Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
100 col i form bacteria isolated from chickens a fter discontinuation of tetracycl ine-supplemented feed . The potent ial for spread of antimicrobial -res i stant ï¿½ &Q1i between farm anima l s and from farm anima l s to farm workers and the environment was further demonstrated in very recent experiments by Levy and Marsha l l . 1 1a In thi s study a ca l f was fed a marked stra in ( conta ining both chromosomal ly mediated nal idixic acid res istance and a large plasmid encoding multiple antimicrob ial res istance including that of tetracycl ine ) capable o f colon i z ing the human as wel l as bovine intestina l tracts . In the absence of any ant imicrob ial admini stration the marked stra in was detected in the feces of the cal f , and of another bovine kept in an adj acent stal l , for at least 3 months . The same market stra in was also present in the excreta of mice kept caged in the sta l l with the cal f and in fl ies trapped in the farm . In addition , two farm workers caring for the bovines began excreting the marked stra in in the stoo l s 4 -7 days a fter the experimental stra in had been fed to the cal f . Colon i z ation o f the intestinal tract o f these two farm workers , who were not rece iving antimicrob i a l s continue for 3 0 - 4 5 days . That antibiotic-res i stant col i form organisms of farm origin somet imes can cause disease in humans was suggested by Humme l and col leagues . 1 0 They studied a pig- farming in a de fined territory in which the streptothricin antibiotic nourseothricin was added to pig feed to promote growth . After 2 years o f nourseothricin use in p ig feed , they reported that col i form organisms conta ining plasmids encoded for nourseothricin res istance were found in 3 3 % of the isolates from fecal cultures from pigs with diarrheal disease , in 18% in those from employees o f the pig farms , 17% among isolates from fami l ies o f employees , and 1 6 % in those of outpatients l iving in nearby communities . Al though no nourseothricin had been used in the human population in the territory , 1 % o f the isolates from urinary tract infect ions of outpati ents were nourseothricin-resistant EÂ· coli . Examinat ion o f cultures from pigs , farm employees , and outpati ents in neighboring territories that did not use nourseothricin in pig feed revealed no nourseothric inÂ res istant EÂ· ï¿½ . Much o f the important informat ion needed to evaluate the results of the study of Hummel et al . is l acking . The dates of the study were not speci fied , nor were demograph ic data on the territories studied wel l de f ined . There was no information on the dens ity o f the pig popu l ation or the human popul ation , and it has been d i f f icult to assess the degree of contact with pigs on d i f ferent farms in the var ious popul at ions studied . S imil arly , the timing o f cul tures and the ir study thereof for nourseothricin res istance in the various populat ions was not spec i f ied . However , the ava i l able data do suggest that nourseothricin-re s istant
101 ï¿½ . QQli was transmitted from p igs t o humans and from humans to other humans . Once the human gastrointestina l tract had been found to be col on i z ed , it was not surpri s ing that the organisms were occasional ly found in urinary tract infect i ons 1 however , it was not determined whether the organi sms were more or less virulent than ï¿½ . ï¿½ not res istant to nourseothricin . Parsonnet and Kass 14 compared the antib iotic-res istance patterns of EÂ· 2211 isolated from the urine of bacteriuric female s laughterhouse workers with those of EÂ· QQli from poultry in the process ing l ine . ï¿½ . ï¿½ was found in 9 5 % of the cultures from poultry 1 9 6 % of them were res istant to antibiotics , and 8 7 % were res istant to more than one antibiot ic . The microorganisms i solated from the bacteriuric women ' s urine , however , only infrequently showed s imilar res istance patterns or ident ical patterns with those o f the microorgani sms from the poultry to which they were heav i ly exposed . Unfortunately , the bacteria in the women ' s feces were not studied , so direct spread from the process ing l ine to the ir gastrointestinal tracts could not be determined . Such a direct analys is o f antimicrobial res i stance patterns in fecal ï¿½ ï¿½ stra ins ( rather than stra ins causing urinary tract infections ) would seem necessary to document spread from animal foodstuff to humans , since ï¿½ QQli stra ins caus ing human urinary tract infections represent only a sma l l nonrandom group of clones , not found with equal probab i l ity among those which colon i z e the intestinal tracts of humans