On Christmas Day 1809 in the backwoods of Kentucky Ephraim McDowell, a surgeon who had studied under the famous John Bell from Edinburgh, was summoned to the assistance of a woman who was sick with an enlarged abdomen and severe pain. Mrs. Jane Todd Crawford was told by McDowell that he could attempt surgical removal of this tumor by performing exploratory surgery of the abdomen. It would be an experimental procedure fraught with dangers and a high likelihood that the patient would die. McDowell made this abundantly clear to the patient.
Mrs. Crawford was a highly religious woman and, believing that God would intercede on her behalf, agreed to undergo the procedure. Her faith was to be put to the test. Dr. McDowell would have to perform the surgery with neither anesthesia (this would be introduced to American surgery by William Morton in 1846) nor antisepsis (Joseph Lister introduced his technique for prevention of infections in 1865). While McDowell worked rapidly with skillful hands, Mrs. Crawford distracted herself by singing hymns during the procedure. Legend has it that during the operation a large crowd of the patient's family and friends surrounded the house with the intention to shoot or hang the practitioner if the patient died. Undoubtedly, the personal injury law offices of Jacob Ellis' ancestors were well represented in this malpractice lynch mob.
Happily for Ephraim McDowell, Mrs. Mary Todd Crawford made a miraculously speedy recovery, no doubt with divine help. She survived the procedure for 32 years eventually dying of old age at 78 years old. McDowell also survived the procedure and became known as the "father of abdominal surgery". Because of the enormous surgical risk to the patient and one must presume to some extent the potentially equally severe retribution to the surgeon, the practice of abdominal surgery was slow to become popular.
Hysterectomy, the removal of the uterus, is a surgery of significantly more complexity than removal of an ovary. For this reason it was not until 1843 that the first reported abdominal hysterectomy was performed by Charles Clay, a physician from Manchester, England. This procedure was performed for an enlarged fibroid uterus. Unfortunately the patient, who did well in the immediate post partum period, eventually died on the 15th day following surgery. It was not until 1853 that the first fully successful abdominal hysterectomy was performed, and that actually by accident. Dr. Walter Burnham was performing exploratory surgery for what he believed to be an enlarged ovarian tumor. During the procedure the patient vomited, pushing an enlarged uterus through the abdominal incision. Dr. Burnham was unable to replace the uterus inside the abdominal cavity and had no alternative but to remove it. His patient became the first woman ever to survive abdominal hysterectomy. The doctor, encouraged by this serendipitous success, subsequently attempted a number of other hysterectomies. Of his next 15 patients only 3 survived the surgery. As late as 1880 at the fifth annual meeting of the American Gynecological Society a review of more than 100 hysterectomies performed prior to that time revealed a mortality rate of 64%. It was actually estimated that the true mortality rate was under-reported, however, and that in reality three of every four women undergoing this procedure before that time had died.
What a difference a century makes. By 1875 a little more than 100 hysterectomies had been reported in the world wide medical literature. One hundred years later, in 1975, 775,000 hysterectomies were performed in a single year in the United States alone. Since that time the number of these procedures performed has declined to a current rate of approximately 500,000 per year. As I have said previously about 10% of these are performed to remove cancer. The indications for the remainder are shown by percentage in table 1. A preponderance of modern authors allege that a significant percentage of hysterectomies are performed unnecessarily. Necessity, like beauty, is in the eye of the beholder. In 1809 Mary Todd Crawford, convinced of the necessity to remove her abdominal tumor, undertook the ultimate risk against staggering odds. Her physician, embracing the time honored tradition of mutual trust with his patient, also placed himself at considerable jeopardy to alleviate the suffering of his patient. The unexpected success of McDowell's surgery rescued him from the awaiting mob who would undoubtedly have declared his efforts unnecessary in the event of the patient's demise.
In a perfect world one would hope that the need for surgery would be determined as the result of a sincere discussion between the patient and her physician. The responsibility of the physician in this scenario is to honestly represent the risks, benefits and alternatives to the procedure in a forthright and unbiased manner. The patient would also be obliged to honestly represent the severity and impact of her symptoms without exaggeration. We do not live in a perfect world. Physicians' incomes are dependent upon the number of surgical procedures which they perform. Patients may also have a variety of alternative motivations for seeking surgery.
