By Christopher Guadagnino, Ph.D.

Delaware County Medical Society President Mark F. Kelly, M.D.

Published November 2000

• React to this article in the Discussion Forum.

Physician leaders recently testified before the Pa. House Insurance Committee that managed care dominance in Pa. is having such a negative impact on the medical practice environment—particularly driving down reimbursements and mandating unfavorable provider contract provisions in a unilateral fashion—that it is driving physicians to leave the state or retire early, making physician recruitment difficult and producing worrisome shortages in some specialties, thereby eroding the quality of patient care.

At the same hearing, Independence Blue Cross (IBC) Vice President for Patient Care Management Gary M. Owens, M.D. disputed those allegations, counter-charging that there is an abundance of physicians in Pa. and an oversupply in southeastern Pa. Owens cited state licensure data, statistical evidence of increased per-member-per-month payments to physicians and a marked increase in physician visits to suggest that an increasing supply of physicians is contributing to higher utilization and health care costs.

An analysis of these two seemingly incommensurate alternatives turns up issues far richer than a mere arithmetic assessment would suggest. Both views can be right, simultaneously, as revealed by the experience of physicians and recruitment firms, and the analysis of health economists, all of which indicates that a managed care-dominated practice environment can have serious impacts on both physician distribution and the practice of medicine.

Physician Experience

On August 10 in Delaware County, during one of four House Insurance Committee hearings on Rep. J. Scot Chadwick’s (R-Bradford) joint physician negotiation bill (HB 2685), several physician leaders decried the imbalance in market power enjoyed by Pa. managed care insurers and described some of the negative impacts they had personally experienced.

John D. Kelly, IV, M.D., past president of the Pennsylvania Orthopedic Society, argued that the lack of health care competition, such as in southeastern Pa. where Kelly said IBC holds an unfair market advantage of 60 percent, has led to rising revenue for insurers and declining physician compensation and patients’ rights, and is causing the physician community in the region to lose young talent.

Independent physician practices are at extreme economic risk when subjected to reimbursements that are under fair market value by HMOs acting with monopolistic power, testified C. Richard Schott, M.D., past president of the Delaware County Medical Society and Chairman of the Division of Internal Medicine and Chief of the Subdivision of Cardiology at Riddle Memorial Hospital in Media, Pa.

Keystone Health Plan East introduced a contact capitation program for cardiologists in 1994 and for gastroenterologists shortly afterward, paying a single fee for three months of care, which Schott said has put physicians at increasing economic risk in the face of increasing costs of state-of-the-art services. Some cardiology groups in suburban Philadelphia, he told committee members, have threatened to drop their contact capitation contract, which could create a disruptive "domino effect" whereby other practices could not afford to take on the increased volume of displaced HMO patients, nor the losses they would incur from under fair market reimbursements.

Access to, and quality of, cardiology and gastroenterology services in southeastern Pa. are approaching a crisis level, Schott argued, as practices are unable to economically compete with those in other regions to attract new quality physicians to replace those which have retired or fled the area. Offers for entry-level salaries elsewhere, he said, are above what many physicians with established practices are now earning in the region.

Schott also noted to Committee members that physicians over age 50 are retiring prematurely. A Merritt, Hawkins & Associates survey of 300 physicians over age 50 nationwide—a category that represents 38 percent of all U.S. physicians—found that 38 percent plan to retire in the next one to three years, while 12 percent plan to seek jobs in non-medical settings and another 28 percent plan to close their practices to new patients, significantly reduce their patient loads or work on a temporary basis. Fifty-six percent of the survey’s respondents cited managed care as the single greatest source of their professional frustration.

"I am certain that I made a mistake in practicing in this state," declared Mark F. Kelly, M.D., an ENT physician in Drexel Hill and president of the Delaware County Medical Society, to the Insurance Committee, referring to his ability to provide excellent care increasingly hindered and questioned by managed care companies over the past six or more years. Kelly relayed a number of anecdotes to the Committee, including that of an anesthesiologist who told him that an average of one or two operating rooms were closed over the past six months due to lack of anesthesiology coverage, resulting in patient surgical delay.

