Rational use of antibiotics

Rational use of antibiotics

Rational Use of Antibiotics HEINZ F. EICHENWALD,M.D. ANDHENRY R. SHINEFIELD, From tbe Department of Pediatrics, Tbe New York Hospital-Cornell Medica...

567KB Sizes 0 Downloads 38 Views


Use of Antibiotics

HEINZ F. EICHENWALD,M.D. ANDHENRY R. SHINEFIELD, From tbe Department of Pediatrics, Tbe New York Hospital-Cornell Medical Center, New York, New York. Tbis work was supported, in part, by Research Grants E-944, E-998 and E-3669 from tbe National Institute of Allergy and Infectious Disease.

New York, New York

to chemotherapy is iIIustrated by the use of certain antibiotic combinations such as erythromycin and chIoramphenico1, tetracycIine and peniciIIin, novobiocin and tetracycIine, and so forth. Combinations of antibiotics have an estabIished place in the treatment of a few diseases such as tubercuIosis, subacute bacteria1 endocarditis, bacteria1 meningitis of undetermined cause and certain serious infections occurring during the newborn period. However, for the vast majority of infectious iIInesses, combinations neither increase the potency of therapy or the rapidity of response, nor inhibit the emergence of drug-resistant variants of the organisms [21--231. NevertheIess, around the country, many physicians use, for exampIe, a combination of erythromycin and chIoramphenico1 in the therapy of staphyIococcaI disease. Aside from the fact that much better therapeutic agents are presentIy avaiIabIe, no evidence exists to prove that this type of treatment offers greater effectiveness than erythromycin aIone [24], nor are data avaiIabIe which unequivocaIIy demonstrate that the emergence of resistant strains is inhibited. In fact, evidence exists that cross resistance of StaphyIococcus aureus can be induced in the Iaboratory by a combination of these two agents 1251. We mention these facts not because we desire to engage in controversy but to iIIustrate a point, nameIy, rationa use of antimicrobia1 therapy depends not onIy on knowIedge of the subject but also on wiIIingness to resist various psychoIogic and socioIogic pressures which often suppIant scientific fact. When one Iooks about to determine how antimicrobia1 therapy is presentIy being used in infants and chiIdren, it becomes immediateIy apparent that a strong tendency exists to appIy data obtained in aduIts directIy to younger patients. As we and others have pointed out in the past, such an approach represents a

UR ASSIGNED topic “The RationaI Use of Antibiotics” is one that we approach with some trepidation since this pIeasant-sounding titIe assumes two facts to be true: (I) the state of knowledge of the subject and of Iaboratory technic has progressed to the point where a compIeteIy rationa approach is possibIe and (2) the physician who decides whether or not some or any antimicrobial agent shouId be given to a particuIar patient usuaIIy, if not aIways, makes this decision in a IogicaI and rationa manner. NeedJess to say, neither one of these assumptions is correct: antimicrobia1 agents are usuaIIy seIected on the basis of unreIiabIe Iaboratory data which are incorrectIy interpreted or bIindIy accepted by the physician [I]. A decision may aIso be made from information derived from advertising copy or because a “routine” was estabIished, which is uncriticaIIy foIIowed. An exceIIent iIIustration of such an irrationa1 approach is represented by the manner in which antimicrobia1 agents are often empIoyed in the pre- and postoperative period. A Iarge number of articles have been pubIished citing experimenta or statistica evidence indicating that the administration of prophyIactic antibiotics in surgica1 practice not onIy faiIs to contro1 wound sepsis but aIso is associated with a higher incidence of infection and iatrogenie disease [2-201. It wouId seem that since antimicrobia1 therapy is often used in this way, there shouId at Ieast be some vaIid data proving the vaIue of systemic prophyJactic therapy. It is, however, aImost impossibIe to find recorded in the Iiterature significant resuits which wouId support its use. A different aspect of an irrationa1 approach


American Journal of Surgery. Volume IO,. March 1964




Use of Antibiotics

to the child [26-281. In t.he first place, the course of a given infectious iIIness ih often radicaIIy different in chiIdn:n from that observed in nduIts; the iIIness may be either more or Iess severe. A specific antimicrobia1 agent may be unpredictabIy toxic to the infant or child, whiIe paradoxicaIIy a different drug may prove reIativeIy harmIess to the baby but too dangerous for use in the aduIt. FinaIIy, certain antibiotics appear to be more effective in patients of one age than ot another. The effectiveness and toxicity of a number of antimicrobia1 agents have been carefuIIy studied in infants and chiIdren. I wiI1 briefly discuss these data, mentioning the drugs in the aIphabetica1 order of their generic names. disservice



