Monday, January 12, 2004
PRACTICE MAKES PERFECT IV
Last week Mr. Holt at The Health Care Blog had a post summarizing this article from The Boston Globe about the error rate in pediatric appendectomies, that is the removal of a normal appendix due to the presumed diagnosis of appendicitis. (The abstract, from the January issue of Pediatrics may be found here).
Overall, at least 8.4 percent of appendectomies -- one of the most common pediatric surgical procedures -- are performed on children who turn out to have a normal appendix, researchers at Children's Hospital in Boston found. That's largely because appendicitis can be difficult to diagnose, since the sharp abdominal pain and nausea of an intestinal virus, gynecological problems in girls, and other conditions produce similar symptoms, and no test for the disease is 100 percent accurate.
The risk of misdiagnosis is much lower -- 4.8 percent -- at hospitals that do many pediatric appendectomies, perhaps because doctors are more skilled at interpreting children's symptoms and more likely to confirm the diagnosis with a CT scan or ultrasound. But appendectomies are so common that the Children's Hospital researchers say it would be impractical to perform most appendectomies in a few major medical centers, a practice common with complex surgeries like coronary bypass operations.
Impractical indeed since the 4.8 percent negative appendectomy (NA) rate was only achieved in 13 hospitals of the 2521 that contributed their 1997 appendectomy statistics (accounting for 5.6 percent of procedures studied). These were done in hospitals that had >3 pediatric appendectomies per week. If your local hospital does 2-3 per week, your error rate jumps to 7.6 percent. Only 21 of the 2521 participating hospitals reach this level. These hospitals accounted for 5.7 percent of appendectomies. The great majority of the recorded appendectomies (75.7%) were performed in hospitals (1023 of 2125) doing anywhere from 4 to 8 a month with a NA rate of 8.9-7.9 percent. Most of the hospitals studied (1060 of 2125) performed less than one pediatric appendectomy per month, accounting for 12.9 percent of procedures and a NA rate of 8.8 percent.
Given its' age the data probably have less to do with practice today than first meets the eye because of one important thing: the increasing use of CT scanning as a tool in the exclusion of appendicitis over the past several years. Before the widespread use of CT, patients with equivocal evaluations for appendicitis were observed and underwent serial exams. If no improvement was noted after 12 hours or so, they were often taken to the operating room for appendectomy. With a negative CT a patient may be managed as an outpatient, or if admitted, the threshold for operation is higher.
While some may not agree with the increased use of imagery to evaluate for appendicitis, if the goal is to reduce the NA rate and the financial and medical complications thereof, why not?
But recent research has questioned that tolerance of error, showing that unnecessary appendectomies are both risky and expensive. A University of Washington study found that, among adults, unnecessary appendectomies typically resulted in a six-day hospital stay, while one patient in 40 develops serious complications and 1.5 percent of the patients die.
Dr. David R. Flum, an assistant professor of surgery at the University of Washington in Seattle who led the 2002 study, estimates that Americans paid $741.5 million for unnecessary appendectomies on adults in 1997, the year on which his study focused.
The basis for the above numbers is this paper published in the July 2002 Archives of Surgery. In the second paragraph the authors describe appendicitis as a "straightforward diagnosis by clinicians" and later in the conclusion as "a common diagnosis, it is by no means a simple one to establish." The paper then presents the increased LOS, hospital charges, case-fatality rate, and infectious complication rate of NA. The rates for NA were higher for women and those <5 and >60 years of age. The authors themselves point out that the higher costs and LOS may be due to further evaluation to determine the cause of the abdominal pain, and the case-fatality rate due to other causes (since it was higher in the very young and very old).
With lifetime risk of 12 percent for men and 25 percent for women, the building of a better mousetrap to diagnose appendicitis is always going to be a goal of surgery. My personal experience is that I have performed about 50 appendectomies over the past 2 years and CT has not been wrong yet. My error was taking the young 20'ish man with guarding and rebound on exam to the OR for an appendectomy despite a negative CT. A generous pathologist called it "early" appendicitis, but it looked normal to me. |
Last week Mr. Holt at The Health Care Blog had a post summarizing this article from The Boston Globe about the error rate in pediatric appendectomies, that is the removal of a normal appendix due to the presumed diagnosis of appendicitis. (The abstract, from the January issue of Pediatrics may be found here).
