Rapid Responses to:
|
|
Rapid Responses: Rapid Responses published:
-
![[Read Rapid Response]](/icons/misc/sectionDOWN.gif)
Sham surgical controls would be better for RCTs of established surgical techniques
- Tim JB Crayford (31 March 2007)
-
Are we too scientifically orientated?
- Celine M Aranjo (26 April 2007)
-
Risk factors for recurrent pharyngitis in adults
- PRASANTA PADHAN (6 May 2007)
-
Methodological flaws and management misconceptions
- Ewen F Flint (8 May 2007)
-
Letter in response to: Tonsillectomy versus watchful waiting in recurrent streptococcal pharyngitis in adults: randomised controlled trial
- Jonathon Joseph, Joanna Stephens, John Almeyda (17 May 2007)
-
Unnecessary tonsillectomy?
- S S Musheer Hussain, Peter Ross (22 May 2007)
-
Confidence interval too narrow
- Michael J Campbell, University of Sheffield S1 4DA (23 May 2007)
-
Is surgery the only alternative?
- Thomas Sandeman, Melbourne, Australia (27 May 2007)
|
|
|||
|
Tim JB Crayford, Director of Public Health Croydon PCT
Send response to journal:
|
Sir Re: Tonsillectomy versus
watchful waiting in recurrent streptococcal pharyngitis in adults: randomised
controlled trial BMJ, doi:10.1136/bmj.39140.632604.55 Proper controls are required for established surgical
techniques The paper from Alho et. al (1) appears
to present some of the better evidence in favour of tonsillectomy seen in some
while. Indeed, the authors conclude: “tonsillectomy
is an effective alternative for adults with a documented history of recurrent
episodes of pharyngitis”. It is
important to counter these unsupported assertions. Problems with randomised controlled trials often become
evident when one considers things from the patient’s perspective. The
next few paragraphs illustrate this point as it applies to recruitment to this trial.
I pose the following questions. Is it possible that
having agreed to have this done to you, you might believe that this
operation is going to do you more good than waiting? If so, then people
in group B may be more likely to report symptoms of a sore throat than those
who had the operation done immediately. Could this problem have accounted
for apparent differences in reported symptoms? Using the data provided, I have examined a few questions
that weren’t directly addressed in the paper: a) People
who were offered early tonsillectomy took fewer swabs than people given late
tonsillectomy (20 out of 34 vs. 5 out of 36, p<0.0001). b) Focussing
on the people who took swabs, there appears to be no apparent difference
between the positivity rates for the primary outcome measure (Strep A
infection) in intervention and control groups (8 out of 20 samples vs. 1 out of
5, p>>0.1). This comparison of
course still biased, just differently so to the one the authors report about
Step A positivity. c) For
those who had the early operation, the combined time with a sore throat
including the operation was actually over 4 days greater for those whose
operation was deferred. d) Despite
the potential for bias demonstrated in a) and b), the authors nonetheless cited
the following statistic in the abstract “At
90 days, streptococcal pharyngitis had recurred in 24% (8/34) in the control
group and 3% (1/36) in the tonsillectomy group (difference 21%; 95% confidence interval
6% to 36%).” Had the authors wished
to compare both groups, they should have taken samples from all patients at
(say) 90 days after recruitment. e) They
conclude “Adults with a history of
documented recurrent episodes of streptococcal pharyngitis were less likely to
have further streptococcal or other throat infections or days with throat pain
if they had their tonsils removed.” Neither
of these conclusions seem to be supported by the data. f) The
authors also point out, “A substantial improvement over time in the rate of
episodes of pharyngitis occurred in the control group during the follow-up,
probably because of the natural course of the disease”. Isn’t this the point? Recurrent
tonsillitis tends to get better over time. We know too little about the natural history of pharyngitis
or recurrent tonsillitis, let alone the benefits of tonsillectomy or
adenotonsillectomy. If there are any
benefits at all, do they last life-long, or just a year or so? Do any benefits steadily decline over the
months, or do they work for a couple of years and then stop suddenly. Do adults who’ve had a tonsillectomy get fewer
sore throats than those who didn’t? Do
they take any more time away from work? What
is the value of any benefit of tonsillectomy as measured in QALYs? For something in which the NHS invests up to
£40M per annum, isn’t it time we knew? Where ‘established’ surgical techniques such as tonsillectomy
for recurrent tonsillitis are to be properly tested, a proper surgical control
is required. The absence of surgery, or delayed surgery, just does not
pass scientific muster. Sham surgery is really the only proper comparison,
but would people consent to this? We therefore
still have little evidence of the benefit, or lack of it, of adult
tonsillectomy. The evidence for
childhood tonsillectomy is not much better.
