Click for next page ( 66


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 65
CONSULTANTS' PAPERS THE APPENDIXES The following papers, Appendixes A through K, were commis- sioned by the Committee to Study the Human Health Effects of Subtherapeutic Antibiotic Use in Animal Feeds. They were used by the committee as working papers and are attached to the committee's report for information only. They do not constitute part of the foregoing report prepared by the committee. All references to these papers should be attributed to the authors, not to the cam- mittee. A transcription of the public meeting held August 23, 1979 also formed part of the working papers used by the committee. These records may be obtained on loan from Dr. Enriqueta C. Bond, National Academy of Sciences, 2101 Constitution Ave., N.W., Room 347, Washington, D.C. 20418. 65

OCR for page 65

OCR for page 65
APPENDIX A THE CLINICAL USE OF ANT IMICROBIALS AND THE DEVELOP1dE:NT OF RESISTANCE John P. Utz Infectious diseases are the most common cause of absence from school or work, are the most frequent of the known causes of birth defects, and, as pneumonia and influenza, are the fourth most c ammo n cause of death in the United States. CONTROL OF DISEASE . Immunization with vaccines is the major technique used in the United States to prevent infectious diseases. The oral polio vaccine, as the ultimate example, is clearly responsible for the world-wide control of poliomyelitis, despite an outbreak of the disease, which spread from the Netherlands to Canada and the United States in 1979 (Center for Disease Control, 1979a). Similarly, the tetanus vaccine is solely responsible for the elimination of tetanus, as observed in the Armed Forces during World War II and the conflicts in Korea and Vietnam. Other live virus vaccines are used primarily in children to prevent rubella, rubeola, and mumps. Vaccination with killed bacteria (pertussis) or their toxic products (diphtheria toxoid) is also an effective preventive measure. Antimicrobials are used to prevent approximately a dozen categories of infection. Some of these uses are described below. Between 1960 and 1976 there were 131 published studies of the prophylactic use of antimicrobials to reduce the possibility of infection after surgical procedures. Although only 24 of these studies describe controlled, prospective, double-blind studies, they report striking statistically significant reductions in various types of wound infections when antimicrobials were administered in conjunction with repair after hip fracture, hip prosthesis, vaginal hysterectomy, and colon, biliary, and gastro- intestinal surgery (Chodak and Plaut, 1977~. These procedures are regarded as clean in contrast to compound fractures and gunshot or stab wounds of the abdomen, which are considered already infected. Thus, antimicrobial use in these instances is therapeutic, rather than prophylactic. School of Medicine, Georgetown University, Washington, D.C. 67

OCR for page 65
68 A second category of antimicrobial prophylaxis was first demonstrated in San Francisco where isoniazid was used to pre- vent overt tuberculosis in populations with delayed cutaneous hypersensitivity to tuberculin (American Thoracic Society, 1974; Ferebee, 1970~. The continuing protection demonstrated in the treated group led to the wide use of this drug in patients with a known positive test. The enthusiasm engendered by these suc- cesses continued until the recognition in the Washington, D.C. area of the hepatotoxicity of isoniazid (Garibaldi et al., 1972~. Antimicrobials administered prophylactically to immunode- ficient patients lacking the benefit of a protected environment have produced variable results. Immunodeficiency may result from a disease state or from the use of drugs. Immunosuppres- sion may be desired, e.g., when drugs are administered to trans- plant recipients to prevent host-versus graft rejection, or may be an undesired consequence of cancer chemotherapy. Protection afforded by antimicrobial prophylaxis is enhanced by a protected environment. Studies conducted over many years, beginning early in the antimicrobial era, have shown that oral sulfonamides and peni- cillin G or intramuscular benzathine penicillin G protect against recurrence of rheumatic fever by preventing infection by Group A Streptococcus pyogenes. Failure of such prophylaxis is not usually related to the development of resistance, but instead to a lack of compliance on the part of patient or parents. Hence, benzathine penicillin G injected by a medical professional is the most successful drug and route, since patient cooperation--except for travel to a doctor's office or clinic--is not required. An even earlier practice was the prevention of ophthalmia neonatorum by the routine use of silver nitrate and, later, peni- cillin G. in all newborn infants. Scrub typhus (tsutsugamushi disease) occurs rarely in this country, and then only in people who have once spent time near or in northern Australia, Pakistan, or Japan. This disease can be prevented by 1 g of chloramphenicol every other day for a month after exposure, but because the disease is infrequent, this pro- phylactic application of chloramphenicol is little known. At various times and during different epidemics, sulfonamides, rifampin, or minocycline given to family members or other groups have prevented the spread of meningococcal infections from patients

