|
| |
Charting the Death of Eleanor Roosevelt
From: Columbia University
| By:
Barron Lerner |
EDITOR'S INTRODUCTION |
The death of Eleanor Roosevelt has long cast a shadow over Columbia-Presbyterian Medical Center. Did the physicians there misdiagnose her illness? Was her death hastened by poor care? Using Roosevelt's own patient chart, Barron Lerner, the Angelica Berrie Gold Foundation Associate Professor of Medicine and Public Health at the Columbia College of Physicians and Surgeons, unraveled the mystery behind Roosevelt's final illness. |
 | |
| Memorial commemorating Eleanor Roosevelt. ITN video archive. | |
 first heard the story about Eleanor Roosevelt's death when I was a medical student at the Columbia College of Physicians and Surgeons. "Did you know that Eleanor Roosevelt died at Columbia-Presbyterian?" someone asked me. "The doctors missed the diagnosis. They thought she had lupus but she really had miliary tuberculosis." Over the years, as a medical student and then an attending physician at Columbia-Presbyterian (now New York-Presbyterian Hospital), I heard other versions of the story. |
Similar tales circulated elsewhere. For example, James Rahal Jr., a New York infectious-diseases specialist training in Boston at the time of Roosevelt's death, in 1962, recalls hearing (and then spreading) "a profound rumor that she had died of miliary tuberculosis misdiagnosed as leukemia." Some even used the term "Hyde Park fever" to connote the mysterious nature of Roosevelt's illness. One common thread among all these stories was that a terrible error had occurred and the diagnosis of tuberculosis had been missed. Somehow, the "First Lady of the World" had died of a curable disease at one of the country's most prestigious medical institutions. |
Imagine my curiosity, then, when I learned that Eleanor Roosevelt's medical record was available to the public, sitting in a box at the Franklin D. Roosevelt Library in Hyde Park, New York. I learned about the chart in a most offhand manner. In December 1999, I was on the phone with Raymond Teichman, the head archivist at the library, discussing one of my research projects in medical history. Hearing that I was from Columbia and a historian of medicine, Teichman asked if I knew that the library housed a photocopy of Eleanor Roosevelt's medical record. It had been sealed for 25 years after her death, he explained, but was now available for research if I was interested. |
Solving medical mysteries
I most certainly was interested. Most of my acquaintances don't understand why I love to travel to often remote archives and painstakingly review old, dusty papers, but it is the excitement of solving a mystery that often inspires historians. Here was a potentially solvable medical mystery that had cast a long shadow over my own institution. Remarkably, no one had examined Roosevelt's medical record until now. |
In 1965, three years after her death, Eleanor Roosevelt's family and Columbia-Presbyterian agreed to seal her medical record for 25 years. In 1990, due to its "historical importance," it was made available to researchers interested in the "medical aspects of great people." |
 | |
| Franklin D. Roosevelt and Eleanor Roosevelt in 1941. | |
Medical records are not commonly found in archives. Confidentiality between a physician and patient--even if that patient is famous--is the cornerstone of medical practice, enabling honest discussion of intensely personal issues. If used properly and after an appropriate interval, however, previously confidential medical information about public figures allows researchers to study the relationship between illness and history. For example, recent scholarship on Woodrow Wilson and Franklin D. Roosevelt demonstrated how their physicians concealed medical information relevant to the ability of these presidents to run the country. |
My discovery of Eleanor Roosevelt's chart came at a propitious time for another reason. At the end of 1999, the Institute of Medicine, a branch of the National Academy of Sciences, issued a stunning and scathing report estimating that as many as 98,000 patients die each year from medical mistakes. Such a figure, if accurate, makes errors the fourth leading cause of death in the United States. Had Eleanor Roosevelt been one of those cases? Only her medical record could reveal the answers to these questions. I sat down to open a piece of history. |
Early diagnosis: aplastic anemia
Eleanor Roosevelt's chart sat in a larger box that had formerly been sealed. The box contained a photocopy of her Columbia-Presbyterian medical record, papers explaining the decision to make the record available and various pieces of correspondence from the family members regarding her illness and death. I turned first to a single-spaced, eight-page case summary that had been prepared by one of her doctors, George Hyman, for a conference after her death. From this document emerged the details of how Eleanor Roosevelt first became ill. |
 | |
| Eleanor Roosevelt in 1961. | |
