Monitoring treatment with aminoglycoside antibiotics.

14 downloads 0 Views 649KB Size Report
3 Working Party of the British Society for Antimicrobial Chemo- therapy. Antibiotic treatment of streptococcal and staphy- lococcal endocarditis. Lancet 1985 ...
EDITOR,-Renewed interest in the activities and professional training of counsellors in general practice is welcome.'2 In the 1990s there will be a concerted move by professional bodies such as the British Association for Counselling and the British Psychological Society to regulate counselling and ensure that those who practise are both competent and qualified to do so. Contrary to some popular wisdom, bad counselling can be damaging to patients. The question of who counsels patients in medical settings needs to be addressed. There is some difference between being a counsellor and having counselling skills. Almost all health care professionals with primary training in medicine, nursing, physiotherapy, or other allied professions counsel people during their work: if the broadest definition of the term is used an episode of counselling occurs in every medical consultation. Health care professionals constantly give patients information, clarify treatment options, and help people to adjust to new, and sometimes unwelcome, circumstances. Specialist counsellors, on the other hand, have usually had advanced training in counselling, psychotherapy, or family therapy, and some may be professionally trained in other disciplines such as medicine, clinical psychology, social work, or nursing. Although specialist training is not a requirement to practise as a counsellor, a professional may occasionally refer a patient to a specialist counsellor in the same way that a doctor may refer medical problems to a specialist colleague.3 Clinical and counselling psychologists do not deal only with mental illness. Our training (at MSc level) emphasises the need for counselling skills for people with many complex medical problems, including HIV infection, management of diabetes, infertility, problems after disasters, pain control after surgery, and neurological problems. In addition, psychologists conduct collaborative research with their colleagues in general practice, provide health education to patients, and conduct psychosocial assessments and may also provide a consultation and liaison service. Though there may currently be an undersupply of qualified and accredited counsellors, this should not deter doctors from referring patients to them when resources permit. ROBERT BOR MARY WATTS

City University, London EC I V OHB 1 Sibbald B, Addington-Hall J, Brenneman D, Freeling P. Counsellors in English and Welsh general practices: their nature and distribution. BMJ 1993;306:29-33. (2 January.) 2 Pringle M, Laverty J. A counsellor in every practice? BMJ 1993;306:2-3. (2 January.) 3 Bor R, Miller R, Goldman E. Theory and practice of systemic HIV counselling London: Cassell, 1992.

EDITOR,-The editorial and paper on counsellors and counselling in primary care are misleading. 2 Mike Pringle and John Laverty's editorial seems to muddle Balint's work in developing the psychotherapeutic skills of general practitioners to aid their work in consultation and the role of a counsellor or psychotherapist working alongside the general practitioner as part of the primary care team.' Pringle and Laverty also state that counsellors who work in primary care should concentrate on non-directive counselling. That is only one form of ctunselling or psychotherapeutic approach. Many others are now being used effectively-for example, brief therapy models, behavioural change techniques, and gestalt-as the counsellors match the patients' needs to therapy and not vice versa. Pringle and Laverty are apparently unaware of the many family health services authorities that have set up approval procedures for employing counsellors in primary care. The imminent publication of guidelines on employing counsellors in primary care prepared by a working party of the

BMJ VOLUME 306

6 FEBRUARY 1993

British Association for Counselling will greatly help those family health services authorities and general practitioners who have yet to develop such guidelines or do not understand how to assess the qualifications and competence of counsellors they wish to employ and work with. Bonnie Sibbald and colleagues conclude that, because general practitioners did not know the qualifications of the counsellors they employed, those counsellors were probably unqualified.2 As most general practitioners would be hard put to describe accurately the qualifications of the nurses they employ, this, I believe, is an unwarranted conclusion. I hope that Sibbald and colleagues will go back to the practices in their survey to examine this important issue and to explain the uneven distribution of counsellors working in general practice. Research by the Counselling in Primary Care Trust has shown that in an admittedly small randomised sample of counsellors working in primary care 24 of 26 had had three or more years of training. Nearly all were undertaking their own personal therapy or had done so in the past, and all were appropriately supervised. GRAHAM CURTIS JENKINS

