The Cost Effectiveness of Surgical Excision of Endometriosis

posted in: Articles | 0

Unpublished Scientific Article

Cindy M. Mosbrucker MD
David B. Redwine MD
Endometriosis Institute of Oregon
2190 NE Professional Court
Bend, Oregon 97701


Endometriosis is a common disease whose treatment consumes a substantial amount of health care resources. As healthcare spending grows as a proportion of GDP, there is more pressure for cost-containment and wiser use of funds. External sources such as third party payors and government regulators are beginning to dictate treatment regimens. The medical community must critically evaluate itself and learn to be fiscally responsible with treatment strategies in order to ensure our autonomy and preserve the physician-patient relationship.

Several papers have been written on the cost-effectiveness of various treatments of endometriosis. (1-7) It has been proposed that medical therapy should be the first-line treatment because it is the most cost-effective. Most papers on this subject focus primarily on the costs involved, but only when we know the true effectiveness of a therapy can we truly adjudicate its actual cost-effectiveness. This paper will review the efficacy of two main therapies for endometriosis: surgical excision and medical treatment with gonadotropin-releasing hormone analogues, specifically leuprolide acetate. The literature was reviewed to ascertain cure rates, duration of pain relief, and variations of these factors by stage of disease. Our calculations indicate that surgical excision of endometriosis provides a much greater chance for long-term cure than does medical therapy with leuprolide acetate. Even though surgery is more costly in the beginning, due to the longevity of response it becomes more cost-effective than medical therapy when measured over time.


Endometriosis is a very common but certainly not the only cause of chronic pelvic pain in women. The two are often confused both by patients and their physicians. The only way to definitively diagnose most cases of endometriosis is by laparoscopy. Chronic pelvic pain is a common condition affecting between 10 – 15% of reproductive aged women, resulting in some 400,000 laparoscopies per year (1). A large proportion of those undergoing laparoscopy for pain are found to have endometriosis, as judged by either biopsy or the surgeon’s impression of pelvic findings.

Current therapy for women with endometriosis includes various types of surgery as well as treatment with medications, some of which are costly. Many women have multiple surgeries and/or courses of medical therapies, including gonadotropin releasing hormone (GnRH) agonists. Assuming a 10% prevalence rate, annual direct and indirect costs of endometriosis in the US were estimated at 22 billion dollars in 2002. (2) Over the last several years some attention has been given to comparative costs associated with medical or surgical treatment of women with endometriosis, but these studies focus primarily on complex cost analyses from an economist’s point of view rather than critically addressing the efficacies of common therapeutic modalities from a clinical perspective. (2,3) As politicians continue to speak of “universal healthcare” and “aggressive cost-containment” in the same breath, it is not hard to envision a time when physicians are told how to care for their patients by governments and insurance companies, with complete disregard for individual patient differences and physician expertise. The best way to avoid this is for the medical community to learn fiscal responsibility and consider the long-term financial and medical consequences of treatments. By determining true cost-effectiveness of various treatments and following our own findings, we can preempt outside meddling into the practice of medicine, and leave the job of deciding on therapeutic strategies to the patient and her physician.

Chronic pelvic pain is a symptom and endometriosis is one of the histologically identifiable diseases that can be its cause. Yet, these are overlapping conditions that are frequently lumped together improperly in many studies. Women can simultaneously have more than one cause of pelvic pain since endometriosis, uterine adenomyosis, interstitial cystitis, levator myalgia, and irritable bowel syndrome can occur in varying combinations and degrees of severity in the same patient. For this reason, studies not using a surgical diagnosis cannot be considered reliable for information regarding endometriosis specifically. Similarly, studies that do not include a second-look surgery to measure the response of endometriosis to treatment may inaccurately estimate the response to treatment since women with an initial proven diagnosis of endometriosis may be relieved of their endometriosis pain following treatment but have persistent or de novo pain from a different source. Unfortunately, some studies of medical therapy for either chronic pelvic pain or endometriosis do not involve any surgical diagnosis, and most studies of either medical or surgical treatments do not include second-look surgery. This lack of surgical control leaves questions about the true efficacy of treatment of endometriosis. Nonetheless, such studies might be considered as high quality evidence and used in the formation and justification of healthcare policy and practice recommendations. (8,13,20)

In order to create a more realistic model for comparing cost-effectiveness between therapies, one must first determine the true efficacy of a treatment both initially and over time, allowing a calculation of true cost-effectiveness per unit of time. Ensuring that results apply to patients with endometriosis specifically and not other etiologies of pelvic pain is necessary in order to avoid making any therapy directed against endometriosis seem less effective than it really is.

