Adventures of a Female Medical Detective

by Mary Guinan

  As of 2008, there was only one documented case in the United States of a healthcare worker transmitting HIV infection to patients: that of a dentist to patients during invasive dental work (7).

  I was the first expert witness to testify. After my first day on the stand, the defense attorney told the judge that he predicted he would need my testimony for another two or three days. The judge refused, stating that this estimate was excessive, and said he would only allow for one more day of my testimony. The defense attorney asked if I had reviewed John Doe’s medical record provided to all expert witnesses in advance by the defense. I had. He asked me what I thought of it. The record contained only two laboratory reports and no physician’s notes. One report contained the results of a blood screen performed at WCMC for new employees, and the results were normal. The other was a test for CD4 lymphocytes, white blood cells that are part of the immune system and that are the particular cells attacked and destroyed by HIV. Because CD4 counts decrease as untreated HIV infection progresses, they are used to follow the status of the immune system in patients with HIV infection. Noticeably absent from the record was an HIV test. The defense attorney asked if I remembered what Doe’s CD4 count was. I replied that I did not. He expressed incredulity.

  “Why wouldn’t you remember?” he asked. I explained that no standardized, Food and Drug Administration−licensed laboratory test for CD4 counts existed and that the laboratory report in Doe’s record had no indication what laboratory performed the test or what the normal values for that laboratory were. As a result, I could not interpret the test results. The attorney discontinued this line of questioning. Most of his subsequent questions were about the published reports of the risk for HIV infection in the healthcare setting.

  On my second day on the stand, the lead counsel said to me, “I give up. Why are you wearing that uniform?” My uniform must have irked him, giving credibility to the idea that the uniform had some positive effect for the plaintiff’s side.

  Before lunch on that day, the defense lawyer stated that published reports showed considerable differences in the estimated risk of HIV transmission in healthcare settings. He referred to the results in two different publications, and I asked to see them. I explained to him that the data from one study were presented in decimals and the other in percentages, but that they were essentially equivalent risks. At lunch, my team colleagues, Ms. Neslund and Dr. Jones, who had been in the courtroom, commented that my explanation about the differences in the two studies was technically correct, but they weren’t sure that my answer was clear enough for the judge to understand. They suggested that, if the issue came up again, I address the judge with my answer and not the lead counsel. (We believed that the lawyers knew the studies gave the same result but were trying to create doubt in the judge’s mind.) The discussions over the publications continued in the courtroom in the afternoon, and this time I addressed the judge, pointing out that the risks found were the same in both studies, only expressed differently. The judge then asked me several questions, and it was clear he understood. It was great to have my colleagues in the courtroom to help me clarify that important issue.

  The last series of questions addressed to me were about how sure I was that a pharmacist with HIV infection could not indirectly infect a patient by inadvertently contaminating parenteral medications with his blood.

  “Could this never happen?” the defense attorney asked. I answered with the adage I had learned in medical school, “Never say never in medicine.” But I pointed out that the likelihood that a pharmacist would prick himself with a needle and not know or see the blood was extremely small, and the likelihood that HIV could live in parenteral solutions was unknown but also very small. (To survive, HIV needs living cells, which parenteral solutions do not have.) I answered that the risk of transmission was extremely small but incalculable. The lawyer asked whether, even with this small risk, the hospital shouldn’t be 100 percent sure, to protect the patients, and thus deny the pharmacist the job.

  Mr. Astrue, general counsel for HHS, objected to the continued “what-if” questions, and to the constructed scenarios of possible routes of transmission. The judge agreed, and my testimony was soon over. I would have loved to be present at the testimony of the other two expert witnesses, but expert witnesses are only permitted in court for their own testimony.

  Judge Kessel decided the case for HHS and determined that federal funding for WCMC should be terminated. He stated that he based his decision on the testimony of HHS expert witnesses because they are “not only medical experts but are charged with public health responsibilities.” The area in which the HHS witnesses and defense witness disagreed was on the public health implications of the evidence. Both Drs. Guinan and Henderson, he said, testified that the risk that Doe might infect a patient with HIV through his preparation of parenteral products was so small as to be unquantifiable. The opposing witness, Dr. Mansell, had agreed that the risk was small but that no matter how small the risk Doe should not be allowed to prepare parenteral products.

  Judge Kessel stated, “I find these [HHS] experts’ conclusions to be credible and to be buttressed strongly by the evidence which they relied on. The evidence establishes that there has never been a single episode documented of a pharmacist transmitting either HIV or hepatitis B through a contaminated parenteral product” (1).