and anima l s . ï¿½ col i stra ins caus ing urinary tract infections in individual s without underlying microbiologic abnormal ities belong to a l imited number o f O . K . serogroups and possess spec i f ic v i rul ence factors . l O a , 2 4 a The ant ibiotic-resi stance plasmids o f the poultry and s laughterhouse workers from the latter study were examined for matching restriction endonuclease digestion- fragment patterns ( T . F . O ' Brien , 19 8 8 , personal communication ) . If a plasmid had been endemic among the poultry i solates , as found earl ier among cattle isolates of Salmonella typhimurium var . copenbaqen , its presence o r absence i n the human i solates would support or argue aga inst the s ï¿½ read of drug res i stance from the poultry to the workers . l O , l In fact , the same plasmid could be found in two isolates in only a few instances , so the result had l ittle power to exclude the pos s ib i l ity of spread . Extens ive spread , however , might have been expected to yield human isolates with h igher rates of res i stance or more ant ibiotypes closely matching those o f the poultry i solates . The human health ha zards attributable to infection with mul t iple-drug-re s istant ï¿½ . QQli of animal origin were studied more directly by Habte-Gabr and col l eagues in Iowa . 8 In 1 9 7 2 - 19 7 3 , they studied 14 8 Iowa fami l ies : 5 1 fami l ies exposed to l ivestock given antibiotic-supplemented feed , 4 3
102 rural fami l ies with n o exposure t o l ivestock , and 5 4 urban fami l ies . Multiple-drug-resistant ï¿½ . QQli was found in 1 5 % o f the stool cultures from members o f animal-exposed fami l ies , in 6% o f those from members o f rural fami l ies not exposed , and in 7 % of those from members of urban fami l ies . A fol l ow-up health survey was conducted 12 years l ater with 1 2 6 o f the original 14 8 fami l ies . The incidence o f serious infections was 6% in members of rura l fami l ies exposed to l ivestock , 1 3 % in members of rural fami l ies not exposed to l ivestock , and 1 2 % in members of urban fami l ies . Thus , colon i z ation by multiple-drug-resistant E . Â£211 o f farm origin did not appear to be a factor in infection in members of those populations . More extens ive bacteriologic studies were not carried out and the populat ions studied were not large , so the study could " detect " only a h igh l evel o f transmiss ion o f mul t iple-drug-res istant E Â· &Qli o f farm . origin that caused serious infections . If such spread o f infection occurs a t a l l , it i s l ikely t o be infrequent . SALMQNELLAE Most evidence l inking human disease to multi -resi stant bacteria of farm origin has been found in salmone l l ae . Data deta i l ing the incidence and associated morbid ity and mortal ity of salmone l l a in fections in farmers , sl aughterhouse workers , and the ir fami l ies are not ava ilab l e . Comparison of case reports on farmers who used subtherapeut ic antibiotics as l ivestock feed additives with those on farmers who did not might be particularly informative . The only information ava i lable i s in the form of case reports or descript ions of sma l l numbers o f outbreaks in farmers and their fami l ies , but not in sl aughterhouse workers . The paucity of reports might suggest that the occurrence o f salmone l l a infection in the rural or urban population is rare , indeed . In a 1 0 -month study o f 2 7 9 second-grade farm chi ldren in a rural county of Virginia , 14 9 episodes o f diarrheal i l l ness occurred i n 9 7 chi ldren ; salmone l l ae were isolated from only one of over 4 0 0 stool samples in the 14 9 cases o f diarrhea . 5 The chi ldren were in two groups : 9 2 l ived o n commerc ial poul try farms , and 1 8 7 did not . The occurrence o f diarrheal episodes was almost ident ical in the two groups . Despite the high preval ence ( 2 7 % ) o f salmonel l a infect ion among the poultry fl ocks , onl y one culture-proven case of salmone l l a gastroenteritis ( antib iot ic suscept ib i l ities not known ) occurred , and it was in a chi l d who d id not l ive o n a poultry farm . Wi l l iams 2 5 described two veterinarians with pustular forearm lesions due to sa lmone l l ae ( ï¿½ . dubl in and ï¿½ . typh imurium) that occurred several days a fter they del ivered an infected sti l lborn cal f or cleaned a cow that had recently
