Rapid Responses to:
|
|
Rapid Responses: Rapid Responses published:
-
![[Read Rapid Response]](/icons/misc/sectionDOWN.gif)
Bacterial load and response to treatment
- Nadaraja Bathirunathan (24 March 2006)
-
The sample
- Wenbin Liang (26 March 2006)
-
Dexamethasone in use?
- Jon Henrik Laake (30 March 2006)
-
Antibiotics in meningococcal disease - we need an urgent review of treatment criteria
- Amanda J Kvalsvig, Michael Baker and Graham Mills (6 April 2006)
-
Re: Antibiotics in meningococcal disease - we need an urgent review of treatment criteria
- Nadaraja Bathirunathan (8 April 2006)
-
British GPs practice excellent medicine
- Jim York (3 June 2006)
-
Administering parenteral penicillin in general practice
- Alfred J Meadows (3 June 2006)
-
Odds ratio uncertain
- Michael W Yung (5 June 2006)
-
Parenteral penicillin for children with meningococcal disease before hospital admission: case-control study
- D Graham Mackenzie, Diptendu Bhattacharyya, Chris Steer, Charles Saunders (5 June 2006)
-
Penicillin in Meningococcal Sepsis
- Philip C HOLLAND (6 June 2006)
-
Pre-hospital penicillin in meningococcal sepsis – good or bad?
- Anjum Gandhi (9 July 2006)
-
Re: Pre-hospital penicillin in meningococcal sepsis – good or bad?
- L S Lewis (10 July 2006)
|
|
|||
|
Nadaraja Bathirunathan, Visiting Professor Department of Transfusion Medicine,MGR, Medical University, Guindy,Chennai,600032
Send response to journal:
|
The most important factor which determines the outcome of an infection is the microbial load at the time of initiation of treatment. A look at Fig.2 makes the results obvious.The number of patients who had penicillin treatment and had a GMSPS score over 6 on admission is 36, while those who had no penicillin treatment is only 3. The severity of the disease and the outcome always correlated with the bacterial load. A lot of rubbish has been talked about the release of endotoxin during treatment and its adverse effect on the outcome. A living, metabolising,multiplying parasite is much more dangerous than a dead organism. Also when organisms are allowed to multiply to higher levels greater will be the amount of endotoxin. There cannot be any doubt that appropriate treament initiated as soon as possible following an infection leads to a superior outcome. I have worked mainly with Klebsiella pneumoniae infection in mice. In order to study the effect of any treatment the time interval and the nitial inoculum has to be chosen very carefully. If too soon after inoculaton all the animals survived with the treatment. If too long all the animals died. Competing interests: None declared |
|||
|
|
|||
|
Wenbin Liang, master candidate School of Public Health Curtin University of Technology
Send response to journal:
|
Dear Editor, Receiving an advance diagnosis of meningococcal disease was essential to be a subject in this study. [1] The diagnosis by a GP as well as penicillin allergy (Subjects with penicillin allergy were included in the study though they were not eligible to receive penicillin.) would only affect the chance of a patient to be exposed to penicillin, but it does not involve in the possible mechanism through which that penicillin affects the disease outcome. However whether a case was diagnosed in advance would be partly determined by the characteristics of the children and the characteristics of the bacteria, which may interact with the possible way that penicillin affects the disease outcome. However any kind of this possible interaction would not be observed after restricting the sample by whether receiving an advanced diagnosis. Given that the advance diagnosis may strongly affect the length of time between the onset of the disease (not the moment for the advanced diagnosis) and hospital admission, which would affect the outcome. It might be better to control the variation of the length of this period rather than using an advance diagnosis of meningococcal disease as selection criteria. Reference: 1. BMJ, doi:10.1136/bmj.38789.723611.55 (published 24 March 2006) Competing interests: None declared |
|||
|
|
|||
|
Jon Henrik Laake, Consultant, Cardiovascular Anaesthesia & Intensive Care Medicine National Hospital, 0027 OSLO, NORWAY
Send response to journal:
|
de Gans et al (1) found that in adults, early treatment with dexamethasone improves the outcome from acute bacterial meningitis. Were steroids used in these children? If so, did you control for this? Reference: 1. de Gans J et al. Dexamethasone in adults with bacterial meningitis. N Engl J Med. 2002 Nov 14;347(20):1549-56. Competing interests: None declared |
|||
|
|
|||
|
Amanda J Kvalsvig, Senior Research Fellow, Public Health Wellington School of Medicine and Health Sciences, Wellington, New Zealand, Michael Baker and Graham Mills
Send response to journal:
|
Harnden et al [1] have highlighted a very interesting problem in the early management of meningococcal disease: that while it seems intuitively obvious that pre-hospital antibiotic treatment should improve outcome, this is extremely difficult to demonstrate objectively. In theory, there are at least three possibilities for the effect of early treatment on outcome, two of which are discussed by Harnden et al. Firstly, they postulate that antibiotic administration is beneficial, but that confounding by severity makes this hard to detect. Secondly, they explore (and reject as unlikely) the possibility that antibiotic administration worsens the outcome by precipitating the release of endotoxin. We would like to suggest a third possibility: that in most cases pre- hospital antibiotic administration has little or no effect on mortality, because the traditional diagnostic