Patient-reported rates of chronic pain and recurrence after groin hernia repair

[William Bryant]

Alephy, can you explain for simpletons such as myself what this means?

“I still believe that if the hernia numbers were not so big we wouldn’t be in this messy substandard situation….”

Is it that there are so many hernia repairs needed worldwide?

[Alephy]

@Watchful indeed this is what made me angry when I found out that most doctors, including hernia specialists, will tell you false or inaccurate information!
Btw I don’t think this site was created with a focus on high volume surgeons and their patients, and many patients do not have the option to go to the best specialists either…it is ironic that part of the mesh argument was to make the surgery simpler therefore making the procedure also less problematic. I still believe that if the hernia numbers were not so big we wouldn’t be in this messy substandard situation….

[Watchful]

This 15% figure for chronic pain after groin hernia surgery is consistent with what studies have shown in other countries as well. It’s not a Sweden-specific issue. It would be reasonable to expect some surgeons to get better results than others, but this is the overall picture, which is not good, and not something that’s commonly known to patients. There’s an incorrect perception that hernia is a minor routine surgery with excellent results.

[pinto]

Nothing wrong with being Swedish if anyone would need reminding. Focus on this Swedish study inevitably puts a spotlight on a nation’s health care system. Although Sweden’s known as world class, one world ranking has it at number 23 (World Population Review 2023). That same index puts South Korea at no. 1. If so, it explains South Korea’s attracting a great deal of medical tourism, something apparent by the large “medical corner” at Seoul’s main international airport Incheon, and understandably the popularity of Dr. Kang for hernia repair. But I digress. This foray about Swedish health care system, while noting the unexpected less-than-stellar ranking, Sweden by the turn of this century had a national reduction of hospital beds of 50% across all medical sectors (Hamberger, 1998), the likely source for the hospital overcrowding in Sweden noted by The Local Europe AB (2018). Tibor (2018) calls the Swedish system in chaos due to funding and doctor availability.

Although Swedish citizens by law have a right for primary care, I wonder for elective surgery such as hernia, how much citizens really get to choose their own surgeons and even the surgical method. Implications for patients surely arise in the case of chronic pain. It reminds us we can’t take national research reports in isolation but rather ought to to place them in their proper circumstances. How much of the chronic pain rate grows out of a cultural milieu. For example some cultures can be distinguished by their members’ tolerance for pain; some others, less so. What about the stresses imposed by systems in “chaos”? Things to ponder about …

• This reply was modified 11 months ago by  pinto.

[MarkT]

@Mark T, are you Swedish btw, is that why you’re arguing: “This study is therefore quite relevant to hernia patients in Sweden” (as you stated in a related thread).

Why would you ask if I am Swedish? I said it is quite relevant to hernia patients there, because the sample covers almost 98% of repairs in their country…while you are suggesting the study is “biased” and “irrelevant” (your words).

They plainly state that nearly half of the 22,000 cases had surgery performed by a surgeon with a load less than 26 a year. That should raise a red flag. Apparently the researchers assume surgeon load is a minor factor as they provide no other information for the upper range of surgeon load. That raises an alarm because the database has all that info.

Why should that raise a ‘red flag’? They are responsibly disclosing specific and relevant details about the surgeons. They clearly do not assume it is a minor factor, because they included surgeon volume as a variable.

How comical that you only now criticize them for not providing the upper load (which is what *I* noted and your response was “The fact that the median surgeon load is 25 means that the range is zero to 50!”. . At least you understand what the median is now, I guess? But now you are just parroting something that I said!

“They describe pure tissue repair cases as outside of the framework of this study apparently because as they say “there were very few recorded in the register.” “Very few” of course is vague. Why not state the number? It actually could be nearly 2,000 cases based on their account how the cases came to total 22,000.

We can deduce that it is nowhere near 2,000 tissue repair cases considering three of the five repair groups had fewer cases than that (Open anterior: 18,034; TEP: 2,688; TAPP: 380; CAP: 1022; and OPPM: 793). You would know this if you actually read the study. Instead, you strangely surmise they might be biased against tissue repairs (?).