Bacitracin. If a staphyIococcus is resistant to peniciIIin, bacitracin, in our experience, is the most potent agent presentIy avaiIabIe. Because of its nephrotoxicity in aduIts, it has not had much use in the past. In infants under one year of age, in whom severe staphyIococca1 disease occurs with considerabIe frequency, the drug has proved to be both singuIarIy effective and virtuaIIy nontoxic when used in a dosage of 1,000 units per kilogram of body weight per day given in two intramuscuIar doses tweIve hours apart. If further treatment is needed after a week to tweIve days of bacitracin therapy, another suitabIe drug, preferabIy one that can be administered oraIIy, is substituted. Infants who respond poorIy to the discontinuation of bacitracin may safeIy be given an additiona week’s course of this drug. An indication of the effectiveness of bacitracin in the treatment of staphyIococca1 pneumonia with empyema is demonstrated by our data. Over one hundred infants were treated with bacitracin at the New York HospitaI and in neighboring institutions over the past five years. Fewer than 5 per cent died of the disease, most of whom died within a few hours after the start of therapy. Despite carefu1 prolonged foIIow-up study of a11 patients, no irreversibIe toxic effects have been noted to foIIow bacitracin therapy. ChIoramphenicoI is wideIy Cblorampbenicol. used in pediatrics as we11 as in interna medicine. This is probably reIated to two factors: (I) its demonstrabIe effectiveness in the


therapy of the most common forms of bacteria1 meningitis, and (2) the elegance and ease with which it can be administered both orally and intramuscuIarIy. UnfortunateIy, the exceIIent rcputa,tion chIoramphenico1 enjoys for wide spectrum potency is not entireIy borne out by the facts. Its efficacy in meningitis is the resuIt of its appearance in spina ffuid in higher concentrations than other avaiIabIe antimicrobiai agents, rather than to innate differences in potency. In other areas of the body, excIuding onIy the interior of the eye, chIoramphenico1 possesses no cIearcut cIinica1 superiority in the treatment of infection with the exception of saImoneIIa and rickettsia1 iIInesses. Many strains of Escherichia coIi of cIinica1 importance are resistant [29]. In the Iaboratory, staphyIococci are often very susceptible to chIoramphenico1, whiIe in practice the drug has given disappointing resuIts, particuIarIy when used aIone [29]_ The risk of chIoramphenico1 administration must aIso be considered. Not onIy does it occasionaIIy produce unpredictable hematoIogic disturbances, but aIso in young infants and sometimes even in babies as oId as two months, a pecuIiar form of cardiovascuIar coIIapse known as the “Gray Syndrome” has been observed after its administration [JO]. Colistimetbate. This antimicrobia1 agent is produced under the trade name of CoIymycin. @ It is an effective agent against pseudomonas, nearIy as much so as polymyxin, and some other gram-negative organisms [31,32]. Its major use Iies in the treatment of infections of the urinary tract caused by susceptibIe organisms. The dosage in infants is about 5 to 6 mg. per kiIogram per day and the onIy toxic effect we have observed in short-term use has been transient aIbuminuria. In the treatment of systemic E. coIi infections, streptomycin or kanamycin wiI1 usuaIIy prove far more effective if the organisms in question are susceptible. Erytbromycin. This antibiotic has a spectrum resembIing that of peniciIIin. Of practica1 importance is its effectiveness against certain strains of staphylococci resistant to peniciIIin. It is one of the Iea.st toxic of known antibiotics; the Iiver toxicity rareIy found in aduIts with the IauryI suIfate preparation has not been observed in chiIdren. A recommended dosage is 25 mg. per kiIogram per day in four equa1 doses. We empioy this substance in chiIdren



