Publication Bias is Common in High Impact Journal Systematic Reviews

A very interesting study was published earlier this month in the Journal of Clinical Epidemiology assessing publication bias reporting in systematic reviews published in high impact factor journals.  Publication bias refers to the phenomenon that statistically significant positive results are more likely to be published than negative results. They also tend to be published more quickly and in more prominent journals. The issue of publication bias is an important one because the goal of a systematic review is to systematically search for and find all studies on a topic (both published and unpublished) so that an unbiased estimate of effect can be determined from including all studies (both positive and negative). If only positive studies, or a preponderance of positive studies, are published and only these are included in the review then a biased estimate of effect will result.

Onishi and Furukawa’s study is the first study to examine the frequency of significant publication bias in systematic reviews published in high impact factor general medical journals. They identified 116 systematic reviews published in the top 10 general medical journals in 2011 and 2012: NEJM, Lancet, JAMA, Annals of Internal Medicine, PLOS Medicine, BMJ, Archives of Internal Medicine, CMAJ, BMC Medicine, and Mayo Clinic Proceedings. They assessed each of the systematic reviews that did not report an assessment of publication bias for publication bias using Egger test of funnel plot asymmetry, contour-enhanced funnel plots, and tunnel effects. RESULTS: The included systematic reviews were of moderate quality as shown in the graph below. About a third of “systematic reviews” didn’t even perform a comprehensive literature search while 20% didn’t  assess study quality. Finally, 31% of systematic reviews didn’t assess for publication bias. How can you call your review a systematic review when you don’t perform a comprehensive literature search and you don’t determine if you missed studies?

Quality of included reviews

From J Clin Epi 2014;67:1320

Of the 36 reviews that did not report an assessment of publication bias, 7 (19.4%) had significant publication bias. Saying this another way, if a systematic review didn’t report an assessment of publication bias there was about a 20% chance publication bias was present. The authors then assessed what impact publication bias had on the estimated pooled results and found that the estimated pooled result was OVERESTIMATED by a median of 50.9% because of publication bias. This makes sense as mostly positive studies are published and negative studies aren’t. Thus, you would expect the estimates to be overly optimistic.

The figure below reports the results for individual journals. JAMA had significant publication bias in 50% of the reviews that didn’t assess publication bias while the Annals had 25% and BMJ 10%. It is concerning that these high impact journals publish “systematic reviews” that are of moderate quality and have a significant number of reviews that don’t report any assessment of publication bias.

Results by journal

From J Clin Epi 2014;67:1320

Bottom Line: Always critically appraise systematic reviews published in high impact journals. Don’t trust that an editor, even of a prestigious journal, did their job….they likely didn’t.

Mix up your EBM teaching

Actually you should mix up all your teaching. Keep it fresh and interesting to your learners. I am getting a Master in Educational Technology and as part of one of my classes I had to design a WebQuest activity. You can learn more about WebQuests from WebQuest.org.

A WebQuest is an inquiry-oriented lesson format in which most or all the information that learners work with comes from the web.

I developed a WebQuest on determining pretest probability. You can look at my WebQuest here. It is geared toward 3rd and 4th year medical students but could be used with residents also. Feel free to use it.

probabilityheader2

While using the Web to teach is nothing new it is good to change up how you teach. Instead of lecturing on determining pretest probability have your learners do this WebQuest on their own time and flip that part of your lesson on diagnostic testing. It can also be done during class time and many of the WebQuests on WebQuest.org are designed for in-class use. While I had to actually develop/code the website there are online WebQuest creators (just look on the WebQuest.org site for the Creating WebQuests tab).

Why do clinicians continue medications with questionable benefit in advanced dementia?

A recent study in JAMA Internal Medicine estimated the prevalence of medications with questionable benefit being used by nursing home residents with advanced dementia. This is an important question because significant healthcare resources are utilized in the last 6 months of life. Furthermore, if there is no benefit then the only possible outcomes can be excess cost with or without harm. As the authors note most patients at this stage just want comfort care and maximization of quality of life.

The researchers studied medication use deemed of questionable benefit by nursing home residents with advanced dementia using a nationwide long-term care pharmacy database.  A panel of geriatricians and palliative medicine physicians defined a list of medications that are of questionable benefit when the patient’s goal of care is comfort and included cholinesterase inhibitors, memantine hydrochloride, antiplatelets agents (except aspirin), lipid-lowering agents, sex hormones, hormone antagonists, leukotriene inhibitors, cytotoxic chemotherapy, and immunomodulators.

53.9% nursing home residents with advanced dementia were prescribed at least 1 questionably beneficial medication during the 90-day observation period with cholinesterase inhibitors (36.4%), memantine hydrochloride (25.2%), and lipid-lowering agents (22.4%) being most commonly prescribed. Patients residing in facilities with a high prevalence of feeding tubes were more likely to be prescribed these questionably beneficial medications.

Of only those residents who used at least 1 questionably beneficial medication, the mean (SD) 90-day drug expenditure was higher ($2317 [$1357]; IQR, $1377-$2968 compared to $1815 for all residents), of which 35.2% was attributable to medications of questionable benefit (mean [SD], $816 [$553]; IQR, $404-$1188).