criteria for meningococcal disease reflect a stage of pathogenesis at which the opportunity for treatment benefit has already passed. Shock, meningism and petechial rash are manifestations of the profound inflammatory response to endotoxin rather than direct effects of the meningococcus itself. [2] It may be argued that antibiotic treatment at this late stage of disease will be of little benefit to the patient, as it fails to address the principal mechanisms of morbidity and mortality operating at this point. We note with interest that the same group of authors has recently identified three early signs of meningococcal disease in children - leg pains, cold hands and feet and abnormal skin colour - which are reliably reported by parents and are often present at the first consultation with a general practitioner. [3] This is a very interesting finding. In New Zealand, as in many countries including the United Kingdom, guidelines for general practitioners emphasise late signs of disease. [4] We suggest that if treatment criteria for meningococcal disease were widened to include these early signs, there may be a greater likelihood of demonstrating – and more importantly, achieving - a beneficial effect from the use of pre-hospital antibiotics. Amanda Kvalsvig
1. Harnden, A., et al., Parenteral penicillin for children with meningococcal disease before hospital admission: case-control study. Bmj, 2006. 2. Kvalsvig, A.J. and D.J. Unsworth, The immunopathogenesis of meningococcal disease. J Clin Pathol, 2003. 56(6): p. 417-22. 3. Thompson, M.J., et al., Clinical recognition of meningococcal disease in children and adolescents. Lancet, 2006. 367(9508): p. 397-403. 4. Ministry of Health, Immunisation Handbook 2002. 2002, Wellington, New Zealand. Competing interests: None declared |
|||
|
|
|||
|
Nadaraja Bathirunathan, Visiting Professor Dept.of Transfusion Medicicine,MGR Medical University,Guindy,Chennai 600032,India
Send response to journal:
|
There cannot be any doubt that delay in the treatment of an infection is harmful and early diagnostic criteria would be immensely beneficial. Also it should be obvious that prehospital or indeed any antibiotic treatment is of little use when the patient has passed the point of no return with a heavy microbial load and the consequent tissue damage. Infection elicits an immediate innate immune response. Toll receptors (TLR} and other microbial component sensors in sentinel cells {macrophages, dendritic cells etc.} detect the microbe by its pathogen associated molecular patterns (PAMPS) and powerful innate immune mechanisms are brought into play to eliminate the infection. We have seen exciting developments in our understanding of innate immunity in the past few years. TLR4 detects endotoxin which is the PAMP of gram negative bacteria. To call the lipopolysaccharide a toxin is a misnomer with disastrous consequences in the past e.g. when the Lancet promoted a preparation of endotoxic antibodies called Centecor with a stupid editorial "Endotoxin bound and gagged.". Inflammation is a powerful and basically a protective innate immune response to infection. Blaming the innate immune response and the components of its mechanism as toxic is misleading. Our understanding of innate immunity is progressing fast and we may soon be able to use it effectively to control infection and its disastrous consequences. Competing interests: None declared |
|||
|
|
|||
|
Jim York, GP, Kingscliff, NSW, Australia Kingscliff Family Medical Centre, Kingscliff, NSW, 2487
Send response to journal:
|
Sir, Harnden et al have produced a piece of research which serves to confirm that a dangerous procedure, administering parenteral antibiotics to children and babies in the community, is done with great clinical acumen by general practitioners in the UK. It would be a pity if an over analysis of this outcome led to any confusion in the mind of a GP registrar faced by a child with a purpuric rash. I am forcibly reminded of the young lad I called in to see on the way to surgery some years ago, he had a rash and his mother was worried. He was sitting up in bed, bright as a button and I couldn't immediately see what she meant, being a little myopic. I got him to come downstairs where the light was brighter and as he skipped in front of me I wondered why I was bothering. The rash developed as I looked at it in the sunlight. He had his penicillin and a quick car ride to the local hospital, I was late for surgery. I called in to see him on the paediatric ward that evening on my way home. I left the ward in tears, profoundly moved; he was still as bright as a button and had remained so, but had a more or less total body purpuric rash; the cultures proved positive for meningococcal disease. I know it is anecdotal, but it is directly observed evidence. By the way, I was lucky, I had initially tried to persuade his mum to bring him to the surgery rather than call in on him. The good teachers I had, something, I don't know what, made me drop in on him. It is probably the best thing I have ever done. This excellent piece of research shows just how well British GPs deal with meningococcal disease. I respectfully ask our hospital and academic colleagues to label interesting arguments about inflammatory responses and release of endotoxins with a large health warning for the benefit of those working at the coal face. If you see a child with a purpuric rash, give penicillin. By all means stay with them and monitor them and give them fluids by any route, intraosseous included. But give them penicillin, please, please, give them penicillin, and without delay. Dr Jim York, NSW, Australia Competing interests: General practitioner involved in acute primary care |