For this reason, this review will use best available evidence to construct a cost-effectiveness study comparing treatment of endometriosis by aggressive laparoscopic excision with medical treatment using leuprolide acetate on an intention-to-treat basis. There is insufficient follow-up data in reoperated patients regarding the response of endometriosis to either laser vaporization or electrocoagulation to include these forms of treatment in this analysis.

Materials and Methods

The medical literature was surveyed to arrive at estimates of surgically proven cure rates, long-term clinical efficacy of both surgical excision and medical therapy with leuprolide, efficacy over time including likelihood of re-treatment, and patient dropout rates.


Cure rates of endometriosis by medical therapy

Cure of a disease is the goal of any clinician, although it is sometimes unattainable.  When used in this paper, the term “cure” of endometriosis is defined in the classical sense: visual and/or biopsy evidence at reoperation that endometriosis is absent.

Medical therapy of endometriosis is historically based on the twin notions that pregnancy and menopause physically eradicate endometriosis by some undetermined cytocidal effect.  No study was ever done to prove these seminal concepts.  Modern thought on endometriosis now accepts that pregnancy and menopause do not eradicate endometriosis but may provide symptomatic improvement due to suppression of the biological activity of the disease by either increased progesterone effect during pregnancy or decreased ovarian estrogen production after menopause.  The initial enthusiasm for the possible cure of endometriosis by any medical treatment was misplaced because of conclusions based on the response of symptoms rather than response of the disease, as well as surgical re-evaluation of patients during therapeutic ovarian suppression when endometriosis might be more visually subtle and therefore more easily missed (16).  Interestingly, it is impossible ever to know the effect of medical therapy on endometriosis because it is impossible to biopsy the same lesion twice, once before and again after medical therapy.

GnRH agonist therapy induces a “pseudo-menopause” by shutting down ovarian hormone production due to a suppressive effect on pituitary FSH production.  Because of the significant decrease in bone mineral density, the length of treatment has been limited to 6 months unless add-back therapy is used.  There are several GnRH agonists, some of which are available only in Europe, and the one most commonly used in the US is leuprolide.  Variations exist regarding the extent to which they achieve ovarian suppression, (17) therefore caution must be used when interpreting studies so as not to mix apples and oranges.

Response of endometriosis to GnRH agonists is unpredictable.  Histologic study shows that suppression of peritoneal endometriosis is variable and incomplete, with one study finding 52% of lesions showing regression, 29% showing progression, and 19% showing both regression and progression in the same biopsy or even within the same gland (18).  Another study of the GnRH agonist buserelin in treating ovarian endometriomas found that at the conclusion of therapy while patients were still hypoestrogenic, 97% of ovarian biopsies were positive for endometriosis and 66% of the patients had histologically active endometriosis (19).  No study has ever shown that any medical therapy destroys any manifestation of endometriosis, including minimal disease.  A histologic cure rate of zero following medical therapy is reasonable to assume and mirrors modern thought on the results of medical therapy.

Long-term clinical efficacy of medical therapy of endometriosis

Waller and Shaw (8) studied 130 women with laparoscopically proven endometriosis treated for 6 months with GnRH agonists and followed them for up to 7 years post treatment. They reported on a questionnaire follow-up of endometriosis patients treated in the United Kingdom with one of three GnRH agonists: buserelin, goserelin or nafarelin.  Their data were incomplete since 68 of their patients (52.3%)  were lost to follow-up. Among the successfully-followed patients, they used life table analysis to estimate a 53% overall clinically suspected recurrence rate by 5 years, with stage-specific recurrence rates of 37%, 67%, 48%, and 74% for revised American Fertility Society stages I, II, III, and IV, respectively.  The results of Waller and Shaw are further clouded by the fact that symptom recurrence was allowed as evidence of recurrent endometriosis in non-reoperated patients, while in reoperated patients biopsies that were negative for endometriosis were allowed as proof of the presence of recurrent endometriosis. The authors commented that the results of treating endometriosis with GnRH agonists was inferior to the results observed following laparoscopic excision of endometriosis of all stages, where Redwine found a 19.6% risk of surgical diagnosis of endometriosis by 5 years after surgery was reported (10).