  Without CDC’s surveillance studies, there would have been insufficient evidence to decide the case. These studies of the risk of HIV transmission in the workplace were started in 1983, two years after the first reported AIDS cases (8), when there was no CDC AIDS budget. In an example of how the public health system worked despite the deplorable lack of HIV funding, the studies were conducted by 10 CDC investigators and cooperating investigators from 335 institutions nationwide, who followed over 1,600 healthcare workers who had needlesticks that were contaminated with blood from HIV/AIDS patients (9). These studies provided information on how HIV is transmitted and how exposures occurred. These data were used to improve healthcare workers’ safety by providing infection-control guidelines and eventually led to the redesign of sharp instruments. These improvements reduced blood exposure, which is the single greatest risk of infection for healthcare workers. So the surveillance data also led to improved prevention of occupationally acquired HIV infection.

  After the decision was announced, the New York Times reported on the case. “This is an important precedent. It is the first civil rights enforcement action based on HIV discrimination by any federal agency,” Astrue was quoted as saying. The decision meant that “the federal government will not tolerate discriminatory activity for the disabled in general, and specifically for individuals with HIV” (10). WCMC appealed to the HHS Civil Rights Reviewing Authority, which reviewed the case and affirmed Judge Kessel’s decision in September 1992 (2).

  On January 11, 1993, HHS announced in a press release that WCMC had agreed to “comply with Section 504 of the Rehabilitation Act of 1973 and all terms of a Compliance Agreement. The medical center will offer Doe an unrestricted full time position as a hospital pharmacist with retroactive seniority to Dec 1986 and will provide him with back pay and a lump sum payment for other employee benefits” (11). Additionally, the report stated, “the Medical Center will adopt a statement of non-discrimination in which it explicitly encourages persons with AIDS and HIV infection to participate as fully as possible in all of the medical center’s programs and activities.” How wonderful!

  A spokesperson for WCMC said it was somewhat disconcerting for the hospital to “be painted by some people as having some sort of phobia about HIV. We are one of the few AIDS management centers designated by the NY State Health Department and the only one in Westchester County. We are very sensitive to the needs of AIDS patients. To intimate anything less is an injustice (10).”

  EPILOGUE

  I recently learned from the narrative in the judge’s decision and the civil rights authority review of the decision that, before applying for the positi
on, Doe had been under the care of a WCMC physician at the infectious diseases clinic. Doe’s personal physician had tested him for HIV, and the positive results were in Doe’s medical record at WCMC. When Doe arrived for his preemployment physical, a healthcare worker recognized him as a clinic patient and told the examining physician. Doe’s records were illegally retrieved and the positive HIV test results found—an egregious break in confidentiality of patient medical records. The examining physician told Doe he had found Doe’s medical record with the positive HIV test results and believed he had AIDS and would not recommend him for employment. Neither Doe’s personal physician nor his medical record supported the AIDS diagnosis, and indeed Doe did not have a condition that fit the CDC case definition for AIDS. This discrepancy—and the fact that Doe’s records had been obtained illegally—may explain why there was no HIV test result in Doe’s medical record that was provided to expert witnesses. These facts also clarify why the attorney was so focused on the CD4 test results during my testimony. Very low CD4 counts are supportive of an AIDS diagnosis. Fortunately, I had not remembered Doe’s CD4 count. Who knows how the defense obtained these test results.

  At the 1987 hearing on Doe’s complaint, Doe v. WCMC, two WCMC physicians—an oncologist and a pathologist—testified that Doe constituted a risk for transmission of HIV to patients (12). But another physician, Gary Wormser, MD—who was chief of the WCMC infectious diseases clinic, the director of the AIDS unit, and a leading AIDS researcher—testified against WCMC at that hearing (13). Dr. Wormser believed that Doe would not be a threat to patient safety if he were hired. WCMC’s decision to disregard its own expert HIV/AIDS researcher and continue the legal action is an example of the fear and superstition that overruled the clinical expertise and published findings of the time. But such discrimination about HIV and AIDS was prevalent at that time. It is comparable to fears today about Ebola virus in West Africa, where an epidemic is occurring, and also in the United States, where the first Ebola cases occurred in 2014. It is difficult to exaggerate the spread of fear and myths in these chaotic situations.

  I recently contacted Dr. Wormser to discuss John Doe’s case. He and I had worked together on some early AIDS cases in New York, and we also were classmates in medical school. (However, I never knew that he had testified at the original hearing.) He said he remembered how unhappy the hospital was about his testimony. Knowing that this was a considerable understatement, I congratulated him on having the courage to testify. But he survived, and still leads the WCMC infectious diseases unit. He is also a professor of medicine and pharmacology at New York Medical College.

  Although Doe said in a statement that he would accept the pharmacist’s position at WCMC, it is not known whether he did. Doe added that the outcome had been worth the fight (14).

  Thank you, John Doe.

  ELEVEN

  The Milk Industry Challenges CDC over the Source of a Listeriosis Outbreak

  BACK in 1985, I had been appointed associate director for science at CDC. One of my responsibilities in that position was to coordinate responses to challenges to the scientific publications that emanated from CDC. During my first year in that position, we faced a formidable challenge from the milk industry about a listeriosis outbreak that CDC had investigated.