Counselling in Primary Care Trust, Staines TW18 4DG 1 Pringle M, Laverty J. A counsellor in every practice? BMJ 1993;306:2-3. (2 January.) 2 Sibbald B, Addington-Hall J, Brenneman D, Freeling P. Counsellors in English and Welsh general practices: their nature and distribution. BMJ 1993;306:29-33. (2 January.)

Monitoring treatment with aminoglycoside antibiotics EDIToR,-Guidelines for monitoring aminoglycoside antibiotics are essential, but those provided by J K Aronson and D J M Reynolds contain many misleading statements. ' Kanamycin is included in the guidelines, but this has long been superseded by other aminoglycosides.2 The article fails to emphasise the two most important tenets of safe prescribing of aminoglycosides-namely, that they should be used only if clinically justified and that they should be stopped as soon as the patient's condition permits. We are concerned that those who prescribe aminoglycosides will be distracted by the minutiae of calculations concerning half lives and body loads from the real problem of deciding whether the agents are indicated at all. Infections requiring prolonged courses (more than seven days,2 or in our opinion five days, not one to two weeks as the authors suggest) are uncommon, are unlikely to be encountered by most clinicians, and require expert supervision. Gram negative endocarditis is exceptionally rare and certainly does not include that due to Streptococcus faecalis, which is a Gram positive organism correctly called Enterococcus faecalis. The case histories quoted highlight the common errors in prescribing aminoglycosides. Firstly, there is ignorance of their synergistic role in the treatment of viridans streptococcal and enterococcal endocarditis. A 120 mg loading dose of gentamicin followed by 80 mg eight hourly is excessive for such infections. National recommendations are clear: 60-80 mg of gentamicin twice a day,2'3and peak concentrations should be between 3 and 5 p.g/ml,' not 9 ,ug/ml as suggested in the first case history. Secondly, adjustment of the aminoglycoside regimen during treatment often leads to the course being prolonged unnecessarily and diverts attention from the selection of suitable alternatives (cases 2 and 3). Finally, it is unrealistic to recommend (desirable though it might be in theory) repeated estimates of auditory and vestibular function and, likewise, to believe that busy house officers will take samples

for measurement of peak concentrations precisely 15 minutes after the end of an infusion and one hour after intramuscular administration as the authors advise. We, and others,2 recommend that a sample should be obtained after one hour for both routes and are pleased to receive one at all, for it immediately involves the medical microbiologist in the care of the patient, and close cooperation is important.4 W R GRANSDEN SUSANNAH J EYKYN

Department of Microbiology, St Thomas's Hospital, London SE 1 7EH 1 Aronson JK, Reynolds DJM. Aminoglycoside antibiotics. BMJ 1992;305:1421-4. (5 December.) 2 BMA and Royal Pharmaceutical Society of Great Britain. British nationalformulary, number 24. London: BMA, RPSGB, 1992. 3 Working Party of the British Society for Antimicrobial Chemotherapy. Antibiotic treatment of streptococcal and staphylococcal endocarditis. Lancet 1985;ii:815-7. 4 Cooke EM. Aminoglycoside toxicity. Journal of the Medical Defence Union 1988:23.