103 aborted . But they did not describe the antimicrob ial susceptibi l it ies of the isolates . Through 1 9 8 0 , f ive outbreaks of human salmonel l os i s d i rectly l inked t o contact with farm anima l s have been reported . In the mid- 1 9 6 0 s a multiple-ant ib iot ic-re s i stant stra in o f ï¿½ . typhimurium ( phage type 2 9 ) caused a large outbreak of bovine infection in Great Britain . 1 Infect ion occurred in farmers , the ir fami l ies , and veterinarians who treated infected calves ; spread from an ima l s to humans was imp l i cated . Spread o f infection to da i ry cows led to 59 human cases o f mi lk-borne salmone l l a gastroenteritis . Prophylactic use , in healthy anima l s , of ant ib iot ics to which the epidemic Salmone l l a stra in was res i stant might have favored infection with the pathogen by reducing the numbers of compet ing nonpathogenic antibiotic-susceptibl e intest inal bacteria . Sa lmone l l a gastroenteritis occurred in a 1 2 -year-old canad ian farm boy who cared for an in fected da i ry cow and its new cal f . ? The stra ins o f ï¿½ - typhimurium isol ated from the cow and the boy were of the same phage type and ant ib ioticÂ res istance pattern ( resi stant to tetracycl ine and chl oramphenicol ) . Spread of infection from the cattle to the child was considered most l ikely . Admini stration of ant ib iot ics to the s ick cow by the farmer 5 - 6 days be fore h i s son became i l l might have led t o select ion o f sa lmonel lae with the a forementioned res istance pattern . However , poss ible use o f subtherapeut ic doses o f ant ibiot ics in feed was not mentioned . An outbreak of salmonel los i s involving several newly arrived calves on a connecticut farm occurred in 1 9 7 6 . 1 2 ï¿½ . he idelberg that was res istant to chloramphenicol , sul famethoxa zole , and tetracycl ine was respons ible . The farmer and his pregnant daughter cared for the s ick anima l s and became infected themselves . The daughter gave b i rth t o a son 9 days a fter the calves arrived on the farm ; 3 days a fter del ivery , her newborn infant developed salmone l l a gastroenterit i s and bacteremia . Infection spread in the nursery to two other bab ies , most l ikely by contact with nursery sta f f . The stra in of ï¿½ . heide lberg isol ated from three calves and the farmer had identical antimicrobial susceptib i l ities , and those i solated from the farmer ' s daughter and the three infants were very s imi l a r ( re s i stant to chloramphen icol , sul famethoxazole , and tetracyc l ine ) , but l acked res istance to neomyc in , streptomyc in , and kanamyc in . Informat ion on subtherapeutic or therapeutic use o f antibiotics in the calves is not ava i l able . In the l ate 1 9 7 0s , numerous outbreaks o f salmone l l o s i s due t o mul tiple-antibiotic-res istant ï¿½ - typh imurium o f phage types 2 04 and 1 9 3 occurred among calves on over 3 0 0 farms throughout Great Britain . 17 , 2 4 The two stra ins o f ï¿½ . typhimurium made up 2 8 % of a l l s . typhimurium isol ates from
1 04 cattle that were sent to the central Publ ic Health Laboratory for evaluation in 1 9 7 8 . The same two stra ins were l ater Â· i solated from 2 1 1 human infections , including one that ended fatal l y , in the British Isles . In most human cases , no apparent connection with cattle could be found , but the same stra ins were l ater isolated from minced meat and sausage , suggest ing entrance at some point into the human food supply . However , 3 0 of the human infect ions occurred in persons on farms where outbreaks of bovine infection with the multipleÂ drug-res istant stra ins of ï¿½ . typhimurium were occurring or had prev iously occurred . 17 Before 1 9 7 7 , the predominant-ant ibiotic res istance pattern in s . typhimurium of phage type 2 0 4 respons ible for several outbreaks of salmonellosis in cattle and humans in Great Britain cons i sted of res istance to sul fonamide ( nontrans ferable ) and tetracycl ine ( not directly trans ferable , but mob i l i z able by F-l ike plasmids ) . In 1 9 7 7 , a stra in o f phage type 2 04 that had ga ined an addit ional transmissible R plasmid ( H2 compatibil ity group ) bearing res istance to chloramphenicol ( C ) , streptomycin ( Sm ) , sul fonamide ( Su ) , and tetracycl ine ( T ) was responsible for a sma l l outbreak of salmonellosis on a farm in Leicestershire . The farm was involved extens ively in cal f-trading , which resulted in wide distribution of calves infected with the multiple-drug-resistant strain . It was thought that acquis ition of the new H2 plasmid probably resulted from select ive pressure introduced by the use of chloramphenicol in treatment o f a cal f infected with a type 2 04 stra in that had the original R plasmid ( SuT ) , which was predominant be fore 1 9 7 7 . Alternatively , the multiple-drug-res istant ( C Sm Su T) plasmid might have been brought in with a newly purchased , already infected animal that could have been introduced into the herd shortly before the outbreak . 