Kelly says he presently knows of 18 physicians from the Main Line Health System who have left the area within the past year, while one of his main referring physicians is moving to a better practice climate in a different state rather than remain in Pa. in a practice that he started.

The surgical subspecialties are the most prominently affected, Kelly said, noting that starting salaries elsewhere exceed established salaries for southeastern Pa., almost exclusively because of the market dominance of IBC and Aetna U.S. Healthcare allowing them to set payment rates.

Physician Licensure Data

These accounts of physician flight seemed to have had little impact on several Committee members, who also heard testimony from Owens of IBC that it is "simply wrong" that physicians are leaving Pa. because they cannot earn enough money. Owens noted that, according to the Pa. Medical Licensing Board, there were 48,729 licensed physicians in 1995 and 53,382 today, with virtually no change in Pa.’s population. In February 1997 there were 13,456 licensed physicians in the five-county southeastern Pa. area, Owens added, increasing to 15,281 in July 2000.

Owens argued that a physician oversupply was driving up health care cost and leading to overutilization, noting that IBC’s Personal Choice PPO data revealed an eight percent increase in per-member-per-month payments to physicians from 1998 to 1999, as well as a 7.5 percent increase in patient visits during that same time period. Owens cited A.M. Best’s 1999 annual report on HMO statistics, which reported an average of seven physician visits per member per year in Pa. while it is only 4.9 visits per year in the nation.

Faced with conflicting accounts of physician supply and impacts on patient care, Committee members, including Rep. William F. Adolph, Jr. (R-Delaware) and Rep. Joseph M. Gladeck, Jr. (R-Montgomery), pressed the physicians who testified to offer more than just anecdotal information to refute Owens’ hard data. The confrontation led to particularly uncomfortable moments during the hearing in which physicians were seemingly unable to offer more than their personal experiences.

The Pennsylvania Medical Society takes issue with the numbers Owens cited. Although the aggregate number of physicians may not have declined over the past few years in Pa., the number of physicians who are directly involved in patient care may have declined, according to Don McCoy, director of regulatory affairs for PMS, which hopes to collect more concrete data on that issue. The PMS has determined that 17,000 of the physicians claimed by Owens to practice in Pa. as of July 2000 had out-of-state addresses, a number which has grown by 4,000 since 1999, McCoy notes.

According to Diane Miller of the Pa. Board of Medicine, as of July 7, 2000, Pa. had 35,761 active physicians, including 31,252 MDs and 4,509 osteopathic surgeons. The number excludes medical trainees but does not necessarily reflect the number of physicians focused on patient care, says Miller. The number of active Pa. physicians in 1995 was 32,360, she adds. While Owens’ figures on the total number of Pa. physicians do not parallel those reported by Miller, his conclusion that the total number of physicians in Pa. has increased between 1995 and 2000, does.

AMA data further suggests that aggregate physician numbers exaggerate the number of actively practicing physicians in Pa. Approximately 80 percent (30,699) of Pa.’s total number of non-federal, active physicians (38,080) were classified as focused on patient care by the American Medical Association in their year 2000 edition of Physician Characteristics and Distribution in the U.S., which reflects 1998 data. The patient care category excludes physicians who practice fewer than 20 hours per week.

Physician density per total population suggests that there is no physician oversupply in southeastern and western Pa., argues Stephen Foreman, Ph.D., J.D., a health care economist and assistant professor at Penn State University, who has also been commissioned by the PMS to study Pa.’s health insurance markets.

According to the U.S. Dept. of Health and Human Services Area Resource File data from 1997, he notes, there was an average of 2.85 active physicians per 1,000 total population for all 314 metropolitan statistical areas (MSAs) in the U.S. The average number of physicians per 1,000 population for the ten largest MSAs in the country was 3.87. The nine-county MSA that includes southeastern Pa. and parts of New Jersey, which is one of the ten largest MSAs in the country, was 3.52. For the six-county MSA that includes Pittsburgh, average number of physicians per 1,000 population was 2.97.