Picture of elderly patient

I think this study demonstrates excess medication usage and excess costs in a population in which costs are already high and for which this added cost is of little benefit. So why are these medications continued in this population? Are physicians unaware of the lack of benefit of these medications in this population? Are they aware but worried that stopping them will make the patient worse? I suspect a little of both is the correct answer.

A major challenge for EBM is getting the E out there. Numerous resources are available but a major step in accessing a resource is recognizing a knowledge deficit. How do you know you don’t know something? Pushing evidence (for example by email) is useful for general knowledge but isn’t useful to answer specific questions. Most of the push email services require enrollment to get the emails and many clinicians probably don’t even know this exists. Maintenance of certification could be a useful tool to improve knowledge if it was designed properly but can’t cover everything for all clinicians.

So we have a dilemma. Studies like this show areas for improvements in knowledge and practice but there are no great practical ways to improve either in a nursing home setting. Clinical reminders still have to be acted upon. Payers could refuse to pay for certain services but docs will likely continue to order them with the patient picking up the bill. Protocols could be put in place but they have to be followed and agreed upon by clinicians. They all will have anecdotal evidence of grandpa getting worse when his cholinesterase inhibitor was stopped. And they will say “What’s the harm in continuing it?”.

Affect Heuristic,COI, or Lack of Knowledge? Why Do Cardiologists Overestimate Benefits of PCI in Stable Angina?

A recent study in JAMA Internal Medicine by Goff and colleagues made me wonder if the Cardiologists studied are uninformed of the limited benefits of stenting (PCI) for chronic stable angina, do they have too strong of a conflict of interest due to economic gain, or is the affect heuristic playing a big part? Probably  a mixture of them all. The COURAGE Trial taught us that PCI was better than medical therapy at reducing anginal symptoms but wasn’t any better for reducing MI and death.

Goff and colleagues reviewed 40 recordings of actual encounters of Cardiologists with patients being considered for cardiac catheterization and PCI. I am unsure if these were video recordings or audio recordings. As best I can tell these were all private practice Cardiologists.  Cardiologists either implicitly or explicitly overstated the benefits of angiography and PCI. They presented medical therapy as being inferior to angiography and PCI (a statement that defies the findings of the COURAGE trial). In fact, in only 2 of the encounters did they state PCI would not reduce the risk of death or MI. These Cardiologists also didn’t use good communications styles that encouraged patient participation in the decision making process.

Why might these Cardiologists do this? They could be uninformed of the limited benefits of PCI in stable angina, but I doubt it. COURAGE was a landmark publication in one of the world’s most prominent medical journals. I find it hard to believe that Cardiologists wouldn’t be aware of the results of this trial.

They certainly have a financial stake in their recommendations.  The image below shows that a diagnostic cath is reimbursed at approximately $9,000 while a PCI with DES is reimbursed at approximately $15,000. That has to have an impact on decision making. I don’t accuse these Cardiologists of doing a procedure only for money but subconsciously this is playing a role. Recommending medical therapy only gets you an office visit reimbursement (maybe $200 or so).

What about the affect heuristic? My colleague Bob Centor writes about this often in his blog. A heuristic is a quick little rule we use to make decisions. The affect heuristic is a particular rule we use that is based on our emotions about a topic. Do I like it? Do I hate it? How strongly do I feel about it? The affect heuristic leads to the answer to an easy question (How do I feel about something?) serving as the answer to a much harder question (What do I think about something?) Its not hard to imagine (and data in the Goff paper supports this) a Cardiologist feeling that PCI is beneficial and should be done. They are emotionally tied to angiography and PCI….they have seen patients “saved” because of this procedure.

So what can be done? The solution is harder than determining the problem (as is often the case). The easiest solution is for insurance companies to stop reimbursing for the procedure in stable angina unless patients have failed optimal medical therapy but this is draconian. I also worry that patients will then receive bills for unreimbursed  catheterization charges. I think using the technology that was used in this study combined with feedback could be useful but logistically impossible. I have always wondered why we don’t use secret shopper fake patients to evaluate physician skills and knowledge (of course the answer is a logistic one) instead of the MOC system. Just publishing a study doesn’t work if physicians don’t read or if they don’t use that study to answer a clinical question. Patient decision aids (like this excellent example) could be very useful but the physician would have to use the tool and many don’t even know they exist.

Some would argue EBM has failed again. A well done study was published and it hasn’t made a difference. The principles of EBM have not failed and in fact, if they were used, could limit the inappropriate use of PCI in stable angina patients. What has failed is the desire to learn and use these skills by the older Cardiologists in this study. Like many physicians, they rely on outdated knowledge and emotions or beliefs. As stated by Bob Centor in a post about the affect heuristicDecision making bodies have biases. Until they understand their biases, we will have the problem of unfortunate, unnecessary and potential dangerous unintended consequences“. In this case the Cardiologists are the decision making bodies and the unintended consequences are the MIs, strokes, renal injury, and death that can and do occur from cardiac catheterization.