A  phase III trial of leuprolide was a prospective, randomized double-blinded study assessing safety and efficacy of 6 months of leuprolide vs. placebo (13).  Three of 31 (9.7%) patients in the treatment arm dropped out; two due to persistent pain,  and one from side effects (hot flushes).  There was a significant decrease in dysmenorrhea (not surprising as leuprolide induces amenorrhea), and slight though statistically significant reductions in both pelvic pain and pelvic tenderness on exam.  There was no significant reduction in dyspareunia.  24 of 28 patients in the treatment arm were subsequently followed for 1 year after completion of therapy.  The recurrence rate was found to be 66% for dysmenorrhea and 63% for non-cyclic pain by 1-year post treatment.  When the three patients who dropped out are added back to these rates on an intention-to-treat basis, the failure rate of leuprolide at 1 year rises to almost 80%.

A 1998 study by Miller et al (20) was a retrospective chart review of patients who had been successfully treated for 6 months, 128 with Danazol and 199 with GnRH agonists (buserelin, nafarelin, goserelin, or leuprolide).  All patients had an initial staging laparoscopy and significant pain relief while taking the medications.  Charts were then reviewed to determine the time lapsed between conclusion of therapy and the recurrence of pain requiring intervention.   No mention was given regarding the numbers of dropouts or those lost to follow-up.  The median time to recurrence of pain was 5.2 months in the GnRH agonist group, and 6.1 months in the Danazol group, a statistically significant difference.  The majority of patients in each group had stage 1 or 2 disease, which theoretically should respond best to medical therapy.  An interesting aside of this study was that 26 of the GnRH agonist patients received add-back therapy, and in these patients the average time to pain recurrence was only 2.6 months.  Their final conclusion was that “The rapidity of pain recurrence is disappointing in both groups.”

Fedele, (21) considered the response to a 6-month course of leuprolide in 15 women with rectovaginal endometriosis and moderate to severe pain.  All were treated with leuprolide for 6 months and then followed for 12 months after completion of therapy.  Two women (13%) dropped out prior to completion of therapy due to persistent pain.  Of the 13 who completed the 6 months of treatment, 11 required further therapy prior to 1 year of follow-up.  The intention to treat failure rate of 87% (13 of 15) was similar to the failure rate in the phase III trial (13).  The authors concluded that medical therapy was inappropriate for patients with disease of the rectovaginal septum.

Leuprolide and “add-back” therapy were studied by Hornstein (22) in a randomized, placebo-controlled double-masked trial.  All women were given leuprolide acetate depot  3.75 mg monthly for one year.  Patients were randomly assigned to one of four treatment groups: Group A received placebos for both estrogen and progestins; Group B received norethindrone acetate 5 mg daily and placebo for estrogen, Group C received norethindrone acetate 5 mg and .625 CEE daily; Group D received norethindrone acetate 5mg and 1.25 mg CEE daily.  The initial study reported at the conclusion of one year of treatment, and a follow-up study (23) reported after 24 months post-treatment.  At the end of the treatment year, all groups showed slight reductions in dysmenorrhea, pelvic pain and pelvic tenderness with no significant differences between the groups (although the 1.25 mg CEE group had the least pain reduction).  Over the subsequent 2 years 99/123 patients dropped out of the follow-up, 32 due to increasing pain, 19 due to pregnancy, and the remainder due to personal request or lost to follow-up. Because of the high numbers of dropouts, the intent-to-treat analysis is the most accurate interpretation of their results.  The only significant differences between groups were that group D with higher estrogen add-back had a faster return to baseline scores for dysmenorrhea, pelvic tenderness, and induration.  All patients had return to baseline symptoms between 1 and 12 months (average of 4.5 months) after the conclusion of therapy, for a 100% short-term failure rate.

In summary, it is likely that the histologic cure rate after GnRH agonist therapy is near zero regardless of disease stage.  Regarding long-term symptom reduction, it is difficult to condense the above studies into a meta-analysis due to differences in drugs, protocols, and disease stage, but, conservatively speaking, patients treated with GnRH agonists alone have less than a 40% chance of pain relief at 1 year after treatment for all stages.  There is minimal pain relief in stage IV patients with obliteration of the cul de sac.  When add-back therapy is included in the analysis, medium-term benefit is significantly decreased regardless of disease stage, with symptomatic failure rates approaching 100%.