who are onIy miIdIy or moderateIy iI with staphyIococca1 disease since it is far Iess effective than bacitracin or newer semisynthetic penicillins. Kanamycin. Kanamycin beIongs to the neomycin-streptomycin famiIy. It is an extremeIy wide-range drug, effective against a variety of gram-positive, gram-negative and acid-fast bacteria. Most importantIy, this agent dispIays activity against aImost a11 strains of E. coIi seen in this area, even those resistant to streptomycin. It is active aIs against saImoneIIa, aerobacter and many strains of proteus. WhiIe kanamycin possesses significant toxicity in aduIts, affecting the kidneys and the eighth crania1 nerve, its use in young infants is associated with Iittle or no hazard [26,33]. Kanamycin when given in a dose of 7.5 mg. per kiIogram every tweIve hours for a period not in excess of ten days is not associated with any detectabIe form of toxicity in the infant under one year of age. We have foIIowed up babies treated in this manner during the neonata period or first few months of life for more than four years and have been unabIe to show any impairment of their renaI or auditory function. The high therapeutic efficacy of this drug is striking. It has a bacteriocida1 action on most gram-negative organisms causing dangerous disease in infants; thus, it has been possibIe to reduce mortaIity to Ievels which we had been unabIe to attain prior to its use. Nitrofurantoin. This drug is soId under the trade name of [email protected] The agent is absorbed from the gastrointestina1 tract but is then rapidly excreted by the kidneys so that onIy insignificant serum IeveIs are found. Thus, the major effect is on the urinary tract, and Furadantin has proved usefu1 onIy in the treatment of infections of the urinary tract in oIder chiIdren and aduIts. The drug has not been adequateIy evaIuated in young infants; it seems Iikely, on the basis of what is known, that its use wouId be associated with hemoIytic anemias and other toxic effects in a smaI1 proportion of patients of this age. Novobiocin. This agent is usefu1 in the treatment of miId and moderateIy severe staphyIococca1 disease caused by strains resistant to peniciIIin and erythromycin, in the therapy of infections of the urinary tract caused 520


by susceptibIe strains of E. coIi and proteus, in the treatment of diarrheas caused by enteropathogenic E. coIi resistant to neomycin 1341, and in occasiona cases of systemic proteus infection. The use of this drug in smaI1 infants is associated with a rise in the biIirubin IeveI of an appreciable proportion of patients [35]; in oIder chiIdren the drug has been associated with a fairIy high incidence of rashes and drug fevers and the occasiona occurrence of signs of Iiver dysfunction. Penicillin. This most remarkabIe substance is too we11 known to warrant extensive discussion. OnIy the most recent modifications of the moIecuIe wiI1 therefore be discussed, namely, the so-caIIed antistaphylococca1 peniciIIins. UnfortunateIy, the high promise of these drugs in the Iaboratory has not been entireIy borne out in cIinica1 practice. The reasons for this appear to be twofoId: (I) the tissue IeveIs achieved are very Iow and (2) these peniciIIins are more toxic than the parent moIecuIe so that the amount of drug which can be safeIy administered is limited. One possibIe toxic effect is on the kidney where aIbuminuria and cyIindruria combined with a rising IeveI of bIood urea nitrogen have been noted [36]. In addition, a few cases of transient depression of bone marrow have been reported, aIthough the exact reIation of the drug to this reaction is not yet estabIished [37]. It seems essentia1, therefore, that these peniciIIins be used in the recommended dosage which in infants is 40 mg. per kiIogram every six hours and 25 mg. per kiIogram every six hours in oIder chiIdren. The therapeutic efficacy in infants is Iess than that of bacitracin; as a resuIt we have continued to empIoy bacitracin in the treatment of severe staphyIococca1 disease in infants. However, in older chiIdren, in whom bacitracin has proved more toxic, these peniciIIins are extremeIy usefu1. We prefer to use the ora form, oxaciIIin, whenever possibIe. The advantages of this route of administration are obvious; in addition, 0xaciIIin is more effective microgram for microgram than the injectabIe drug methiciIIin. Streptomycin. Aside from its use in tubercuIosis, streptomycin is effective against a variety of gram-negative organisms, particuIarIy E. coIi. UnfortunateIy, many strains of this enterobacteria have acquired resistance whereas kanamycin is we11 toIerated. This is one reason why we have substituted kanamycin in


Use of Antibiotics growth /~cI]. Such a reaction is obvir~usIy undesirabIe in children, and these antimicrobial agents shouId therefore probably not be used repeatedly or for prolonged periods. Another t.roubIescme side effect is related to potency of action. I‘ctrncyclinc an d wide spectrum regularly denudes the skin and mucous membranes of their norma flora, permitting overgrowth of highIy resistant organisms. Such a change in flora. is obviousIy an undesirabIe effect particuIarIy in patients who wiI1 or have undergone surgica1 procedures.