Overall, at least 8.4 percent of appendectomies -- one of the most common pediatric surgical procedures -- are performed on children who turn out to have a normal appendix, researchers at Children's Hospital in Boston found. That's largely because appendicitis can be difficult to diagnose, since the sharp abdominal pain and nausea of an intestinal virus, gynecological problems in girls, and other conditions produce similar symptoms, and no test for the disease is 100 percent accurate.
The risk of misdiagnosis is much lower -- 4.8 percent -- at hospitals that do many pediatric appendectomies, perhaps because doctors are more skilled at interpreting children's symptoms and more likely to confirm the diagnosis with a CT scan or ultrasound. But appendectomies are so common that the Children's Hospital researchers say it would be impractical to perform most appendectomies in a few major medical centers, a practice common with complex surgeries like coronary bypass operations.
Impractical indeed since the 4.8 percent negative appendectomy (NA) rate was only achieved in 13 hospitals of the 2521 that contributed their 1997 appendectomy statistics (accounting for 5.6 percent of procedures studied). These were done in hospitals that had >3 pediatric appendectomies per week. If your local hospital does 2-3 per week, your error rate jumps to 7.6 percent. Only 21 of the 2521 participating hospitals reach this level. These hospitals accounted for 5.7 percent of appendectomies. The great majority of the recorded appendectomies (75.7%) were performed in hospitals (1023 of 2125) doing anywhere from 4 to 8 a month with a NA rate of 8.9-7.9 percent. Most of the hospitals studied (1060 of 2125) performed less than one pediatric appendectomy per month, accounting for 12.9 percent of procedures and a NA rate of 8.8 percent.
Given its' age the data probably have less to do with practice today than first meets the eye because of one important thing: the increasing use of CT scanning as a tool in the exclusion of appendicitis over the past several years. Before the widespread use of CT, patients with equivocal evaluations for appendicitis were observed and underwent serial exams. If no improvement was noted after 12 hours or so, they were often taken to the operating room for appendectomy. With a negative CT a patient may be managed as an outpatient, or if admitted, the threshold for operation is higher.
While some may not agree with the increased use of imagery to evaluate for appendicitis, if the goal is to reduce the NA rate and the financial and medical complications thereof, why not?
But recent research has questioned that tolerance of error, showing that unnecessary appendectomies are both risky and expensive. A University of Washington study found that, among adults, unnecessary appendectomies typically resulted in a six-day hospital stay, while one patient in 40 develops serious complications and 1.5 percent of the patients die.
Dr. David R. Flum, an assistant professor of surgery at the University of Washington in Seattle who led the 2002 study, estimates that Americans paid $741.5 million for unnecessary appendectomies on adults in 1997, the year on which his study focused.
The basis for the above numbers is this paper published in the July 2002 Archives of Surgery. In the second paragraph the authors describe appendicitis as a "straightforward diagnosis by clinicians" and later in the conclusion as "a common diagnosis, it is by no means a simple one to establish." The paper then presents the increased LOS, hospital charges, case-fatality rate, and infectious complication rate of NA. The rates for NA were higher for women and those <5 and >60 years of age. The authors themselves point out that the higher costs and LOS may be due to further evaluation to determine the cause of the abdominal pain, and the case-fatality rate due to other causes (since it was higher in the very young and very old).
With lifetime risk of 12 percent for men and 25 percent for women, the building of a better mousetrap to diagnose appendicitis is always going to be a goal of surgery. My personal experience is that I have performed about 50 appendectomies over the past 2 years and CT has not been wrong yet. My error was taking the young 20'ish man with guarding and rebound on exam to the OR for an appendectomy despite a negative CT. A generous pathologist called it "early" appendicitis, but it looked normal to me. |