Cure rates following excision of endometriosis

The pioneering follow-up study of conservative surgical excision at laparotomy was by Wheeler and Malinak (11) in 1987.  They meticulously excised all visible endometriotic implants at laparotomy in 423 patients, then determined long-term outcome by direct office follow-up, reoperation, or contact by researchers.  Life-table analysis was used to analyze their data.  Stage of disease was not specified.   62 patients had recurrence of pain severe enough to have repeat surgery, and of those, only 21 had recurrent or persistent endometriosis at surgery, for a surgically proven cure rate of 66%.  Using the life-table data, by 5 years postop there was a 19% risk of a surgical diagnosis of recurrent or persistent endometriosis.  350 (82.7%) patients were lost to followup during the study.

Four years later, Redwine published a similar follow-up study (10). Using life-table analysis, 359 patients were followed for up to 7 years after aggressive laparoscopic excision of endometriosis.  Eighty-one patients had recurrence of pain resulting in reoperation, and 35 of those women who were reoperated were found to have endometriosis, for a surgically proven cure rate of 57%.  Similar to the study by Wheeler and Malinak, a 19.6% risk was found for a surgical diagnosis of new endometriosis by 5 years after surgery.  The revised American Fertility Society stages of the patients were as follows: 62% stage I, 26% stage II, 8% stage III, and 4% stage IV.  Only 23 patients (6.4%) were lost to follow-up.

More recently, Abbott et al (12) in Australia and the UK compared conservative laparoscopic excision to sham surgery in a prospective randomized fashion.  The immediate surgery group consisted of 39 patients randomized to laparoscopy with excision, while the delayed surgical group underwent diagnostic laparoscopy only.  The patients were followed for 6 months, after which time the groups were crossed over so that the delayed group underwent excisional surgery, and the immediate surgery group had diagnostic laparoscopy with biopsies of any suspicious areas.  80% of the immediate and 32% of the delayed surgery groups reported pain relief 6 months after the first surgery.  Seven of 16 patients in the immediate surgery group who underwent the second laparoscopy had biopsy proven disease, for a surgically proven cure rate of 56%.  In contrast to the Redwine study, these patients had more severe disease, with 2% stage I, 44% stage II, 10% stage III, and 44% stage IV.  Four of 20 patients (20%) were lost to follow-up in the immediate surgery group.  Although this study had a limited followup period, the histologic cure rate observed is remarkably similar to the two previous studies of excisional surgery.

These published cure rates include treatment of all disease stages.  It is highly unlikely that such similar cure rates obtained in different eras and different centers around the world are due to pure coincidence, but rather must be a function of endometriosis itself responding predictably to similar techniques used for excision regardless of an open or laparoscopic approach.  Indeed, the concept of cure of endometriosis by conservative excision is not new.  Meigs, writing in 1953, commented that “recurrence is uncommon and…cure of the lesion by conservative surgery is usual” (24).

Studies of long-term symptom reduction following excision

Vignali (25) studied 115 patients who underwent surgical excision for deeply infiltrating endometriosis, with a mean follow-up of 34 months.  28 patients (24%) had recurrence of pain, and 12 (10%) underwent reoperation.  The only predictor of reoperation was the incompleteness of the initial surgery as judged by the operating surgeon at the time, and the only predictor of pain recurrence was complete obliteration of the cul de sac.

Regarding endometriosis of the cul de sac and rectovaginal septum, Redwine and Wright (26) studied women with complete obliteration of the cul de sac who had undergone laparoscopic en-bloc resection.  84 patients underwent excision of the cul de sac with retention of the uterus and a least one ovary, and 67 were available for follow-up via questionnaire at a mean interval of 4.24 years.  Between 59-78% of those who responded to questionnaires had improvement in dyschezia, low back pain, dysmenorrhea, dyspareunia, menstrual pain other than cramps, and non-menstrual pelvic pain (ordinally).  Only 19% required further surgery.

Busacca et al (27) found recurrent pain in 31 of 141 women (22%) undergoing conservative excision of deeply infiltrating lesions and stripping cystectomies of ovarian endometriomas during a follow-up period of 6-24 months.