infants. I II oIder chiidren, streptnm~c~in is particularly usefu1 in the treatment of infections of the urinary tract and peritonitis. When it is given, its dosage should not exceed 40 tug. per kilogram per day. When urinary output is decreased for any reason, the daily dose should be reduced to 30 or even 20 mg. per kilogram per day. The drug shouId not be administered for periods Ionger than ten to unless tubercuIosis is being twelve days, trcatecl. This group of drugs shouId Suljonamides. not be empIoyed in newborn or premature infants because its administration may resuIt in an increased incidence of kernicterus [?a]. Sulfonamides remain heIpfu1 in the treatment of infections of the urinary tract in oIder children and represent the treatment of choice in meningococcal disease. The form with which we have had most experience is sulfisosazoIe (Gantrisi$). The so-caIIed Iong acting sulfonamicles have proved disappointing and more toxic, and therefore have not been wideIy empIoyed by us. The Tetracyclines. These represent a family of cIoseIy reIated compounds incIuding tetracycIine itseIf, chIortetracycIine (AureomycirP), oxytetracycIine ([email protected]) and demethylChIortetracycIine (DecIomycirP). These four drugs have more or Iess identica1 antibacteria action, and exhibit essentiaIIy complete cross resistance. They shouId be administered orally, or if this route is not avaiIabIe, intravenousIy. IntramuscuIar administration in chiIdren using any materia1 presentIy on the market is foIlowed by erratic and unpredictabIe blood IeveIs. When tetracycIines are administered oraIIy, the blood IeveIs attained are reIatec1 to the amount of fluid administered concurrentIy. At Ieast I cc. of water per miIIigram of drug wiI1 permit maxima1 absorption; if Iess fluid is used, Ievels are Iower. In young infants, the rates of absorption and excretion of these agents are unpredictabIe [26]. Inadequate or excessiveIy high bIood IeveIs are often encountered after administration of these drugs, and thus no adequate dose information for use in babies exists. In infants and chiIdren, at Ieast one and probabIy a11 of these drugs are associated with an unusual long-term toxicity. The antibiotic is deposited in the bones and denta ename1, and it has been shown that this deposition is associated with a decrease in the rate of bone younger


The inteIIigent use of antimicrobia1 agents of the activity of depends on knowIedge various chemotherapeutic drugs, their relative effectiveness and toxicity in a variety of infectious conditions in patients of various ages and on the proper utiIization and interpretation of laboratory data. Finally, the physician must cIearIy differentiate in his own mind between the desire to do something truIy useful for the patient and the magic of administering a drug; it is often better to do nothing than to use these potent and dangerous agents in an empiric or bIind fashion. REFERENCES I.



5. 5.


E. The physician and the disc. Am. J. Dis. Cbild., 102: 291, 1961. h~ckrTRIcK, L. S. and WHEELOCK, F. C. The routine use of antibiotics in cIective abdominal surgery. Surg. Gynec. e’? Obst., gg: 376, 1954. WEINSTEIN, L., GOLDFIELD, IZI.and CHANG, T. W. Infections occurring during chcmotherapv.. 1 New England J. Med., 25 I : 247, '954. WEINSTEIN. L. Chemom-oohvIaxis of infection. Ann. Int. Med., 43: 287, 1955. HASSLINGER, W. Pi. Stelfungnahme zur Penicillin Anwcndung in der Magenchirugie. Zentralbl. Cbir., 81: 1566, 1936. HOGIIAN, C. F. and SAIILIX, 0. Infections complicating gastric surgery. Acta cbir. scundinav., I I 2:


27i, 1957. 7. HOWE, C. W. The problem of postoperative wound infection caused by Staphylococcus aureus. Ann. Surg., 146: 384, 1957. 8. LINDER, F. Discussion on the problem of resistant organism and chemotherapeutic sr,nsitivity in surgery. Proc. Roy. Sot. Med., 30: 1.53, 1947. g. PETERSDORF, R. G. et al. Study of antibiotic prophyIaxis in unconscious patients. Nelv Engkmd J. Med., 257: 1001, 1957. IO. TACHDJIAK, M. 0. and COMPERE, E. L. Postoperative wound infection in orthopedic surgery; an evaIuation of prophyIactic antibodies. J. Internat. 611. Surgeons, 28: 797, 1937.