If the data is age-adjusted, argues Foreman, one would be hard-pressed to claim a pattern of physician oversupply, given the greater need of physician services by an elderly urban population. For the top 25 MSAs in the U.S. in 1997, there was an average of 23.16 physicians per 1,000 population over age 65, he notes, while the figure for the Philadelphia-area MSA was 25.12 and the figure for the Pittsburgh-area MSA was 16.66.

The number of teaching hospitals and physician training programs in the Philadelphia area, as well as the relatively high indigent population, which both require a high level of intensity of medical care, also significantly skews the physician density data for the region, Foreman notes.

Physician Recruitment Experience

Even if physicians cannot make the claim that a bad practice environment has reduced the overall physician supply, that does not mean that there are not recruitment problems for certain specialties caused, in part, by low reimbursement and credentialing problems, as suggested by physician recruitment experience.

Physician recruitment demand by hospitals in western Pa. is not as robust as it was a few years ago, while demand in Pittsburgh has dropped by as much as 30 to 50 percent in the past three years, observes Daniel Stern, president of Daniel Stern & Associates in Pittsburgh. Stern attributes the downturn to significant hospital losses on operations since the Balanced Budget Act of 1997 and to a dramatic increase in medical malpractice insurance premiums in the past few years.

An unattractive practice environment is also influencing physician location decisions. Some private physician groups in western Pa. are spinning off junior partners from their practices because of pressures from reduced practice reimbursements, according to Stern. For every physician who wants to come to Pa., Stern observes, there are eight or nine who want to go elsewhere. Many resident physicians who Stern has spoken with say they rather not come into the region.

Recruiting to rural parts of the state, on the other hand, continues to be a challenge and all of the rural hospitals are scrambling either for replacement physicians or to fill the needs of a growing town, says Stern.

Recruitment in Pa. has drastically shifted to surgical specialty areas such as cardiology, radiology, orthopedics, urology and nephrology, notes John Lawlor, president of Lawlor & Associates in West Chester. While recruitment demand has decreased overall in the past few years, particularly in the greater Philadelphia and Pittsburgh metropolitan areas, there are pockets of unmet recruitment demand for quality candidates in both markets for those specialties, at least partly because of low reimbursements, Lawlor says.

There is no longer as great a demand for primary care physicians in metropolitan areas, as competing hospital networks that have aggressively recruited them have created a false demand before becoming unravelled from economic losses, he adds.

Another fallout of the practice purchase feeding frenzy by hospital networks is a sizable number of primary care physicians in the Philadelphia region with several years of experience who are reluctant to move and have accepted income reductions and less attractive positions to stay in the region, notes Robert Templin, a principal of Longshore.Simmons.Garofolo in Philadelphia. Over the past year and a half, some 40 percent of those physicians have opted to leave the area, estimates Templin, who has been placing those physicians in areas such as New Jersey, New York and Maryland. Twice as many out-of-area primary care physician placements came from the Philadelphia region this year compared to the previous year, says Templin.

Most specialty physician placements for clinical positions, he says, have been in New Jersey and central and upstate Pa. It has become difficult to convince specialists to come to the Philadelphia area, says Templin, particularly non-invasive cardiologists, pediatric specialists, gastroenterologists and internal medicine specialists.

Unmet recruitment demand may be occurring despite an adequate supply of physicians in the aggregate, says Templin, who notes that shifts in reimbursements and contractual provisions have hurt Philadelphia-area physician practices that have overexpanded and now have more providers than they can support. Non-compete obstacles faced by departing associates, combined with an unattractive reimbursement environment, have produced a re-positioning of specialists to more attractive practice locations, Templin believes. Templin has had good success recruiting physicians to places such as Harrisburg, Lancaster, Ephrata, Scranton and Wilkes-Barre.

A significant obstacle to recruiting physicians to southeastern Pa. is the inability of practices to offer a competitive salary and benefits package to a new associate. Northern New Jersey, for example, offers 70 to 80 percent higher salaries to emergency physicians than can be matched by George Walters, M.D., director of the Emergency Department of Riddle Hospital in Media, Pa. Walters’ department needed to move to double coverage, requiring six full-time physicians. Walters says it has taken him two years to attract the fifth physician, and he is still trying to attract a sixth. His practice currently has one and a half full-time slots unfilled.

