Guidelines or Physicians' Judgment?

Milliman USA Health Care Guidelines Come Under Fire

Texas Medicine Logo(1)

Cover Story -- July 2001  

By Walt Borges
Associate Editor  

In several of its presentations to promote its health care guidelines, the actuarial and consulting firm Milliman USA (formerly Milliman & Robertson [M&R]) uses a funnel to illustrate the future of health care. Through the funnel drips two drops, an image that lends itself to several interpretations.

As used in the presentations, the funnel illustrates Milliman's belief that its guidelines will be used to limit care to medically necessary care. But another interpretation -- and one that is consistent with growing physician frustration over insurers' use of guidelines -- is that the funnel is a model for the economic component of the health care system, where large amounts of employer and employee premiums are fed through an insurer's funnel to produce a small drip of payments to physicians.

Care guidelines are a lightning rod for physician discontent as insurance companies, hospitals, third-party administrators, and physician groups use the guidelines to determine what services will be paid for. Milliman's guidelines have generated a significant share of doctor criticism, largely because the company seeks to specify the appropriate length of hospital stays for various medical conditions. In fact, Milliman's insistence that a one-day hospital stay was appropriate for women after giving birth led to the term "drive-through deliveries" and to federal legislation designed to guarantee women and their physicians a longer hospital stay.

What riles many physicians is the belief that Milliman's guidelines, which the firm says are designed to designate the optimum lengths of stay for patients with uncomplicated conditions, are used by insurers to deny payments for hospital stays that are medically necessary. Not only do the Milliman guidelines provide untrained claims representatives justification to deny valid claims, the guidelines also undermine the standards of care that are traditionally based on medical standards, critics charge.  

Milliman's guidelines are the focus of two lawsuits of importance to Texas physicians. In a Houston courtroom, four current and former membersof the pediatrics department at The University of Texas-Houston Health Science Center are suing the Seattle-based actuarial and consulting firm and a fellow faculty member for including them as contributing authors of Milliman's 1998 pediatrics guidelines. They allege Milliman defamed them by listing them in a document regarded as clinically suspect by many pediatricians.

Thomas G. Cleary, MD, a tenured professor of pediatrics at UT-Houston who specializes in treating infectious diseases, says he and the other doctors initiated the lawsuit because their names appeared as contributing authors in a book that was medically suspect. "The guidelines represent practice inconsistent with what physicians believe to be accepted practice and standard of care," Dr. Cleary said.

Jan Woodward Fox, JD, the Houston lawyer who represents Dr. Cleary and the other doctors, says the lawsuit alleges that Milliman "is in the service of the insurance industry and is trying to capitalize on the credibility of academic and clinical opinion leaders.

"M&R has publicly stated that its guidelines aid HMOs [health maintenance organizations] to control physician performance and limit medical reimbursement," Ms. Fox continued. "To use for this purpose the names of doctors who have devoted their careers to the welfare of children gives the guidelines a credibility they do not deserve, hurts the doctors, and poses a risk to more than 50 million children in plans using the M&R guidelines."

Farther afield, the use of the Milliman guidelines by health insurers and HMOs is expected to provide key evidence in a federal racketeering lawsuit in federal district court in Miami. The lawsuit, which the Texas Medical Association recently joined, alleges that HMOs and insurers conspired and operated together to deprive physicians of prompt and full payment for services. (See " Court of Last Resort," June 2001 Texas Medicine.)

"M&R is one of many examples of methods that this industry uses to usurp the medical doctors' opinion and expertise strictly for purposes of making money," said Archie Lamb, JD, one of the lead attorneys in the case. "What the industry has done through M&R and other similar vehicles is attach like a leech to the medical doctors' and hospitals' income stream. What they do through the guidelines is divert income and deny necessary care to patients."

It will be up to Mr. Lamb and a team of attorneys to develop how insurers have used the guidelines to serve their own purposes as alleged in the suit. But TMA leaders already are concerned enough about that possibility to begin collecting information about denials attributed to utilization guidelines.

 "We're upset that economic considerations, rather than considerations based on care, are being used to create the guidelines," said Robert Gunby, MD, the Dallas gynecologist who chairs TMA's Council on Socioeconomics. He encouraged Texas physicians to provide TMA General Counsel Donald P. Wilcox, JD, "with copies of any written documentation that care was delayed or denied because of adherence to medical treatment guidelines such as M&R."