EichenwaId and ShinefieId I I. MONCRIEF, J. A. and RIVERA, J. A. The probIem of infection in burns by resistant microorganisms with a note on the use of bacitracin. Ann. Surg., 147: 295, 1958. I 2. MARSHALL, A. ProphyIactic antimicrobia1 therapy in retropubic prostatectomy. &it. J. Ural., 31:

26. EICHENWALD, H. F. and SHINEFIELD, H. R. AntimicrobiaI therapy in the neonata1 period. Pediat. Clin. North America, 8: 509, 1961. 27. SUTHERLAND, J. M. and MOHLMAN, Y. M. Antimicrobia1 therapy in the newborn infant. Pediat. Ch. Nortb America, 8: I 143, 1961. 28. SUTHERLAND, J. Al. Hazards of antimicrobial therapy in newborn infants. Postgrad. Med., 32: 127, 1962. 29. WOODWARD, T. E. and WISSERMAN, C. L., JR. ChIoromycetin. New York, 1958. MedicaI EncycIopedia Publishers. 30. SUTHERLAND, J. M. FataI CardiovascuIar coIIapse in infants receiving Iarge doses of ChIoramphenico1. Am. J. Dis. Cbild., 97: 761, 1959. 3 I. MARSDEN, H. B. and HYDE, W. A. CoIistin methane suIphonate in chiIdhood infections. Lancet, 2: 740, 1962. 32. MCMILLAN, M. et a1. Pseudomonas pyocyaneainfection treated with colistin methane suIphonate. Lancet, 2: 737, 1962. 33. Yow, M. D. and TENGG, N. E. The use of kanamycin suIfate as a therapeutic agent in infants and chiIdren. J. Pediat., 58: 538, 1961. 34. OLBING, H. and KUSTER, F. Erfahrungen in der Antibiotischen BehandIung von SaeugIingsenteritiden unter Besonderer Berucksichtigung des Novobiocin. Arch. Kindetb., 166: 216. 1962. 35. SUTHERLAND,J. M. and KELLER, W. H. Novobiocin and neonatal hyperbiIitubinemia. Am. J. Dis. Cbild., IOI: 447, 1961. 36. ALLEN, J. D., ROBERTS, C. E., JR. and KIRBY, W. M. M. StaphyIococcaI septicemia treated with methiciIIin. New England J. Med., 266: I I I.

431, 1959. 13. TAYLOR, G. W. Preventive use of antibiotics in surgerv. Brit. M. Bull.. 16: 4 I. 1060. 14. FEINS%N, A. R., PETER&&RF; R. G. and BROWDER, A. A. ProphyIaxis of poststreptococca1 sequelae and bacteria1 endocarditis. J. Pediat., 58: 164, 1961. 15. FORBES, G. B. A report on staphyIococca1 infection of operative wounds with speciaI reference to topical antibiotic prophylaxis. Lancet, 2: 505, 1961. 16. GAYLOW, D. W. Preoperative BoweI “Sterilization.” A DoubIe BIind Study. AntimicrobiaI Agents Annual. New York, 1961. PIenum Press. 17. KING, G. D. The case against antibiotic prophyIaxis in major head and neck surgery. Laryngoscope. 71: 647, 1961. 18. LORD, J. W. et a1. Endocarditis compIicating open heart surgery. Circulation, 23: 489, 1961. 19. PETERSDORF, R. G., BROWDER, A. A. and FEINSTEIN, A. R. Protection against infection in susceptibIe individuaIs. J. Pediat., 58: 174, 1961. 20. JOHNSTONE,F. R. C. An assessment of prophyIactic antibiotics in genera1 surgery. Surg. Gynec. & Obst., 116: I, 1963. 21. PRYLES, C. V. Current status of combined antibiotic therapy. Pediatrics, 21: IOO, 1958. 22. THOMPSETT, R. AntimicrobiaI therapy: the concurrent administration of two drugs. Am. J. M. SC., 235: 585, 1958. 23. LACEY, B. W. The rationale and management of combined therapy. Brit. M. Bull., 16: 42, 1960. 24. JONES, W. F. and FINLAND, M. Antibiotic combinations: antistreptococca1 and antistaphyIococca1 activity of pIasma of norma subjects after ingestion of erythromycin or chIoramphenico1 or both. New England J. Med., 257: 744, 1958. 25. BARBER, M., CSILLAG, A. and MEDWAY, A. J. StaphyIococcaI infection resistant to chIoramphenico1, erythromycin and novobiocin. Brit. M. J., 2: 1377, 1968.

1962. 37. MCELFRESH, A. E. and HUANG, N. N. Bone marrow deDression resulting from the administration of m&hiciIIin. New England J. Med., 266: 246, 1962. 38. SILVERMAN, W. A. et a1. A difference in mortality rate and incidence of kernicterus among premature infants aIIotted to two prophylactic antibacteria1 regimens. Pediatrics, 18: 614, 1956. 39. COHLAN, S. Q., BEVELA~DER, G. and TEOMSIC, T. Growth inhibition of the developing skeIeton due to tetracycIine deposition in bone. Proceedings of the American Pediatric Society, AtIantic City, 1962.