Walters says that salaries in his practice flow almost entirely from what the physicians receive from health insurers, 40 percent of which is determined by Aetna and IBC. According to Walters, emergency departments throughout southeastern Pa. are having difficulty recruiting for the same reason.

The situation may lead to a degradation of the quality of care. Walters says he can hire a non-board-certified, non-residency-trained emergency physician for the salary he is able to offer, but adds that he and his partners have decided only to hire boarded candidates. He says he is not sure how long his practice can stick to that policy, given the need for physicians in his department spurred by the volume of patients it is seeing.

Another key obstacle to recruiting in western Pa., argues Lawlor, is a pattern of health insurer credentialing delays, especially by Highmark Blue Cross Blue Shield. Hospitals are looking for physicians who are already credentialed, thereby shrinking the candidate pool, because of credentialing delays of up to five to six months, Lawlor notes. If a new hire cannot see and bill for 30 to 40 percent of a practice’s patients—i.e., Highmark enrollees—for up to half a year, that makes it difficult for physician groups to remain competitive if they bring on a new associate, Lawlor maintains.

Highmark says that it adjudicates clean credentialing applications in 60 days or less, with delays occurring only when there is a problem on the application, such as insufficient verification of a physician’s hospital training, explains Highmark’s Medical Director of Quality Improvement Carey Vinson, M.D.

According to Highmark, 50 percent of its physician credentialing applications this year have been "clean," while 50 percent are in some manner incomplete and require a followup fax, phone call or signature. The amount of lag time in those cases depends on the responsiveness of the practice in supplying the needed information, Highmark notes.

Highmark is working to accommodate NCQA’s recently introduced provisional credentialing plan for new physician graduates who have completed fellowship or residency training, whereby such a physician can commence seeing a health plan’s patients while the plan reviews the physician’s credentialing criteria, says Vinson. Until HCFA officially recognizes the provisional credentialing plan, it will not be available for any of Highmark’s managed care products, Vinson explains. Highmark hopes to persuade HCFA to accept the NCQA plan in the next few months, he adds.

A health insurer’s network needs may further control the flow of physician entry into a region. Highmark has a network development process to analyze whether it has enough physicians for its panels, Vinson explains. Under Highmark policy, all primary care and ob/gyn physicians are approved for network development and can go through the credentialing process, as can any physician joining an existing group in his or her same specialty.

Highmark scrutinizes other physicians seeking to be credentialed according to ratios of practitioners to its enrollees, practitioners to hospital staff, and practitioners to mileage, the latter to ensure that a specialist is available within 30 miles or less to enrollees, says Vinson.

Highmark typically denies credentialing to fewer than three physicians per month, while approving 100 to 200 physicians per month, notes Vinson. From Highmark’s network development standpoint, urban areas are the most difficult for physicians to enter, particularly specialists, Vinson adds.

Highmark has an internal work group that works collaboratively with physician groups and hospitals to help them anticipate their needs and has made exceptions to its network development criteria, according to Vinson. Highmark has agreed, for example, to approve the credentialing process at rural hospitals for additional general surgeons, and has similarly approved a solo ENT physician sought by a hospital in one of Allegheny’s surrounding counties whose other two ENTs were approaching retirement age, Vinson explains.

Physician Location Research

Preliminary national studies of the impact on managed care on physician practice location choices seem to support the experiences of physicians and physician recruiters in Pa., namely, that a managed care environment negatively impacts physician practice location choices.

One study published in the November, 1998 Medical Care by RAND health program economist José J. Escarce, M.D., Ph.D. and colleagues found that younger physicians were much less likely to establish a new practice in an environment with high HMO penetration, holding other variables equal. The study looked at all U.S. physicians between 1989 and 1994 in major metropolitan areas whose main professional activity 2.5 years after completing residency or fellowship was direct patient care and found that HMO penetration had a strong negative effect on practice location decisions of new specialists and a weak but significant negative effect on new generalists. The authors posited that higher HMO penetration decreases earning opportunities for specialists and may also reduce physicians’ quality of life by eroding professional autonomy. HMO penetration in the markets studied averaged 27 percent and ranged from zero percent to 61 percent.