Yet the NHS continues to do around 40,000 procedures in adults and
children each year. Until the position
is clearer, isn’t there an argument to stop offering this intervention on the NHS
for recurrent tonsillitis outside of a properly conducted clinical study? Yours sincerely Dr. Tim Crayford MB.BS MSc. FFPH Director of Public Health References Tonsillectomy versus watchful
waiting in recurrent streptococcal pharyngitis in adults: randomised controlled
trial BMJ, doi:10.1136/bmj.39140.632604.55 Competing interests: None declared |
|||
|
|
|||
|
Celine M Aranjo, Senior G.P. NSW, Australia 2208
Send response to journal:
|
Acute Strep.Tonsillitis: common in children and effectively treated with appropriate antibiotics.However, the adult sources of infection in children go unrecognised very often, as well as the cleanliness and hygiene of drink-ware and other eating gadgets for adults and children. This RCT addresses the adult infection factor as well as it possibly can,however,it is regretted that the importance of adult oral and oro- pharyngeal hygiene has not been emphasised in this study nor the drinkware factors. Probably due to the long-held belief that Acute tonsillitis is not 'contagious'? Competing interests: None declared |
|||
|
|
|||
|
PRASANTA PADHAN, MD,POST DOCTORAL FELLOW DEPARTMENT OF CLINICAL IMMUNOLOGY AND RHEUMATOLOGY,CHRISTIAN MEDICAL COLLEGE ,VELLORE,INDIA.632004
Send response to journal:
|
Dear Editor, Olli-Pekka et al adress the very important issue regarding the role of tonsillectomy in adult patients with recurrent streptococcal pharyngitis(1).However it is also important to rule out any underlying predisposing conditions such as primary immunodeficiency diseases,diabetes,sicca syndrome in such patients- which has not been mentioned in this study. Grumach et al describe recurrent infections of the upper respiratory tract in partial complement factor I deficiency in a Brazilian family(2). References: (1)Olli-Pekka Alho, Petri Koivunen, Tomi Penna, Heikki Teppo, Markku Koskela, and Jukka Luotonen.Tonsillectomy versus watchful waiting in recurrent streptococcal pharyngitis in adults: randomised controlled trial.BMJ 2007; 334: 939 (2)A. S. Grumach,M. F. Leitão,V. G. Arruk,M. Kirschfink and A. Condino-Neto.Recurrent infections in partial complement factor I deficiency:evaluation of three generations of a Brazilian family.Clinical and Experimental Immunology 2005, 143:297–304 Competing interests: None declared |
|||
|
|
|||
|
Ewen F Flint, Consultant ENT surgeon Dumfries & Galloway Royal Infirmary DG1 4AP
Send response to journal:
|
Ths Scottish Intercollegiate Guidelines Network (SIGN) has produced guidance on the management of sore throat and indications for tonsillectomy (SIGN publication 34). I have compared some of the statements in this paper with the relevant sections from the SIGN guideline. "Recommended treatment is antibiotics to prevent rheumatic fever to prevent suppurative complications to ameliorate symptoms to decrease contagion" - SIGN 34 specifically refutes all four of these reasons "Tonsillectomy has been used to prevent recurrent streptococcal throat infections" - Tonsillectomy should only be used to prevent tonsillitis "Patients were referred for tonsillectomy because of recurrent episodes of streptococcal pharyngitis.The symptoms and signs had to be typical of streptococcal pharyngitis. At least one episode had to be group A strep. proved by culture or rapid antigen test" - SIGN states that (a)clinical examination should not be relied upon to differentiate between viral and bacterial sore throat (b)throat swabs should not be carried out routinely in sore throat (c)rapid antigen testing should not be carried out routinely The primary end point was the proportion of patients with an acute episode of group A strep. pharyngitis during 90 days follow-up. Since none of the patients were carriers, these must have been new infections, so they are of no relevance. I also take issue with the authors' use of statistics; a primary outcome of 8 strep infections in one group against 1 in the other,i.e. 24% vs 3% is reported as a 21% difference, when it is in fact an 87.5% difference. Similarly a reduction in pharyngitis of any kind from 56% in one group to 31% in the other is a difference of 45%, not 25%. Prof Little comments that "The trial is underpowered to quantify the risks of these complications accurately" but this trial is not about the risks of tonsillectomy, which are already well known. In my opinion the take-home message(s) for clinicians should be: that antibiotics are not indicated for most sore throats that no attempt need be made to determine the cause that tonsillectomy is only indicated for the treatment of recurring, debilitating tonsillitis the only guarantee is that the patient will never have tonsillitis again Competing interests: I have been a consultant ENT surgeon for 23years, and I developed the use bipolar diathermy in tonsillectomy |
|||
|
|
|||
|
Jonathon Joseph, SHO ENT West Middlesex Hospital, Twickenham Road, Middlesex, TW7 6AF, Joanna Stephens, John Almeyda