With regard to endometriomas, Saleh and Tulandi (28) compared reoperation rates after laparoscopic treatment of ovarian endometriomas by either excision or fenestration and ablation.  Of 70 women treated with fenestration and ablation, and 161 treated with excision, the probability of reoperation at 42 months were 57.8% and 23.6%, respectively.

To summarize the surgical data, conservative excision of endometriosis by either laparotomy or laparoscopy results in a surgically proven cure rate for all stages of over 50% as proven in those patients who required additional surgery.  By 5 years after excision the risk of a surgical diagnosis of new endometriosis is 19%.  These results are consistent even in patients with more severe disease such as endometriomas and cul de sac obliteration.  Unlike medical therapy, excision of endometriosis is not associated with patient dropout, since at the conclusion of surgery, 100% of the patients would have received complete treatment.

Cost-comparison analysis

It is difficult to compare the costs of treating endometriosis by medical treatment with leuprolide versus surgical excision because there are such different assessments of the outcomes.  Most studies of medical therapy use only pain relief as the primary outcome. Surgical studies use histology and visualization of the site of original disease to generate true cure rates, as well as tabulation of symptom recurrences. Our financial assumptions use Winkel’s dollar amounts for costs of both medical and surgical therapy, with the realization that these costs are higher today.  Our focus is on patients with stage I and II disease, since it is unreasonable to use leuprolide for higher stage disease.

The most favorable scenario for empiric therapy with leuprolide is that 60% of patients with stage I and II disease will have recurrence of pain by 12 months after the conclusion of treatment.  Most likely, those with recurrence of pain will require subsequent treatment with either medicine or surgery by 1 year post treatment.  If we assume that half of the failures will choose surgery and half will choose re-treatment with leuprolide, and that the initial cost of leuprolide is $2400, then costs will be as follows: 40% will cost $2400 (one course of leuprolide), 30% will cost $4800 (two courses of leuprolide), and 30% will cost $8200 (one course leuprolide plus one surgery).  This results in an average cost of $4860 for 18 months of benefit (6 months of treatment and 12 months of follow-up) or $270.00 per pain free month.  Even with this favorable short-term scenario, 70% of the patients (those with medical therapy alone) have no reasonable expectation of long-term pain relief, as the histologic cure rate after medical treatment is zero. For this reason, the total costs associated with medical treatment are probably much higher than our conservative calculations.

Surgical therapy is more costly in the short run, with a cost of $5800 for a laparoscopy done on a same-day surgery basis, which could accommodate virtually all patients with stage I or II disease.  Given the studies previously examined, over 50% of patients will need no further treatment because their endometriosis has been cured.  The cure rate may be higher than this in patients with lower stage disease.  There is a reasonable expectation of long-term resolution of endometriosis in about 80% of patients.  If the 20% with persistent symptoms due to recurrent endometriosis require a second surgery, then the surgical cost are as follows:  $5800 for 80% (one surgery), and $11600 for 20% (two surgeries) for an average cost of $6960 for at least 5 years of benefit. For financial calculations, we will assume a risk of complications requiring return to the operating room to be 1%, with an increased cost of an estimate of  $15,000.00 (including an additional surgery as well as an extended inpatient stay) for these patients.  If we add in $150 per patient for complications requiring return to surgery and extended hospital stay (1% x $15000), then average costs increase to $7110, again expecting at least 5 years of benefit.  This translates to $118 per month for surgically induced pain relief. For those patients who have permanent pain relief after surgery, the monthly cost for surgical pain relief drops to pennies per day.


Winkel (1) studied the expenses associated with treatment of chronic pelvic pain and asserted that a 6-month course of leuprolide acetate is more cost-effective than laparoscopic surgery for the treatment of chronic pelvic pain.  The direct costs were estimated at $5805 for laparoscopy, including hospital and surgeon’s fees, compared to $2400 for 6 months of leuprolide including office visits and medication.  There are several unsupported assumptions that are problematic with this analysis.  First, that 6 months of treatment with leuprolide is therapeutically equivalent to any laparoscopic surgery, both treatments resulting in 100% pain relief for 5 years, after which time all pain recurs.  Studies regarding efficacy of pain relief from GnRH agonists were misinterpreted to support the above conclusion; these will be discussed in detail later in this paper.  Second, no contemporary reference from any surgical treatment series was cited; therefore his assumption that all surgical therapies for endometriosis eventually fail equally is unsubstantiated. Data regarding surgical cure rates of 56 – 66% documented among reoperated patients were ignored (10-12).

