Physician profiling: trends and implications.

Every professional is subject to profiling, and most are profiled more rigorously, more thoroughly, and more dispassionately than are physicians. Profiling occurs when society has the technologies and the economic incentives to perform profiling. In the first of this two-part article, we will consider trends in profiling all professionals and those in profiling physicians. We conclude in the second part of the article with a discussion of the specific methodologies and vendors used in profiling the practice habits of physicians.

Professional once meant a person working in a learned discipline (medicine, law, teaching) that maintained standards of quality and performance for its members and that organized continuing education for them to keep their skills current. Now professional means anyone paid to perform a set of tasks associated with a vocation.

We are accustomed to profiling professional athletes. We measure the performances of our athletes on the field completely. We use carefully compiled statistics to select them for our teams, label the best performers as all-stars, calculate their incomes, rank their accomplishments for posterity, admit them to various halls of fame, and plot their gradual senescence to retirement with declining statistics. We argue about who was the best batter with lifetime batting and slugging averages; runs batted in; and counts of singles, doubles, triples, and home runs. We debate about the best professional running back with measurements of average yards per carry, average carries per game, numbers of touchdowns, and success of the teams on which he played in winning the Super Bowl. We rank professional hockey forwards by their average goals and assists per game, their total goals and assists, and the records of their teams in competition for division titles and the Stanley Cup.

How are litigators measured? Clearly, by the proportion of cases they litigate to victory for their clients and by the income they generate for their firms. Some attorneys are measured by the number of articles they publish in legal journals and by the number of clients they bring to their firms. Management consultants and accountants are measured in a similar way. Legal and accounting firms face increasing competition, and more of their customers negotiate fixed price contracts. How are the partners responsible for those fixed revenue accounts evaluated? Their firms know what they negotiated as fees and can estimate the costs of servicing those contracts. The partners are evaluated on the basis of the profits (or losses) of their firms in servicing those accounts. Engineering and advertising firms have more experience with fixed price contracts. They routinely propose fixed price contracts to their potential customers in soliciting business. The partners who craft those proposals are measured by the business they generate and by the profits (or losses) their firms incur in meeting their obligations under those contracts.

How are movie directors measured? By the profits (or losses) the movies they direct generate. Behind the profits (or losses) are the skills of the director to capture footage that will interest enough people to pay money to see the film so that the revenue generated by the movie exceeds the expenses of making it. Of course, the director has to be able to motivate people; plan complex logistics of casting, staging, and film production; and control budgets. But his or her performance is defined by the profits (or losses) of the films he or she directs.

Consider the managers of mutual funds. How are they evaluated by their employers? By the growth of their mutual funds, which, in turn, is directly related to the total annual financial return of their funds and the performance of their funds in bull and bear markets. There are many details that go into successful selection of financial goods stocks and bonds - for mutual funds, but the measures of success are growth and annual financial return of the fund.

Profiles of Professionals Are Measures of Their Outcomes

What do all these measures have in common? There is one common characteristic of runs batted in. slugging averages, goals and assists per game, free throw percentages, points and assists per game, numbers of new customers per year, total revenue produced per year, ratio of revenue to cost of films, and total annual return of a mutual fund. They are all measures of the key outcomes of the persons producing them, and the better those outcomes, the more valuable are the persons producing them. The education, experience, training, and hard work that went into producing those outcomes do not matter in the marketplace. The customer has no interest in paying for the processes used to produce movies, contracts, engagements, and annual returns. The customer wants and pays for outcomes. We measure golfers by the number of tournaments they have won. We count the basketball player's assists, points, blocked shots, steals, and free throws made per game. All the training that goes into an athlete we take for granted. We know basketball, football, baseball, and hockey players, golfers, investment managers, corporate executives, engineers, lawyers, and movie directors have invested decades of hard work to perfect their skills in the marketplace, and we may enjoy watching them perform their work, but we determine their incomes and bonuses, and their eligibility for their respective halls of fame, by their outcomes.

Competition and commerce depend on standardized measures of the goods and services traded. An owner of a professional football team will insist on knowing the measures of a prospective quarterback's performance in college and will pay more for one with a higher pass completion average, larger number of yards covered on offense per game, and higher percentage of games won than he or she will for a quarterback with lower scores on those key outcomes, provided those outcomes are adjusted for the competition the quarterback's team faced.

When you consider buying stocks or bonds, you rely on generally accepted accounting principles (GAAP) to define the financial performance of the corporation whose financial assets you may purchase. The marketplace for stocks and bonds utterly depends on trust that the profits and losses, net income, and assets and liabilities of all corporations allowed to sell such assets are calculated in the same way. We measure the relative value of stocks by their price-to-earnings (P/E) ratios and their current prices per share. The price is determined in the marketplace. Earnings are determined by accountants and independent auditors, who define earnings as the excess of all revenue after all expenses have been paid. Earnings thus depend on an accurate, and standard, accounting of revenue and expenses. How can you compare the P/E ratio of the common stocks of two corporations if the E is determined in a different way for each firm?

Now, let's consider the practice of medicine. Medicine may be more difficult to practice than any of these other professions. Medicine may require more intellectual talent and longer years of training than these other professions. Medicine may include more process variables and more outcomes of importance, and those outcomes may be harder to measure. Nevertheless, we are learning to measure the processes and outcomes of medicine better every day. Learning to measure the processes will help us codify them into practice guidelines. Learning the outcomes will allow us to define the performance, and determine the market value, of providers and health plans.