OCR for page 65
69 or carriers of Neisseria meningitidis. - frequently a commensal of humans, it has virtually always re- sulted in later recolonization and the carrier state following prophylaxis. Since this species is Chemoprophylaxis is probably effective in the prevention of syphilis and shigellosis, but is rarely used for this purpose. More commonly it is practiced in the hope of preventing bacterial endocarditis, but there are no data demonstrating the efficacy of this use. Controlled trials indicate that either a tetracycline or ampi- cillin is effective in preventing febrile disease and loss of time from work for patients with chronic obstructive pulmonary disease (Batten, 1976~. However, the magnitude of that population and the cost-benefit ratio are so great that this method of preventing recurrent infection has not been widespread. The elimination of smallpox (variola) by the use of vaccina- tion (cowpox or vaccinia) has overshadowed the demonstration of methisazone as a chemoprophylactic agent. Field trials in Madras, India showed that the use of methisezone in 1,101 exposed persons resulted in only three cases of mild smallpox, in contrast to 78 cases (with 12 deaths) among 1,126 controls who did not receive the drug (Bauer et al., 1963~. Enthusiasm for amantadine in the prevention of influenzas has been tempered by the limited spectrum of its antiviral activity-- type A2 (Asian) influenza only--and by the limited period of effec- tiveness (immediately before or after the exposure, the time of which is always uncertain). Since World War II chemoprophylaxis has been an accepted prac- tice in malarial areas. The development of chloroquine-resistant strains in Southeast Asia and Africa has lessened the effectiveness of that antimalarial drug and has led to the requirement for other drugs (Anonymous, 1978a). Specific Agents With the exception of the arsenicals and bismuth, the anti- microbial era began in the mid-1930's with the use of the sulfo- namides. Even today members of this family are the first choice in the treatment of uncomplicated urinary tract infections,

OCR for page 65
70 nocardiosis, some Chlamydia trachomatis infections, and chan- croid. Resistance of N. meningitidis to the sulfonamides has been encountered principally in prophylaxis of family members or in active treatment of meningitis. The more recent combina- tion of a sulfonamide, sulfamethoxazole, with trimethoprim has resulted in an oral preparation that is useful in the treatment of an even wider range of infections, e.g., pneumonia caused by Pneumocystis carinii. The antibiotic era began in the mid-1940's with the redis- covery of the activity of penicillin G and its use in severe human disease. This drug remains the first choice in the treat- ment of infection by Group A S. pyogenes (e.g., rheumatic fever), S. pneumonias (pneumonia), N. meningitidis (meningococcal dis- eases), Pasteurella multocida (e.g., skin ulcer, osteomyelitis, pleuritis, sinusitis, leptomeningitis), Treponema pallidum (syphilis), Actinomyces Israelis (actinomycosis), Leptospira spp. (leptospirosis), Bacillus anthracis (anthrax), and Strep- tobacillus moniliformis (streptobacillary rat-bite fever). However, by 1947 Weinstein had reported superinfection by Haemophilus influenzas, which produced septicemia in a patient treated with penicillin G for pneumonia. This first report is the prototype of the truism that the use of any antimicrobial agent ultimately results in the appearance of, or colonization by, microorganisms that are resistant to or not susceptible to that agent. However, this is not always detrimental if the pathogenicity of the new microorganism is less or nonexistent. Indeed, were this not usually the case, antimicrobial therapy could rarely, if ever, be justified. Streptomycin, first used in 1945, remains as a single drug treatment for one infection by Francisella tularensis (tularemia). In combination with penicillin G it is optimal therapy for Entero- coccus (Group D streptococcal infections), with a tetracycline for Malleomyces mallet ("landers) and Yersinia pestis (plague), with ampicillin for more resistant Listeria monocytogenes (listeriosis), and with other antituberculous drugs for tuberculosis. The tetracyclines and chloramphenicol became available almost simultaneously in 1948. The former are active against many of the Gram-negative Enterobacteriaceae that are resistant to streptomycin and are optimal therapy for Calymmatobacterium granulomatis (granu- loma inguinale), Brucella spp. (brucellosis), Vibrio cholera e (cholera), and Borrelia recurrentis (relapsing fever). Tetracy- clines are also the first choice in the treatment of another group of infections caused by Rickettsia. Infections caused by Chlamydia, both _. psittaci and C. trachomatis, are similarly treated._

