a financial disincentive to perform cochlear implant surgery for most, and perhaps almost all, physicians.
Data from the hospital survey suggest that even private insurance can fail to cover device costs; 15 of 27 hospitals reported higher average purchase prices than average private reimbursement levels.
Regarding outpatient surgery under Medicare, the midpoint of the range of 1999 reimbursement rates for a device ($14 440) was about $6400 and $5300 less than the median and mean reported purchase prices, respectively (Table 6). The 2002 payment rates under OPPS are about $3773 higher than that midpoint, with a little more than half of that difference attributable to transitional pass-through payments that will expire at the end of 2002. Regarding inpatient surgery, the average Medicare reimbursement per hospital discharge ($8982.58) in 1997 was less than half of both the median and mean purchase prices (Table 6). Thus, Medicare diagnosis related group payment levels often confronted hospitals with financial losses of more than $10 000 for each inpatient cochlear implant surgery.
Information in Table 7 indicates that in at least 18 states Medicaid payment policies usually or often failed to cover hospitals' costs of purchasing cochlear implant devices; these 18 states account for 44% of national Medicaid enrollment.12 In contrast, Table 7 also indicates that Medicaid payments usually or often covered costs in at least 8 other states; however, these states account for only 8% of national Medicaid enrollment.12 For the other 25 jurisdictions, survey information is not specific enough to draw either inference.
Suppose that many additional publicly insured candidates for cochlear implant surgery had been referred to otolaryngologists in 1999. Might lack of sufficient reimbursement have impeded many of them from being served? Consider some "back-of-the-envelope" calculations.
About 1100 Medicare and Medicaid fee-for-service enrollees received cochlear implants in 1999 (Table 1). Approximately 30% of the estimated 460 000 to 740 000 severely to profoundly hearing-impaired individuals in the United States have only public health insurance.5 Increasing implantation rates 10-fold for Medicare and Medicaid enrollees would involve 11 000 implants per year or 5% to 8% of the publicly insured, severely to profoundly hearing-impaired, an unknown fraction of whom are clinically appropriate and potentially enthusiastic candidates for cochlear implant surgery.
Findings from our surveys indicate that an additional 10 000 cochlear implants per year would require an additional 180 000 hours per year from audiologists just for aural rehabilitation and unreimbursed activity, or 340 hours per year for each audiologist currently working with patients receiving cochlear implants. Similarly, an additional 70 000 hours per year would be required from physicians for surgery and unreimbursed time, or 200 hours per year for each of approximately 350 physicians who currently perform cochlear implant surgery. Given the financial disincentives discussed already, additional time requirements of these magnitudes would involve considerable financial sacrifice by many audiologists and physicians.
Under Medicare, in 1999 hospitals lost about $6000 per outpatient surgery and more than $10 000 per inpatient surgery, respectively, on the device. Under Medicaid, the extent of any losses varies considerably, with potential losses in many states ranging from approximately $5000 to $20 000 per device. It seems conservative, then, to assume that on average hospitals would have lost $5000 for each of the hypothetical 10 000 additional publicly insured patients who would have received implants in 1999. This implies an additional loss in 1999 of $50 million or about $250 000 for each hospital that provided access to cochlear implant surgery.
As 5 physicians suggested in response to the open-ended question, professional (and institutional) providers may be willing to serve some patients receiving cochlear implants despite financial sacrifices. Motivations include professional responsibility, personal satisfaction from making profound differences to people's lives, a desire to be at the professional cutting edge, and the possibility of generating publicity. However, current rates of access may largely satisfy motives for providing unprofitable services, and therefore such motives may not suffice to support substantial expansion of access under current financial arrangements.
We shall discuss herein some qualifications and comment on some factors and issues not already addressed.
First, consider potential survey response biases. Directions of any biases in the audiologist and physician surveys are difficult to gauge. Providers facing relatively low reimbursement rates may be underrepresented in our samples if, as a result, they serve fewer cochlear implant recipients and are thus less motivated to respond. Alternatively, providers facing relatively low rates may be more concerned about payment issues and thus more motivated to respond. Participation by hospitals is unlikely to be related to the key items in our survey, because cochlear implants generally represent tiny fractions of purchasing activity or revenues for hospitals. Finally, we expect that participation by Medicaid agencies is determined almost entirely by agency policies or attitudes of individuals regarding cooperation with research studies and by time availability.