The concentration on outcomes in medicine coincides with medicine's maturing into a business where contracts between purchasers of care (employers and insurers) and delivery systems of providers determine which providers treat which patients. The outcomes movement in medicine also coincides with our emerging ability to measure and standardize the processes and outcomes of medical care. We have standardized nomenclatures for diagnoses and procedures (CPT-4 and ICD9-CM), for patients' functional status (Short Form-36), for physicians' findings on examinations before and after surgery (TyPE scales), and for health risk (health risk appraisals). We have inexpensive means of storing all those data in electronic form, in relational databases on personal computers, to permit rapid analysis of clinical and financial data. We have learned how to adjust for the prior probabilities of specific outcomes in patients - before clinicians' treatments increase the likelihood of good outcomes and reduce the probability of undesirable outcomes.

Profiling Medical Professionals and Their Firms

Competition and commerce have come to medicine. Providers will not be paid much longer by fee-for-service methods financed by cost-based indemnity insurance. Providers' customers are corporations that wish to purchase health care services for their beneficiaries in large contracts with organized delivery systems that share some, or take all, of the financial risk for the care of those defined populations. Providers are aggregating into competing delivery systems, most of them defined by large insurers that have organized HMOs and PPOs that contract with individual hospitals and physicians. But some organizations of providers contract directly with employers. Employers want the means to compare the virtues of various health care delivery systems that they may offer to their beneficiaries, considering outcomes in addition to premium price.

In the absence of differentiating outcomes among health plans, most employers select health plans (and their networks of providers) on the basis of premium price. But some beneficiaries successfully sue for damages when health care insurers and employers restrict their access to medical care to specific delivery systems that deliver inadequate care. Not all delivery systems are alike. Characteristics other than price are important to define the "quality" of a health care delivery system. Employers and insurers want to measure those characteristics. Providers who want to earn more money for better work will want those characteristics measured.

The era of profiling hospitals arrived in the 1980s, when the Health Care Financing Administration began paying hospitals fixed case rates for Medicare patients. In turn, hospitals began to study the practice habits of physicians, because their orders for services for Medicare inpatients determined whether or not the hospitals profited from the care they delivered. The profiling was limited to inpatients, and most of the profiling was kept from the purview of the physicians profiled.

Public comparisons of outcomes were limited to hospitals after states began collecting standardized outcomes measures of inpatients in the 1980s. Pennsylvania mandates that medical records of all inpatients be abstracted by the MedisGroups methodology. California and New York require hospitals to report on inpatient admissions using the standard UB-92 data set but risk-adjust those data by proprietary methodologies developed by each state. Virginia requires the same data collection from hospitals but uses the DATIS methodology for risk adjustment.

Now more than 30 states collect a limited, standardized data set on inpatients. While most states collect the uniform billing abstract (UB-92) for every inpatient - which includes up to nine diagnosis codes and six procedure codes, birth date, admission date, discharge data, and discharge status of every patient - no states collect detailed data from outpatient visits, except for claims data from their own Medicaid populations. States are not creating large databases of claims from inpatient and outpatient services with which to profile physicians' practice habits over episodes of illness. Efforts to profile physicians' practice habits over episodes of illness are limited to insurers with large claims databases, and insurers are beginning to use those databases to invite individual physicians and group practices for their "select" PPO networks, meant to replace their open preferred provider panels that include most providers in any given community.

Economic Credentialing 

Hospitals use the discharge abstract (UB-92) to profile physicians practice habits for economic credentialing, deciding, in a few well-documented cases that have led to litigation, that certain clinicians, do not warrant admitting privileges because their practice habits, adversely affect the economic performance of the hospital. A substantial number of legal cases appear to have established a clear precedent in favor of a hospital's right to withhold admitting privileges on the basis of economic issues, in the absence of discrimination based on age, race, sex, ethnic background, or creed of a physician. A hospital may apply economic credentialing to all physicians and set arbitrary standards that may lead the hospital to withhold admitting privileges from some physicians.

Physicians have argued in court that such behavior violates antitrust statutes, but the courts have found that physicians are sellers of services to patients (and hospitals) and not consumers, and antitrust statutes are meant to protect consumers, so they do not apply in most situations of loss of admitting privileges. Courts have determined that the board of directors of a hospital has ultimate authority over the credentialing process and may unilaterally rule against the interests of a member of the medical staff. In fact, in spite of the publicity given to economic credentialing, very few physicians lose their admitting privileges at hospitals, because every physician represents potential fee-for-service income to each hospital where he or she has admitting privileges. Physicians who do not admit patients-pathologists, anesthesiologists, radiologists - are more likely to lose admitting privileges when other physicians of the same specialty win exclusive contracts with their hospital where the other hand, insurers now readily use claims databases 146 profile physicians' practice habits before winnowing their networks. Profiling by insurers, if not discriminatory by race, sex, age, or religion (creed), is deemed legal by the courts, because clinicians are considered vendors, not customers, and excluding individual physicians from provider networks does not diminish competition. Insurers may eliminate one-third to one-half of physicians from their older PPOs, preferring more restricted networks of physicians with more "efficient" practice profiles. Profiling by insurers is a much greater economic threat to most physicians than profiling by hospitals.

In the concluding part of this two-part article in the November 1995 issue, we will study specific trends in profiling of physicians by insurers, including profiling by visit, by episode, and by population of patients over time, and the methodologies and characteristics of specific profiling information systems used by health plans and insurers.

Marshall Ruffin, MD, MBA, MPH, FACPE, is President and CEO of The Informatics Institute, Falls Church, Va.