Send response to journal:
|
Dear Sir, We read your article with interest, but noticed several points which we felt warranted discussion. Firstly, the choice of presence of a positive throat swab as a primary outcome measure. This is commendably an objective measure, however is not as useful a marker in day to day practice, as throat swabs are not routinely used in the UK to diagnose throat infections. Also a useful outcome measure would have been the number of days taken away from work or schooling as a direct result of the throat infection, as this is an important marker which has a significant impact on patients. We felt that the length of follow up was also an issue, as 90 days is a very short period. It is expected that patients will suffer some pain and discomfort following their operation, and to allow only a 90 day period puts an undue emphasis on this, which would be eradicated by including a longer follow up period. In addition you estimated that patients suffer 13 days of severe pain post operatively, which is somewhat different to our experience. We find patients often return to normal diet and activities after approximately a week, with much lower pain scores during the second week of recovery. So in summary, we support the findings of your study, but feel that it could have a much greater impact and more widespread acceptance with a few changes to the methodology. These will make the findings more robust and less subject to criticism. Competing interests: None declared |
|||
|
|
|||
|
S S Musheer Hussain, Consultant Otolaryngologist Ninewells Hospital & Medical School, Dundee DD1 9SY, Peter Ross
Send response to journal:
|
Crayford(1 )has questioned Alho et al's(2) assertion that adults with documented recurrent episodes of streptococcal pharyngitis are less likely to have further throat infections if they had their tonsils removed. Tomkinson et al(3) have reinforced our belief that “the jury is still out” on the role of tonsillectomy for recurrent sore throats in modern medical practice, however, they have referred to their earlier paper(4) on the considerable morbidity in patients awaiting tonsillectomy. Freeland and Curley(5) on the other hand have shown that at least 20 percent of children “grow out” of their problem while waiting for tonsillectomy. We believe it is the time to seriously consider two interventions. Firstly, and in the medium term, a definitive multi-centred randomised control study is required. We disagree with statements that a placebo or no intervention group would be unethical(3). We look to the experiences of recently completed and soon to be published multi-centre RCT for the treatment of Bells Palsy (Scottish Bells Palsy Study) which had a true placebo group. This national multi-centred trial had the support of ENT departments all over Scotland and has produced a study which is a credit to the authors. In the case of tonsillectomy for recurrent tonsillitis, the recent controversy only strengthens the ethical argument for an RCT. If this study is done at a national multicentre level over 2 to 3 years, the large numbers enrolled will compensate for the inevitable dropout rate. Surely we owe it to our patients to ascertain the answer to this question? Secondly, and in the short term, we need to urgently revisit current guidelines. Bisset and Russell(6) have demonstrated that areas with less deprivation (social class 1 and 2) tended to have lower tonsillectomy rates than those with more deprivation (classes 3 and 4). In addition the study demonstrated surgical practice varied greatly among health regions. The subsequent SIGN guidelines(7) set a standard to overcome some of these issues and to which ENT Surgeons have practiced to. It is for this reason that the Dutch study(8) showing adenotonsillectomy to have no major clinical benefits over watchful waiting in children with mild symptoms of throat infections is probably not applicable to current UK practice. However these guidelines are now 8 years old, there is much that is new and relevant and the need to update is compelling. Peter Ross S S Musheer Hussain Department of Otolaryngology, Ninewells Hospital & Medical School, Dundee DD1 9SY Email musheer.hussain@nhs.net References 1. Crayford TJB. Time to stop doing tonsillectomies. BMJ 2007; 334: 1019 2. Alho OP, Koivunen P, Penna T, Teppo H, Koskela M, Luotonen J. Tonsillectomy versus watchful waiting in recurrent streptococcal pharyngitis in adults: randomised controlled trial. BMJ 2007;334:939-41 3. Tomkinson A et al. More advice to clinicians. BMJ 2007; 334: 1019 4. Fox R, Tomkinson A, Myers P. Morbidity in patients waiting for tonsillectomy in Cardiff: a cross sectional study. J Laryngol Otol 2006;120:214-8 5. Freeland AP, Curley JW The consequences of delay in tonsil surgery. Otolaryngol Clin North Am. 1987 May;20:405-8. 6. Bisset AF, Russell D. Grommets, tonsillectomies, and deprivation in Scotland. BMJ 1994;308:1129-32. 7. SIGN Publication No. 34 Management of Sore Throat and Indications for Tonsillectomy 1999 http://www.sign.ac.uk/guidelines/fulltext/34/ index.html Accessed 22.05.07 8. Van Staaij BK, van den Akker, EH, Rovers MM, Hordijk GJ, Hoes AW, Schilder AGM. Effectiveness of adenotonsillectomy in children with mild symptoms of throat infections or adenotonsillar hypertrophy: open, randomised controlled trial. BMJ 2004; 329:651 Competing interests: The senior author (SSMH) is a member of the Council of British Association of Otolaryngology Head & Neck Surgery, The Scottish Otolaryngological Society and The Tonsil and Adenoid subcommittee of the American Academy of Otolaryngology-Head & Neck Surgery. The views expressed are the authors own and do not represent those of the above organisations. |
