Changes in the General Surgical Workload, 1991-1999 FREE

THE GENERAL surgical workload has importance for many groups—students considering general surgery as a profession, deans determining medical school curricula, department heads developing training programs, certifying agencies such as the American Board of Surgery, hospital boards determining privileges, and specialty societies planning the future and direction of continuing medical education.

General surgery has been the main discipline from which every major surgical specialty has derived, with the exception of ophthalmology. At the beginning, the discipline of general surgery was merely identified as "surgery," and perhaps this explains why the American Board of Surgery, which certifies general surgeons, does not include "general" in its name. Until recently, the American College of Surgeons called its advisory council on general surgery the Advisory Council for Surgery, but the name has been amended to the Advisory Council for General Surgery. In the past, multiple subspecialties have split off from general surgery or "surgery," and today there is the possibility of further fragmentation, as surgeons restrict themselves to trauma, vascular, oncologic, laparoscopic, colorectal, or breast surgery.

In addition to competing specialties, other factors that have affected the general surgeon's workload are technological changes such as laparoscopic surgery and percutaneous biopsies, changes in the population, and changing indications for surgery, such as those for carotid endarterectomy.

The purpose of this study was to evaluate recent trends in the general surgical workload, based on Medicare data from 1991 through 1999. Using a high general surgical frequency criterion, selected procedures representing families of procedures were identified to evaluate changes.

The data used in this study were derived from 2 sources. The first source was the National Claims History Procedure Summary Files from June 30, 1991, 1994, 1995, and 1999, published by the Centers for Medicare and Medicaid Services, US Department of Health and Human Services. For each procedure, these files include the total Medicare volume performed by all surgeons and the percentage performed by each surgical specialty. The second source was the American Medical Association's Relative Value Scale Upgrade Committee (RUC) 2001 database, version 4.02, based on the 2000 Physician/Supplier Procedure Summary Master File from the Center for Medicare and Medicaid Services. The RUC meets several times a year to assign or reevaluate the values of procedures. This RUC data set covered 1993 through 1999, but general surgical percentages were only available for 1999. Specialty percentages are based on the specialty designation that the health care provider indicates when making a claim. The designation is not verified to determine whether the surgeon is board certified or has a certificate in vascular surgery; hence, in the case of vascular surgery, a self-reported vascular surgeon could be a certified general surgeon with a certificate in vascular surgery, a certified general surgeon without a vascular certificate, or a surgeon who is not board certified.

The overall general surgical workload was divided into major and minor procedures, using my criteria. One criterion for a major procedure was that it be an operative procedure within a serous cavity, excluding percutaneous procedures such as needle biopsies. Major procedures also included operations on major bones, endocrine glands, and major blood vessels.

Changes in the total volumes of major and minor general surgical procedures were derived from the RUC database. Based on this database, the 1994 and 1999 total Medicare volumes and the 1999 general surgical volumes for major and minor procedures were tabulated and are presented in Table 1. Total Medicare volume for each procedure was the sum of the number of Medicare claims by general surgeons and all other physicians for that code. The total volume for major procedures included the total Medicare volume of all major surgical procedures that met the minimum volume requirement; likewise for the total volume of minor procedures. Only those procedures performed at least a thousand times annually by general surgeons were included. To identify which procedures met this minimum volume requirement, all procedures performed by general surgeons were identified, and the yearly general surgical volume for each of these procedures was calculated by multiplying the total annual volume by the percentage performed by general surgeons. If the general surgical volume was fewer than 1000 cases yearly, then the procedure was excluded. The evaluation and management codes were also excluded.

The changes in general surgical volume for major and minor surgery were evaluated using 2 different data sets. The first, used for major procedures, was composed of "representative" procedures and codes that represented families of codes. These codes and definitions were derived from the Current Procedural Terminology (CPT). For example, partial colectomy (code 44140) was selected as the representative procedure within the family of colon resections. An exception was made for the cholecystectomies—the 2 open procedures were considered as 1 representative procedure and the 2 laparoscopic cholecystectomies were considered as 1 procedure. The data set for major procedures included the total volumes and the general surgical volumes and percentages for the years 1991, 1995, and 1999 (Table 2).