OCR for page 65
71 Despite the rare, but commonly fatal, idiosyncratic pancyto- penic reactions, chloramphenicol is indicated for H. influenzee (meningitis) and Salmonella typhi (typhoid). Decisions on whether to use antimicrobials for salmonellosis, often acquired from meats and eggs, should be based upon the studies of Woodward and Smadel (1964~. These authors defined the precise benefits of such therapy as well as the unaffected factors, e.g., metastases, carrier state, and bowel perforations, and its ill effects, e.g., enhanced relapse rate. Tricarcillin, usually in combination with an aminoglycoside, is used to treat Pseudomonas aeruginosa infections. Among the aminoglycoside antibiotics, tobramycin appears to have both greater efficacy and less toxicity than gentamicin, kanamycin, or amikacin. The development of the penicillinase-resistant penicillins methicillin, oxacillin, cloxacillin, dicloxacillin, and nafcillin has resulted in the preferential use of the last two for any Staphyloccus aureus infection when the sensitivity to penicillin G is unknown. In addition to streptomycin, the most active and best toler- ated among the antituberculous drugs are rifampin, ethambutol, and isoniazid. The latter three drugs are generally preferred because they can be administered orally, which is more convenient than the intramuscular route for both long-term use and therapy administered in the home or to ambulatory patients. A study by Phillipon et al. (1977) indicates that rifampin may be preferable to tetracycline in the treatment of brucellos~s, especially in its ability to reduce residual infections. But further confirmatory studies are needed before its widespread use can be recommended. Erythromycin is indicated for the Corynebacterium diphtheriae carrier state (but is not adequate for active diphtheria, which re- quires antitoxin) and for Mycoplasma spp. and Legionella pneumophila infections because the drug lacks serious side-effects. For the anaerobic infections, which commonly cause intracere- bral, pulmonary, peritoneal, or pelvic infections (usually with abscesses), one has a choice of four drugs: chloramphenicol, clin- damycin, metronidazole, or cefoxitin, depending upon which side- effect is less disturbing to the physician, e.g., enterocolitis from clindamycin or the idiosyncratic pancytopenic reactions from chloramphenicol.

OCR for page 65
72 Although the activity of amoxicillin or ampicillin is similar to that of penicillin G. the drugs are especially effective in the treatment of infection from Proteus mirabilis, Salmonella (except typhi), Shigella spp., and Listeria monocytogenes. When none of the preceding microorganisms is the pathogen, or occasionally, when one after another has been progressively selected, fungal infection or superinfection may occur. Agents for these are amphotericin B. flucytosine, or miconazole, depending on the species or occasionally on the sensitivity of the isolated strain. Other useful agents include vancomycin, which is enjoying a resurgence of interest because it is so helpful in patients on dialysis, with Clostridium difficile in necrotizing, pseudomem- branous enterocolitis and in Gram-positive coccal infections that are resistant to penicillin. A special mention must be made of the cephalosporins because there are so many of them, because they are remarkably active against many Gram-negative and Gram-positive bacteria, and because minute chemical changes have resulted in such great activity against previously unsusceptible organisms, e.g., Enterococcus and Pseudomonas aeruginosa. BACTERIAL RESISTANCE TO ANTIMICROBIALS Failure of antimicrobial therapy can be attributed to a number of causes: untreatable infections (e.g., pneumonia due to measles virus), improper dosage (e.g., overdosage increasing risk of super- infection or too low a dose resulting in failure to achieve bacter- icidal concentrations, as might occur in bacterial meningitis where the level in cerebrospinal fluid may be much lower than that in the blood), improper duration (e.g., failure to treat a Group A strepto- coccal infection for 10 days), omission of surgical drainage of an intraabdominal or pelvic abscess, or the emergence of microorganisms that are resistant to the antimicrobial. The development of S. aureus that is resistant to penicillin G by means of a penicillinase is of historical importance. From 1955 to 1960 severe disease was caused by such bacteria. These outbreaks were ended by the development of penicillinase-resistant penicillins. The second instance of resistance to antimicrobial agents that was of major importance was the development of Streptococcus pneumon- iae, which became resistant to tetracyclines, drugs that were useful at that time for treating pneumonias of uncertain origin, i.e., those now known to to be caused by Mycoplasma pneumonias, Cox~ella burnetii,