Other modeling studies of endometriosis cost-effectiveness have also suggested that medical management is more cost-effective than surgery, however they too have their flaws. Surrey (4) suggests that women should be treated empirically with GnRH-agonists and only undergo laparoscopy if they fail to have pain relief after 3-5 months of therapy.  His rationale is that “laparoscopy does not provide a diagnosis in 39% of women with chronic pelvic pain”. Sutton’s study of laser ablation was quoted that found 38% pain relief for those with stage I disease 6 months postoperatively.  This was compared to Waller and Shaw’s study of medical management showing that 63% of those with stage I endometriosis had pain relief 5 years after completing 6 months of GnRH agonists. For patients with stage II disease, however, the results are nearly opposite with 69% and 33% having pain relief after surgery and GnRH agonists, respectively.  There is no mention of how to differentiate stage I from stage II endo without laparoscopy.  Other modeling studies (5,7) have the same problem of leaving out any studies of excision and using Sutton’s laser ablation study as the only surgical data.  Common to most of these papers is the implication that all surgical treatments are equivalent, and that all are flawed because very few gynecologists are able to perform endometriosis surgery well, thus the need for a treatment that can be prescribed by anyone.

Two recent papers from the economist’s viewpoint are also notable.  Gao, et al (3) reviewed studies of direct costs related to endometriosis as well as analyzing national databases to obtain overall costs.  The most concerning aspect of this review is that the top 4 surgical procedures performed for endometriosis in 2002 were hysterectomies (TAH, LAVH, TVH, and SCAH), comprising nearly 90% of all surgeries done for endometriosis.  Abdominal hysterectomies comprised 74% of all hysterectomies, and even for adnexectomies and other laparoscopic procedures, the average LOS was greater than 2 days.  This is not representative of modern surgical practice, and it brings up a point of caution with regards to the interpretation of surgical cost data.  Given the relative miniscule numbers of outpatient laparoscopies in this data, it seems that surgical costs will be grossly overestimated compared to a more ideal situation of 90% laparoscopy, 10% laparotomy with hysterectomy.  Simoens, et al, (2) is also a cost-analysis, but recognizes the lack of clinical data contributing to the efficacy analysis.  They pose the question “whether economic savings can be generated by non-surgical diagnosis and medical treatment that allows patients to be cared for in primary care and that prevents hospitalization and a surgical approach?  The answer is no.”  They go on to say that based on evidence from existing studies “to date, it is unclear whether a medical approach is less expensive than a surgical approach to treating endometriosis in patients presenting with chronic pelvic pain.”  This begs the question of whether a randomized prospective study of excision vs GnRH agonists needs to be done in order to prove both efficacy and cost-effectiveness of the two treatments, or whether there is currently enough data to answer these questions. A related question has to do with the quality of the data currently in the literature, and if cohort studies of surgical excision patients will be accepted by those who reject anything but a “randomized, prospective, double-blinded study”.

It is a fact that health care expenses are at an all time high, and that resources are limited.  It is worthwhile to adjudicate both the most effective and the most cost effective ways to treat common conditions such as endometriosis, however we must keep in mind that the most cost effective treatment is not necessarily the cheapest initially. Response to therapy is often subjective, and there are no standardized treatment protocols or outcome measures, although tabulation of the presence and amount of disease at reoperation is the most accurate measure of successful treatment.  Regardless, it is incumbent upon physicians, both authors and those doing patient care, to accurately interpret the existing data. Those promoting empiric treatment with leuprolide often reference the Waller and Shaw paper, even though leuprolide was not used in the study. Nonetheless this data gets “borrowed” inappropriately to describe results anticipated after treatment of endometriosis with leuprolide.  Despite a high dropout rate (lost to follow-up and dropouts due to side effects) and the acceptance of patients with pain but negative biopsies to have a recurrence, both of which are significant faults of the study, this paper is still used to promote medical therapy.  Many studies of medical therapy do not describe the percentage of patients with significant pain relief or stratify them by stage, but instead group all patients together to state the average amount of pain relief of the entire group, resulting in outcomes that are more subjective than objective.  One study of GnRH agonists accompanied by “add-back” therapy treated patients for up to 2 years with a combination of GnRH agonist, estrogen, and progestin, thus more than quadrupling the costs estimated herein. (29)

Our very conservative analysis of treatment of endometriosis by either leuprolide acetate or laparoscopic excision suggests that the effective monthly cost of medical therapy is at least 233% of the cost of expert laparoscopic excision to achieve symptomatic relief.  Since our medical therapy model covered 18 months of surveillance, while our surgical model covered 60 months of surveillance, it is likely that the relative cost per month of medical therapy is even higher than our extremely conservative estimate.   Reflection on these results leads to a consideration of informed consent.