According to 1999 Interstudy data, HMO penetration in the the Philadelphia-area MSA is 52.2 percent, while in the Pittsburgh-area MSA it is 30.8 percent, notes Foreman. These figures would make the two regions contenders for the effects found by Escarce and colleagues.

Another national study published in Vol. 19 (2000) of the Journal of Health Economics co-authored by Escarce investigated whether HMO penetration induced established clinicians to relocate their practices or leave patient care. The study found that physicians in early career—40 years of age or younger—were twice as likely as physicians in mid-career to relocate their practices and were also more likely to leave patient care. Hospital-based specialists were more likely than either generalists or medical-surgical specialists to relocate and to leave patient care. More than two-fifths of physicians who relocated chose markets with lower HMO penetration levels and about one-third chose markets with the same levels. Further, more than five-sixths of the physicians who relocated moved to a different metropolitan area.

A third study published in the October, 2000 Health Services Research by Escarce and colleagues assessed the impact of the growth in HMO penetration between 1986 and 1996 on the change in the number of generalists, specialists and total physicians in metropolitan areas. HMO penetration did not affect the number of generalist physicians or hospital-based specialists, but faster HMO growth led to smaller increases in the numbers of medical-surgical specialists and total physicians, as well as to larger increases in the proportion of physicians who were generalists. The findings imply that, during the past decade, there has been a redistribution of physicians, particularly medical-surgical specialists, from metropolitan areas with high HMO penetration to low-penetration areas.

The study noted that PPOs probably have weaker effects than HMOs on physician behavior, but suggested that, because PPOs discount fees, they could affect physician incomes and hence the attractiveness of different market areas to physicians.

Escarce is the coauthor of another study, in press, that shows a moderate effect of premature retirement by physicians in markets with high HMO penetration.

The findings of these studies strongly indicate that managed care presence alters the geographic distribution of physician practices at the market level and upsets a market’s historical balance of physician age profile. A market dominated by managed care, says Escarce, seems to result in fewer physicians of all types, particularly specialists, and produces a remaining physician profile that is disproportionately middle-aged, with fewer younger and older physicians.

Patient care would suffer, he maintains, inasmuch as a market is both stripped of its more experienced older physicians who have large, active patient practices, and is left with a shortage of younger physicians who are trained in new technologies and who can create a base for future physicians in the market.

The urban markets in Pa. may take those effects to an even higher level. Escarce’s studies focused on markets with high overall HMO penetration, but did not investigate potentially anticompetitive behavior of dominant managed care companies within a single market and their impacts on physician location choices.

Health insurers in southeastern and western Pa. appear to be behaving as though there were a physician oversupply—negotiating fees and contract terms that are not favorable to physicians—which could have catastrophic results if they are wrong. As the research suggests, physicians who have invested several years establishing a patient base for their practice, as well as family and personal ties to a community, are the least likely to leave a market, even one dominated by managed care. Those physicians may tolerate reimbursements below cost for some time.

Interstudy data from 1999 reveals that physician payment per-member-per-month in the Philadelphia-area MSA is the second-lowest among the top 25 MSAs in the U.S., notes Foreman. From an economic standpoint known as "Target Income Hypothesis," physicians in such a situation may increase their number of patient visits, lab tests and procedure utilizations in an attempt to compensate for low reimbursements. Such an outcome would be consistent with the experience of IBC’s PPO product as reported by Gary Owens.

If a health plan’s physician profiling apparatus cannot isolate overutilizers and the plan compensates for alleged overutilization and higher health care costs by further ratcheting down reimbursements across the board, as Pa. experience bears out, the additional squeeze may only intensify system-gaming behavior of physicians whose only other recourse is to leave the region. Health care cost increases aside, overutilization of health care services increases the chance of harm to the patient.

Short of accepting a lower standard of living, the other alternatives left to physicians are to leave the area or leave patient care, producing a shortage of physicians that even IBC cannot dispute.