  How could pasteurized milk be the source of a Listeria monocytogenes outbreak? The process of pasteurization, bringing milk to a high temperature for a short period of time, was specifically designed to kill bacterial organisms. But that is what a medical detective—CDC Epidemic Intelligence Service (EIS) officer David Fleming, MD—and his team found in a 1983 outbreak of listeriosis in Massachusetts (1).

  The outbreak included forty-nine patients. All were hospitalized, and fourteen of them died. Seven of the cases were in mother/infant pairs, and forty-two occurred in immunosuppressed adults. This cohort of patients was characteristic of human Listeria infection, which attacks those with a compromised immune system. Listeria organisms also infect cows, causing encephalitis and uterine infection, which can result in an aborted pregnancy.

  Infected cows can shed the bacteria in their milk, but until Dr. Fleming’s study, human Listeria infection had not been shown to be transmitted through pasteurized milk. The results of Fleming’s investigation were published in the New England Journal of Medicine, one of the most prestigious medical journals in the world (1). Yet the outbreak was puzzling. Although the epidemiologic investigation included two case-control studies that linked the infection to drinking a specific brand of pasteurized milk, exactly how the milk got contaminated was unclear. And although many samples were tested, the organism was never found in the suspected brand of milk, so there was no incontrovertible evidence, or “smoking gun.” But by the time the milk was implicated as the source and the milk samples were tested, the epidemic was over, and the milk was unlikely to still be contaminated.

  The milk lobby, which consists of a formidable coalition of giant milk companies and dairies, decided along with their congressional allies to challenge CDC on the study. They essentially created a “dream team” to work against CDC.

  First, they retained Elliot Richardson as one of their lawyers. Richardson had served as the US attorney general in 1973, when President Nixon ordered him to fire the special prosecutor investigating the Watergate scandal. Richardson resigned rather than carry out the president’s orders, and became one of the few who emerged from the Nixon administration as a man of integrity. He also had excellent credentials, having graduated from Harvard University and Harvard Law School, where he was editor and president of the Harvard Law Review. Between college and law school, Richardson served in the US Army infantry during the Normandy Invasion of World War II, receiving numerous decorations, including the Purple Heart. But this wasn’t all. Richardson had also served in four different cabinet-level positions within the US government, including secretary of the Department of Health Education and Welfare (1970–73). And did I mention that he looked like Gregory Peck?

  The medical team was led by a nationally known infectious diseases expert, epidemiologist and Harvard professor of medicine Ed Kass, MD, PhD, and the equally famous Alex Langmuir, MD, who had been CDC’s chief epidemiologist from 1948 to 1970. In 1951, Dr. Langmuir had convinced Congress to fund the EIS to train physicians and other scientists to identify, investigate, and control epidemics throughout the nation. Langmuir had developed the whole concept of infectious disease surveillance, epidemic investigation, and control. Imagine having Langmuir on the side of the milk industry opposing CDC. Quite a lineup!

  And who was at CDC to face this dream team? By then, both Dave Fleming and Art Reingold, MD, the lead and senior authors, respectively, of the New England Journal of Medicine article, had left CDC. Claire Broome, MD, also a senior author and chief of the special pathogens unit that conducted the Listeria investigation, became the lead scientist for the defense. James Mason, MD, DrPH, director of CDC, and I rounded out the CDC team.

  Through a Freedom of Information Act request, the milk industries’ medical team had asked CDC to provide all original data from the case-control studies, including the original forms completed for each case and control and the calculations used for the results in the published study. They then hired a group of expert consultants to review the data and check all calculations.

  In late 1986 and early 1987, the Health and Human Services (HHS) secretary summoned the CDC director to Washington, DC, to discuss the study with representatives of the milk industry. Claire Broome and Walter Dowdle, PhD, deputy director of CDC, attended the meeting at the office of Assistant Secretary of Health Robert Windom, MD.

  Eliot Richardson and his entourage presented to Dr. Windom and the two CDC scientists the results of their team’s review of the study that linked pasteurized milk to the listeriosis outbreak. Richardson stated that errors were found in the coding of data that would invalidate the study and concluded that the “data were obviously cooked.” Richardson then asked Windom to have CDC recall the paper
from the New England Journal of Medicine and implied that the coding errors were deliberate. (If true, it would imply that the scientists were guilty of misconduct in science, a very serious charge.) This was a proverbial bombshell for CDC. Because the details of the errors were not provided, Dowdle and Broome could do little but sit silently through the accusations. On their return to Atlanta, I was briefed on the situation.

  Dr. Broome requested the details of the medical team’s review. She and her staff agreed that two coding errors had occurred: two of the controls in the case-control studies had not answered the question of whether they had drunk milk. Instead of being coded as “not answered,” they had instead been coded as “no.” Broome recalculated the data using the correct coding and found that the results did not change; the odds ratios remained the same. Despite the coding error, drinking pasteurized milk from this particular brand of milk was still statistically significantly related to developing Listeria infection.