We have seen that in a little more than a century surgery has progressed from a uniformly life-threatening venture to a point where its risks compares favorably with other elective activities which people choose in their daily lives (motorcycle riding and down-hill skiing to mention only two). Naturally, as the risk of a surgical procedure declines, it is likely that the indications for its use will become more liberal. A procedure which carries a high expectation of fatality can only reasonably be used to treat an immediately life-threatening illness. It is the perception that hysterectomy carries little or no risk to the patient which has enabled it to become popular in the treatment of conditions which are in no way life-threatening, but whose consequences nevertheless negatively impact the patient's quality of life.
The question "Are unnecessary hysterectomies being performed?" is not by any means a recent one. This debate dates back at least to 1946 when Miller published a review of pathologic findings in hysterectomy specimens. In this study and subsequently in others of its type the author reviewed the medical record to see why a given surgery was performed. He compared this "preoperative diagnosis" with the final diagnosis obtained after the pathologist had carefully examined the surgical specimen in his laboratory. Miller discovered that 33% of the uteruses showed no concrete evidence of the condition which the gynecologist believed to be the cause of the patient's symptoms. His conclusion, therefore, was that since pathologic findings were absent in these patients, their hysterectomies were "unnecessary".
Miller's conclusion that one-third of all hysterectomies were unnecessary was followed by similarly structured studies performed between 1953 and 1971 all of which concluded that one- third or more of hysterectomies were not necessary because the final pathologic diagnosis revealed that a "normal" organ had been removed.
It is said that anytime you have two doctors you are guaranteed at least three opinions. In 1977 two studies reviewing the appropriateness of indications for hysterectomies found that the majority of these procedures had been performed for appropriate reasons. In one of these studies, performed by the quality care committee of the New York Medical Society, 3.1% of the indications were regarded as "unacceptable." The other study, published by Emerson, found that only 10 of 1,900 hysterectomy procedures (less than 1%) were "unjustified."
Whether or not a surgical procedure is "necessary" depends on the criteria by which these procedures are judged. Few people would argue that surgery performed to correct or eliminate life- threatening conditions is unnecessary. With regard to hysterectomy life-threatening indications would include cancer of the uterus, cervix or ovaries. It would also be reasonable to include a number of precancerous conditions if they presented a significant risk of future cancer and if there were no reliable alternative therapeutic approach to eliminate them. There is a condition known as endometrial hyperplasia which is sometimes severe enough that gynecologists believe surgical removal of the uterus is justified in order to prevent a future malignant transformation. The severest degrees of dysplasia of the cervix which are capable of progressing to cervical cancer might also qualify for this treatment.
Occasionally, uterine hemorrhage is severe enough to require blood transfusion. Most commonly in older women this is associated with a large fibroid uterus. Obviously, transfusion itself carries significant risks of infection with hepatitis and even HIV. Minimizing the need for transfusion by hysterectomy may, therefore, be regarded as a potentially life saving intervention. In younger women occasionally severe uncontrollable hemorrhage immediately following childbirth results in an emergency hysterectomy.
From time to time a severe pelvic infection occurs which cannot be controlled with antibiotics. This condition, known as pelvic inflammatory disease, has been discussed in detail in an earlier chapter. In its most extreme form, large abscesses can form inside the pelvis causing destruction of the tubes and ovaries. A pelvic infection of this type is in many ways similar to a ruptured appendicitis. Definitive surgical therapy through hysterectomy is considered by most people a reasonable option in the patient who has no further desire for fertility.
We have seen that cancers comprise only about 10% of the diagnoses for hysterectomy. Endometrial hyperplasia accounts for approximately another 6%. The remaining life-threatening indications discussed above each contribute only a fraction of a percent of the nationwide hysterectomy rate. If our yardstick for the necessity of a surgical procedure depends on its ability to preserve or prolong life, we will be obliged to conclude that 80% of hysterectomies are "unnecessary." Of course, if we use this exacting yardstick to determine the necessity of hysterectomy, we must also apply this criteria to all other surgical procedures. It is immediately apparent that in the United States the majority of surgical procedures are elective. For example, hundreds of thousands of arthroscopic procedures are performed each year to correct athletic injuries.
The way in which surgical necessity is currently justified for hysterectomy is the tissue diagnosis approach. This is an institutionalization of Miller's view from the mid 1940's that a hysterectomy is only necessary if the pathology suggested by the preoperative diagnosis is found to be present in the hysterectomy specimen in the pathology laboratory.