The upshot is that the elephant in the room is ignored by not considering the effects of surgeon experience.

They did consider this. Re-read my posts. I included some concerns/questions about how it was considered though.

I don’t know what else to say…you simply don’t understand the methodology and are not taking the study within its proper context.

[pinto]

@Mark T, are you Swedish btw, is that why you’re arguing: “This study is therefore quite relevant to hernia patients in Sweden” (as you stated in a related thread).

The issue has nothing to do about relevancy within or for Sweden. Frankly I couldn’t care less. My point has been the researchers generally were vague by not providing information more fully and fail to pursue an important avenue: the relation between surgeon experience and outcome. They plainly state that nearly half of the 22,000 cases had surgery performed by a surgeon with a load less than 26 a year. That should raise a red flag. Apparently the researchers assume surgeon load is a minor factor as they provide no other information for the upper range of surgeon load. That raises an alarm because the database has all that info.

They describe pure tissue repair cases as outside of the framework of this study apparently because as they say “there were very few recorded in the register.” “Very few” of course is vague. Why not state the number? It actually could be nearly 2,000 cases based on their account how the cases came to total 22,000. Whether 500, 1,000, 1,500 tissue repair cases, outcome data at least would be of some use comparatively speaking; not definitively of course but possibly suggestive. Being vague about this invites thinking the researchers are biased against tissue repair. One bias can lead to others. Being cavalier with some aspects or details can suggest a lack of uniform consistency in method.

The upshot is that the elephant in the room is ignored by not considering the effects of surgeon experience. Well and good they only want to present the amount of chronic pains complaints by patients. However no reflection is made as to the type of hospitals or medical services involved or surgeon experience possibly involved. Given surgeon loads of less than 26 so widespread, I must think that most of the surgeries were at general hospitals—-not at independent hernia specialists. I grant that some general hospitals may have some amazing surgical wizards, my own direct experience with such hospitals, however, tells me not in the main. Even specialist hernia surgeons are doing other surgeries preventing them doing hernias exclusively.

Mark T, I hope this post satisfies your curiosity, for the trail is getting stale. From your concern that the study has value in Sweden or misreading/misquoting my “narrow” descriptor or falsely claiming I have disparaged the Swedish databank (Not. My concern is how used, presented), I believe it is plainly seen that the 15% chronic pain rate or whatever is compromised for the reasons stated. Thank you for your interest.

[pinto]

You misread what I wrote despite multiple mentions. I was not referring to the database per se.

As far as your 98% figure, apparently you do not know but it is an estimate made nearly 25 years ago (coming from a 1998 publication!). Such shows how the researchers you are defending have under-reported information, which entails yourself as well.

Let me point out also the database, though national, is purely voluntary and privately run (though having some national support). Its member medical units number approximately 90, which is only about 10% of the medical clinics nationally in Sweden. Thus we must caution the notion that nearly every hernia repair has been recorded. More importantly, we need to know whatever the percentage given what it actually covers.

[pinto]

@Mark T, again if you don’t like the term bias as applied then the study is narrow for based on presumably general hospitals.

@Good intentions, you will need to elaborate because I don’t see any relation to my choice of Dr. Kang as surgeon nor to “the very common mesh repair methods.” Thank for such interest in my many posts here at HT but apparently you overlooked my reviews for Dr. Kang, which include my rationale. Thank you again for your kind interest.

[Good intentions]

Pinto, if you could supply the information sources that you used to choose a pure tissue repair from Dr. Kang instead of one of the very common mesh repair methods it would help you make your points.

I looked back through your posts but can’t find an explanation of your rationale. What information led you to a pure tissue repair from Dr. Kang?

[pinto]

Mark T, I beg to differ with you variously: First contrary what you think—-we do know the distribution. The fact that the median surgeon load is 25 means that the range is zero to 50! From the perspective of HT all of the cases involved low-volume surgeons!

