A funny thing happened on the way to growing up. While I was in medical school, we got a new chairman of Internal Medicine at the University of Tennessee. He was a big guy, both in persona and in medical reputation – not the kind of person you’d expect at a State Medical School in Memphis on the banks of the Mississippi. But he was a determined Streptococcus researcher, and the Mississippi Delta and eastern Arkansas were at the epicenter of the streptococcal related diseases in the world – Acute Glomerulonephritis and Rheumatic Fever – so Memphis is where he wanted to be. He never said it outright, but we had a contract, Dr. Stollerman and those of us he trained. We cultured the throats and skin lesions of every patient that came near the hospital to feed his huge frozen locker full of the bugs he loved to type. In return, he showed us what a doctor was.
He was at every Morning Report, every Grand Rounds, made rounds weekly on every ward [and he knew if we were there too]. If we got a case at night that we didn’t understand and didn’t call him, he raised so much hell that we never made that mistake again. I recall a late night when he was along side as we were heading down the long ward aisle to restrain a Lupus patient psychotic on steroids, throwing the contents of her bed pan at us and the other patients. On another night, I admitted a case of Beri-beri, a case of Scurvy, and a woman with Tetanus. I felt like the best diagnostician in North America. After a brief acknowledgement in the morning, he marched our entire residency class from case to case showing us all the things about each case that we would never have imagined anyone knew. None of us could ever figure out how he knew so much.
But that wasn’t the most important part. When he made rounds on the wards, we lined up around the bed, but he talked to and looked at the patient. If he turned to question one of us, he’d explain why to the patient. And when he came back the next week, he’d stop by and speak to the ones he’d seen before. While we picked the patients to present, frequently he’d scan the ward and move us to the bedside of the patient he thought needed seeing the most. He taught us that the most important thing about being a doctor was knowing when you don’t know something. And if one of us asked a question he couldn’t answer, the next time you saw him, he’d tell you what he found out.
I sometimes find myself wondering what Dr. Stollerman might think about all the tangles in psychiatry I write about today. I think he would say something like, "What does that have to do with taking care of patients? What’s wrong with those people?" But I bet that sometime later that week, he’d be talking to the Chairmen of Psychiatry he knew around the country trying to figure it out. I’m pretty sure he’d really like both Ben Goldacre and David Healy, but caution them both not to get too stuck in a rut of knowing.
That brings me to why I thought about him today. I was thinking about clinical judgement. In some of the comments, there’s a somewhat predictable dichotomy. Clinical judgement as medical imperialism, shooting from the hip versis clinical judgement as the highest order patient focused wisdom. I know that I way overstated that – the right of literary device. But I thought of a story from the dawn of time. One of the things Dr. Stollerman was vigilant about was overuse of medication – antibiotics, opiates, tranquillizers, and particularly steroids. Steroids are both miracle and curse drugs. You can do wonderful things or really blow it. One summer’s evening, we were at a resident picnic. My wife was great with child [I mean greeaat]. She had weeded our back yard in her nesting fervor and had a rip roaring case of poison ivy. She was miserable. I hadn’t put her on steroids, having taken on the Chief’s conservative approach to their use. He called me aside, smiled and said, "You’ve learned well," and then he handed me a prescription for the steroids that melted her poison ivy overnight [after that, my wife loved the guy almost as much as I did]. For Dr. Stollerman, there were no rules – only cases. He didn’t preach the art of medicine, he practiced it.
” I was thinking about clinical judgement. In some of the comments, there’s a somewhat predictable dichotomy. Clinical judgement as medical imperialism, shooting from the hip versis clinical judgement as the highest order patient focused wisdom. I know that I way overstated that – the right of literary device.”
I have enormous respect for your blog, but I consider that a misstatement more than an overstatement.
You say “He taught us that the most important thing about being a doctor was knowing when you don’t know something.”
I think that Goldacre would agree:
“As Robert M Pirsig said, in Zen and the Art of Motorcycle Maintenance: ‘the real purpose of the scientific method is to make sure nature hasn’t misled you into thinking you know something you actually don’t know’.”
http://www.testingtreatments.org/tt-main-text/background/foreword/
Ian Chalmers makes the point in his emas to Healey that different kinds of clinical issues may lend themselves better to different types of trials, including n of 1 trials.
I don’t think that mainstream medicine should pervert itself just to distinguish itself from CAM. But the question of how to systematically support holding on to knowing what you don’t actually know, whether through science or art, is important. Even alongside good intentions and rooting out gaming of the system.
I don’t believe that is simply trotting out old tropes.
A doctor with good clinical judgment is very hard to find these days.
Unfortunately, when it comes to the vast majority in the filed, the practice of modern psychiatry means never having to say you’re sorry.