OCR for page 65
73 and Legionella pneumophila. In their studies of Australian aborigines and New Guineans, Hansman et al. (1971) found the first isolates of Streptococcus pneumonias with resistance to penicillin. Although strains with intermediate resistance were first reported in New Guinea and Australia and highly resistant ones first in South Africa, resistant organisms were soon there- after identified in Europe and the United States. Curiously, the highly resistant strains have not been further reported in states other than Minnesota (Center for Disease Control, 1979b). Fortunately, alternative chemotherapy exists for those bacteria with resistance to both tetracyclines and penicillin. Examples of alternatives to the first-choice drugs usually used for various diseases will be found in the literature (Anonymous, 1978b). The disadvantages in having to use alternatives vary from case to case and are related to the organisms to be com- batted, the alternative doughs), and the patient to be treated. An additional aspect of importance is the 10% to 15% frequency of patients with alleged sensitivity to penicillin and to whom the drug cannot be given. An examination of alternative drugs to combat penicillin- and tetracycline-resistant pneumococci reveals problems in therapy caused by antibiotic resistance quite well. In order of decreasing desirability, alternatives include the cephalosporins, which can be administered orally or parenterally, but are less active; chloramphenicol, which has the danger of fatal aplastic anemia in approximately 1 in 30,000 individuals; vanc~mycin, which is the most active but can only be administered intravenously and is ototoxic; and erythromycin, which, while being the least toxic, is also the least active. Resistance to the tetracyclines and penicillins has not been a recognized problem in those patients more prone to infec- tion or more severe disease, e.g., those who have congenital or acquired impairment of antibody- (hypogammaglobulinemia) or cellular- (Hodgkin's disease, sarcoidosis) mediated immunity, or both (chronic lymphatic leukemia). Other patients are susceptible because their defenses are deliberately compromised by azathio- prine or prednisone to prevent host-versus-graft reaction, as in kidney or heart transplant patients. Lastly, there are other patients, e.g., those with malignancy, whose antineoplastic ther- apy, e.g., chemotherapy or radiologic therapy, has the undesired and unpreventable side-effect of such compromise. Although infants and the elderly are considered fragile and susceptible, one can contend that either group has an advantage over populations of less extreme ages: the infants because they

OCR for page 65
74 handle stressful procedures well and possess transplacental material antibody, and the elderly because of their past expo- sure over many years to infectious agents and their development of an imposing immune globulin or other nonantibody defenses. However, one would not question the assertion that there is a distinctive pattern of susceptibility to infections in each group: the infant to the Gram-negative Enterobacteriaceae, Haemophilus influenzas, and the Group B S. pyogenes, and the elderly, notably those hospitalized, to S. pneumonias and the Gram-negative Enterobacteriaceae. Increased frequency of infection, especially in those colonized with resistant bacteria, had been anticipated. There have been many studies of two groups of patients that have re- ceived antimicrobials daily for many years in doses considered suboptimal for treatment of active disease. One group consists of cystic fibrotic patients in whom duration and quality of life have been improved with such drugs as a tetracycline or chloram- phenicol (Batten, 1976~. Recurrent infection with P. aeruginosa, especially the highly mucous variant, is an acknowledged problem, but it is difficult to attribute this directly to chemotherapy (Mearns _ al., 1972~. The second group is composed of patients taking tetracycline for either acne vulgaris (adolescents) or acne rosacea (middle-aged and elderly patients). Although tetra- cycline-resistant organisms can be isolated readily from these patients, more frequent or more resistant disease has not been documented (Schmidt et al., 1973~. Resistance in microorganisms occurs in four patterns, which may overlap: Bacteria are unaltered in their pathogenicity or other characteristics (e.g., resistance to streptomycin or sulfonamide) Some bacteria lose special enzymes (e.g., catalase and peroxidase) and pathogenicity (e.g., resistance to isoniazid). ~ Plasmid-mediated (nonchromosomal) antibiotic resistance may be acquired by and transferred among many Enterobacteriaceae. This is the most recently observed pattern and the most alarming one. Development of a specific enzyme (beta-lactamase) by Enterobacteriaceae and S. aureus that inactivates penicillin.

