Wednesday, August 28, 2013

Discrediting the paper about discredited practices

According to a study in the August 2013 issue of the Mayo Clinic Proceedings, 146 papers appearing in the New England Journal of Medicine over the first decade of this century contradicted medical practices previously thought to be effective.

The paper's findings were widely publicized. There was talk in the New York Times of inertia hindering change and allowing ineffective treatments to continue for years.

The full text of the paper and a supplement containing a brief summary of all 146 discredited practices are available on line.

I decided to see for myself if any practices relating to general surgery were included in the paper and found 11.

Two of them seemed somewhat debatable to me.

Number 43 on the list was a comparison of open mesh to laparoscopic mesh inguinal hernia repair that appeared in NEJM on 4/29/04. This was the critique:

"A laparoscopic approach to repair inguinal hernias with mesh was thought to have lower hernia recurrence rates and less post-operative pain. This multicenter, randomized trial in a VA population found that the laparoscopic approach led to a higher rate of complications and a higher rate of recurrences when repairing primary inguinal hernias."

It definitively closed the door on laparoscopic inguinal hernia repair. Or did it?

In the 4/30/04 issue of NEJM, letters to the editor pointed out that the laparoscopic recurrence rate of 10% in the VA study was much higher than in other reported series, and the size of the mesh (~8.0 cm) used in the laparoscopic cohort was much smaller than the 10 cm x 15 cm that most experts recommended.

How has the VA paper affected surgeons' choice of technique for hernia repair?

A report from the American Journal of Surgery in 2012 found that as of 2008 at the Mayo Clinic, 41% of inguinal herniorrhaphies were performed laparoscopically.

Looking at national resident case logs data for 2012 from the ACGME, 35% of all groin hernia repairs were done laparoscopically.

Despite having been "discredited" in NEJM, laparoscopic inguinal hernia repair is quite alive and well.

The Mayo Clinic Proceedings paper also stated that preoperative biliary drainage for patients with cancer of the head of the pancreas was discredited by another NEJM paper for 1/14/10. Here is what they said about number 131 on their list:

"Jaundice in surgical patients is postulated to increase the rate of postoperative complications. Many surgical centers have employed biliary drainage prior to surgical intervention for cancer of the head of the pancreas, but there is conflicting evidence regarding its effect on morbidity and mortality. This multicenter, randomized trial found that routine preoperative biliary drainage increases the rate of serious complications without a mortality benefit."

Subsequent letters in the 4/8/10 issue criticized the study because patients were drained for 6 weeks prior to surgery which was not the norm of 2 weeks, patients with bilirubin levels above 14.6 gm/dL who were most likely to benefit from preop drainage had been excluded, the wrong type of stent was used and prophylactic antibiotics for ERCP were not uniformly administered.

Is preop biliary drainage still being used?

A randomized trial from South Korea in the July 2013 American Journal of Surgery showed that preoperative biliary drainage for longer than 2 weeks resulted in twice as many complications as drainage for less than 2 weeks, 25.9% vs. 9.1% respectively. This compares favorably to the 74% complication rate of 6 weeks of drainage found in the 2010 NEJM study.

In the July 2013 American Journal of Gastroenterology, a group from Memorial Sloan Kettering Cancer Center published a retrospective review of over 500 pancreaticoduodenectomy patients, 220 of whom had preop stents. The overall complication rates did not differ whether a stent was used or not.

Again despite being "discredited," the use of preoperative biliary drainage continues to be very common.

So what happened here?

The only surgeon among the authors of the 146 discredited practices paper is a third-year general surgery resident. Maybe he did not have enough experience to evaluate these papers and their impact.

Or maybe one should not necessarily base an opinion about whether a practice has been discredited or not on a single paper in one journal.

The findings about these two topics, hernia repair and biliary drainage, lead me to question just how many of the other discredited practices are really no longer indicated or used.


3 comments:

Anonymous said...

It takes more than just one study to refute common practice. On the other hand, multiple studies have legitimately taken down Swan-Ganz's and routine HRT, and saved lives doing so.

Is laporospic inguinal hernia repair worse than open repair? Probably not. Is it better (in the sense of long-term outcome, pain, out-the-door time)?

A big problem with studies involving procedural interventions is that *every* proceduralist considers himself above average, and thus the study results do not apply to him. "So there is no advantage shown, but in my hands, because I used 3 inches instead of 2.6....."




Peter Nigos said...

In your eagerness to discredit Prasad et al, you missed a perceptive editorial in the same issue by Ioannidis, who pointed out several difficulties with the methods used. However, he repeated the assertion that in all areas of clinical medicine, primary care, dermatology and surgery more frequently lack evidence to support their treatments when compared with the current standard of internal medicine interventions.

Skeptical Scalpel said...

Anon, I am not sure that laparoscopic hernia repair is better. It might not be. The point is that the NEJM paper saying it was worse was flawed and has not deter surgeons from performing laparoscopic hernia repairs.

Peter,

I had seen the Ioannidis editorial. I'm a big fan of his. In the interest of space, I did not include his comments.

I agree that surgery often does not have Grade I evidence to support a lot of what we do. That type of evidence is much harder to acquire in surgery. Again, my point in writing this post was that a single paper rarely changes practice and some of the 146 "discredited" practices were not really discredited by a single paper in NEJM.

Post a Comment

Note: Only a member of this blog may post a comment.