A similar data set was composed for minor procedures, but it differed in that the years studied were 1994 and 1999. In a few instances, the 1994 data for minor procedures were not available, and 1995 data were substituted. A few new procedures had to be excluded because there were no data for 1994 or 1995. This data set was composed of "selected" procedures rather than representative procedures. The reason for using selected procedures was that for many of the large categories of minor procedures, such as skin excisions, it was difficult to find a representative procedure. Thus, within each family, several procedures with high volumes were selected. These selected procedures, with their associated total volumes, general surgical percentages, and general surgical volumes for 1994 and 1999, are shown in Table 3. The major and minor summary tables (Table 4 and Table 5) compile data from the corresponding families and associated subsets. Total volumes and general surgical volumes in these summary tables (Table 2, Table 3, Table 4, and Table 5) are less than the totals shown in Table 1 because the major and minor summary tables are based on representative or selected cases and not the entire family. Based on the representative major data set, a table on laparoscopic procedures was constructed (Table 6). Tables for breast biopsies (Table 7) and endoscopic procedures (Table 8) were constructed based on the minor procedure data set.

According to the Medicare database, 3088 different procedures, each represented by a CPT code, were performed at least once by a general surgeon. Using the RUC database and the criterion that the procedure had been performed at least 1000 times by general surgeons in 1999, the large list was narrowed to 335 minor and major procedures or CPT codes (Table 1). In 1999, surgeons, including all specialists, performed more than 32 million procedures from this list of 335 codes. There were 165 major procedures that met the frequency criterion, and the total surgical volume for all these major procedures in 1999 was more than 1.8 million cases; 170 minor procedures were identified, and the 1999 total surgical volume exceeded 30 million cases. Of the overall total of 32 million major and minor cases, general surgeons performed more than 1.2 million major and 1.9 million minor procedures in 1999.

The total volume of major procedures performed by all surgical specialists, more than 1.8 million cases in 1994, fell by 101 000 cases (−6%) in 1999, based on a summation of all major cases in the 2001 RUC database (Table 1). The representative data from 1991 to 1999 show a gain in the total volume of procedures performed but a 7% loss from 1995 to 1999 (Table 4). This trend is also shown in Figure 1, which is based on the total volumes in the 2001 RUC database.

During the period 1994 to 1999, the total volume of minor procedures increased by 12 million cases (68%) (Table 1). This increase is distorted by the fact that procedures with the CPT code 17003 were included. Code 17003, which represents destruction of each subsequent lesion after destruction of the first skin lesion, was not accepted until 1998. Consequently, the 1994 volume for this code was not available, but by 1999, the volume for this procedure exceeded 10 million. When procedures with the CPT code 17003 are excluded, there was a 19% increase in the total volume of minor surgical procedures (Table 5).

Changes in the volume of major surgical procedures performed by general surgeons from 1991 to 1999 are shown in Table 4. There was a 1.8% decrease in the volume of procedures performed by general surgeons based on data from 1991 to 1999 but a 6% decrease based on data from 1995 through 1999. In either case, the volume of general surgical procedures is declining slightly. The RUC database from 1993 through 1999 could not be used to evaluate changes in the general surgical volume because it did not contain the general surgical percentages for the year 1994.

The changes in volume for various families of procedures are also shown in Table 4. From 1991 to 1999, there were increases in both the total and general surgical volumes for cholecystectomies and appendectomies. There were 4 families—splenectomies, partial colon resections, mastectomies, and inguinal hernia repairs—in which there was a decrease in the general surgical volume, which paralleled the decrease in the total volumes for these families. Because partial colectomy is a small part of the family of colectomies, the entire family was analyzed. For the entire family, the total number of cases decreased by 4053. The change in general surgical volume was calculated using a general percentage rate of 84% noted both in 1991 and 1999 in the data set of representative procedures. Using this percentage, the general surgical volume for the colectomy family decreased by 3530 cases (3%) during the period 1993 to 1999. The total volumes for major lower extremity amputations, lung resections, and vascular procedures increased, but the general surgical volumes fell. The general surgical volume for vascular procedures decreased by 1548 cases (−10%). This occurred despite the fact that the total volume of vascular procedures performed by all specialists increased by more than 28 000 cases. A similar pattern was seen with lung resections and major lower extremity amputations.