Cooking with M&R  

Milliman has been producing guidelines governing the efficient treatment of various types of medical conditions since 1989. The guidelines, usually called Health Management Guidelines (HMGs), are licensed to hospitals, insurers, and physicians groups as part of Milliman's consulting business. Copies also can be accessed over the Internet by licensed users. Milliman estimated that more than 100 million Americans have been covered by insurers using the guidelines. Prudential, CIGNA, Aetna US Healthcare, Kaiser Permanente, and some of the Blue Cross and Blue Shield plans have used the guidelines.

Milliman spokesperson Jim Loughman said the guidelines "are prepared by doctors, not actuaries." The firm, which changed its name to Milliman USA on June 1, employs a full-time team of physicians and nurses to research and write the guidelines, Mr. Loughman says.

"There is an evidence basis for the guidelines, which has changed over the years from a consensus basis," Mr. Loughman said. The hierarchy of evidence includes randomized controlled trials, published research, and unpublished research and studies, he says.

"These guidelines are written for doctors by doctors," Mr. Loughman said.

M&R says the guidelines provide descriptions of observed "best practices" and observed hospital lengths of stay. Those practices and lengths of stay are drawn from "efficient health care delivery systems," which Mr. Loughman translates as "a place where everything works without generating unnecessary days in the hospital." The guidelines apply to stays for patients under age 65 who do not have medical complications and who recover "as well as can be expected." (See "What Milliman Says About its Guidelines.")

The HMGs have earned the reputation among some physicians as "cookbooks" for practicing medicine. Controversy has been widespread among physicians who disagree with the guidelines. The most criticized guidelines include:

  • Hospital stays are denied for mastectomies;
  • Cataract removals are limited to one eye, unless the patient is fairly young and needs both eyes for work;
  • Hospital stays are limited to one day for a normal delivery and two days for a cesarean delivery;
  • Length of stay for a stroke is limited to three days, even if the patient cannot walk out of the hospital; and
  • Tonsillectomies can be performed only if the patient has a possible cancer, has six cases of tonsillitis within a year despite treatment with antibiotics, or has blocked breathing during sleep.

According to the testimony of Milliman officials in depositions, the HMGs and the pediatric guidelines, which formally are called Health Status Improvement and Management (HSIM) guidelines, are based on reviews of scientific literature and medical charts, and on consultations with physicians in various fields of practice. And there's the rub in the Texas case: Not all of the pediatricians on the UT-Houston faculty believe that the consultation and writing of the HSIM guidelines met the correct medical standards.

Unappreciated appropriations  

The Texas challenge to Milliman is not a direct challenge to guideline use or the appropriateness of the guidelines.

The lawsuit was originally brought by Dr. Cleary and William J. Riley, MD, of Corpus Christi, a former professor of pediatrics who is now vice president of medical education for Driscoll Children's Hospital.

The two physicians allege that Milliman and a fellow professor who served as chief editor of the HSIM guidelines, Robert Yetman, MD, appropriated their names and reputations for commercial gain. Drs. Cleary and Riley were later joined in the suit by Steven J. Culbert, MD, an associate professor, and Steven B. Wolfe, MD, a former assistant professor.

Each physician says his name was placed on the guidelines even though none of them participated in writing the guidelines. Dr. Cleary had even criticized the guidelines in the Wall Street Journal before they were published.

The physicians say they were not contributors and that the listing of their names "exposes them to ridicule and contempt among the medical and health care community. It impeaches their honesty and integrity and reputation within that community by suggesting that [the four doctors] are placing cost containment above the safety and well-being of their patients," the doctors said in their pleadings.

The plaintiff physicians say their involvement was claimed "solely to gain credibility and physician acceptance within the medical community."

According to testimony gathered for the plaintiffs in depositions by a team of lawyers headed by Ms. Fox, Milliman agreed to pay Dr. Yetman $40,000 for preparing the guidelines as primary author and pay UT-Houston $100,000 for its cooperation. After the lawsuit was filed, the health science center backed off its involvement and asked that its name not be listed as a co-presenter with Milliman in future editions.

Milliman and Dr. Yetman contend the four faculty members failed to protest when the entire faculty was asked if they had problems with the school's participation in the project. They provided input for the project and were not injured when their names were included in the guidelines, Milliman and Dr. Yetman say.

The suit is still in the discovery phase, and it is set for trial on Oct. 1, 2001.