Second, financial incentives depend on costs and revenues totaled over all services, but our analyses for professionals focused on selected services. Nonetheless, our conclusions appear valid, because we focused on key services, and additional data suggest that payments for other services rarely, if ever, make up for the losses we have analyzed. For audiologists, we italicasized aural rehabilitation (CPT 92510). While per-hour payment is higher for some other services (Table 4), the other service standing out as requiring considerable time is CPT 92506 (Table 2), which involves even lower per-hour payment (Table 4). Moreover, our analysis of audiologist incentives ignored their unreimbursed time. For physicians, we italicasized reimbursement for surgery (CPT 69930). Relatively few physicians reported using other codes to bill for their time, and payment levels for other codes are relatively low. Under Medicaid, for example, medians for these other payments were all less than $100, and only one was more than $60.
Third, our analysis for hospitals considered only purchase prices and reimbursement rates for devices. It seems implausible, however, that hospital costs other than device costs are often, if ever, reimbursed sufficiently to compensate for losses on the devices, which often amount to $5000 or more.
Fourth, we have not addressed various issues of interest. For example, we did not investigate in detail differences in outpatient and inpatient reimbursement policies under Medicaid, nor how differences in payment policies may affect whether surgery is performed on an inpatient or outpatient basis. We have also not investigated how payment rates vary with the organizational forms of payers (health maintenance organization, preferred provider organization, etc). Nor did we collect information concerning retrospective payment denials; our surveys requested average payment rates when services were reimbursed, and thus our analysis does not account for potential effects of payment denials on financial incentives. We also have not surveyed speech therapists or analyzed payment rates for speech therapy (CPT 92507). Although CPT 92507 was included in the audiologist survey, only 16, 10, and 8 audiologists reported sufficient information to calculate hourly payment rates for private payers, Medicare, and Medicaid, respectively.
Fifth, we have not analyzed potential sources of funding to augment insurance payments such as state programs other than Medicaid, private donations, and foundation grants to cochlear implant programs.
Sixth, our ability to draw inferences is limited by the nature of available information concerning earnings and practice costs of audiologists and physicians. For example, we do not know whether the average incomes of audiologists and otolaryngologists who work with cochlear implant recipients differ substantially from the average incomes of all Certificate of Clinical Competence in Audiology audiologists and self-employed otolaryngologists, respectively.
Seventh, the time and device costs used in analyzing potential financial burdens of expanding Medicare and Medicaid access may be somewhat high. Major increases in cochlear implant production might reduce unit production costs and device prices because of economies of scale and learning in production. Moreover, one third of our sample physicians reported that their practices had implanted 5 or fewer cochlear implants in the past 12 months; additional experience with patients receiving cochlear implants might decrease audiologist and physician time costs per patient (and improve patient outcomes).18
Regarding the scope of our analysis, we have analyzed financial incentives only under fee-for-service insurance. Capitated public insurance arrangements, such as Medicare + Choice (Part C) plans and Medicaid full-risk plans, are becoming more widespread. However, in 1998 shares of revenues of otolaryngologists from capitated contracts were low (mean, 7.6%; median, 0%; 75th percentile, 5.0%).17 We also have not analyzed exclusion of cochlear implants in private health insurance contracts, which existed, but was not common, during the early 1990s,19 or the stringency of medical-necessity criteria and preauthorization procedures.
A troubling picture emerges from this analysis. Insurance payment levels that do not cover the costs of physicians, audiologists, and hospitals could be an important factor limiting access to cochlear implants, and gaps between payment levels and costs are particularly striking for Medicaid and Medicare. This is of substantial concern because many potential candidates for cochlear implants are insured only through these programs.
Many factors other than insurance-based, financial disincentives are likely to limit access to cochlear implants, and to increase access substantially, various potential barriers would need to be addressed. Our data point to specific elements of financial incentives that seem particularly important. For patients to benefit from cochlear implants, audiologists, physicians, and hospitals must all be willing to provide services; the access chain is only as strong as its weakest link. While hospital reimbursement appears to be the weakest link in most cases, in some instances hospitals may be willing to allow more cochlear implant surgeries than physicians want to perform.
To hope that providers will serve candidates for cochlear implant surgery despite financial sacrifices may not suffice to achieve socially desirable rates of access. A key message for policymakers is that facilitating access to cochlear implants for qualified candidates enrolled in Medicare and Medicaid could require changes in payment policies at both the federal and state levels. Expanding access to cochlear implants for qualified candidates poses complex challenges to providers and manufacturers of cochlear implant systems as well as to policymakers.
Accepted for publication March 21, 2002.
This study was supported by Advanced Bionics Corporation, a manufacturer of cochlear implant prosthetic systems.
The research was conducted independently, and RAND retains all rights to publish data, interpretations, and conclusions.
We thank Mary Lou Gilbert, JD, and Meg Bernhardt, MS, for research assistance, and Jay Bhattacharya, MD, PhD, for helpful comments on a previous draft.
Corresponding author: Steven Garber, PhD, RAND, 1700 Main St, Santa Monica, CA 90407 (e-mail: Steven_Garber@rand.org).