Surgical therapy of endometriosis by excision remains the gold standard.  Not only are there objective data indicating histologically documented cure rates over 50% but pain relief is long lasting or permanent in the majority of patients, regardless of whether endometriosis was the source of all of their original pain.  Surgical management has several other significant benefits over medical management, including the ability to accurately diagnose the cause of pelvic pain and rule out malignancies and other more dangerous causes of pain, as well as the avoidance of long-term side effects of leuprolide (menopausal symptoms, mood alteration, bone loss, etc.)  Surgery is more appropriate for those women attempting fertility.  Winkel’s final statement is “the efficacy of surgery is largely based on the skill of the surgeon.  Medical treatment becomes the great equalizer…”  Some might worry that the gynecological profession is populated by poor surgeons. The fact that a skilled and well-trained surgeon is required to successfully excise endometriosis should not influence the discussion of cost-effectiveness but speaks to the need for better gynecological surgical training.  Excision of endometriosis is the most difficult gynecological surgery, especially in higher stages of disease and with multi-organ system involvement. Additionally, endometriosis is more prevalent than all gynecological malignancies combined.  For these reasons, serious consideration should be given to the establishment of a subspecialty for surgical treatment of endometriosis and to the establishment of centers of excellence for surgical treatment of this disease.


Endometriosis is a common disease, affecting 10-15% of reproductive aged women, and encompasses a large portion of most gynecologic practices.  Healthcare is becoming increasingly expensive, and in order to ensure that resources are not prematurely exhausted, it is necessary and prudent to determine the most cost-effective ways to treat common disease states.  The facts in the literature can be summarized by the following statements.  Medical therapy with GnRH agonists is reasonably effective for pain relief during therapy for early stage endometriosis.  It is not effective for long-term pain relief, nor does it provide a histologic cure.  Medical therapy is inappropriate for higher-stage and invasive disease. Conservative surgical excision is the gold standard of endometriosis therapy, providing the best chance of a long-term cure while preserving good reproductive function.

It is true that one surgery is more expensive than one course of leuprolide, yet surgery is the most cost-effective way to treat endometriosis due to the duration of symptom relief and the likelihood of long-term cure.  Our calculations indicate that surgical therapy is less than half the price of medical therapy when measured as cost per month of pain relief.  Gynecologic surgeons should strive to achieve results close to those published in the literature, and patients should be counseled on the above results and given true and scientifically based informed consent regarding treatment options.