|||
|
|
|||
|
Alfred J Meadows, GP Principal and trainer Chessel Practice, 4 Chessel Avenue, Southampton, SO19 4AA
Send response to journal:
|
Before anyone gives serious consideration to trying to conduct a randomised controlled trial of parenteral administration of penicillin prior to transfer to hospital, may I suggest some further thought is given to attempting to analyse the most important therapeutic intervention of all: a presumptive diagnosis of the possibility of meningococcal disease accompanied by a decision to refer for urgent secondary care. Some of the following insights may prove useful: 1 This decision is the result of a process that dichotomises a long continuous spectrum, from presentation of a child with a very mild URTI to severe systemic illness and circulatory collapse. 2. This dichotomisation is probably made substantially more effectively by some GP's than others. One cannot function as a GP referring every child with a cold, and one is truly dangerous ignoring clear signs of systemic illness. Somewhere on the line between the two must lie an ideal point for dichotomisation, where the pick-up rate of serious illness at an early but diagnosable stage approaches 100%, without compromising one's ability to deal with the daily multitude of non-serious illness. It is likely that some GP's hit this point much more nearly than others. 3. Use of non-specific markers of systemic illness such as pulse rate, respiratory rate and capillary return, not to mention the notoriously difficult-to-quantify 'gut instinct', will normally be confined to the middle of the spectrum. It is quite unrealistic to expect GP's to be recording these observations in every child with a minor cold, nor is it relevant to a GP when the decision to refer is made on first sight of a child in a critical condition. But my own experience also would suggest that not enough use of thes markers is made around the point of dichotomisation, and such use (whether recorded or not, that is another issue) is likely to be a useful marker of the GP who can most effectively make diagnoses of serious illness at an early stage. I would also suggest that the manner of their use (whether algorithmic, Bayesian-like, or simply as pattern recognition) is probably not so important as the fact that they are used. 4 While the decision to make explicit a suspicion of meningococcal disease should (currently) be tightly linked to the decision to give parenteral penicillin, the latter decision is not necessarily made at the same point as the decision 'this is potentially serious illness requiring referral to hospital to secure an optimum outcome'. I would suggest that it would not be surprising if GP's referring to hospital urgently, but not giving penicillin, did not sometimes have the possibilty of meningococcal disease somewhere in mind, but preferred simply to disclose a diagnosis of 'potentially serious systemic illness', for disclosing the former would push them into giving penicillin which their own process of dichotomisation did not recommend to them. Intra-muscular penicillin injection is generally poorly recieved by a child who is not yet in a critical state, and the collateral panic reaction of the mother is not negligeable. These considerations are presently rightly ignored where the possibility of meningococcal disease becomes high enough, but where it is extremely speculative they are not inconsequential considerations. I would suggest that these dimensions to real life diagnosis and treatment decisions should be amenable to prospective quantitative analysis of some description, which may go some way to explaining the likely residual confounding with severity of disease at the point of decision to give (or not to give) penicillin, which I, as I suspect many others, would hold to be the most likely explanation for the results of this most interesting analysis. Competing interests: None declared |
|||
|
|
|||
|
Michael W Yung, paediatric intensivist Women's and Children's Hosptial , Adelaide, South Australia 5006
Send response to journal:
|
Dear Sir/Madam Harnden et al set out to address the important issue of the efficacy and safety of pre-hospital parenteral penicillin for meningococcal disease. They found an odds ratio of 5.96 for death when penicillin was given, a greater effect than severity assessed by the general practitioner. They offer two explanations: severity and endotoxin release, favouring the former. However the six cases for whom penicillin administration could not be determined are important. Including these in a "worst case scenario" analysis: assuming both who died had not received penicillin, and the four who survived received penicillin, the unadjusted odds ratio becomes 2.84, 95% confidence interval 0.88 to 12.0. Although 2.84 is still impressive, this should have been mentioned in the paper. Michael Yung Competing interests: None declared |
|||
|
|
|||
|
D Graham Mackenzie, Specialist Registrar in Public Health Public Health Department, Cameron House, Cameron Bridge, Windygates, Leven, KY8 5RG, Diptendu Bhattacharyya, Chris Steer, Charles Saunders
Send response to journal:
|
EDITOR: Harnden et al’s study1 raises important questions about the role of pre-hospital parenteral penicillin in children with meningococcal disease (MCD). Their paper and accompanying statistical comment2 have shown the potential importance of excluding patients who would never have been considered for treatment. However their study design is likely to have excluded a group of children with suspected MCD who were given pre-hospital parenteral penicillin.
