Much to my own personal amusement, Peter Saunders has attempted to respond to my recent post on the inane pile of drivel that is the Christian Medical Fellowship’s submission to the RCPsych consultation on its systematic review of the evidence relating to the mental health impact of induced abortion.
Much of Saunder’s reply is taken up with banal willy-waving and pointless ad hominems which do nothing whatsoever to address any of the substantive points I made in my own post.
We’re told, for example, that the CMF ‘put together a working group of academic psychiatrists and gynaecologists’ in order to produce its submission, to which I can only respond by asking where, exactly, these academic psychiatrists are based so I can be sure to give their lectures a miss as they’re clearly incapable of putting together a halfway decent critique of a systematic review. Saunders’ of course, knows nothing of my own academic background, despite posting comments on both his own blog and at Liberal Conspiracy asking for personal information about me (naughty, naughty) but without going overboard what I can say is that I am, at the very least, much better qualified to comment on a systematic review based on epidemiological data relating to the prevalence of psychiatric disorders than some of his working group members and rather more honest – and, dare I say it, competent – than other given the dreadful arguments that the CMF are trying to put forward.
And yes, I did post my comments on a blog, because I really don’t have too much add to the review that isn’t already there, although there is a recently published paper based on research undertaken in Brazil relating to risk of post natal depression in women after an unsuccessful attempted abortion, which I could chuck into the mix, but as that’s been published in the British Journal of Psychiatry, why bother telling RCPsych about a paper which appears in their own journal?
I’m also, apparently, the “infamous pseudonymous Liberal Conspiracy troll-blogger ‘Unity’” although, truth be told, I haven’t done any serious trolling in a long time, and my online alter-ego from those days (long retired) is just about the last ‘person’ that Saunders would ever want to encounter out here is cyberspace.
But let’s cut the bullshit and get down to business, because this is about substance and on that score, Saunders has pathetically little to bring to the table.
His first effort at mounting as substantive argument runs like this:
First, Unity claims that the CMF is wrong to suggest there is insufficient transparency in the RC Psych’s selection, exclusion and rating of research papers, and instead asserts that the methodology section in the consultation document contains all the elements one would expect from a rigorously conducted systematic review.
However, there is actually no publication of the data extraction tool used by the review group, nor is there an appendix with the data extraction table which summarizes the data that has been extracted, nor is there a table showing how the criteria for the quality items for each paper were scored and how the overall score was derived. But all of these are actually common practice in Cochrane Reviews published in the Cochrane Library and are in fact required in order to understand the final conclusions drawn by the review team. The methods may be accurately described in the review document (as Unity correctly asserts) but the selection process is not made clear. Decisions for judgments made regarding a source of bias have to be made explicit.
So, he’s whining about RCPsych’s failure to spend several thousand pound on installing the same data extraction tool on its website that the Cochrane Library has built into its website – and this is a guy he happily peddles the biased dreck put out by Priscilla Coleman and David Reardon and has the gall to call it evidence. There’s a handy list of the papers that Saunders and his anti-abortion buddies have been peddling as evidence of late in this post – please feel free to cross-reference this list against the papers in the RCPsych review and look at the quality ratings they’ve been given by RCPsych, you’ll find plenty of exclusions and almost nothing that rates above ‘Fair’.
As Saunder’s concedes, the inclusion/exclusion criteria and selection methods are accurately described in the review paper, and that’s more than enough to allow for replication of the study and easily as much, if not more, than you’ll find in hundreds of other published systematic reviews. The absence of a Cochrane-style data extraction tool just means that Saunders and his buddies have to put in a bit of work if they want to argue that the selection criteria used by RCPsych is biased and back that argument up with evidence. Whining about the lack of data extraction tool suggests nothing more than desperation on their par. The review does not support their arguments and they’re now casting around for a means of trying to discredit the review without having to go to all the messy business of putting together their own analysis and put up supporting evidence for their preferred position where it could be shot at by others.
Its neither credible nor an honest line of argument, and one that I fully expect RCPsych will reject out of hand.
Saunders’ only other sunstantive argument is this one:
Second, Unity takes issue with CMF’s suggestion that, given the limitations in the current data, other methods to seek the views and experiences of those involved in the care of women who have had an abortion should be considered as a valid source of evidence. He/she complains, ¬ somewhat absurdly, ¬that qualitative studies almost always provide no usable evidence for a review of this kind.
However, the inclusion of qualitative data, particularly in evaluating complex health care issues is of increasing importance to policy makers and health care professionals. The Cochrane Collaboration and leading academics such as Tricia Greenhalgh, Andrew Booth, Geoff Wong and Jennie Popay are developing ways of incorporating and synthesizing qualitative data in systematic reviews. It is simply naive to dismiss the rigorously collected narrative data qualitative that researchers can bring to an understanding of such an issue.
The RCPsych review is an epidemiological study which, first and foremost, aims to evaluate the extent to which there is evidence for a causal relationship between induced abortion and subsequent mental health problems in women and, if so, the extent to which women face an elevated risk of developing such problems relative – ideally – to the prevalence of similar problems in women who unwillingly carry an unwanted pregnancy to term.
Qualitative studies, which tend to have very few participants, often as few as five to ten, and provide little or no data that could reliably be generalised to a much large population – remembering that around 189,000 women currently have abortions every year in England and Wales, offer nothing of value to studies of this kind, for all that they can offer useful – and useable – evidence to policy makers and health care professionals in so far as the evidence they provide can, and does, information practice on the ground.
That the Cochrane Collaboration is currently developing new methods of incorporating data from qualitative studies into its systematic reviews in interesting but irrelevant. Those methods still have to be tested and validated before they can reasonably be used in a live setting and there’s no guarantee that their use, now, would offer anything of value to the current review or, in any way, enhance or alter its core findings.
Noticeably, Saunders has nothing whatsoever to say on the subject of RCPsych’s reevaluation of the data from David Fergusson’s two papers, where the CMF’s arguments are not just wrong, they’re not even wrong, but then I wouldn’t expect him to call too much attention to that on his own blog as this would require him to admit that his organisation has had its arse handed to it on that one.
Nor does Saunders’ address the issue of the CMF’s cherry-picking of data from entirely the wrong results table to make an incorrect argument and, again, that hardly surprising given that getting caught making such a basic error rather call into question the merits of all the willy-waving nonsense earlier is his response.
Ultimately, the proof of the pudding is in the eating, and the test of who’s right and wrong here will be decided when RCPsych publishes the final version of its review and we get to see whether its accepted any of the arguments put forward by the CMF.
When that happens, I’ll be sure to revisit this issue and settle up with Peter once and for all.