In April 1960, at age 75, she was found to have anemia, an insufficient number of red blood cells. Until this time, Roosevelt had maintained a remarkably vigorous schedule, traveling around the world to publicize the plight of the disadvantaged. To learn the cause of the anemia, her physician and close friend David Gurewitsch referred her to a hematologist for a bone marrow aspiration. This test, which involves the removal of cells from the pelvic bone, revealed that Roosevelt had aplastic anemia. Her bone marrow, responsible for producing red blood cells, was simply not making enough. |
Despite her anemia, Roosevelt continued to travel, remarking that she was "too busy to be sick." But in September 1961 her anemia worsened, and she required two blood transfusions. While receiving the blood, Roosevelt suffered the first of many transfusion reactions, experiencing a high fever and chills. By April 1962 she had not only worsening anemia but also an inadequate number of the two other products of bone marrow, white blood cells and platelets. White blood cells help fight infection and platelets enable the blood to clot. |
Given this development, Gurewitsch and Hyman, who was a hematologist on the staff of Columbia-Presbyterian, decided to treat the aplastic anemia and low platelet count. They prescribed prednisone, a synthetic version of a corticosteroid produced by the adrenal glands. Prednisone was known to stimulate the bone marrow to produce more cells. But it also had a downside: it made patients less able to combat infections. |
Despite prednisone therapy, Roosevelt continued to require periodic blood transfusions during the summer of 1962. When a post-transfusion fever reached 102 degrees, on August 3, she was admitted to Columbia-Presbyterian. At this point, I was able to supplement my review of Hyman's summary with the actual medical record. |
Roosevelt's record looked remarkably similar to the Columbia-Presbyterian charts in which I began to write at the start of my medical career in the early 1980s, down to the abbreviations and code phrases. Most telling were the daily "progress notes," in which doctors recorded how Roosevelt felt, her physical examination, the results of tests and the doctors' tentative conclusions regarding her condition. |
Investigation of tuberculosis
Upon admission, Roosevelt told her doctors, including Gurewitsch, Hyman and an internist named Alfred Gellhorn, that she had been experiencing a cough for several weeks. Given the combination of fever and a cough, they considered the diagnosis of tuberculosis. Once known as the "white plague," tuberculosis had been the leading cause of death in New York in 1900. Yet rates of the disease had declined considerably by 1962, due in part to the introduction of the first antimicrobial agents to treat the disease, in the late 1940s. |
Tuberculosis typically caused an infection in the lung that showed up as a localized density on a chest X-ray. But Roosevelt's X-ray was clear, suggesting that no active tuberculosis was present. There were old scars, however, indicating past exposure to the infection, perhaps when she had been told of "pleurisy" in 1919. A that time, presumably, Roosevelt's immune system had been able to contain the infection, keeping the tuberculosis bacteria safely dormant in her lungs. When her fever declined after five days, Roosevelt was discharged from Columbia-Presbyterian to continue prednisone therapy to treat the anemia. |
Fever of unexplained origin
The next note in the chart was dated September 26, 1962, six weeks later. On that day, Roosevelt was readmitted to the hospital. Her condition had continued to worsen after her August discharge, having persistent fever and now passing blood in her stool. At this point, Roosevelt clearly had more than just aplastic anemia. Doctors termed her case a "fever of unexplained origin," or FUO, characterized by a temperature of at least 101 degrees lasting for three or more weeks. But FUOs provoke anxiety among doctors. What if you can never discover the cause of the fever? And what if your undiagnosed patient happens to be a national treasure? |
Since one of the possible causes of an FUO is tuberculosis, Roosevelt's doctors once again entertained this possibility. But now they were focusing on a rarer form of the disease, so-called miliary tuberculosis, which occurs when tuberculosis bacteria spread through the bloodstream to the lungs, bone marrow and other organs. Because Roosevelt had evidence of an old tuberculosis infection, it was possible that her debilitated condition, plus the side effects of the prednisone, had caused the disease to reactivate and spread. The diagnosis of miliary tuberculosis was often made by chest X-ray, which typically showed tiny, discrete nodules the size of millet seeds--hence the name miliary. In Roosevelt's case, the X-ray showed an "ill-defined nodularity," but not one that looked like miliary tuberculosis. |
To pursue the diagnosis of tuberculosis, doctors performed another bone marrow aspiration on September 27, this time to look for the characteristic bacteria of the disease. When they found none, they sent the bone marrow specimen to a laboratory to be placed on a special culture plate. If Eleanor Roosevelt had tuberculosis, colonies would grow on the plate. But this process would take four to six weeks. |
