This is getting a lot of coverage. Unfortunately, with headlines like:
This is getting a lot of coverage. Unfortunately, with headlines like:
"Don't Finish the Course of Antibiotics ..."
"Completing antibiotic courses is a medical advice myth that may make bacterial resistance worse ..."
"Rule that patients must finish antibiotics course is wrong, study says"
"You need not always complete full course of antibiotics ..."
Here's the problem (ok, problems):
1. It's a think piece, not a study. The hypothesis may be valid, but you don't test hypotheses by throwing them out into the wild and just seeing what happens.
2. We already have problems with people taking medication as prescribed. Even for short, 3-day courses of medication people can't stick to it.
3. Those headlines are going to lead people to simply choose not to take antibiotics as prescribed, if at all. They are going to lead to both random supplies of unused antibiotics being saved and used later without consulting a doctor, and people are going to not get better, or get sicker.
4. Starting and stopping antibiotics has previously been shown to be one way to select for resistance. They argue both for and against this in their essay, then try to hedge by saying well maybe with some organisms. Again, it would be helpful to have science, here.
5. We cannot truly customize medical care per individual at this point. The tech does not yet exist. We should ideally be able to choose a specific antibiotic, for a specific length of time, for a specific disease, in a specific individual. Science is working on improving that.
6. Yes, antibiotics have side effects. Yes, this is a risk/benefit balancing act.
7. Yes, every time we treat one bug, we select for resistance in other bugs that may simply be commensals.
8. Yes, we should do prospective, randomized trials to determine the best average course for treatment of various conditions.
9. I disagree about the one stated reason they give for why it is difficult to do such trials.
Reasons may include: the potential to do harm to patients, the difficulty of observing patients who are not in a controlled setting so you have to rely on their reporting (which is a known problem), the number of people you would need to enroll, how do you control for co-morbid conditions like diabetes, do you bother to control for prior exposure to antibiotics, if so is that lifetime or specific antibiotics or diseases, do you have multiple arms for the same disease to compare different antibiotics, how do you decide how many days should be tested (3, 5, 7, all of the above?), do you have multiple arms for multiple time periods, do you do invasive testing to get objective info on critters and their numbers and activity or do you just sub end points like fever or 'malaise', what end points overall do you use if it's a question of feeling better versus whatever 'clearing the infection' is defined to mean? ETC.
The problem isn't that people wouldn't do it. It's that it's hard to design the studies, it's expensive, and you will never answer all the questions for all the antibiotics for all the diseases so you will still end up extrapolating.
10. The uncontrolled use of antibiotics in agriculture, both animal and plant, is still a huge contributor to resistance and one that could have a greater impact than any of this. Not to distract from the point, but this is really a big deal and we're not doing much to change it, if the fear is overall antibiotic resistance.
It's not that I don't think this is a question worthy of study. I do. I just think this is a little glib and careless and it doesn't look like much effort was made to control the media and its disaster movie clickbait approach to covering science.
http://www.bmj.com/content/bmj/358/bmj.j3418.full.pdf
http://www.bmj.com/content/bmj/358/bmj.j3418.full.pdf
"Don't Finish the Course of Antibiotics ..."
"Completing antibiotic courses is a medical advice myth that may make bacterial resistance worse ..."
"Rule that patients must finish antibiotics course is wrong, study says"
"You need not always complete full course of antibiotics ..."
Here's the problem (ok, problems):
1. It's a think piece, not a study. The hypothesis may be valid, but you don't test hypotheses by throwing them out into the wild and just seeing what happens.
2. We already have problems with people taking medication as prescribed. Even for short, 3-day courses of medication people can't stick to it.
3. Those headlines are going to lead people to simply choose not to take antibiotics as prescribed, if at all. They are going to lead to both random supplies of unused antibiotics being saved and used later without consulting a doctor, and people are going to not get better, or get sicker.
4. Starting and stopping antibiotics has previously been shown to be one way to select for resistance. They argue both for and against this in their essay, then try to hedge by saying well maybe with some organisms. Again, it would be helpful to have science, here.
5. We cannot truly customize medical care per individual at this point. The tech does not yet exist. We should ideally be able to choose a specific antibiotic, for a specific length of time, for a specific disease, in a specific individual. Science is working on improving that.
6. Yes, antibiotics have side effects. Yes, this is a risk/benefit balancing act.
7. Yes, every time we treat one bug, we select for resistance in other bugs that may simply be commensals.
8. Yes, we should do prospective, randomized trials to determine the best average course for treatment of various conditions.
9. I disagree about the one stated reason they give for why it is difficult to do such trials.
Reasons may include: the potential to do harm to patients, the difficulty of observing patients who are not in a controlled setting so you have to rely on their reporting (which is a known problem), the number of people you would need to enroll, how do you control for co-morbid conditions like diabetes, do you bother to control for prior exposure to antibiotics, if so is that lifetime or specific antibiotics or diseases, do you have multiple arms for the same disease to compare different antibiotics, how do you decide how many days should be tested (3, 5, 7, all of the above?), do you have multiple arms for multiple time periods, do you do invasive testing to get objective info on critters and their numbers and activity or do you just sub end points like fever or 'malaise', what end points overall do you use if it's a question of feeling better versus whatever 'clearing the infection' is defined to mean? ETC.
The problem isn't that people wouldn't do it. It's that it's hard to design the studies, it's expensive, and you will never answer all the questions for all the antibiotics for all the diseases so you will still end up extrapolating.
10. The uncontrolled use of antibiotics in agriculture, both animal and plant, is still a huge contributor to resistance and one that could have a greater impact than any of this. Not to distract from the point, but this is really a big deal and we're not doing much to change it, if the fear is overall antibiotic resistance.
It's not that I don't think this is a question worthy of study. I do. I just think this is a little glib and careless and it doesn't look like much effort was made to control the media and its disaster movie clickbait approach to covering science.
http://www.bmj.com/content/bmj/358/bmj.j3418.full.pdf
http://www.bmj.com/content/bmj/358/bmj.j3418.full.pdf
Comments
Post a Comment