  1. Winkel CA. Modeling of medical and surgical treatment costs of chronic pelvic pain: New paradigms for making clinical decisions. Am J Managed Care 1999;5:S276-90.
  2. Simoens S, Hummelshoj L, D’Hooghe T. Endometriosis: cost estimates and methodological perspective.  Hum Reprod Update 2007;13:395-404.
  3. Gao X, Outley J, Botteman M et al.  Economic burden of endometriosis.  Fertil Steril 2006;86:1561-72.
  4. Surrey E. An economically rational method of managing early-stage endometriosis.  Med Interface 1997;10:119-124.
  5. Glazer M. The clinical and economic benefits of GnRH agonist in treating endometriosis. Am J Managed Care 1999;5:S316-S325.
  6. Heinrichs WL, Henzl MR. Human issues and medical economics of endometriosis. Three- vs. six-month GnRH-agonist therapy.  J Reprod Med 1998;43:299-308.
  7. Kephart W. Evaluation of Lovelace Health Systems chronic pelvic pain protocol.  Am J Managed Care 1999;5:S309-S315.
  8. Waller KG, Shaw RW. Gonadotropin-releasing hormone analogues for the treatment of endometriosis: long-term follow-up. Fertil Steril 1993;59:511-5.
  9. American Fertility Society. Revised American Fertility Society classification of endometriosis: 1985. Fertil Steril 1985;43:351
  10. Redwine DB. Conservative laparoscopic excision of endometriosis by sharp dissection: life table analysis of reoperation and persistent or recurrent disease. Fertil Steril 1991;56:628-34.
  11. Wheeler JM, Malinak LR. Recurrent Endometriosis. Contr Gynecol Obstet; Vol 16, pp13-21.
  12. Abbott J, Hawe J, Hunter D, Holmes M, Finn P, Garry R. Laparoscopic excision of endometriosis: a randomized, placebo-controlled trial.  Fertil Steril 2004;82:878-884.
  13. Dlugi AM, Miller JD, Knittle J, Lupron Study Group. Lupron* depot (leuprolide acetate for depot suspension) in the treatment of endometriosis: a randomized, placebo-controlled, double blind study. Fertil Steril 1990;54:419 – 27.
  14. Ling FW, for the Pelvic Pain Study Group. Randomized controlled trial of depot leuprolide in patients with chronic pelvic pain and clinically suspected endometriosis.  Obstet Gynecol 1999;93:51-8.
  15. American College of Obstetricians and Gynecologists. Medical management of endometriosis. ACOG Practice Bulletin no. 11. Washington DC: ACOG, 1999.
  16. Evers J. The second-look laparoscopy for evaluation of the result of medical treatment of endometriosis should not be performed during ovarian suppression. Fertil Steril. 1987 Mar;47(3):502-4.
  17. Schroder AK, Diedrich K, Ludwig M. Medical management of endometriosis: A    systematic review.  IDrugs 2004 7(5):451-463.
  18. Ochs H, Cirkel U, Schweppe K-W, Schneider HPG. Residualendometriose nach LHRH-Analogtherapie. Geburtsh u Frauenheilk. 1990;50:140-3. (English translation of abstract).
  19. Nisolle-Pochet M, Casanas-Roux F, Donnez J. Histologic study of ovarian endometriosis after hormonal therapy. Fertil Steril 1988;49:423-6.
  20. Miller JD, Shaw RW, Casper RFJ, Rock JA, Thomas EJ, Dmowski WP, et al. Historical prospective cohort study of the recurrence of pain after discontinuation of treatment with danazol or a gonadotropin-releasing hormone agonist.  Fertil Steril 1998;70:293-6.
  21. Fedele L, Bianchi S, Zanconato G, Tozzi L, Raffaelli R. Gonadotropin-releasing hormone agonist treatment for endometriosis of the rectovaginal septum.  Am J Obstet Gynecol 2000;183:1462-7
  22. Hornstein MD, Surrey ES, Weisberg GW, Casino LA, for the Lupron add-back study group. Leuprolide acetate depot and hormonal add-back in endometriosis.  Obstet Gynecol 1998;91:16-24
  23. Surrey ES, Hornstein MD, for the add-back study group. Prolonged GnRH Agonist and Add-Back Therapy for Symptomatic Endometriosis:  Long-term Follow-up.  Obstet Gynecol 2002;99:709-19.
  24. Meigs JV. Endometriosis. Etiologic role of marriage age and parity. Obstet Gynecol 1953;2:46-53.
  25. Vignali M, Bianchi S, Candiani M, Spadaccini G, Oggioni G, Busacca M. Surgical treatment of deep endometriosis and risk of recurrence. J Minimally Invasive Gynecology 2005;12;508-13.
  26. Redwine DB, Wright JT. Laparoscopic treatment of complete obliteration of the cul-de-sac associated with endometriosis: long-term follow-up of en bloc resection.  Fertil Steril 2001; 76:358-65.
  27. Busacca M, Bianchi S, Agnoli B, Candiani M, Calia C, DeMarinis S, et al. Follow-up of laparoscopic treatment of Stage III-IV endometriosis. J Am Assoc Gynecol Laparosc 1999;6:55-8.
  28. Saleh A, Tulandi T. Reoperation after laparoscopic treatment of ovarian endometriomas by excision and by fenestration. Fertil Steril 1999; 72:322-4.
  29. Pierce SJ, Gazvani MR, Farquharson RG. Long-term use of gonadotropin-releasing hormone analogs and hormone replacement therapy in the management of endometriosis: a randomized trial with a 6-year follow-up.  Fertil Steril 2000;74:964-8.