BRITISH MEDICAL JOURNAL VOLUME 294 9 MAY 1987 1181

probably because its reintroduction becomes ever more aureus.' Broad spectrum such as , Br Med J (Clin Res Ed): first published as 10.1136/bmj.294.6581.1181 on 9 May 1987. Downloaded from unlikely. Nevertheless, the ethical issues raised by the , , and do not always Amnesty report concern doctors throughout the world and produce adequate activity against pseudomonas in res- not merely those in the United States. Doctors have an piratory secretions,g" and they are best used with an important part to play in abolishing what is cruel, inhuman, aminoglycoside in patients with serious infection. Cepha- and degrading punishment. Firstly, they must articulate lexin is active against Str pneumoniae in sputum, but the and implement ethical codes that unambiguously prohibit concentrations are not likely to be inhibitory for Haemophilus doctors participating in executions, and, secondly, they must influenzae. " Injectable such as , widen the discussion to include the-broader ethical issues of , and achieve higher serum concentra- the death penalty. tions, and peak concentrations in sputum are at least four ANTHONY W CLARE times higher than those resulting from oral agents.'013 Professor ofPsychological Medicine, Erythromycin is widely used in treating respiratory in- St Bartholomew's Hospital Medical College, London ECIA 7BE fectionsbecauseofitsactivityagainstmycoplasma, legionella, and various other bacteria-pneumococci and branhamella I Curran WJ, Casscells W. The ethics ofmedical participation in capital punishment by intravenous are very susceptible, but concentrations needed to inhibit drug injection.N EnglJMed 1980;302:226-30. H influenzae are higher. Erythromycin produces good but 2 Amnesty International. UnitedStates ofAmenca: the death penalty. London: Amnesty International Publications, 1987. variable sputum concentrations when given intravenously'4 3 Casscells W, Curran WJ. Doctors, the death penalty and lethal injections. N EngI 7 Med 1982;307:1532-3. though much lower concentrations (which may be sub- 4 Amnesty International. The death penalty in the United States: an issue for health professionals. inhibitory for haemophilus) after a 500 mg oral dose.'5 Much London: Amnesty Intemnational Publications, 1987:13. (Quotation from Jomal da Associacao greater activity is detected in lung tissue after oral and Medica Brasileira 1985 November: 12.) intravenous erythromycin. Clindamycin and rifampicin readily attain good sputum activity against Staph aureus,"' and rifampicin is a suitable adjunct to such as or in staphylococcal pneumonia. Anaerobes (implicatedinaspirationpneumonia)areinhibited Penetration of antibiotics into by bronchial concentrations of metronidazole after 400 mg oral doses." the respiratory tract Tetracycline concentrations in bronchial secretions are inhibitory to most strains ofStrpneumoniae, though activity Few antibiotics penetrate well into broncial secretions, and against H influenzae is not always adequate.'7 Sputum yet most respiratory- infections respond to treatment. The concentrations and clinical results may correlate poor penetration seems to matter only in patients with poorly: Maesen and colleagues found that haemophilus chronic suppurative airways disease (bronchitis, bronchi- strains with minimum inhibitory concentrations of doxy- ectasis, and cystic fibrosis) and in those infected with less cycline exceeding 2 mg/l were rarely eradicated by con- sensitive organisms and may then contribute to recurrent ventional doses of doxycycline in patients with chronic sepsis. In these patients theinitial response to antibiotics may bronchitis, although almost two thirds ofisolates with lower be good because the drugs penetrate the mucosa better than minimum inhibitory concentrations responded, despite a the secretions, but subtherapeutic concentrations in--the mean peak sputum concentration of only 0-3 mg/1.1s Con- mucus may lead to relapse. Furthermore, penetration may siderably higher doxycycline concentrations are, however, http://www.bmj.com/ become worse as tissue damage progresses.' For these reached in the bronchial wall and lung tissue.'9 reasons doctors, and particularly those looking after patients Treatment of respiratory infections caused by Gram with chronic suppurative lung diseases, need to know negative organisms with gentamicin is most likely to succeed something about the penetration of antibiotics into the ifpeak serum concentrations exceed 8 mg/I.2" Studies in dogs respiratory tract. have shown that peak concentrations in bronchial mucus are The penetration of (3 lactam antibiotics is modest-peak about one quarter of those in the serum and that even high sputum concentrations ofpenicillins are only 5-20% ofthose doses of gentamicin may fail to reach therapeutic concentra- in serum. Even 1 g of oral will not always attain tions against Pseudomonus aeruginosa in respiratory secre- on 29 September 2021 by guest. Protected copyright. inhibitory concentrations for haemophilus, although activity tions.21 Adequate concentrations may not be readily achieved against more sensitive bacteria such as Streptococcus pneu- in the elderly or in patients with renal impairment, when moniae is readily