Dr. Yetman is represented by Peter Plotts, JD, an assistant state attorney general. Because of the litigation, Mr. Plotts said he and Dr. Yetman would not discuss the allegations.

Mr. Loughman also said the firm would not comment on the litigation.

Potential for disaster  

Although Dr. Cleary will not comment about the disagreements among department faculty and administration over the guidelines and the appropriateness of the health science center's deal with Milliman, he is convinced that he and the other doctors were right to ask that their names be removed from the HSIM guidelines. Milliman issued a corrected credit sheet, but Ms. Fox says the wording still incorrectly implied that the four doctors participated in writing and reviewing the guidelines.

Dr. Cleary and the other plaintiffs allege that the "goal lengths of stay" contained in the HSIM guidelines are dangerous to sick children. He says he communicated that to Dr. Yetman in no uncertain terms during his reviews.

In his deposition, Dr. Cleary said he told Dr. Yetman that the guidelines were dangerous and that children could die.

To make his point, Dr. Cleary uses the example of the guidelines for treating endocarditis, a disease that results in the inflammation of the heart and its valves. It is usually treated in children with antibiotics given intravenously in the hospital, Dr. Cleary says.

"Even with treatment, the disease is 20 to 40 percent fatal for children," Dr. Cleary said. "The usual treatment is four to six weeks in the hospital, but M&R recommends three days and then the families of the children are supposed to take the antibiotics home and administer them."

Dr. Cleary said Milliman's recommendation doesn't give the treating physician enough time to be sure the right antibiotic is being given and the right course of treatment is being followed.

While the home intravenous treatment may be accomplished by a family member who is a nurse, Dr. Cleary says, few families want the responsibility and the uncertainty of monitoring the treatments.

Milliman admits that its HSIM recommendations are controversial. In one 1998 letter to a Michigan physicians group that was asked to review the HSIM guidelines, M&R principal Jim Turner blamed the controversy on those physicians who "do not 'think outside the box' or who remember only their worst cases rather than their uncomplicated ones."

"When we start talking about sending home babies for home IV therapy to complete a course of antibiotics, people will again stop thinking outside the box and say 'it can't be done,'" the Milliman document said.

Dr. Cleary takes an opposite tack: With the risks of endocarditis so high, even for patients undergoing treatment, is there really an uncomplicated case of endocarditis?

How low can you go?  

So far, Milliman has had little success in calming physicians' fears that medically untrained insurance claims reviewers will use the guidelines simply to deny care.

 The American Academy of Pediatrics has not endorsed the HSIM guidelines and the American Medical Association rejected Milliman's other HMGs as the clinical standard of care in May 2000.

One danger of the guidelines, notes Dr. Cleary, is that insurers understand that the practice of medicine is tied to the purse strings. "They understand that you change the practice by stopping payment to the hospitals for some levels of treatments," he said.

A number of issues are emerging from the five-year debate over the guidelines. Many physicians are now questioning whether Milliman has tried to push too far to cut back on expensive hospital stays and to push "best practices" at the expense of more accepted medical practice.

Former Milliman executive David Axene confirmed in the Texas case that the company estimates 60 percent of the care provided nationally in hospitals could be avoided. As much as 35 percent of ambulatory care may be medically unnecessary, Mr. Axene projected in one document.

Reflecting this assumption, Milliman focuses its guidelines on defining a length of stay for the 10 percent of patients with the shortest hospital stays. Milliman critics say that puts at risk the treatment of the other 90 percent of patients, some of which may be medically necessary but which will fall victim to insurance claims reviewers.

The sun never rises  

One thing that critics find misleading for users of the guidelines is the definition of what constitutes a day. In the parlance of hospital billing, a day occurs when a patient is in the hospital at midnight, whether he or she has been there for 24 hours or was admitted at 11 p.m.      

But Milliman says its guidelines don't reflect the billing and insurance definition. A "day" in Milliman parlance is the point when certain treatment goals have been met. However, while pleading that the definition of "days" in the HMGs and HSIM guidelines don't reflect the conventional or billing definitions, Milliman resists labeling the "days" in the guidelines as treatment stages.

However, the guidelines' author and guru, Richard Doyle, MD, says in his deposition in the Texas case that he gives the standard meaning for billing to the term.

Other concerns were sounded by the AMA Council on Scientific Affairs in a 1996 report, which notes that early Milliman Optimal Recovery Guidelines were developed "by a panel of physician consultants as idealized 'best practices' for uncomplicated cases, and are not based on actuarial data."