Treatment for tuberculosis
What was to be done in the interim? Here Roosevelt's chart both provides and withholds information. After the aspiration, her physicians did decide to treat her with two antimicrobials, isoniazid and streptomycin, to fight tuberculosis, on the chance that the disease was present. Indeed, the record shows, aside from a five-day period during which the streptomycin was stopped due to a question of a drug allergy, Roosevelt received two-drug treatment for her tuberculosis for the rest of her life. |
 | |
| Eleanor Roosevelt wearing her wedding dress, New York City, January 20, 1905. | |
Yet the process that led Roosevelt's physician to reach this decision remains somewhat opaque. Gurewitsch later claimed that he had pushed the hardest for this treatment, stemming in part from a conversation he had had about the case with a visiting South African hematologist named Moses Suzman. Some of Gurewitsch's attention to tuberculosis may also have resulted from his own bout with the disease in 1947, when he was hospitalized in a sanitarium in Switzerland. Although the chart cannot confirm or refute Gurewitsch's account, it does reveal another important fact. One individual who remained skeptical about the diagnosis of tuberculosis was J. Burns Amberson, the renowned lung specialist from New York's Bellevue Hospital, who formally consulted on Roosevelt on October 2, 1962. Although Amberson agreed that the treatment should continue, he believed the case was not characteristic of miliary tuberculosis. |
Although the fever briefly waned after the medication was begun, it returned within five days. By October 12, it again reached 105 degrees. How was Roosevelt dealing with her continued illness? Since the patient's voice is generally absent from medical records, information about how the patient is feeling may be found in the nurses' notes. Roosevelt, these notes reveal, was increasingly unhappy with being hospitalized. "Patient very miserable with temperature rising," wrote a nurse on October 10. |
Last days: returning home
Prior to her August admission, Roosevelt had made Gurewitsch promise that she would not die in the hospital. Her children, notably her daughter Anna, agreed, and on October 18 she was discharged. "Mrs. Roosevelt is determined to go home," staff physician Randolph Bailey wrote in her chart, "and has a right to make her own decision." The Columbia physicians had been unable to make a definitive diagnosis. Her doctors agreed that her prognosis was poor, but planned to follow her closely at home. |
It is here that the narrative of Roosevelt's illness, as told by her medical record, ends. To learn what happened after her discharge, I once again turned to the summary prepared by Hyman. Roosevelt's biographer and friend Joseph Lash has also described the former First Lady's last days. |
Roosevelt's illness remained a mystery until October 26, one week after her discharge. On that day, the bone marrow culture grew tuberculosis organisms, suggesting that the old dormant infection had indeed reactivated and spread throughout her body. No one was more encouraged at the news than Gurewitsch, who estimated that her chances for survival had increased by "5,000 percent." Although Roosevelt had trusted Gurewitsch to ease her death, now, perhaps because he had suspected tuberculosis from the start, he pushed the hardest for continued treatment. The family was less enthusiastic. The "prolonged suffering," the family told Lash, "was exactly the way our mother did not want to go." This discussion became moot on November 4, 1962, when Roosevelt suffered an apparent stroke that left her comatose. She died on November 7. She was 78 years old. An autopsy was performed the next day. |
Postmortem
From the family's perspective, the Columbia physicians--while not curing Roosevelt--had done an excellent job. In a letter accompanying the medical record, James Halsted, Anna's husband and himself a physician, wrote, "I have personally never seen anyone receive better medical care in an extraordinarily complicated and exceedingly grave illness." But Gurewitsch had misgivings, arguing that the diagnosis of miliary tuberculosis should have been made earlier, at least at the time of the August admission. "It was left for me to make that diagnosis 40 days ago," he had told Lash the day before Roosevelt's death. "Others should have made it. The dirty linen will come out." |
On December 12, 1962, members of the Columbia-Presbyterian community attended a clinical pathological conference (CPC) at which Hyman presented the case. Physicians hold CPCs, which involve a discussion of a patient's medical course followed by autopsy findings, to learn from their experiences and examine treatment decisions. |
Based on Hyman's summary and the extensive autopsy report, which accompanies Roosevelt's record, I discovered just how widespread Roosevelt's tuberculosis had been. Columbia's pathologists had found a remarkable amount of tuberculosis bacteria throughout her body, including her lungs, liver, kidneys and brain. The characteristic collections of cells and proteins--known as granulomas--that most patients produce to fight off the infection were almost entirely absent. The final diagnosis of the autopsy was "disseminated tuberculosis acutissima," an extremely severe and rarely seen form of miliary tuberculosis. Doctors hypothesized that Roosevelt had been unable to effectively combat her disease in part due to the prednisone treatment. |