achieved-2 Parenteral ampicillin gives dosage must be carefully controlled to avoid toxicity: com- higher concentrations in both the serum and sputum. bination with a broad spectrum may produce Amoxycillin is more completely absorbed from the gastro- synergy, but newer antipseudomonal agents such as cefta- intestinal tract and achieves higher serum and sputum zidime are safer. concentrations than oral ampicillin,34 though even 750 mg Trimethoprim passes readily into bronchial secretions, will not always produce inhibitory sputum activity against and concentrations often exceed those in serum,22 though haemophilus.5 Cole and colleagues have reported longer sulphamethoxazole activity after oral co-trimoxazole may be remissions in patients with chronic bronchitis after short subtherapeutic. Brumfitt and colleagues detected no sulpha- courses ofhigh dose oral amoxycillin (3 g 12 hourly),-though methoxazole and variable trimethoprim concentrations in peak sputum concentrations varied widely.6 Davies and sputum in 24 patients given co-trimoxazole, though both Maesen found higher sputum ampicillin concentrations after drugs were equally effective clinically.23 Quinolones have bacampicillin (800 mg) than after ampicillin (1 g),5 although renewed interest in antibiotic in the lung eight hourly 400 mg or 800 mg doses of bacampicillin because, despite effective diffusion in bronchi and good controlled haemophilus infections in bronchitis.7 Standard antimicrobial activity including against haemophilus and oral doses of in patients with cystic fibrosis branhamella, they are only moderately active against Str scarcely exceed inhibitory concentrations for StaphyJlococcus pneumoiae. Peak sputum concentrations exceedhalfofthose 1182 BRITISH MEDICAL JOURNAL VOLUME 294 9 MAY 1987

in the serum,24 but less susceptible pneumococci may not be casualties of bystanders being trained in cardiopulmonary Br Med J (Clin Res Ed): first published as 10.1136/bmj.294.6581.1181 on 9 May 1987. Downloaded from eradicated despite lung tissue concentrations exceeding those resuscitation,"4 and empirical calculations suggest that a in serum.25 trained bystander can improve the survival chance of some- Even if permanently eradicating infection from patients body with ventricular fibrillation from 21% to 43%.15 Up to with advanced chronic suppurative lung-disease remains a two lives for every 10000 people could thus be saved forlorn hope, betterrecognitionofthebehaviourofantibiotics annually.'i'8 Such calculations also suggest that a trained lay in the respiratory tract should help in assessing new thera- person will meet a casualty between once in 25 years to once peutic regimens. in over 112 years.'&2' JANE SYMONDS Some of these uncertainties have been incorporated into a Consultant Microbiologist, model of the cost effectiveness of training programmes in Russells Hall Hospital, cardiopulmonary resuscitation.2' Important questions are Dudley DYI 2HQ how often people should be retrained, whether resources should be concentrated on training key groups, and what I Hallstrom 0, Keyrilainen 0, Markhula H. Ampicillin concentration in normal and pathological lung tissues afteroral administration ofbacampicillin. Infection 1979;7(suppl 5):469-71. sort of people should be trained. Over 40 studies have 2 Stewart SM, FisherM, Young JE, Lutz W. Ampicillin levels in sputum, serum and saliva. Thorax shown that the skills ofcardiopulmonary resuscitation decay 1970;25:304-1 1. 3 Stewart SM, Anderson IME, Jones GR, Calder MA. Amoxycillin levels in sputum, serum and rapidly.'92223 Research is equivocal whether'training should saliva. Thorax 1974;29:110-4. be concentrated in medical and paramedical groups.2;27 4 Ingold A. Sputum and serun levels ofamoxycillin in chronic bronchial infections. BrJ Dis Chest 1975;69:211-6. A surveyofover 3000 people showed individual differences 5 Davies B, Maesen F. Serum and sputum antibiotic levels after ampicillin, amoxycdillin and in willingness to attempt cardiopulmonary resuscitation, but bacampicillin in chronic bronchitis patients. Infection 1979;7(suppl 5):465-71. 6 Cole PJ, Roberts DE, Davies SF, Knight RK. A simple oral antimicrobial regimen effective in over 40% reported that they would do something.28 Reported severe chronic bronchial suppuration associated with culturable Haemophilus infuenzae. JAntnimcrob Chemother 1983;11:109-13. willingness and actually carrying out cardiopulmonary re- 7 Maesen FPV, Beeuwkes H, Davies BI, et al. Bacampicillin in acute exacerbations of chronic suscitation are, however, different things: when medical bronchitis-a dose range study.JAntiicrob Chemother 1976;2:279-85. 8 Saggers BA,'Lawson D. In vivo penetration ofantibiotics into sputum in cystic fibrosis. ArchDis or paramedical people witnessed a collapse then cardio- Child 1968;43:404-9. pulmonary resuscitation was performed in one third ofcases, 9 Marlin GE, Burgess KR, Burgoyne J, Funnel GR, Guinness MDG. Penetration ofpiperacillin to bronchial mucosa and sputum. Thorax 1981;36:774-80. but when only non-medical people were present then resusci- 10 Smith BR, LeFrock JL. Bronchial tree penetration ofantibiotics. Chest 1983;6:904-8. tation was performed in about one in every 25 cases." 