General surgical involvement remained between 83% and 93% from 1991 to 1999 for cholecystectomies, appendectomies, splenectomies, colon resections, mastectomies, and hernia repairs (Table 4). These high and consistent percentages suggest that these procedures are the core of general surgery. General surgical percentages for vascular procedures, lung resections, and amputations were lower, in the range of 19% to 63%, and each dropped by about 10 percentage points during this period. Despite the low percentages, general surgeons were the leading specialists for all 3 vascular procedures and the 2 major amputations. Thoracic surgeons were the dominant group for all lung resections.

Changes in laparoscopic procedures are shown in Table 6. These results are somewhat limited by incomplete data, because the CPT panel developed codes for various laparoscopic procedures at different times between 1991 and 1999; until a code was established, no data were collected. The leading minimally invasive procedures were laparoscopic cholecystectomies (n = 155 293), laparoscopic hernia repairs (n = 10 860), and laparoscopic Nissen fundoplasties (n = 6813). Families with the highest percentage of procedures performed laparoscopically were cholecystectomies (79%), Nissen fundoplasty (77%), appendectomy (32%), and thoracoscopic lung resections (17%). The procedures with the greatest increase from 1995 to 1999, as evidenced by the rate of change, were laparoscopic appendectomy (62%) and thoracoscopic lung resections (30%). Laparoscopic Nissen fundoplasties have also increased, but the increase could not be calculated because only data for 2000 were available.

Minor surgical volumes, both total and general surgical, increased from 1994 to 1999 (Table 5). For most families, the increases and decreases in total and general surgical volumes paralleled each other. Exceptions included skin repairs, breast surgery, orthopedic surgery, and gastrostomies, for which the general surgical volume decreased in spite of an increase in the total volume. The explanations for these exceptions are exemplified in Table 7, which shows an increase in the total number of breast biopsies but a decrease in the general surgical volume. Part of the decrease can be attributed to the decreasing general surgical percentage for incisional biopsies. Another factor was that the volume of excisional biopsies, which have a high general surgical percentage, fell dramatically.

Because endoscopic procedures have become a significant portion of the general surgeon's workload, the changes in endoscopic procedures were tabulated (Table 8). Gastroenterologists performed the most procedures for all the families within gastrointestinal endoscopy, except for rigid proctoscopies and anoscopies, for which either general surgeons or proctologists were the leading specialists. Although general surgical involvement was low (8%-28%), because the overall volume of endoscopic procedures was so great, the total general surgical volume was more than 490 000 cases in 1999. From 1994 to 1999, the number of endoscopic procedures performed by general surgeons increased by more than 16 000 (4%), whereas the total number of endoscopic procedures increased by more than 360 000 (9%).

The major general surgical workload, based on both data sets, would appear to be shrinking, despite the fact that the Medicare population increased by 13% during this period. Based on changes in the total number of representative cases, the volume of procedures performed by general surgeons declined by 1.8% from 1991 to 1999 and by 6% from 1995 to 1999 (Table 4). This trend is supported by the decrease in the total number of general surgical procedures performed by all specialists from 1994 to 1999, based on the entire summation of high-volume general surgical cases in the 2001 RUC database. During this time, the total volume dropped by more than 100 000 cases (−6%) (Table 1; Figure 1).

The percentage of general surgical involvement in cholecystectomies, appendectomies, colon resections, hernia repair, and splenectomies remained stable throughout the study period (Table 4). On the other hand, procedures that overlapped with other specialties, such as lung resections, vascular procedures, and major amputations, had diminishing general surgical involvement. Despite this decline, general surgeons were the leading specialists for amputations and vascular procedures. In conclusion, in those areas where there is no subspecialty overlap, general surgical involvement is very high and steady; however, in the subspecialty areas, there has been a reduction in the percentage of procedures performed by general surgeons.

The decrease in the number of major vascular procedures performed by general surgeons, despite an increase in the total number of vascular procedures, relates to the decreasing percentage of specialty procedures performed by general surgeons. The fall in the general surgical volume occurred despite an increase in total carotid endarterectomies of more than 34 000 cases from 1991 to 1999. Because the percentage of procedures performed by general surgeons dropped from 47% to 35%, the increase in general surgical volume for carotid endarterectomies was less than 5000 cases (Table 2).