EDITOR,-J K Aronson and M Hardman do not do justice to the range of antimiocrobial drugs for which monitoring of serum concentrations is necessary,' and Aronson and D J M Reynolds include several statements that we find unacceptable in their review of monitoring of aminoglycoside antibiotics.2 Serum estimations are of proved value for all aminoglycosides and for the closely related agent streptomycin. Serum concentrations should also be assayed in all patients receiving vancomycin, flucytosine, or cycloserine; neonates and perhaps those under 4 years old receiving chloramphenicol; those with severe sepsis receiving teicoplanin'; and patients receiving prophylactic itraconazole.4 With all these agents serum concentrations have been putatively related to toxicity or clinical efficacy or there are inconsistencies between the dosage and serum concentrations. In selected patients assays of penicillin, co-trimoxazole, flucloxacillin, ciprofloxacin, metronidazole, and rifampicin may also be of clinical value. Serum aminoglycoside concentrations should be measured one hour after the dose, not 15 minutes after as stated.56 One hour after the dose the phase of rapid equilibration between the blood and tissues is generally complete and the elimination phase becomes dominant, and thus the observed concentration more accurately reflects the concentration in tissue, where the infection is most likely to be. The serum concentration is falling rapidly at 15 minutes, and small changes in the timing of the sample will greatly affect the observed result, which could result in day to day inconsistencies and unwarranted changes in dosage. We suspect that the high concentration in the first clinical vignette was due to sampling too early. There is little evidence, however, to correlate serum concentrations after the dose with toxicity, and indeed experimental nephrotoxicity caused by gentamicin is more severe when the total daily dose is divided than when it is given by a single bolus, when concentrations after the dose are higher.7 Two other inaccuracies are perpetuated. One is that a loading dose is required in patients with apparently normal renal function, and the other is that the "standard" dose is 80 mg intravenously eight hourly. We disagree strongly with any calculation of dosage being based on plasma half life derived from the difference between only two observations, especially if the first of them is taken early in the distribution and not in the elimination phase. A half life calculated in this way would be falsely short. It is also important to realise that alternative and less toxic antimicrobials were probably available for the patient who had sepsis due to Escherichia coli and renal impairment. Finally, aminoglycoside concentrations are

391

measured in mg/I, not ,ug/l or ,ug/ml as given in Reynolds and Aronson's article on making the most of measurements of plasma drug concentrations.8 ALASDAIR P MAcGOWAN DAVID S REEVES Regional Antimicrobial Reference Laboratory, Department ofMedical Microbiology, Southmead Health Services NHS Trust, Bristol BS IO 5NB

1 Aronson JK, Hardman M. Measuring plasma drug concentrations. BMJ 1992;305:1078-80. (31 October.) 2 Aronson JK, Reynolds DJM. Aminoglycoside antibiotics. BMJ 1992;305:1421-4. (5 December.) 3 MacGowan AP, McMullin CM, White LO, Reeves DS, Davis E, Speller DCE. Antimicrob Chemother 1992;30:399-402. 4 British Society for Antimicrobial Chemotherapy Working Party. Laboratory monitoring of antifungal chemotherapy. Lancet 199 1;337: 1577-9. 5 Reeves DS. Treatment of bacterial endocarditis. Lancet 1985;ii: 1420-1. 6 BMA and Royal Pharmaceutical Society of Great Britain. British national formulary, number 24. London: BMA, RPSGB, 1992: 218. 7 Mattie H, Craig WA, Pecheve JC. Determinants of efficacy and toxicity of aminoglycosides. J Antimicrob Chemother 1989;24: 281-93. 8 Reynolds DJM, Aronson JK. Making the most of plasma drug concentration measurements. BMJ 1993;306:48-51.

(2 January.)

EDITOR,-We believe that some of the guidelines on the use of gentamicin in J K Aronson and D J M Reynolds's article are potentially misleading.' The use of a "standard" maintenance dose will inevitably cause either subtherapeutic or toxic concentrations in an appreciable proportion of patients. The dose should be calculated on an individual basis, with age, body weight, and renal function being used as a guide. In particular, the frequency of the dose should be reduced in patients with renal impairment, including elderly people. An audit of prescribing of gentamicin in Leicester showed that this approach considerably increased the number of gentamicin concentration that were within the therapeutic range when measured after one to two days' treatment. MVICKERS A VICCA RA SWANN Public Health Laboratory Service, Leicester Royal Infirmary, Leicester LEl 5WW 1 Aronson JK, Reynolds DJM. Aminoglycoside antibiotics. BMJ 1992;305:1421-4. (5 December.)