OCR for page 65
75 Resistant strains become more prevalent when antibiotics are used frequently and/or in high doses (Finland, 1979~. Often a resistant strain may develop during the course of anti- biotic treatment given to a patient originally infected with a sensitive strain. It is well known that a patient entering a hospital with a urinary tract infection usually has sensitive E. colt, whereas the patient hospitalized for only 3 or 4 days who develops such an infection has a much more resistant isolate, presumably acquired from the hospital flora which has been sub- jected to selection pressure by the use of antimicrobials. To a considerable degree this is a reflection of in-hospital therapy. Surprising, however, is the fact that extensive out-of-hospital use of a drug, e.g., for acne, has not resulted in noticeable increases in diseases from the emergence of resistant organisms. Primary or emergent resistance has not been attributable to antibiotic use in animal feeds. Nor does it seem possible to attribute a hospital outbreak to therapeutic or subtherapeutic use in those feeds. Lack of response to an antimicrobial may be the result of an inappropriate choice of agent. This is most likely when the causative organism for a disease has not been fully characterized. A number of examples of such diseases whose cause has only recently or has not been recognized come to mind. They include Legionnaires disease (Legionella pneumophila), nongonococcal urethritis (Chlamy- dia sppe ~ ~ infantile pneumonia (Chlamydia trachomatis), Lassa fever (Lassa virus), Marburg disease (Hamburg virus), infantile botulism (Clostridium botulinum), hepatitis (non-A/non-B hepatitis viruses), Norwalk gastroenteritis (Norwalk virus), and enterocolitis caused by Yersinia enterocolitica, Clostridium difficile, or Campylobacter fetus. To choose the optimal therapeutic agent, one must know the cause of a disease or infection and its antibiotic susceptibility. For these new diseases as well as the older diseases, the selection of antimicrobial agents must be continually reviewed; the pattern of sensitivities of isolates in the hospital or com- munity microbiological laboratory must be evaluated at least yearly, laboratory by laboratory; and the practice of chemotherapy must be monitored, adjusted, and changed continuously. Since the discovery of antimicrobials approximately 40 years ago, the changes in antimi- crobial use have been far too numerous to list, but some examples can be cited: the drift toward tetracycline from penicillin to treat N. gonhorrheae; the increased use of chloramphenicol and the decreased use of penicillin for acute otitis media owing to the frequency of resistant H. influenza; the resurgence of vancomycin after a hiatus of almost 20 years because of newer uses and more

OCR for page 65
76 resistant microorganisms; and the changing pattern in the immediate treatment of meningitis (i.e., treatment before the causative species are identified). Sulfonamide was the first agent of choice, but it was replaced by penicillin G which was itself replaced by a combination of penicillin, sulfonamide, and chloramphenicol. This combination gave way to ampicillin, then to a combination of ampicillin and chloramphenicol, which is commonly used today. In many instances the reason for change has been develop- ment of a more efficacious and/or less toxic drug by industry. Two examples of such improved drugs are the aminoglycosides and the cephalosporins. The development of the former spans almost 35 years, beginning with streptomycin, neomycin, and dihydrostreptomycin, progressing to kanamycin, then to amika- cin, gentamicin, tobramycin, and sisomicin. The history of the cephalosporins covers about half as many years, but the number of new agents is far greater. The two earliest agents, e.g., ~ ~ ~ ~ ~ But other drugs, e.g., novobiocin, paromomycin, paraminosalicylic acid, colist~methate, bacitracin, and ristocetin, have disappeared from systemic formularies. But who could predict that they will never reemerge? streptomycin and cephalothin, retain some usefulness. There is no reason to anticipate a radical departure from the past in the development of emergent and primary resistant organisms, of new and challenging diseases, of better and better drugs, and of ebbs and flows in drug selection and usage. Thus, we can expect a continuing production of newer vaccines, most imminently for hepatitis B antigen. Most likely vaccines will not replace the need for antimicrobial therapy of overt disease. Rather, both the prophylaxis against and treatment of infectious diseases will continue to be employed, neither replacing the other. r