2 4 An outbreak of multiple-drug-resistant salmone l l a infect ions involving three o f four members of a family who worked on a da i ry farm in Kentucky occurred in 1 9 7 7 . 4 Infection appeared to have been transmitted through ingestion o f unpasteuri z ed milk . These data are not sufficient to support any conclus ions concerning the relative incidence of infect ions with salmonel lae ( e ither ant ib iotic-susceptible or -re s i stant ) in farm workers or the ir fami l ies , compared with other populat ion groups . Data are not ava i l able to this committee on the frequency or severity of infections with salmonel lae ( e ither antibiotic-susceptible or -resistant ) in s laughterhouse workers . In the f ive salmone l l a outbreaks described above , there are no data on the role o f subtherapeutic use of antibiot ics in feed , although the use of therapeutic dosages of antibiotics in the f i rst four was cons idered to have important e f fects . S ince 1 9 8 0 , several additional outbreaks of multiple-
105 druq-resistant salmonel losis provide some evidence , and in one case compel l inq evidence , that the res istant salmonel l ae oriqinated in farm animals fed antimicrobial druqs . Holmberq et al . 9 reported on an outbreak of ï¿½ . newport res istant to ampici l l in , carbenici l l in , and tetracycl ine that occurred in several midwestern states . Food histories and plasmid profiles of the orqanisms isolated from both humans and animals l ed the authors to conclude that the resistant orqanisms infectinq the patients were of animal oriqin and that the probable source was contaminated hamburqer , the meat of which was derived from a sinqle herd . The subtherapeutic use of chlortetracycl ine in this herd was admitted by the farmer , but thi s has not been analyzed or proven . Althouqh the editorial that accompanied the report of Holmberq et al . 9 suqqested that the study provided the " important miss inq l ink" between human disease and res istance in the infectinq bacteria due to the feedinq of subtherapeutic antibiotics to anima l s , the evidence is incomplete . ( Note : Dr . Holmberq ' s comments , in personal communication , about this article are inserted parenthetica l ly below . ) First , as pointed out by DuPont and Steele , 5 the pathoqenic bacterial strain was not recovered from the s lauqhterhouse or from the hamburqer ( al l the hamburqer meat had already been consumed and none of the s lauqhterhouse animals were ava i l able for study ) , and no cases of ï¿½ - newport disease occurred in the cattl e or in the people associated with the farm that reared the animals or the process inq plants ( l ivinq cows rema ininq on the farm were excretinq ï¿½ newport ) . Second , another processinq plant in another state rece ived hal f the carcasses from thi s herd of cattle and had no apparent problem ( actual ly there were only 12 anima l s out of the 1 0 5 animal herd sent to another state : that cases traceable to these were not uncovered may only mean that some or a l l of these 12 animal s were not infected or that i l l persons were not ascertained or reported ) . Third , the only ï¿½ . newport isol ated from an animal and of a stra in identical with the outbreak strain was isolated from a cal f that died in an adj acent dairy herd . That cal f miqht have been the source of the infection ( this cal f was not the only animal from which ï¿½ newport was isol ated , as stated above , some of the cows on the farm were excretinq the bacterium ) . More recently , an outbreak of multiple-druq-resistant ï¿½ . newport in Cal i fornia in 1 9 8 5 convincinqly demonstrated the entire cha in of transmiss ion . 2 2 The outbreak strain was res i stant to chl oramphenicol , tetracycl ine , kanamycin , ampici l l in , and sul fisoxazole and was characterized by a s i nqle larqe plasmid . Epidemioloqic studies ident i f ied qround bee f as the suspect food vehicle , and many of the patients had consumed the qround bee f at fast- food restaurants . Microb ioloqic and epidemioloqic studies traced