From that standpoint the study is quite biased in the sense the data is skewed by presumably based solely on general hospitals. (Because the surgeon load was so low, we can assume these cases came from general hospitals; thus the data skewed. Typically hernia surgeons at general hospitals even if “specialized” must also perform a variety of abdominal surgeries.) If you disprefer “biased” then the study is narrow by only including general hospitals ergo low-volume surgeons.

Relevancy? It all depends on purpose. This thread concerns rates of chronic pain and recurrence. Thus the study having excluded highly experienced surgeons (presumably), it hardly can be relevant for patients truely seeking hernia specialists, surely the purpose of most HT members. Usefulness? I will not say the study is absolutely not useful. Surely it is useful in showing research weaknesses or mistakes. It might even be used for baseline data.

I believe you picked up the fact that the study appears not to have considered (or did it?!) effects of surgeon load on surgery outcomes. It surely could have been done, so quite concerning that these researchers did not reveal that information.

• This reply was modified 11 months ago by  pinto.

[MarkT]

It is not biased or irrelevant, per se…it just is not relevant for our purposes, since we would all go (or have gone) to a high-volume specialist.

Pretty typical in that they ultimately got data from 65% of people (the 75% figure is only the survey response rate) who had repairs over that 2.5 year period…and they report characteristics of responders and non-responders to check for differences.

“Sutured repairs were excluded as there were very few recorded in the register”…that makes me think very few are performed there, since the register is said to include almost 98% of all repairs performed (it would be nice to know the breakdown of the 2% that aren’t covered by the registry though).

Interesting notes:

– Data is entered in real-time by the surgeons, so there is presumably very little missing data or inconsistencies in measurement between surgeons/sites.

– Open anterior mesh was by far the dominant repair method (78.7%)…much smaller group sizes for all other repairs (as low as 1.66% for TAPP). Important to keep in mind.

– One of their data points was surgeon annual volume, and a little less than half (48%) of respondents’ repairs were performed by a ‘low-volume surgeon’, defined as performing less than 26 repairs per year…but we need more info on that distribution of volume…i.e., is it roughly normal or does it skew upward at some point (and at what point) beyond the median? That would affect the interpretation of results on the influence (or lack thereof) of that variable.

– While chronic pain was defined as “pain persisting for more than 3 months, affecting everyday activities”, they actually looked at groin pain persisting after one year…however, this is still *much* too short of a timeframe and complication rates are definitely understated here. Another thread has a study were WAY longer timeframes are demonstrated to be necessary.

– There wasn’t that much difference in pain by technique, despite the statistically significant result of ‘less pain with TEP’ vs. open anterior mesh, and TEP was associated with much higher risk of recurrence.

What I found most interesting is that they don’t discuss surgeon volume in their analysis or their results. They initially included a bunch of variables in their initial model, then used a ‘backward elimination process’ (i.e. progressively removing variables in a series of models, based upon which ones offer the least explanatory power on the outcome variables of pain and recurrence) to eventually arrive at their final model that only included repair type. While this implies that repair type explained more variance than did surgeon volume, it doesn’t necessarily mean surgeon volume was insignificant (though presumably if it was, it would have been retained). On the surface this is surprising given that nearly half of all surgeries were performed by low-volume surgeons, but we still don’t know what the distribution looks like, so it’s hard to comment.

Annual volume was included as a dichotomous variable of ‘above or below median’. Normally we would like to at least know what the median is, but because almost half were ‘low-volume’ (<26/yr), it gives us *some* idea…even still, it would be nice to see what the distribution looks like…a normal vs. skewed distribution would affect interpretation.

Interesting study, on the whole. The follow-up times are still much too short, unfortunately…but the presumed consistency of the registry data with real-time entry, its inclusion of about 98% of all repairs performed, and at least some accounting for surgeon volume (even if that variable and its influence could be more clear), seem to be advantages over many other studies.

But yes, it doesn’t really help us here though…we are primarily interested in data coming from high-volume specialists. It is relevant for the general population though, since most people don’t go to high volume specialists for their repair.