OCR for page 65
77 REFERENCES American Thoracic Society. 1974. Preventive treatment of tuber- culosis. A general review. Pp. 29-106 in G. Canetti, ea., H. Birkhauser and H. Block, co-eds. Advances in Tuberculosis Research. Volume 17. S. Karger, N.Y. Anonymous. 1978a. Malaria (plasmodia). The Medical Letter 20~1) (Issue 496~:20-21. Anonymous. 1978b. The choice of antimicrobial drugs. The Medical Letter 20~1) (Issue 496~:1-8. Batten, J. 1976. Chemoprophylaxis of respiratory infections. Postgrad. Med. J. 52:571-575. Bauer, D. J., L. St. Vincent, C. H. Kempe, and A. W. Downie. 1963. Prophylactic treatment of smallpox contacts with N-methylisatin 6-thiosemicarbazone. Lancet 2:494-496. Center for Disease Control. 1979a. Follow-up on poliomyelitis-- United States, Canada, Netherlands. Reported by A. van Wezel; ino first initial] van Zermarel; S. Acres; State epidemiolo- gists from Iowa, Missouri, Pennsylvania, and Wisconsin; and the Center for Disease Control. Morbid. Mortal. Weekly Rep. 28:345- 346. Center for Disease Control. 1979b. Isolation of drug-resistant pneumococci--New York. Reported by S. Landesman, V. Ahonkahai, M. Sierra, H. Bernheimer, R. Goetz, A. Josephson, G. Pringle, G. Schiffman, P. Steiner, J. S. Marr, and the Center for Dis- ease Control. Morbid. Mortal. Weekly Rep. 28:225-226. Chodak, G. W., and M. E. Plaut. 1977. Use of systemic antibiotics for prophylaxis in surgery: A critical review. Arch. Surg. 112:326-334. Ferebee, S. H. 1970. Controlled chemoprophylaxis trials in tuber- culosis. A general review. Pp. 29-106 in G. Canetti, ea., H. Birkhauser and H. Bloch, co-eds. Advances in Tuberculosis Research. Volume 17. S. Karger, N.Y. Finland, M. 1979. Emergence of antibiotic resistance in hospitals, 1935-1975. Rev. Infect. Dis. 1:4-21.

OCR for page 65
78 Garibaldi, R. A., R. E. Drusin, S. H. Ferebee, and M. B. Gregg. 1972. Isoniazid-associated hepatitis. Report of an out- break. Am. Rev. Resp. Dis. 106:357-365. Hansman, D., H. Glasgow, J. Sturt, L. Devitt, and R. Douglas. 1971. Increased resistance to penicillin of pneumococci isolated from man. N. Engl. J. Med. 284:175-177. Mearns, M. B., G. H. Hunt, and R. Rushworth. 1972. Bacterial flora of respiratory tract in patients with cystic fibrosis, 1950-1971. Arch. Dis. Childhood 47:902-907. Philippon, A. M., M. G. Plommet, A. Kazmierczak, J. L. Marly, and P. A. Nevot. 1977. Rifampin in the treatment of experi- mental brucellosis in mice and guinea pigs. J. Infect. Dis. 136:482-488. Schmidt, H., E. From, and G. Heydenreich. 1973. Bacteriological examination of rectal specimens during long-term oxytetracy- cline treatment for acne vulgaris. Acta Denmatol. Venerol. 53:153-156. Weinstein, L. 1947. The spontaneous occurrence of new bacterial infections during the course of treatment with streptomycin or penicillin. Am. J. Med. Sci. 214:56-63. Woodward, T. E., and J. E. Smadel. 1964. Management of typhoid fever and its complications. Ann. Intern. Med. 60:144-157.