Public Health guidelines recommend that General Practitioners (GPs) administer pre-hospital parenteral penicillin to patients with suspected MCD, though the guidelines do not make specific recommendations about clinical criteria on which GPs should base their decision3. Recent research shows that clinical features other than haemorrhagic rash can be important in identifying MCD4, reinforcing how difficult clinical decisions can be when treating an acutely unwell child. In contrast, the formal case definition for MCD is decided after consultation between hospital clinician, microbiologist and Consultant in Public Health Medicine, usually hours or days after hospital admission3. According to the methods section in an earlier paper5 this appears to have been the starting point for Harnden et al’s study1. Not all children treated with pre-hospital antibiotics for suspected MCD would meet the formal case definition, partly because of incomplete application of microbiological tests. For example, when we conducted an audit of cases of MCD notified to Public Health in our region between 2000 and 2001, only 18/36 cases (50%) had undergone adequate microbiological testing to confirm MCD. A randomised controlled trial of pre-hospital parenteral penicillin in children with suspected meningococcal disease would be a useful next step, and would address concerns about confounding. However, as demonstrated above, the inclusion criteria for such a study would need to be considered carefully for the findings to be relevant to GPs making decisions about acutely ill children in the community. D Graham Mackenzie, Specialist Registrar in Public Health, Cameron House, Cameron Bridge, Windygates, Leven, KY8 5RG, UK Diptendu N Bhattacharyya, Infectious Disease Consultant, Hayfield House, Hayfield Road, Kirkcaldy, KY2 5AH Chris R Steer, Consultant Paediatrician, Hayfield House, Hayfield Road, Kirkcaldy, KY2 5AH Charles JP Saunders, Consultant in Public Health Medicine, Cameron House, Cameron Bridge, Windygates, Leven, KY8 5RG References 1. Harnden A, Ninis N, Thompson M et al. Parenteral penicillin for children with meningococcal disease before hospital admission: case-control study. BMJ 2006; 332: 1295-8. 2. Perera R. Commentary: Statistics and death from meningococcal disease in children. BMJ 2006; 332: 1297-8. 3. Public Health Laboratory Service Forum. Guidelines for public health management of meningococcal disease in the UK. Communicable Disease and Public Health 2002; 5: 187-204. 4. Thompson MJ, Ninis N, Perera R et al. Clinical recognition of meningococcal disease in children and adolescents. Lancet 2006; 367: 397-403. 5. Ninis N, Phillips C, Bailey L et al. The role of healthcare delivery in the outcome of meningococcal disease in children: case-control study of fatal and non-fatal cases. BMJ 2005; 330: 1475-8. Competing interests: None declared |
|||
|
|
|||
|
Philip C HOLLAND, consultant paediatrician Leeds General Infirmary LS2 9NS
Send response to journal:
|
Sir The publication by Harden et al (1) shows the importance of follow up studies on simple clinical interventions such as intramuscular penicillin in suspected meningococcal sepsis. The importance of giving penicillin has been stressed to primary care so much that in A/E departments it is not infrequent to see children who have been given IM penicillin on relatively flimsy criteria, due to the fear engendered if meningococcal sepsis is missed. Publication of studies such as this, where the results are surprising, leads to a natural questioning of the findings in an attempt to support the accepted practice. However the higher mortality rate with odds ratio of 7.4, along with higher complication rates in survivors who received IM penicillin, suggest that we should immediately stop giving penicillin IM at suspected diagnosis. As the authors suggest, if there is still doubt, a clinical trial should be designed. However the continuation of IM penicillin when meningococcal sepsis is suspected would be wrong. Such findings should make us question the original hypothesis that the rate of bacterial growth is attenuated if penicillin is given immediately thus reducing the bacterial load and that of lipopolysaccaride (LPS). One of the principal characteristics of meningococcal septicaemia with a rash is the onset of shock attributed to capillary leak. All who deal with such patients will recognise the signs of extremely cold extremities and reduced capillary return. The difference between giving penicillin in A/E departments, compared with it being given at home, is that along with the administration of penicillin, it is usual for resuscitation to be initiated with fluid boluses, minimising the severity of shock. The early treatment of shock may be as important as the giving of antibiotics. The authors looked at the time from being seen at home to arrival at hospital, along with the use or otherwise of penicillin. They