EDrrOR,-J K Aronson and D J M Reynolds's article on monitoring treatment with aminoglycoside antibiotics does not adequately address optimal dosing with these agents or the timing of assays, nor does it specify which patients need monitoring.' Serum aminoglycoside concentrations should be assayed in all patients with renal impairment, patients receiving other nephrotoxic agents, elderly patients, and patients who have recently received aminoglycosides, since these groups are at increased risk of toxic side effects. The indications thus include most patients treated with these drugs except those receiving only one to three doses-predominantly those in whom aminoglycosides are prescribed for antimicrobial chemoprophylaxis. Aminoglycosides are often given in insufficiently high doses. In Nottingham we encourage the use of a nomogram that takes account of age, sex, renal function, and body weight, factors that influence the body's handling of these agents.23 An alternative policy advocates a single large daily dose of aminoglycosides, such as 240 mg instead of 80 mg eight hourly, since the antibacterial activity is related to the dose and these agents exert a relatively long antibiotic effect, at least in vitro. Such regimens reduce the cost of administration and may be less toxic.4 Larger doses of gentamicin given less often result in higher peak concentrations

392

of 10-15 mg/l and low trough concentrations. A trough concentration of 1-5 mg/l in a patient receiving 300 mg daily does not, however, have the same significance in terms of predicting toxicity as a trough of 1-5 mg/l in a patient receiving 100 mg eight hourly.5 With the single daily dose it may be more appropriate to assay the concentration eight or 12 hours after administration of the dose or to accept that a lower trough (predose) concentration may be more appropriate. Blood samples for assay of the aminoglycoside concentration should be taken at around the third or fourth dose, when the steady state has been reached. In critically ill patients or those with changing renal function, assays may be necessary earlier than this so that the dosage can be adjusted to avoid tissue accumulation. Subsequently, concentrations should be measured every 48-72 hours, depending on the clinical context, previous results, and likely duration of treatment. H HUMPHREYS D GREENWOOD

Department of Microbiology and Public Health Laboratory, University Hospital, Queen's Medical Centre, Nottingham NG7 2UH 1 Aronson JK, Reynolds DJM. Aminoglycoside antibiotics. BMJ 1992;305:1421-4. (5 December.) 2 Mawer GE, Ahmad R, Dobbs SM, McGough JG, Lucas SB, Tooth JA. Prescribing aids for gentamicin. BrJ Clin Pharmacol 1974;1:45-50. 3 Slack RCB. Antibiotic assay. In: Greenwood D, ed. Antimicrobial chemotherapy. 2nd ed. Oxford: Oxford Medical Publications, 1989:101-8. 4 Levison ME. New dosing regimens for aminoglycoside antibiotics. Ann Intern Med 1992;117:693-4. 5 Humphreys H, Reeves D. Aminoglycoside assays. Reviews in Medical Microbiology 1991;2: 13-21.

EDITOR,-We are concerned that streptomycin is not mentioned in the article on monitoring treatment with aminoglycoside antibiotics.' Although the use of streptomycin to treat tuberculosis has declined in Britain, the drug remains an essential component of treatment for some patients. These patients are generally those infected with drug resistant Mycobacterium tuberculosis or patients with severe disease that has been partially treated (due to poor compliance or intermittent treatment overseas) in whom the results of tests of bacterial susceptibility are not yet available. Drug resistant M tuberculosis is now a major problem throughout the world, and in response to this the production of streptomycin has been increased.2 We believe that good clinical practice should include monitoring the concentrations of streptomycin during treatment, both to avoid potentially toxic concentrations and to ensure adequate dosage. Unfortunately, the assay kit for this drug previously produced by Abbott Laboratories has recently been withdrawn. We are currently seeking alternative rapid methods, but the lack of a suitable rapid assay must not be taken as an indication that therapeutic monitoring is unnecessary. C D SHELDON H GAYA N C BARNES