11 LodeH,GruhlkeG, HallermannW, DzwilloG. Significanceofpleural and spututnconcentrations for antibiotic therapy ofbronchopulmonary infections. Infectim 1980;8(suppl 1):49-53. Whether people help depends on how clearly they under- 12 Halprin GM, McMahon SM. Cephalexin concentrations in sputum during acute respiratory are infections. Anamicrob Agems Chemother 1973;3:703-7. stand what is happening and'on whether other people 13 Bergogne-Berezn E. Antibiotics in therespirasory tree.JAntinicrob Chemodher 1981;8: 1714. present.29 This study also found that women helped less often 14 Naeverson MA. Intravenous administration of erythromycin: serum, sputum and urine levels. CwrMedRes Opin 1976;4:359-64. than men and that, though training did not raise the 15 Marlin GE, Davies PR, Rutland J,Berend N. Plasmaand sputum erythromycin concentrations in intervention rate, it did increase dramatically the effective- chronic bronchitis. Thorax 1980;35:441-5. 16 Seigler D, Kaye CM, Reilley S, Willis AT, Sankey MG. Serum, saliva and sputum levels of ness ofhelp given. metronidazole in acute exacerbations ofchronic bronchitis. Thorax 1981;36:781-3. Some have argued, however, that even when a rescuer 17 MacCulloch D, Richardson RA, Allwood GK. The penetration of doxycycline, oxytetracycline and minocycline into sputum. NZMedj 1974;80:300-2. - does little or performs cardiopulmonary resuscitation in- 18 Maesen FPV, Davies BI, Van Noord JA. Doxycycline in respiratory infections: a re-assessment after 17 years.J7Annmicrob Chemother 1986;18:531-6. adequately the survival chances still improve."' Others have 19 Gartmann J. Doxycycline concentrations in lung tissue, bronchial wall and bronchial secretion. questioned whether "retention of classroom skills is related Chemotherapy 1975;21:19-26. 20 Noone P, Parsons TMC, Pattison JR, et al. Experience in monitoring gentamicin therapy during to performance during actual resuscitation attempts or treatment ofserious Gram-negative sepsis. BrMedJ 1974;i:477-81. to eventual clinical outcomes."22 The dearth of empirical 21 Pennington JE, Reynolds MY. Concentrations of gentamicin and carbenicillin in bronchial studies comparing "classroom skills" with actual perform-

secretions. IInfectDis 1973;128:63-8. http://www.bmj.com/ 22 Hughes DTD. The use of combinations of trimethoprim and sulphonamides in the treatment of ance means that this assertion remains untested. In studies of chest infections. J Antimicrob Chemother 1983;12:423-34. 23 Brumfitt W, Hamilton-Miller JMT, Howard CW, Tansley H. Trimethoprim alone compared to medical students and hospital staffit has been suggested that co-trimoxazole in lower respiratory infections: pharmacokinetics and clinical effectiveness. ScandJ InfectDis 1985;17:99-105. some may have performed better in real emergencies and 24 Bergogne-Berezin E, Berthelot G, Even P, Stern M, Reynaud P. Penetration of ciprofloxacin into some worse.3' 32 bronchial secretions. EurJ ClinMicrobiol 1986;5: 197-200. 25 Schlenkhoff D, Knopf J, Dalhhoff A. Penetration of ciprofioxacin into human lung tissue. In: Many cases are required to determine the effectiveness of Neu HC, Weuta H, eds. Proceedings ofThe IstInuenatiownal ciprojioxacin workshop. Amsterdam: interventions. Assessment is complicated by the nature ofthe Excerpta Medica, 1985:157-9. (Current Clinical Practice Series No 34). incident that causes breathing to fail and the heart to stop. In some cases cardiopulmonary resuscitation would be unsuc- on 29 September 2021 by guest. Protected copyright. cessful however well performed, but developing and main- taining the blood pressure and circulation of oxygen for adequate tissue perfusion and continuing brain function Evaluating mass training in demand a high level of skill. Therefore, there is no basis for assigning low importance to initial and refresher training in cardiopulmonary resuscitation cardiopulmonary resuscitation. As rescuers will not require the full repertoire ofskills in every incident excellent training The Save a Life campaign, which was started in October is required for trainees to have adequate knowledge and skills 1986 to stimulate mass training in emergency first aid, rightly from which to draw should the need arise. emphasised cardiopulmonary resuscitation-the most com- Criticisms of mass training in cardiopulmonary resuscita- plex first aid skill. Such campaigns are not new, and tion are that trainees develop a false sense of competence20 the teaching of rescue breathing has been compulsory in and that resuscitation might be performed unnecessarily or Norwegian schools since the early 1960s.' Recommendations hazardously."3' High drop out rates among volunteer in- have been made for including training on resuscitation in structors have also been encountered.'6 Some of these prob- schools,2 and there is advice on organising community or lems wvould be overcome by thorough traininlg and regular mass training.3-7 But do the benefits of these schemes justify refresher training, with particular emphasis on diagnosis. the costs or could the resources be better used? Evaluation of mass training should also take account of Many researchers have explored the benefits for real possible hidden benefits. For example, we have shown that