Hopefully you are now aware of what the affect heuristic is and how it impacts decision making. Acknowledge it and separate your feelings about a topic from the data. Your patients will benefit.

Underrepresentation of certain patient populations in RCTs informing guidelines: Does it matter?

This week JAMA Internal Medicine published a research letter reporting data on the underrepresentation of women, elderly patients, and racial minorities in RCTs used to inform cardiovascular guidelines. The authors state that RCTs are considered to be the highest level of evidence that should be used to inform guideline development. I would argue systematic reviews would even be better but I understand that questions to be addressed in guidelines often need individual RCTs to answer them.  They then state that “RCTs can have limited external validity”. What do you think?

The authors evaluated all references and then focused on RCTs that were cited in the ACC/AHA guidelines on atrial fibrillation, heart failure, and acute coronary syndromes. They extracted data on age, gender, ethnicity, and continents from which subjects were recruited. What did they find?

Female representation was highest in RCTs in atrial fibrillation (33%) followed by ACS (29%) and heart failure (29%). The next question you should ask is how does this compare to the actual gender representations of people affected by these diseases? In US registries of atrial fibrillation women make up 55% of patients, 42% in ACS registries, and 47% in heart failure registries. Thus women are underrepresented by up to 22%  in these studies but does this affect guideline recommendations? Another way to think about this is would more data change recommendations for women? Hard to know for sure but I suspect not. If enrollment is properly conducted I would think that those enrolled would be a sample of all women with atrial fibrillation, ACS and heart failure. Even though the sampling fraction is smaller as long as they are representative of all women with those problems there should be no bias. The statistical inferences could be affected due to small sample sizes though but the overall qualitative findings (ie benefit or harm) should not be affected.

As expected the majority of patients enrolled in these studies were white. Black patients constituted 19% of heart failure RCT patients and 6% of both afib and ACS patients. In US registries of heart failure, afib and ACS black patients make up 6%, 21%, and 11% respectively. Again I don’t have a problem with this if sampling was done properly.

Elderly (defined as those >75 yrs of age) are very underrepresented constituting only 2% of patients in all the RCTs combined. In this case guideline developers will have to rely on observational data or use expert opinion to inform recommendations.

Finally, the authors point out that 94% of enrolled patients came from North America or Europe. Is this a problem? I don’t think so for the US as ACC/AHA guidelines are developed to guide treatment of American patients. Patients from other underrepresented continents will have less direct evidence informing recommendations on their care. Consequently, those recommendations will be based more on expert opinion.

Using a Personal Learning Network for Professional Development

I’ve been absent awhile and its good to be back. I started a new degree program (Master in Educational Technology) and I just haven’t had a lot of extra time to do other things. Why, you might ask, am I getting yet another degree (will be my 4th)? When I developed my online course in EBM for the medical school I became very interested in online teaching which then led to an interest in mobile teaching which then led me to use a variety of tech tools for teaching. I realized I had no real reason to use one tool over another except for its “coolness factor”. I then decided I wanted formal training in educational technology. All this rambling is to set up the fact that I learned about educational technology through my personal learning network. I made the video below explaining what this is and how to develop one using social media. While I focus on education I do list some EBM thought leaders you should follow on Twitter (in addition to me of course!).

Knowing In Medicine

Or perhaps a more apt title would be “Not Knowing that we Don’t Know in Medicine”. A colleague and I gave a lecture last week on EBM in the context of how do we know what to do in medicine. We pointed out that there are 4 general ways that we know what we know in medicine: authority, clinical experience, pathophysiological rationale, and systematic investigation. I serendipitously read an article late last week in a great new series in JAMA called JAMA Diagnostic Test Interpretation.  I thought about all the times I used ammonia levels incorrectly and all the times my colleagues and residents used ammonia incorrectly. Why? Was I too lazy to evaluate the literature in this area? Admittedly I hadn’t but it didn’t even occur to me to do so because during my training my senior residents and attendings told me to check ammonia levels in patients who we suspected had hepatic encephalopathy as it “would help make the diagnosis”. 20 years later an epiphany has made me realize I have been doing the wrong thing for all these years. I didn’t know jack about the limitations of ammonia in chronic liver disease.

Why did I trust authority? Was it because John Mellencamp sang “I fought authority and authority always wins” so I didn’t question what I was told to do?  I wonder how many other things like this I do with no idea that I am doing it all wrong. Likely a fair amount. The hard part is there isn’t time to go back and review literature on everything we do. Even if I only reviewed the information in a preappraised resource like Dynamed or UpToDate I wouldn’t have enough time as I can barely find time to keep up with newer things without having to go through all the things I “KNOW” already.

I hope articles like this simple review in JAMA will help educate us all. I hope other journals follow JAMA’s lead and make distilled evidence summaries that we can quickly digest to improve the cost-effective care that we provide. I hope each of us will occasionally, maybe just once a month, question how we “KNOW” something and if we can’t give a good enough answer that we will take the time to  find that answer. We likely will be surprised by how much we don’t know in medicine.