Resistant tuberculosis
But the story does not end there. At the CPC, Hyman also revealed an unexpected finding. The Columbia laboratory had discovered that Eleanor Roosevelt's strain of tuberculosis was resistant to the two drugs she had received. That is, neither the isoniazid nor the streptomycin appeared to have been effective in treating her infection. I was able to confirm this result because someone had wisely appended a copy of the resistance testing to Roosevelt's chart. |
 | |
| Eleanor Roosevelt's funeral: John F. Kennedy, Harry S. Truman, Lyndon B. Johnson and Dwight D. Eisenhower in Hyde Park, November 10, 1962. | |
But how could she have developed a resistant organism? Shortly after drugs had been introduced to treat tuberculosis, in the 1940s, physicians learned that bacteria could develop resistance to them. This situation most often occurred when a patient took medication erratically. But persons could also develop drug-resistant tuberculosis through contact with other individuals who had active, drug-resistant disease. In Roosevelt's case, therefore, she had most likely not undergone a reactivation of old tuberculosis but had probably come into contact with someone with active, drug-resistant disease. On prednisone, which made her susceptible, she had become reinfected with an untreatable form of tuberculosis. No one was more relieved at this finding than Gurewitsch, telling Lash that "an enormous sense of relaxation" had come over him. "Nothing could have been done to save her." |
A redeemed reputation
If the specter of Eleanor Roosevelt's death has indeed haunted Columbia-Presbyterian Medical Center--and its successor, New York-Presbyterian Hospital--for almost 40 years, does the medical record finally allow us to wash Gurewitsch's dirty linen? In one sense, his misgivings about the case had some validity. Roosevelt almost certainly had disseminated tuberculosis when she entered Columbia-Presbyterian with fever and cough in August 1962. Her physician might also have more aggressively pursued the diagnosis during her final admission by performing a bone marrow biopsy, which involves removing actual tissue, as opposed to an aspiration. |
But, in practice, making such a diagnosis remained difficult. Reports in the literature suggested that physicians diagnosed miliary tuberculosis prior to death in only 25 percent of cases. Physicians attending a CPC at Yale-New Haven Hospital in 1962 had been unable to identify miliary tuberculosis in a patient on prednisone. The severe form of the disease seen at Roosevelt's autopsy was even harder to diagnose during life because no granulomas were produced. Moreover, as the medical record revealed, Roosevelt's physicians had treated for tuberculosis even as they pursued dozens of other possible causes of her FUO. And, as Gurewitsch suggested, the fact that the bacteria had turned out to be drug-resistant did indicate that her disease was almost certainly untreatable at that time. |
The reality of medical errors
Like my colleagues at New York-Presbyterian Hospital, I welcome the increased scrutiny into the problem of medical errors. The Institute of Medicine report has identified a series of problems that urgently need to be addressed, ranging from improper dispensing of drugs to surgery being performed on the incorrect body part. But as this important investigation proceeds, Eleanor Roosevelt's medical record demonstrates an important point about more subtle "mistakes." Caring for gravely ill patients is never quite as straightforward as it may appear in retrospect. Even the best medical care may not follow a linear progression; false starts and continual reassessments are inherent in the process. Rather than revealing some type of fatal error, Roosevelt's medical record is mostly a story illustrating the complexities that characterize the diagnosis and treatment of disease. |
My review of the record raised another crucial question: Why, if the autopsy showed drug-resistant tuberculosis, has the perception so long persisted that Columbia's medical staff could have saved Eleanor Roosevelt? |
Whether in the 1960s or today, Americans too often view death as a failure. If someone dies, especially of a disease that is often treatable, we assume that a mistake must have been made. |
But Roosevelt herself had realized the fallacy of this argument. Even if her tuberculosis had responded to treatment, she was seriously ill with progressive aplastic anemia and intestinal bleeding and would likely never returned to the active, independent life she had always led. Indeed, when Gurewitsch had excitably told Roosevelt on October 27 that she had potentially curable tuberculosis, she told him, "I want to die." |
The case of Eleanor Roosevelt reminds us that unfortunate outcomes are not always preventable. Some patients die because of medical mistakes, but other patients just die. |
|
| |