The decreasing general surgical percentage for major vascular procedures raises many difficult issues. The relationship of vascular surgery to general surgery is an ongoing struggle. Although the American Board of Surgery has continually tried to keep vascular surgery within general surgery, the vascular societies have made attempts to bolt. In many large institutions, there are separate departments, but in smaller hospitals, there is no separation. Another issue is the minimum yearly volume for critical procedures, such as carotid endarterectomies or repairs of aortic aneurysms. Watchdog agencies, hospital boards, and others are recommending that certain vascular procedures not be undertaken unless the surgeon's workload exceeds the minimum. Some hospitals do not have this volume, yet they have a need for a surgeon capable of handling vascular emergencies. There is also the issue of cross-coverage in smaller hospitals, where general surgeons cover for vascular surgeons and vice versa. Under those circumstances, the vascular surgeon must be knowledgeable about general surgery, and the general surgeon must be capable of dealing with vascular emergencies. If smaller hospitals cannot support a vascular surgeon, then general surgical residencies must continue to offer training in basic vascular surgery so that the general surgeon can handle emergencies.

The reduction in general surgical involvement in pulmonary surgery raises many of the same issues. Moderate to small community hospitals without board-certified thoracic surgeons need surgeons to handle thoracic trauma, spontaneous pneumothoraces, and complications of thoracentesis. The debate about lung resections will depend on factors such as whether the volume in smaller hospitals is adequate to maintain proficiency and the desire and ability of smaller hospitals to support thoracic surgery.

The introduction of laparoscopic surgery probably had the biggest effect on general surgery during this period. The first laparoscopic cholecystectomy was performed in the United States in 1989, yet in 1991, more than 75 000 laparoscopic cholecystectomies were performed on Medicare patients. There were 24 000 more combined open and laparoscopic cholecystectomies performed in 1999 compared with 1991 (Table 2). This increase is probably due the fact that laparoscopic cholecystectomy has reduced pain, length of hospital stay, and disability. As a result, there seems to be greater acceptance of the operation by referring physicians and patients. Based on the 1995 and 1999 volumes, the increase in laparoscopic cholecystectomies was only 10 608 cases, suggesting that this growth may not continue. Slightly more than 10 000 laparoscopic inguinal hernia repairs were performed in 1999, but this represented only 10% of primary repairs (Table 6). The volume of laparoscopic fundoplasties was 6813 cases, which represented more than three quarters of the fundoplasties. The percentages of laparoscopic and thoracoscopic procedures for the year 1999 indicate that minimally invasive surgery is also having a significant effect on appendiceal and lung surgery, for which the percentages for usage are 32% and 17%.

From 1995 through 1999, there was a 30% increase in thoracoscopic lung resections and a 62% increase in laparoscopic appendectomies. Laparoscopic cholecystectomies only increased by 7% during this period, but the small increase can be explained by that fact that 79% of all cholecystectomies were performed laparoscopically in 1999. On the other hand, in 1999, laparoscopic colon resections, splenectomies, and inguinal hernia repairs were 10% or less of the combined open and laparoscopic procedures performed for these corresponding representative codes. Two factors may account for the low percentage of laparoscopic colon resections. One is that the learning curve for laparoscopic colon resections is very steep, and the other is the concern about port site recurrences, if the laparoscopic technique is used for cancer resections. Because of this concern, efforts are being made to protect the port sites. Recent reports that patients whose colon cancers are resected laparoscopically have better survival may be an incentive for surgeons to consider greater use of laparoscopic colon resections.¹

The effect of changing surgical indications can be seen within the families of mastectomy and carotid endarterectomy. The recent emphasis on breast conservation is reflected in the reduction of modified radical mastectomies from more than 51 000 in 1991 to about 32 000 in 1999 (Table 2). This was offset by an increase in lumpectomy and node dissection of about 9000 cases. The total volume of major breast resections decreased by 10 000 cases from 1991 to 1999. This decrease in volume occurred while the incidence of invasive cancer in women remained stable or increased slightly.² One explanation for this discrepancy may be that small cancers treated by lumpectomy only are being billed as excisional biopsies rather than lumpectomies. The number of carotid endarterectomies increased from 61 000 to 96 000 during this 8-year period. The indications for this operation have been extended from a symptomatic patient with stenosis to include asymptomatic patients as well. Also during this period, the indications for laparoscopic cholecystectomies have been expanded to include both chronic and acute cholecystitis, whereas, previously, some had recommended that only patients with chronic cholecystitis undergo the procedure.