comment that there was no increase in severity up to 3 hours following administration of penicillin but that there was also no significant deterioration in those not given penicillin, leading the authors to suggest ‘that a short delay without penicillin is not deleterious’. In conclusion the authors propose that a clinical trial is necessary but do not comment on whether penicillin should or should not be given until a clinical trial resolves the issue. The diagnosis of meningococcal sepsis should result in the immediate transfer of the child to hospital; this study suggests the present practise of giving IM penicillin may not be correct. By stopping the administration of IM penicillin, the death rate may not only be reduced but also the number of children receiving IM penicillin unnecessarily will be lowered. From the data provided it seems unlikely that this will make the situation worse. Dr. Philip Holland Consultant Paediatrician Leeds General Infirmary References 1. Parenteral penicillin for children with meningococcal disease before hospital admission: case controlled study. Anthony Harden, Nelly Ninis, Mathew Thompson, Rafael Perera, Michael Levin, David Mant, Richard Mayon- White. BMJ 2006:332;1295-8. Competing interests: None declared |
|||
|
|
|||
|
Anjum Gandhi, Consultant Paediatrician Royal Glamorgan Hospital
Send response to journal:
|
The paper by Anthony Harnden, Nelly Ninis, Matthew Thompson et al poses a very difficult clinical question. If their interpretations are correct a fundamental change in the current practice may be required. We retrospectively reviewed all children admitted to the paediatric intensive care unit (PICU) at the University hospital of Wales with a diagnosis of meningococcal sepsis over a 5 year period from March 1999 to February 2004. The PICU caters for all children requiring intensive care in south and west Wales. A total of 114 cases were identified over this period although case notes were available for only 101 of these. The diagnosis was microbiologically confirmed in 85 cases by blood culture and/or PCR. The rest were clinically classified as probable menigococcal sepsis. All of these children were extremely sick and required aggressive fluid resuscitation. The mean volume of fluid for resuscitation was 142 ml/kg in the first 24 hours. Ninety three of these cases required ventilation and 76 required a combination of inotropes. There were however only 4 deaths all in children under 2 years of age. We also looked at the number of children who received prehospital parenteral penicillin. This was given to 84 children while 17 did not receive it. All 4 deaths occurred in the cohort who were given penicillin prior to hospitalisation. There were no deaths in the 17 children who did not receive penicillin. While the numbers are small and not statistically significant, our observations do seem to support those made by the authors. More over the argument that penicillin was given to the sicker of the patients does not apply to this population as all children were unwell enough to require intensive care. Certainly there is a case for carrying out a large randomised trial to resolve this issue. Competing interests: None declared |
|||
|
|
|||
|
L S Lewis, GP Surgery, Newport, Pembrokeshire, SA42 0TJ
Send response to journal:
|
This BMJ paper shows that more serious disease ( A ) is associated with early penicillin ( B ) - but being an observational study, it is incapable of answering the imponderable 'does early Penicillin make matters better or worse ?? ' Either: A causes B seriously ill meningococcaemic cause GPs to notice and give Penicillin.. OR : B causes A: giving Penicillin ( bactericidal by cell wall disruption ) causes massive release of endotoxin - and makes shock and death MORE likely .. OR : C causes A and B ?? Such questions require RCT, but it would be difficult indeed to gain ethical approval of a randomised controlled trial of early-Penicillin vs. no Penicillin !! In sub-saharan Africa regular massive epidemics of meningocococcal disease have been extensively studied and reported, since the days of LaPeyssonie. 30 years ago, in Zaria, Northern Nigeria, we asked the ethically testable question of whether Cidal antibiotics were better or worse than Static antibiotics (assuming the endotoxin-release hypothesis). The bacteriostatic, cheap and available-in-bucketloads Oral Chloramphenicol proved to be at least as good as parenteral penicillin, and had considerable ancillary advantages in that environment -- but even with 50 patients in the randomisation, no superiority over penicillin was noted.. suggesting that any real difference must be small. Thus we may continue with some evidence-based justification to give pre-hospital penicillin .. but maybe we should consider also giving some form of anti- toxin ? see : http://adc.bmjjournals.com/cgi/content/abstract/54/1/44 Competing interests: None declared |
|||