Helen Salisbury: When policy doesn’t match evidenceBMJ 2019; 364 doi: https://doi.org/10.1136/bmj.l54 (Published 09 January 2019) Cite this as: BMJ 2019;364:l54
All rapid responses
Rapid recommendations series and EBM: hasten slowly?
Salisbury’s view about guidelines and EBM like the accompanying Editor's choice deserve questions.(1,2)
First, as a Frenchman I am not sure I understand the meaning of “GPs, at least the ones I know, are conscientious people who try to keep up to date with developments in medicine.“(1) Is it an understatement about a bi-modal implementation of EBM, some conscientious and others not? I am more comfortable with Javid and Anderson’s humble acknowledgment “(too many surgeons) are slow to adopt major practice changes” when commenting on a study showing that a substantial number of women with breast cancer should have not been exposed to the morbidity of axillary lymph node dissection.(3)
Second, the Journal’s Rapid Recommendations series is welcome but is it the change of culture needed? The issue seems deeper: The prerequisite for improving implementation of EBM seems first about improving the quality of research. Having been a digestive diseases specialist in the '80s I fear there is not much more evidence now than at this time but there has been an uncontrolled flow of problematic recommendations, even for the most serious frequent conditions. Although PubMed retrieved 4,128 articles on Barrett's esophagus, including 157 clinical trials over the past decade, recommendations (n=45) are based on weak evidence or expert opinion: 12 rated “conditional”, all but one with low or very low level of evidence and 33 rated “strong” but the level of evidence being high for only 3, moderate for 9, low or very low level evidence for the 21 others.(4) Similarly, the American College of Gastroenterology claims, “Patients with severe alcoholic hepatitis should be treated with corticosteroids if there are no contraindications for their use (Strong recommendation, moderate level of evidence)”.(5) However, the results of 16 randomised clinical trials have been very contradictory and the Cochrane review found no evidence of a difference between glucocorticosteroids and placebo.(6) This saga began in 1971.(7)
Last, issuing guidelines without quality assurance programmes is relying on expectations, not on evidence. Quality assurance is about providing tools and resources for implementation, monitoring implementation and also integrity. However, the majority of authors of guidelines for the American College of Gastroenterology have financial conflicts of interest with the industry but only 34% of them are disclosed.(8)
1 Salisbury H. Helen Salisbury: When policy doesn't match evidence. BMJ. 2019;364:l54.
2 Godlee F. We can change practice—can we also change culture? BMJ 2019;364:l108
3 Javid SH, Anderson BO. Delayed adoption of evidence-based breast cancer surgical practices: History repeats itself. JAMA Oncol 2018;4:1517-1518.
4 Braillon A Screening and surveillance for Barrett's Esophagus: When will we reach the horizon? Am J Gastroenterol 2016;111:899-900.
5 Singal AK, Bataller R, Ahn J, Kamath PS, Shah VH. ACG Clinical Guideline: Alcoholic Liver Disease. Am J Gastroenterol 2018;113:175-194.
6 Pavlov CS, Varganova DL, Casazza G, Tsochatzis E, Nikolova D, Gluud C. Glucocorticosteroids for people with alcoholic hepatitis. Cochrane Database Syst Rev 2017 ;11:CD001511.
7 Helman RA, Temko MH, Nye SW, Fallon HJ. Alcoholic hepatitis. Natural history and evaluation of prednisolone therapy. Ann Intern Med1971;7:311-21.
8 Combs TR, Scott J, Jorski A, Heavener T, Vassar M. Evaluation of industry relationships among authors of clinical practice guidelines in Gastroenterology. JAMA Intern Med 2018;178:1711-1712.
Competing interests: No competing interests
I understand and share Dr Salisbury’s dilemma and concerns about choosing between NHS policies vs the evidence available. As previously stated in another platform, I have pointed out the irrationality of embarking on ‘extensive infrastructure building and promote expansion of ancillary support services’ to new policies and programmes with very little evidence of cost effectiveness (ref 1).
Despite some reassurance by various enthusiasts (ref 2,3), that my reservations are misplaced regarding the cost effectiveness of Low Dose CT (LDCT) screening for lung cancer, the announcement of another study (SUMMIT study by the UCLH Cancer Collaborative) ‘to detect lung cancer early amongst at-risk Londoners when the chance of successful treatment and survival from Britain’s biggest cancer killer is greatest; to support the development of a new blood test for the early detection of multiple cancer types, including lung cancer’, as well as to provide evidence to inform a potential national lung cancer screening programme (ref 4) demonstrated that evidence for national lung cancer screening using LDCT is still very much lacking.
The long awaited analysis of the pooled data of UKLS (UK Lung Cancer Screening Trial) with NELSON (Nederlands Leuvens Longkanker Screenings Onderzoek: Dutch-Belgian Randomised Lung Cancer Screening Trial) appears to remain in limbo. It is also clear that the enthusiasm and activitism of LDCT proponents may be unfounded and based on premature and selectively biased conclusions and assumptions if we are to rely on current available evidence (ref 5,6). This is a dangerous trend in which clinicians are driving consumption of new technology and creation of large programmes when the evidence base at this time for this offers very little in support for this.
The efficacy of LDCT in picking up early stages of lung cancer is never in doubt, nor the evidence that some lung cancer when picked up in early stages are treatable and even curable. However the main uncertainties are what kind of screening programme, including the age group for recruitment, the frequency of screening, the outcome for those whose lung cancer are detected early versus the complication associated with false positive and ultimately the true cost effectiveness of the lung cancer screening using LDCT.
It is quite possible that LDCT lung cancer screening can later be proven to be cost effective; and my current concerns may very well be allayed by future studies, but until that day, there is really very little concrete evidence that support such a population based lung cancer screening by this modality. Caution has been advised in this matter (ref 7) yet it is perplexing (and frustrating) to find NHS leaders to jump on the bandwagon of this lung screening-program (ref 8) when they should know better.
The controversy, the vested opinions, and the conflicts of interests involved by the breast cancer screening programmes by mammography are now well known. Now that we have been consciously and actively inculcated in principles of evidence based medicine for the last 20 years, we should not allow the same to happen to lung cancer screening without careful consideration.
Not on our watch.
Not when we have greater responsibilities to defend other evidence based policies in austere times.
5. doi: 10.21037/jtd.2018.07.08
Competing interests: No competing interests