The effect of a greater number of thoracic and vascular surgeons is evidenced by the decreasing percentages of procedures performed by general surgeons in these areas. The effect of colorectal surgeons seems to be minimal at this time because the percentage of colon resections performed by general surgeons was stable at 84%. With an increasing number of colorectal surgery trainees, there will be more competition not only for the complicated cases but also for all colon cases. Today, a few general surgeons have restricted their practices to breast procedures or other fields within general surgery and have established themselves as specialists. If these trends continue, it is conceivable that other areas of general surgery will be cleaved off, unless general surgeons can convince the public that, based on their training and results, they are indeed specialists in these areas. This image problem for general surgeons may be in part due to the phrase "general" and the perception by a segment of the public that general surgeons are "generalists" and not specialists.

Changing disease patterns may be affecting the decreasing colectomy totals. The incidence of colon cancer declined slightly from 1991 to 1999.³ The increases in appendectomy may be due in part to the increasing Medicare population, but the increase was twice the growth in the Medicare population, suggesting that appendicitis is increasing among elderly persons.

Minor surgery is a major portion of general surgeons' workloads, based on the fact that in 1999, general surgeons performed more than 1.9 million minor procedures (Table 1). From 1994 through 1999, the number of minor surgical procedures performed by general surgeons increased slightly (Table 5). Buried within these families of codes are several important codes for general surgery because of the volume and the high general surgical specialty percentage. The field of gastrointestinal endoscopy continues to be dominated by gastroenterologists, except for rigid sigmoidoscopies, which are performed more frequently by colorectal and general surgeons (Table 8). The combined total of malignant and benign skin lesion excisions (selected cases) exceeded 700 000 cases in 1994, but, in 1999, the combined total declined by more than 100 000 cases (Table 5). Dermatologists were the highest-ranking specialists for most skin excisions. The median percentage of procedures performed by general surgeons was 22% for benign lesion excisions and 18% for malignant lesion excisions. The fact that general surgeons are less involved in malignant lesion excisions may come as a surprise to general surgeons, especially older surgeons, who trained at a time when dermatologists were not performing many excisions. Within the family of breast procedures, there were significant changes in the types and volumes of biopsies performed (Table 7). Although there was an increase in the total number of breast biopsies, the total number performed by general surgeons decreased, presumably due to the fact that more biopsies are being performed under image guidance by radiologists. The CPT codes for image-directed breast biopsies were only accepted in 2000. In 1999, radiologists performed 56% of the open incisional breast biopsies, suggesting that radiologists were using the incisional code for image-directed biopsies. Percutaneous placement of central lines remains a high-volume procedure for general surgeons; general surgeons placed 138 000 central lines in 1999, and there was an increase in the general surgical volume of nearly 4000 cases. Although anesthesiologists are the leading specialist for this procedure, general surgeons rank second for central lines. The other high-volume procedure for general surgeons is placement of venous access devices with reservoirs. In 1991, fewer than 1000 of these cases were recorded. By 1999, the general surgical volume had reached nearly 100 000.

As for growth, extrapolating from the current study, it might be predicted that the total volume of major general-surgical–related procedures will remain about the same or decline and that the general surgical volume will decline. This lack of growth has occurred at a time when the Medicare population has grown by 13%. Two of the biggest winners for general surgeons between 1991 and 1999 were carotid endarterectomies and laparoscopic cholecystectomies, but from 1995 to 1999, there was a slight decrease in the number of endarterectomies and only a slight increase for laparoscopic cholecystectomies, suggesting that the volume for each has leveled. There was an increase in appendectomies among elderly persons, perhaps due to a changing disease pattern. During this time, although the number of partial mastectomies with node dissection doubled, there was also a marked reduction in the volume of modified radical mastectomies. The net result was a decrease in the total number of major breast surgeries. The decline in colon surgery certainly may reflect changes in patterns of both benign and malignant colon diseases.

The minor surgical workload has increased slightly. As for the future of minor general surgical volume, implantation of venous access ports, insertion of central lines, and endoscopies will likely continue to increase, and the trend toward greater use of image-directed breast biopsies and the reduction of excisional biopsies will probably continue. These changes have reduced and will continue to reduce the surgeon's role in the treatment of breast cancer.