Duane
oops, typo – field
http://www.billiamjames.com/wp-content/uploads/2013/04/OpenGSK.pdf
In contrast to GSK’s approach:
“Journals can commit to a similar policy undertaken by the BMJ, whereby ‘trials of drugs and medical devices will be considered for publication only if the authors commit to making the relevant anonymised patient level data available on reasonable request.’”
http://www.alltrials.net/wp-content/uploads/2013/01/Missing-trials-briefing-note.pdf
“Existing promises to share clinical trial data
GSK has indicated it will share data on Relenza, its influenza treatment, with the Cochrane Acute Respiratory Infections Group working on Tamiflu. It has been widely and incorrectly reported that this data has already been shared. In fact, no data has yet been shared, negotiations continue, and these have already taken many days of work, requiring senior research contracts specialist and lawyers at the University of Oxford. The company has expressed the expectation that material will be kept confidential and secret by the Oxford research team. Such a stance prevents “reproducibility,” which calls for data sets to be made available for verifying published findings and conducting alternative analyses.
GSK has also offered to share more data from its earlier phases of drug development, to enhance collaboration and innovation. Again, this has not yet happened, and it is unclear what the processes will entail.
The EMA has indicated that it intends to share individual patient data given to them as part of the licensing process for a drug (which is not all trials on that drug). It is presently unclear how this will work. At the EMA meeting to discuss options, industry representatives made clear that they felt industry should be allowed to control access to data and to decide who would be allowed to inspect the data. It was also suggested that this transparency should only be permissible for trials starting in 2014 or later.
Because drugs continue to be used for many decades, this will do nothing for the evidence base of currently used treatments, and will have little impact for a decade, if not longer. This is especially the case since many of the most widely used drugs have been on the market for some time, and good treatments continue to be used for as long as they are the best in their class.”
http://www.alltrials.net/wp-content/uploads/2013/01/Missing-trials-briefing-note.pdf
Also, I think it needs to be made clear to the public as much as it possibly can be that one study un-replicated by an independent agency or agencies doesn’t really mean that much and should not be cited as “proof” of anything— it can only be scientifically construed as “evidence” that is by its very nature questionable.
FYI
Psychiatric Drug Development: Diagnosing a Crisis
By Steven E. Hyman
April 02, 2013
http://www.dana.org/news/cerebrum/detail.aspx?id=41290
A part of the issue at hand:
“Given the fact that negative trials are frequently seen, even for antidepressant drugs that we know are effective, we agree that it would not be useful to describe these negative trials in labeling.” – Russell Katz
http://dida.library.ucsf.edu/pdf/xqu38h10
“negative trials are frequently seen” “even for antidepressant drugs we know are effective.”
we know are effective
Because there was a meta-analysis of all the data from the positive and negative trials combined supported this idea? Because of the clinical judgment of the doctors using it?
If it’s the later then what? There is good reason to believe that Dr. Ryan was absolutely convinced on the basis of his mentor’s, and his own, clinical experience, that SSRIs (and why would paroxetine be any different) were of benefit to adolescents with depression. This may be why he appears to have viewed trial data to the contrary as primarily an unfair impediment to proselytizing this.
Study 329 is a cautionary tale in a number of ways. My interpretation of its lessons drives a number of my questions. Not, at least I believe not, some over-idealized view of what physician decision making is supposed to be. Of course there is clinical judgment. Of course physicians must in many ways view the patient and their individual story as paramount.
1BOM, you appear to feel that if psychiatrists were paying more attention to their patients, to individual cases, that there would be an enormous improvement in our current situation.
In many ways true.
On the other hand, if you look at Dr. Ryan, and particularly his mentor Joaquim Puig-Antich, you see people who it could be argued were (at least in large part) driven very much by what they believed they were seeing in the patients they cared for. In the cases. Not by bowing in front of the altars of RCTs. Keller may or may not have been primarily driven by his courting of drug companies, but would you argue Ryan and his primary mentor Puig0Antich were not true believers?
Could it not be argued that it might have been preferable that Dr. Ryan (and his colleagues) had placed more weight on what he was being told by the RCT he were running (no separation from placebo on 2/2 primary endpoints, 6/6 pre-specified secondary endpoints, at least 15/19 post-hoc secondary endpoints, and increased frequency of self-harm ideation), and less weight on his own clinical judgment derived from his own clinical experience and hopes?
This is a quote that shows up in Goldacre’s Bad Science:
“If you can believe fervently in your treatment, even though controlled test show that it is quite useless, then your results are much better, your patients are much better, and your income is much better too. I believe your patients are much better, and your income is much better too. I believe this accounts for the remarkable success of some of the less gifted, but more credulous members of our profession, and also for the violent dislike of statistics and controlled tests which fashionable and successful doctors and accustomed to display.” – Richard Asher, Talking Sense, Pitman Medical, 1972
I think that this can apply in some ways to some of the extremely gifted and in many respects not credulous, members of the profession.
