David, I would, however, point out that stem cell transplant for breast cancer had been massively introduced in practice based on RRR=72% (using standardized
data set matched according to the most important prognostic features) only to be shown not to work when proper RCTs were done at later time. While ago I surveyed this group re question “ When non-RCTs are not needed?”
See:
http://personal.health.usf.edu/bdjulbeg/oncology/NON-RCT-practice-change.htm. Paul Glasziou had also a nice piece in BMJ (see the link) Perhaps, we should re-do the survey?
Best
ben
From: Evidence based health (EBH) [mailto:[log in to unmask]]
On Behalf Of David Braunholtz
Sent: Wednesday, March 28, 2012 7:54 AM
To: [log in to unmask]
Subject: Re: Critical Appraisal of RCTs Using "RAAMbo" acronym
On the substantive issue of how to treat such patients, I would maintain that in such high mortality acute conditions, given the promising results of the trial (& maybe previous cases),
randomising further patients to 'control' treatment in an RCT would be unethical. If you were such a patient (or their parent / partner) would you prefer the novel treatment or control, or be ambivalent ? I wouldn't. If the novel treatment really does reduce
mortality by anything like 60% then this will be evident in a simple case series after a few dozen patients. If the hoped for improvement in mortality is not so obvious after a few dozen patients, then perhaps a further RCT could be contemplated. Somewhat
similar thing happened with respect to patients with NVCJD aka mad cow disease - where it seems unfortunately the novel treatment (Quinacrine) eventually turned out not to be effective:
David Braunholtz
From: Ruwan Parakramawansha <[log in to unmask]>
To: [log in to unmask]
Sent: Tuesday, 27 March 2012, 22:52
Subject: Critical Appraisal of RCTs Using "RAAMbo" acronym
Dear All,
I am a participant in the ongoing three day EBM workshop at Oxford. I would be grateful if I can have your opinion/evidence on a specific issue related to
critical appraisal of RCTs.
Few years ago I came across the following RCT for the treatment of paraquat poisoning which is a killer and yet a condition where conducting a large RCT
will be very difficult.
Copied as fair use.
Repeated pulse of methylprednisolone and cyclophosphamide with
continuous dexamethasone therapy for patients with severe
paraquat poisoning
Ja-Liang Lin, MD; Dan-Tzu Lin-Tan, RN; Kuan-Hsing Chen, MD; Wen-Hung Huang, MD
Objective: Paraquat is widely used in the world, and all treatments
for paraquat poisoning have been unsuccessful. Many
patients have died of paraquat poisoning in developing countries.
A novel anti-inflammation method was developed to treat severe
paraquat-poisoned patients with >50% to <90% predictive mortality:
initial pulse therapy with methylprednisolone (1 g/day for 3
days) and cyclophosphamide (15 mg/kg/day for 2 days), followed
by dexamethasone 20 mg/day until PaO2 was >11.5 kPa (80 mm
Hg) and repeated pulse therapy with methylprednisolone (1 g/day
for 3 days) and cyclophosphamide (15 mg/kg/day for 1 day),
which was repeated if PaO2 was <8.64 kPa (60 mm Hg).
Design: Randomized controlled trial.
Setting: Academic medical center in Taiwan.
Patients: Twenty-three paraquat-poisoned patients with
>50% and <90% predictive mortality assessed by plasma paraquat
levels were prospectively and randomly assigned to the
control and study groups at a proportion of 1:2.
Interventions: The control group received conventional therapy
and the study group received the novel repeated pulse treatment
with long-term steroid therapy.
Measurements and Main Results: We measured patient mortality
during the study period. There was not a different distribution
of basal variables between the two study groups. The mortality
rate (85.7%, six of seven) of the control group was higher
than that of the study group (31.3%, five of 16; p .0272).
Conclusions: The novel anti-inflammatory therapy reduces the
mortality rate for patients with severe paraquat poisoning.
(Crit
Care Med 2006; 34:368–373)
At that time, given the constraints I felt that this is an acceptable study and started treating paraquat poisoning patients with the immunosuppressive treatment
used in the RCT.
Few months later a letter written to the editor highlighted that the study is not adequately powered to detect such a difference in the out come.(Gunawardena
G et al. Randomized control trial of immunosuppression in paraquat poisoning. Crit Care Med. 2007;35(1):330-1.)
I also noted there is some overlap between the CONSORT checklist for researchers doing RCTs(attached) and RAAMbo check list.However sample size is mentioned
in CONSORT list only.
I am wondering why such an important component of an RCT that can invalidate a trial results even when rest of "RAAMbo"validity criteria are fulfilled is
not taken into consideration during the critical appraisal process of a RCT.
Greatly value your input as it will enrich my learning experience at the EBM course.
Ruwan Parakramawansha
Trainee in Clinical Pharmacology, Therapeutics and Toxicology
University Hospital Llandough