The CREST study in 1981 examined how frequently the pathologic diagnosis confirmed the preoperative diagnosis. For hyperplasia of the endometrium and severe dysplasia of the cervix the pathologic findings confirmed the preoperative diagnosis in 95% and 89% of cases respectively. Uterine fibroids were present 84% of the time when they had been predicted. Pelvic inflammatory disease was found 75% of the time. For the more elusive entities causing pain in the pelvis, endometriosis and adenomyosis the combined predictive rate of the preoperative diagnosis was slightly less than 50%. Overall pathologic diagnosis confirmed preoperative diagnosis in approximately 80% of cases. If we consider a hysterectomy to be justified, if we are able to find pathology which confirms the preoperative diagnosis, then only 20% of hysterectomies would be considered "unnecessary".
As we have discussed previously, however, on many occasions benign lesions of the uterus can remain asymptomatic for long periods of time. Simply confirming that fibroids are present in the uterus does not justify the performance of a hysterectomy if the patient's symptoms did not merit such a drastic intervention. Conversely, it is arrogant to assume that our current state of knowledge enables us to define all of the conditions of the uterus which would result in symptoms which justify hysterectomy. I have no doubt that some women experience daily excruciating pain from the uterus when no identifiable pathologic lesion is present. Do we really know all of the causes of uterine pain? If a uterus causes so much pain that a woman is unable to function normally, or if she is unable to participate in a normal sexual relationship because of pelvic pain which arises during this activity, should we tell her that it is all in her head? The complete absence of endometriosis, adenomyosis, pelvic inflammatory disease, uterine fibroids or any other identifiable pathology in the uterus does not guarantee that a woman will be pain free.
To assume that we have already identified all of the pathologic conditions which can give rise to debilitating pelvic pain and dyspareunia is to repeat the mistake of the man who suggested in 1900 that we should close the patent office because everything had already been invented. I do not believe it is unusual for gynecologists to perform surgery on patients who they sincerely believe to have a specific defined pathologic condition only to find after surgery has been completed that the uterus was essentially normal. My experience has been that in the vast majority of these cases the patient's symptoms completely resolved following the surgery. A woman who is freed from the daily aggravation of pelvic pain or who finds herself able to once again participate in a meaningful sexual relationship without fear of painful consequences is usually delighted by the outcome. She is indifferent to the news that the pathologist was unable to identify endometriosis, adenomyosis, or any other pathology. For her the surgery has been an unmitigated success.
Neither the pathologist nor I are able to receive the news of the absence of pathologic findings with the same equanimity demonstrated by the patient. We are both aware that there exists a "tissue committee" whose job it is to review hysterectomy specimens for legitimate pathologic findings. Absence of pathology in a given specimen raises a flag indicating that the surgeon has performed an unnecessary procedure. A surgeon who is found by the tissue committee to have a significant proportion of uterine specimens which lack objective pathologic findings is likely to find himself officially censored by his peers and by the hospital. He will become the object of close scrutiny and will be under constant suspicion of performing unnecessary surgery for personal benefit. The tissue committee does not care about the patient, her symptoms, or her quality of life.
This explains why when you go to your gynecologist because you are having debilitating pelvic pain and ask him to "Just take it out." he is likely to respond with the suggestion that you adopt a stepwise approach. He will, of course, examine you and it is likely that he will find that your uterus is just as tender as you told him it was. Perhaps the examination will be uncomfortable enough to bring tears to your eyes. Surely he cannot doubt your sincerity in this matter. A gynecologist will be likely to suggest two or three possible reasons why you are having the discomfort. He will suggest a pelvic ultrasound. If you have fibroids, a pelvic ultrasound will show them up right away, but endometriosis is more insidious and is not usually visible on any of the currently available radiologic tests. If nothing obvious appears on the ultrasound, your doctor will probably offer to perform a diagnostic laparoscopy. This is an outpatient surgery which is usually performed under general anesthesia.
In diagnostic laparoscopy a fiberoptic device is introduced into the abdomen through a small incision (usually at the umbilicus) enabling the abdominal contents to be viewed on a video screen. One or more additional smaller incisions may also be made through which tiny instruments can be introduced to manipulate the abdominal and pelvic organs. If endometriosis or scar tissue (adhesions) is present, these instruments can be used to remove this tissue using a variety of surgical techniques. Often, diagnostic laparoscopy will reveal no explanation for the patient's pain. In this situation, the incisions are usually closed and the patient is sent home. Occasionally, for unexplained reasons the pain will improve or even resolve completely. Usually, however, the woman will continue to have the same symptoms and her doctor will try to persuade her to undergo a course of "medical therapy." For unexplained pain of presumed gynecologic origin, two types of medications may be prescribed. Birth control pills are often effective in reducing or eliminating pelvic pain. Otherwise, nonsteroidal anti- inflammatory drugs or a combination of these two types of medication may be tried. If a trial of medical therapy is not successful in alleviating the patient's symptoms, then the gynecologist will usually feel comfortable proceeding to hysterectomy, if this is what the patient wishes.