London Chest Hospital, London E2 9jX 1 Aronson JK, Reynolds DJM. Aminoglycoside antibiotics. BMJ 1992;305:1421-4. (5 December.) 2 Pfizer boosts streptomycin production. Scrip-World PharmaceuticalNews 1992;1764:1 1.

AuTHoRs' REPLY,-W R Gransden and Susannah Eykyn say that we failed to emphasise the need for safe prescribing of aminoglycosides, and they mention two truisms applicable to all drugs. In the articles in the ABC series we have addressed the ways in which drug treatment should be monitored after it has been started. It was beyond our brief to

discuss whether or not treatment is required and for how long. Gransden and Eykyn chide us for mentioning kanamycin, which has been superseded in Britain by newer drugs. It would have been parochial of us not to have mentioned it as it is still used in many other countries. Alasdair P McGowan and David S Reeves discuss measurement of serum concentrations of other anti-infective drugs in special circumstances. We agree that there is a case for this for some of the drugs they mention, particularly to avoid toxicity in renal failure or, in the case of itraconazole, to ensure that adequate dosages are used. We apologise for having used old fashioned terminology in misnaming Enterococcus faecalis as Streptococcus faecalis, but the apparent redesignation of the organism as Gram negative was not intentional and arose because of the eradication of one too many "organisms" from the text, which should read: "Important indications for this category include infections due to Gram negative organisms; organisms, such as Enterococcus faecalis, causing infective endocarditis; osteomyelitis; and infections ofvascular grafts." We are surprised that Gransden and Eykyn consider our back of an envelope calculations to be minutiae in aminoglycoside treatment. Such simple calculations are essential in using aminoglycosides safely and effectively. We agree with MacGowan and Reeves that the half life may be inaccurate if it is calcuated from only two values; but perhaps that is why they have never seen a patient with normal renal function in whom the half life is as long as 7-6 hours,' since in clinical practice two measurements are all one has to go on. In these circumstances one makes the best of limited information. We are disturbed that atiometry is regarded as unrealistic. The number of occasions when prolonged treatment with aminoglycosides is required is small, but in these cases the risks of ototoxicity are high and the costs of failing to monitor for this disabling condition can be considerable. We agree with the recommendation that blood samples should be taken one hour after an intravenous bolus dose. Our comments about sampling 15 minutes after the end of a dose applied to intravenous infusions. Nowhere did we make any fixed dosage recommendations, and we certainly did not write that "standard" doses should be used. We described what we believe usually happens. Thus we stated that the initial maintenance dose is "usually . . . about 80 mg three times a day"; this is so, and there is leeway for altering dosages in patients of low body weight. Loading doses are used, and, because of the large variability in half life, even when renal function is thought to be normal we think that that is sensible. Although there is now some interest in the use of large single daily doses of aminoglycosides, it is far too soon to recommend such regimens with confidence. The review quoted by H Humphreys and D Greenwood makes clear that this type of regimen has mostly been studied in infections that generally respond well to treatment and not in patients with complicated illnesses (in particular, renal disease) andthat its use is currently limited.2 Some of the correspondents have misunderstood the point of the case histories in the articles in this series: far from being examples of perfect treatment, they are examples of problems that commonly arise. For example, we clearly stated in the text that serum concentrations should be measured 24 hours after the start of treatment; the case to which MacGowan and Reeves refer was one in which the clinician measured the serum concentration after 48 hours. We also made it clear that both dose and frequency should be adjusted as dictated by serum concentrations. Finally, we emphasise that close cooperation

BMJ VOLUME 306

6 FEBRUARY 1993