Laparoscopic surgery, because it is an alternative to open surgery, might have been predicted not to increase the total volume of procedures, but there has been an increase in the combined total of both open and laparoscopic cholecystectomies. Perhaps this is due to changing disease patterns or greater acceptance of laparoscopic cholecystectomies by patients and referring physicians. Based on the data from 1995 to 1999, it appears that the number of laparoscopic cholecystectomies performed will continue to increase slowly. As for other laparoscopic procedures, the RUC data are incomplete, but comparing the number of laparoscopic procedures with the total volumes shows that Nissen fundoplasty, thoracoscopy, and laparoscopic appendectomy have seen significant acceptance, but laparoscopic colon resections, hernia repairs, and splenectomies only account for about 10% of the total volumes. As more surgeons become trained in advanced laparoscopic surgery, laparoscopic procedures may increase. Another factor that may increase laparoscopic colon resections is the report that patients who undergo laparoscopic colon resections have had better survival.

Because of the increasing numbers of vascular surgeons, thoracic surgeons, and gastroenterologists, there will be more competition, and it can be anticipated that there will be a reduction in the general surgical volume in these areas. Despite these projected declines, the general surgical volumes in these areas are still high, and surgeons working at community hospitals still perform many of these subspecialty procedures; surgeons who locate in smaller hospitals will be expected to have expertise in these areas.

In the future, there will be technological and other changes. Surgeons were slow to accept flexible endoscopy but quick to accept laparoscopic cholecystectomies. If general surgery wishes to preserve its turf, individual surgeons and surgical leaders must be alert to new innovations and react quickly. If reports that laparoscopic colon resection is superior to open colon resection are confirmed, general surgeons will need to become trained in advanced laparoscopic surgery or face further reductions in their workload.

Corresponding author and reprints: Charles P. Shoemaker, Jr, 96 Washington St, Newport, RI 02840 (e-mail: drcpshoe@ids.net).

Accepted for publication June 6, 2002.

This study was presented at the 83rd Annual Meeting of the New England Surgical Society, Dixville Notch, NH, September 27, 2001, and is published after peer review and revision. The discussions that follow this article are based on the originally submitted manuscript and not the revised manuscript.

I thank Jan Martin, MS, who abstracted the 1991, 1995, and 1999 data from the National Claims History Procedure Summary Files, and Patrick Gallagher, MBA, and Todd Klemp, MBA, who provided the 2001 RUC data.

Nick Perencevich, MD, Concord, NH: When looking at future trends here, would it make sense to anticipate that 8 years from now the very first baby boomers will hit 65, and we may see demographically a lot more people, and would that make sense in planning, correlating the trends here?

David E. Clark, MD, Portland, Me: In your data, how is a general surgeon defined? Specifically, does that include oncologic surgeons, trauma surgeons, etc, as possible subspecialists, or not?

John C. Russell, MD, New Britain, Conn: Just a question in terms of the appendectomy. You suggested, Dr Shoemaker, that maybe the disease process is somehow changing, but I wonder if you looked at it from the perspective of the percentage of laparoscopic cases. A number of papers in the past several years have come out that have shown despite increased use of laparoscopy and imaging studies, such as computed tomography, our accuracy rates and diagnoses rates of perforation, etc, really have not changed very much over the past 10 to 12 years. I am wondering if really what you aren't seeing is because we now perhaps have a lower threshold of doing laparoscopy rather than waiting, and when we do laparoscopy, we usually will take out the appendix, whether that mightn't be the explanation for that and probably is associated with a higher rate of negative laparoscopic appendectomies.

Dr Shoemaker: As for your first question regarding the expansion of the Medicare population, my sense is that decreases in some of these operations like hernia repairs or splenectomies will be offset by increases in the Medicare population.

The definition of general surgeons—this is based on Medicare data. It is based on what you put in for billing. If your office puts in that you are a vascular surgeon, it goes in as a vascular surgeon, even though you might consider yourself a general surgeon or a thoracic surgeon, it is so-called self-designated. It has nothing to do with boards, and there is no verification of it. As for subspecialists, there is a small group of oncologic surgeons that appears in a few areas like breast surgery; they may be 1% of the specialists, but it is a very small percentage who identify themselves as oncologic surgeons. As far as endocrine, again it is a small number. When you look overall, the general surgeon is the big player, and that percentage I don't think changes too much from year to year.

Appendectomy—you might be right. Again, this is Medicare data. It is based on your billing. It has nothing to do with what the pathology report looks like, so your opinion is as good as mine as to whether this represented a lot of normal appendices coming out, but it is one piece of data that does not quite fit with all the others, and maybe there is a change. Perhaps something different is happening more than just people lowering their threshold and doing more appendectomies.