For the patient with the gynecologic symptoms related to her uterus and ovaries who strongly desires a hysterectomy, this stepwise approach is frustrating, to say the least. Many women will have had friends in a similar situation who have undergone laparoscopy or hysteroscopy with dilatation and curettage (D&C) or both and subsequently needed to have a hysterectomy anyway. The patient will not understand why she cannot simply elect to have a hysterectomy, thus preventing the need for multiple minor procedures in this same area. Furthermore, a good proportion of women in their 30's in the United States have already had tubal sterilization to prevent pregnancy. When a woman with a previous tubal receives the suggestion that she use oral contraceptives for relief of her gynecologic symptoms, she is understandably perplexed. She will often say "If I had wanted to take birth control pills, I would not have had my tubes tied!"
The reason for the charade described above is that, before your physician can acquiesce to your request to perform a hysterectomy, he must prove that he has exhausted all other more conservative modalities of treatment. He will not be able to justify the procedure unless he has documented that more conservative methods have failed or he can produce convincing pathologic evidence such as a large fibroid uterus or cancer. The concept that a hysterectomy is unnecessary unless demonstrable surgical pathology is present as introduced by Miller in 1946 has stuck. Insurance companies simply refuse to pay for hysterectomy unless a well recognized pathologic entity is present. Consequently, surgeons are reluctant to perform hysterectomies on women who request them even when they are certain that the woman's symptoms will be alleviated by the procedure. They are fearful that they will submit to the pathologist a uterus which is entirely normal in its appearance and will, therefore, be subject to the scrutiny of the tissue committee.
The fallacy of this "evidence based" approach to the justification of pelvic surgery is that it assumes that all patients with clinically significant symptoms will also have a recognized, identifiable pathological lesion. This is almost certainly not true. Significant pain is actually characteristic of the normal uterus. This is demonstrated by the fact that 75% of women experience pain with their menstrual periods beginning when they are in their early teenage years. This pain, which is debilitating in 15% of women, cannot arise from an abnormality in the uterus in so many people at such a young age. The truth is simply that, even under normal circumstances, the uterus has the capacity to generate severe pain. Painful menstruation is probably the most common reason for school absence in adolescent girls and is also a highly significant cause of work absence in young women. We must conclude, therefore, that the uterus itself has the potential to cause dramatically painful episodes while maintaining an entirely normal appearance. The pain associated with the uterus is probably provoked by relatively minor biochemical and hormonal changes.
But recognizing the uterus as an inherently painful organ should not lead us to dismiss a woman's symptoms of pain simply because there is "nothing wrong" with the uterus. We should rather accept that we do not have an adequate diagnostic, radiologic or pathologic test to confirm the patient's reported symptoms. In effect, by pursuing an evidence based practice which justifies only removal of the uterus with proven pathologic abnormalities, we have divided women into two distinct groups. One group can experience pain from time to time with varying degrees of severity from a uterus which clearly contains some pathological entity such as uterine fibroids. We have seen in an earlier chapter that uterine fibroids and, indeed, all of the "benign" entities which make up the indications for most hysterectomies, are highly variable in the symptoms which they produce. The second group of women may experience pain of varying severity and varying frequency from a uterus which is, to all intents and purposes, normal. Is there a difference between the pain which women experience in these two groups? I repeat my assertion that fibroids are not inherently painful. There are many women whose endometriosis remains asymptomatic for years and even decades. There are also frequent findings of deep and extensive adenomyosis in the uteri of women who underwent hysterectomy for reasons other than pain. In other words, the relationship between pain and the pathologic conditions which are supposed to give rise to that pain is a highly unpredictable one: one might say tenuous. On the one hand the normal uterus may be the source of excruciating and intolerable pain and on the other the distinctly abnormal uterus may remain symptom free. We are aware that the benign conditions discussed are only rarely dangerous to the patient.
Astonishingly, despite the inconsistent relationship between pathologic findings and symptoms, we have used these criteria as a means to determine which woman will receive definitive surgical treatment for her symptoms and which will be denied. A physician is given license to operate on the "abnormal" uterus, but the woman with debilitating symptoms who cannot demonstrate such an abnormality is told that she is normal. She is encouraged to pursue less definitive and often ineffective surgical and medical options. Often, she will eventually arrive at hysterectomy as a definitive treatment, but many times she will have been obligated to undergo several lesser procedures first.