The report continued, "The M&R premise is that there should be only limited variation in treatment patterns of resources used and interventions performed on the generic uncomplicated 'best patient.' … The profile of the generic, uncomplicated 'best patient' does not adequately describe the range of individual patients cared for by physicians."

That strikes a chord with Mr. Lamb, the lawyer for the physicians in the racketeering lawsuit. "They find the healthiest patient with a particular diagnosis and use the treatment plan for the healthy patient as their 'guideline,'" Mr. Lamb said. "This is obviously inadequate for sicker patients."

The AMA paper also made the first call for the Milliman guidelines to take into account the stability of the patients who were being assessed for stays in excess of the guidelines. Dr. Cleary also emphasizes this as a shortcoming of the HSIM guidelines. He notes that a child may show short-term measurements that meet the criteria for hospital discharge under the guidelines, but fluctuating measurements over a longer period of time may indicate that a treatment is not working. Under the Milliman guidelines, the patient may have already been discharged when the instability manifests itself.

Actuaries playing God?  

Of all the issues undermining Milliman acceptance, the issue of clinical documentation of the guidelines stands out.

While Milliman disputes the charge that its various guidelines are generated by actuarial data rather than clinical judgments of physicians, the depositions taken in the Texas case have raised questions about how well the guidelines have been documented.

Under questioning by Ms. Fox, Mr. Axene admitted there was not a systematic attempt to document public data to support the length-of-stay guidelines before publication.

Ms. Fox asked Mr. Axene if public data sources and utilization reviews "were only ever used in a retrospective way to look back at the guidelines?"

"I believe that is correct, yes," Mr. Axene answered.

The plaintiffs' attorney also asked Dr. Doyle to explain how he came up with the three-day stay for an endocarditis patient. He said the call would be based "on a series of physicians with whom I had had conversations for a number of years, maybe over a 10-year period."

"Would you have any data to back that up?" Ms. Fox asked

"Probably not," Dr. Doyle answered.

"Would there be anything in writing from those physicians or from any source that would say that three days is an appropriate length of time for endocarditis in our example?" Ms. Fox continued.

"There probably would be nothing in writing," Dr. Doyle said, "but again, to the best of my recollection about the example you've given, there would be content in the guidelines about the fact that continued antibiotic treatment through home health care would be part of the quality treatment."

The lack of documentation to back the guidelines is many a physician's worst nightmare.

"To say that [the HSIM guidelines] should be the standard of care in the absence of clinical studies is unconscionable," Dr. Cleary said. "I think these recommendations do not reflect the medical literature or good medical practice. [The HSIM guidelines] were the result of nothing other than guys in suits telling us how to practice medicine."

Walt Borges can be reached at (880) 880-1300, ext. 1385, or (512) 370-1385; or by email at Walt Borges.  


What Milliman Says About Its Guidelines

"The Milliman USA care guidelines provide descriptions of real patients who are being treated by their doctors and health care systems using 'best practices.' The guidelines contained in Milliman USA's Inpatient and Surgical Care volume include goal length of stay (GLOS) tables. GLOS provides a snapshot of the recovery times possible in the most efficient health care delivery systems. Inpatient and Surgical Care guidelines begin with a description of the training and treatment a patient scheduled to enter a hospital should receive before admission.

"Once the patient is admitted, guidelines help track a patient's day-to-day progress. Care managers are able to identify omissions in treatment and bring them to the attention of the attending physician, as well as plan for the next step in care and prepare for discharge. In this way, the guidelines function much like a safety checklist.

"As part of the health care delivery process, the guidelines also are used to coordinate care with other disciplines, such as ambulatory case management, disease management, provider relations, and quality improvement. Each of these 'best practice' steps is important in making the patient's transition from a hospital setting to another level of care in a safe and effective manner.

"The goal lengths of stay cover the entire spectrum of medical and surgical patients -- regardless of the severity of condition -- so long as the patient does not develop complications. If the patient has an optimal recovery, if an extended stay is not required because of a complication, and if adequate discharge planning has occurred, the patient can usually go home within the goal length of stay and, thus, recover as well as can be expected. Milliman USA insists that the individual circumstances of each patient must be evaluated in the context of the attending health care professional's clinical judgment in determining treatment patterns and goal lengths of stay."

                                    -- Milliman spokesperson Jim Loughman



July 2001 Texas Medicine Contents
Texas Medicine Back Issues