|||
|
|
|||
|
Michael J Campbell, Professor of Medical Statistics ScHARR, University of Sheffield S1 4DA
Send response to journal:
|
Alho et al (1) have demonstrated an absolute risk reduction of Group A streptococcal pharyngitis of 21%. This is clear (although the editorial refers to it as an adjusted relative risk). I see no particular advantage in Flint’s suggestion (2) that the difference be expressed as a percentage of the control risk, (i.e one minus the relative risk or a relative risk reduction) since the absolute risk reduction is usually what readers need. The numbers are small, and if they had used the method recommended in CIA (3), the confidence interval would be found to be 5% to 37%, not 6% to 36% as given, thus implying even greater uncertainty about the true effect. In the longer text they described a clinically important effect as a reduction in risk of 25% and this they patently failed to achieve and so their conclusions are necessarily tentative. 1. Alho OP, Koivunen P, Penna T, Teppo H, Koskela M, Luotonen J. Tonsillectomy versus watchful waiting in recurrent streptococcal pharyngitis in adults: randomised controlled trial. BMJ 2007;334:939-41 2. Flint EF Methodological flaws and management misconceptions BMJ Rapid responses bmj.com 8 May 2007 3. Altman DG, Machin D, Bryant TN and Gardner MJ. Statistics with Confidence 2nd Ed BMJ Books 2000. Competing interests: None declared |
|||
|
|
|||
|
Thomas Sandeman, Retired radiation oncologist ex Peter MacCallum Cancer Institute,, Melbourne, Australia
Send response to journal:
|
The article by Alho et al (BMJ 2007;334:939-41) on a randomised controlled trial in 70 patients, of tonsillectomy versus watchful waiting in recurrent streptococcal pharyngitis in adults, contains no details of their age other than an average of about 27. It is assumed that no patient was under 18. The 70 involved were selected from 298 candidates in three and a half years so the problem is by no means rare. The accompanying editorial by Professor Paul Little (BMJ 2007;334:909)shows he has some reservations about the claimed benefits. He highlights the 13 days of post-operative pain and associated risks of tonsillectomy as well as the limited statistical reliability and lack of long term results. The alternatives he would like to present to sufferers for them to choose would be very difficult for the average patient to assess. Their choice would be affected by the number of attacks, how recent and how severe the last episode was. The underlying pathology of recurrent tonsillitis seems to be the continuing residence of bacteria in the crypts in the lymphoid tissue of the enlarged pharyngeal tonsil. Whether adenoidal tissue may be involved is not canvassed but in the days of guillotining the tonsils of childhood, the adenoidal area was usually scraped for good measure. Radiation had a dreadful reputation in years gone by when used for anything from shoe fitting to vain attempts to rescue surgical failures. Today it has established a role in many malignancies but its effectiveness in shrinking lymphoid tissue seems to have been forgotten. A Medline search has failed to find any reference to morbidity following low dose radiotherapy for benign local conditions such as keloid, pterygium or recalcitrant nose bleeding. Supraspinatus syndrome caused by lymphatic accumulation around inflamed shoulder tendons was often symptomatically relieved but has never been reported as harmful. There are no references in the past half century to its use in chronic tonsillar enlargement although I have memories of it being mentioned during my training in the specialty. It should not be more of a hazard to run a trial of low dose radiation to the region than many of the potent pharmaceutical remedies being researched these days. It is certainly less of a potential danger and a great deal more comfortable than tonsillectomy. The tissue is normally so sensitive that any mucosal reaction would be very mild. Confining the treatment to adults should allay any trepidation about its eventual carcinogenic risks, which at any rate would be extremely small. Whether strained radiotherapy facilities would be able to cope with an additional indication for therapy is another matter, but an old- fashioned "deep X-ray therapy" unit could provide an adequate dose to an area comparatively close to the surface. T.F.Sandeman, (sandeman@alphalink.com.au) Competing interests: None declared |
|||