The ultimate test of whether a surgical procedure is necessary or unnecessary in an age where we are concerned with more than just life and death should be whether or not the surgery was successful in eliminating the symptoms for which it was performed. A number of recent studies have focused on just this point. The researchers have looked at the symptoms for which women requested hysterectomy. They followed these women with surveys at various intervals following their procedures to determine whether or not the outcome was satisfactory to the patients. These studies demonstrated overwhelmingly that women were satisfied with the outcomes of their hysterectomy with regard to the relief of their preoperative symptoms.
Perhaps the best known of these studies is the Maine Women's Health Study which was published in 1993. For this study, more than 400 women in the practices of 63 different physicians were interviewed at the time of their hysterectomy and at intervals of three, six and twelve months following the surgery. The questions focused on relief of symptoms, changes in the quality of life, and the development of new symptoms. Of those women who reported pelvic pain prior to surgery, only 5% reported pelvic pain following their hysterectomies. This indicated that hysterectomy was successful in eliminating pelvic pain in 95% of cases. The majority of women in this study stated that they were "delighted" with the outcome of their hysterectomy.
What we have found then is that when the pathologist determines whether or not a hysterectomy is necessary on the basis of the presence of abnormal pathology in the hysterectomy specimen, hysterectomies may be "unnecessary" up to 50% of the time. This arbitrary definition of necessity based on objective pathologic findings which bear little or no relationship to the patient's symptoms has been embraced by the insurance companies because they can use it as a weapon with which to intimidate gynecologists. The medical community seems to have accepted the idea that there is an epidemic of unnecessary hysterectomies despite the fact that no such consensus is present among women. Indeed, women who have had hysterectomies appear, in general, to be quite satisfied with the outcome of this surgery. The most recent objective studies using relief of symptoms and improvement of the quality of life confirm the anecdotal views of most gynecologists that their patients find this procedure very helpful.
Who then is to decide whether hysterectomies are necessary or not? At present, when I schedule a hysterectomy my office staff must preapprove the procedure with the patient's insurance company. Frequently, a clerk from the insurance company will call back asking questions regarding the size of the uterus, the presence or absence of uterine fibroids, and the results of endometrial biopsy and the findings at previous laparoscopic surgery. On the basis of these findings, a doctor working for the insurance company (What kind of a doctor is that?) who has never seen the patient or spoken with her will decide whether or not the procedure is justified. At no time in the evaluation of this patient's surgery does any representative of the insurance company show any interest in the patient's symptoms. They are using the approach that if she does not have fibroids, then her uterus is normal. Her pain is also normal. Let her take some Motrin!
Another interesting point revealed by the Maine Women's Health Study was that there was no correlation between the findings of objective pathology and the relief of symptoms. A woman was just as likely to have complete resolution of her symptoms when no abnormality was detected. Conversely, those women with demonstrable pathology were equally likely to have persistent symptoms as those whose uteri were normal in appearance.
The current policy of performing hysterectomies only for women with demonstrable pathologic lesions is insulting to women. Women are not ignorant victims of the health care system who must be protected from unnecessary surgery. In general, they are making informed choices and should be supported in doing so. Justification for performing surgery should be on the basis of a woman's symptoms as she reports them and on the belief that these symptoms can be alleviated by this procedure on the part of the physician. The physician should make it clear to his patient if there are doubts regarding the success of the procedure for alleviating symptoms. He should try to make a realistic appraisal of the probability of success and communicate this in an unambiguous fashion allowing the patient to make an informed decision regarding her desire to proceed. When either party has significant doubt a second opinion should be obtained from a qualified physician. Women should also be encouraged to discuss their options with other women who have undergone similar procedures and with those who have elected more conservative management. In most communities there is no shortage of qualified women who may act as a tremendous resource to the patient considering surgery.
The process of judging the need for surgery on the basis of pathologic findings should be abandoned! It does not have general applicability and results in inferior treatment for a group of women whose symptoms may be socially debilitating. I believe that the current policies are maintained in order to limit the number of hysterectomies performed. Furthermore, I contest that this limitation is not based on the altruistic motivations of their proponents as they contend, but rather to reduce the payments of insurance companies and improve their bottom line.
At the best, this demonstrates a lack of caring and an insensitivity to women by our current medical establishment, the insurance companies and the government.
But at worst it is an insidious perpetuation of the sexism which has suppressed women for millenia.