The doctor you describe from your earlier training sounds like an incredible human being who had an enormous lasting positive impact on the patients and peers (young and old) with whom he came in contact. He also apparently managed to avoid the pitfall of becoming too enamored of any one aspect of his work, of falling into the trap of dogmatically attributing too much of his positive impact to any single aspect of his work. Sounds like he was an incredible scientist, …etc.
I suspect that sometimes there is a lot of damage wrought by those might be gifted, both with patients and their colleagues, and who start to view things too much through one particular lens. E.g., it’s the fact I’m running PET scans on all my patients that accounts for my clinical success with them. Your guy didn’t do this.
You can imagine the kind of impact the doctor you describe in your post could have had if for whatever reason his views had started to ossifiy into one particular direction like that. NOT saying that he wasn’t incredibly well grounded in the trials (he appears to have been), but more that his charisma would have lent incredibly weight to whatever he tried to promote.
I truly don’t know if he is the exception or the rule. Truly. Whether or not listening so closely to patients was protective against such creeping cognitive biases, in addition to all the other wonderful benefits it confers. It sometimes feels though, that it has been the physicians most driven by the suffering of their patients, the gifted ones, who sometimes can most fall prey to their cognitive biases. Or perhaps, those are the ones who have the most impact when they do fall prey. Because there is so much else to respect about them. The ones that don’t as easily fit within the rogues gallery.
I think that, at least in the eyes of the idealistic, one of the greatest value of trials in their various forms is to help as a corrective when the clinical judgment of the gifted and perceptive skews into thinking it has hit upon something that may not actually be the case.
To do so is to accord trials, particularly RCTs, enormous power. We have seen some of the damage that this has wrought. It also seems important to be able to minimize the damage that shooting from the hip can do in some instances. E.g., Those who treat thousands of kids with neuroleptic cocktails and are going to be convinced from their clinical experience that this is essential no matter what any trial ever tells them. They most definitely don’t need algorithms to tell them to do this. Many seem to feel from their own clinical experience that this is more than justified.
I believe in this value. I am NOT saying that the RCT is the only way to achieve it. However, to diminish the power of randomized clinical trials, it seems important to suggest in some detail how the other alternatives (trials run other ways, n of 1 trials run by individuals, data mining, crowd sourcing decisions, whatever) would work to achieve this. So that those can be dissected with the same rigor that we should apply to the benefits and limitations of RCTs. It would seem that better fleshing out these alternatives is needed (beyond individual anecdotes – ironically), in addition to pointing out the limitations of RCTs, would help reduce the chance of an unjustified hegemony of the RCT.
Focusing more on listening to patients and the uniqueness of each case, while of enormous importance, would not seem on its own to be enough.
In regards paroxetine and its use for adolescents in depression: I don’t think it was just GSK footing the bill that interfered with Drs. Ryan et al viewing Study 329 as the corrective it should have been.
I knew Joaquim Puig-Antich very well, both when he worked in New York and in Pittsburgh. I spoke at his funeral after he died tragically at a young age from asthma in Pittsburgh, years before the SSRIs appeared. I do not recall that Kim was as Annonymous describes him. Kim followed the data. Among other things he was quite clear about the fact that imipramine didn’t work in depressed children despite being the gold standard for depressed adults. There is no way he could have primed Dr. Ryan’s beliefs about the efficacy of SSRIs in depressed youth.
In terms of Puig-Antich I meant that he was a true believer in a good way. I did not know him, but I am aware that he was among those who moved psychiatry towards a fuller understanding that children are capable of emotional suffering just as much as adults. For that he should be layded. I was pointing to him as an example of someone who would not fit within a rogues gallery.
It was a supposition on my part that perhaps Ryan had carried forward the appreciation of that suffering and that might have driven, in part, his so wishing for an effective tool. Ryan pushed hard for using imipramine in Study 329, out his shared belief that imipramine was not of utility for depressed adolescents.
I wa not attempting to impune Dr. Puig-Antich. More to say that I am convinced that dr. Ryan would have behaved differently in response to the data from Study 329 even if GSK had not been involved. It does seem that sometimes physicians come to so believe in the utility of a treatment from their clinical experience they trial data be damned. Did not appear to be the case for 1BOM’s example. Or in the example of your colleague.
But it seem to be what is happening to some extent in paychiatry and since RCTs are a limited corrective, what are alternatives?
I’m just trying to say that perhaps even the best intentioned of practitioners need systematic correctives to address biases. Being rigorous and principled would not seem to be enough. At least this is Goldacre’s argument for the importance of trials as I understand it.
Correction: I am NOT convinced Dr. Ryan would have behaved any differently even if GSk had not been involved.