[pinto]

Would you wish to have hernia surgery by a surgeon who only does 50 surgeries or less a year?
Think about that.
What if your surgeon only did 25?!
You probably would say, “thank you, but no thank you.”
Well not for these 22,000 patients!! They apparently were ok with having a surgeon with such limited experience—-50 or less done a year. As I posted elsewhere, this study is biased in that it concerns a population of surgeries done at general hospitals. We (I presume) at HT consider a specialist, an expert surgeon necessary, one who has much if not great experience. I don’t think doing two or less surgeries a month cuts it!! (That is, for the nearly half of the 22,000 concerned)
In short, we must conclude this “research” study is irrelevant.

• This reply was modified 11 months ago by  pinto.

[Watchful]

There’s a huge disconnect between what surgeons tell patients about this, and what the studies show. It’s pretty awful that groin hernia repair leads to so much chronic pain. 15% is not a reasonable incidence for persistent serious pain after 1 year (pain present, cannot be ignored, and interferes with concentration on everyday activities in the past week).

Interestingly, this paper shows similar results for laparoscopic and open in terms of persistent pain. Most other studies that I’ve seen show a lower incidence of chronic pain with lap than open.

[MarkT]

Yes…and the insufficient follow-up time (1yr for pain, 2.5yr for recurrence), which is a common issue across many studies, remains a concern too.

Since we know that many post-op complications will materialize beyond those stated timeframes, the pain and recurrence rates here are almost certainly understated.

It would be great for a study like this one, where such a high % of hernia repairs carried out in the country are covered by the registry, where everyone uses an easily trackable personal identifier to access healthcare, etc. to continue to track this sample for a longer period of time. There would presumably be much less patient attrition vs. a similar study carried out in the U.S., for example.

Of course that is expensive, time consuming, etc…but it could yield some useful and more accurate data.

[MarkT]

Yes, the 98% figure was from a ’98 study, and the data from the study here is from 2012-2015; however, “Today more than 90 units are members and almost 100% of all hernia repairs performed in Sweden are registered. A database covering more than 200 000 repairs has been assembled.”

Keep in mind that residents all use a national id card to obtain universal healthcare there…and that nearly the *vast* majority of people fall exclusively under the public system…so it would presumably be rather easy to debunk this statement if it were inaccurate.

If the 90 clinics covered only represent 10% of all medical clinics in Sweden, then it seems more logical to conclude that the other clinics are not performing hernias (or are performing *very* few of them) than to assume this statement misrepresents the % of repairs covered.

Yes, it’s voluntary database…but consider the context of the country’s healthcare model, size/population, gov’t structure, patient mentality, etc., all of which are extremely different than the U.S., for example, where a similarly voluntary database is surely going to result in a much lower participation rate among providers for several of reasons.

I don’t see a great reason to doubt the claims of the registry…certainly not to the extent that I would characterize the study as ‘biased’ or ‘narrow’.

[MarkT]

The study used data from a national registry that covered nearly 98% of all hernia repairs performed in the country during that time period.

It is the opposite of ‘narrow’.

[MarkT]

Pinto, how is that correct?

The median is simply the middle data point…it does not tell us anything about the range of the data or the shape of the distribution. I’ll give examples if you want.

I’m not sure how it is biased since the study includes nearly ALL repair data for the entire country. As I stated (from the paper), the Swedish Hernia Registry includes data from almost 98% of hernia repairs performed in Sweden during that time period. The likelihood of bias when only 2% of repairs are excluded is rather small (but, as I said, it would be very nice to know the nature of that 2% not covered).

They did not exclude surgeon volume from their analysis…they excluded it from their final model (very different). They fit a series of models and progressively removed variables that didn’t offer explanatory value. By virtue of not including it in their final model, the implication is that surgeon volume did not offer as much explanatory power in differentiating between groups as did repair type. I would still like to see some discussion of the models that included surgeon volume…some details on how little the final results changed by that variable’s inclusion and exclusion.

Again, my real concern is they way they operationalized that variable…they chose to code it as ‘above or below median’…however, again, the median is merely the middle data point and tells us nothing about the range of surgeon volume or the shape of the distribution of surgeon volume. All we know is that roughly half had repairs from their definition of low-volume. That is not enough info, IMHO, to judge whether that was the best way to treat that variable.