106 the epidemic strain through the hamburger , back t o meat process ing plants , and ultimately back to the farms from which the animals were sent for s laughter . The isolates were from i l l calves and cows at a number of dairies in important dairy- farming areas . I solation of chloramphenicol-resistant salmonel lae was associated with chloramphenicol use at those dairies . Such use of chl oramphenicol as a feed additive is not approved by the Food and Drug Administration . Several recent milk-borne outbreaks of multiple-drugÂ res istant salmonel lae provide additional information , but do not directly l ink the organism to a farm source , or to subtherapeutic use of antibiotics . Tackett et al . 2 3 reported an outbreak of multiple-drug-resistant s . typhimurium that occurred in Ari zona caused by the ingestion of raw mi l k . This bacterial strain was isol ated from the raw milk samples . Further investigations into the source of contamination was not done because the impl icated da iry withdrew the product from the market and would not permit any examinat ion o f the fac il ity , its employees , or its anima l s . The largest outbreak o f salmonellosis ever recorded in the United States occurred in I l l inois and Wiscons in in 1 9 8 5 and involved over 1 6 , 0 0 0 bacterial -culture-confirmed cases . 18 In these studies , the est imates of cases derived from telephone surveys estimated the actual number of people infected was about 1 7 5 , 0 0 0 . The epidemic strain was res istant to amp ic i l l in , tetracycl ine , carbenici l l in , streptomycin , sul fisoxazole , erythromycin , and penic i l l in . The outbreak ultimately was traced to two brands o f pasteurized m i l k produced by a s ingl e dairy plant . The source of the infecting bacterial organism was presumed to be the da iry cattle , although this could not be demonstrated conclus ively because no isol ates e ither from the da iryÂ anima l s or the farm had exactly the same plasmid profile as the stra in isol ated from the mi lk . One might speculate on whether those outbreaks due to mul tiple-drug-res istant salmone l l ae might sti l l have occurred had the salmonel lae been ful ly susceptible . The cause of each of the outbreaks appeared to be de fects in food process ing or inappropriate food preparation , rather than be ing due to the fact that the salmone l l ae were multipleÂ drug-resistant . Such de fects would a l l ow the pers istence of any salmone l l ae , whether ant ibiotic susceptibl e or res istant . Hence , it would be diff icult to argue that the outbreaks would not have occurred at a l l had the salmonel l ae been fully susceptibl e . A recently reported outbreak of egg-associated ful lyÂ drug-suscept ibl e ï¿½. enteritidis infections underscores the ease with which salmone l l ae can enter the food chain , in spite of the usual food processing and food preparation sa feguards . Epidemiologic data suggested that , rather than the usual mechanism of contamination of the shel l s ,
107 especia l ly cracked shel l s , by salmonella-conta ining chicken feces , the mode of transmission was transovarial , with infection of the yolk be fore shel l depos ition . That mechanism would thwart the usual method to decontaminate eggs ; and any use of such eggs that involved l ittle or no cooking--e . g . , use with hol landaise sauce , eggnog , or Caesar salad dress ing--would l ikely result in cases of human salmonel losis . The evidence of farm-to-human spread of salmonel lae derived from the study of outbreaks should be put into the broader context of the overal l epidemiology of human salmonel losis . In a study of farm children regularly exposed to poultry , Marx 1 3 could find no evidence of a greater occurrence of salmonel losis and diarrheal i l lness than in a control group . In a multivariate analys i s o f cl inical and epidemiologic features of multi-drug-resistant salmone l l ae caus ing salmone l los is in humans , Riley and col l eagues 1 6 found no evidence that exposure to anima l s or pets was a s ignificant risk factor . Risk factors identi fied in their investigation included the recent use of antimicrobial agents by patients , a Hispanic ethnic background , regular antacid use , and age over 60 years . Thus , although transmiss ion of salmonel lae from farm anima l s to humans has been documented in several instances , it is not frequently recognized . It would appear that the best protection aga inst multiple-drugÂ res istant salmonel lae is the same as that aga inst ful ly suscept ibl e salmonel lae ; that is , accepted sanitat ion and steril i zation ( cooking ) techniques of food processing and food preparation . OTUER ENTERIC PATHOGENS The ava i lable information on three enteric pathogens-Â enterohemorrhagic ï¿½. 2211 , Yersinia enterocolitica , and campyl9bacter spp . --is sparse , but they are respons ible for important cl inical infections and should be mentioned . The committee did not search the data files of cl inical or diagnostic laboratories in the United States for information on those organisms , but has rel ied on publ ished summaries . The committee acknowledges that the hazard associated with transmiss ion to humans of these bacteria that might have originated on the farm cannot be evaluated . Enterohemorrhagic ï¿½ . 2211 was recogni z ed in 1 9 8 2 as a maj or etiologic agent of the syndrome of hemorrhagic col itis , a diarrheal syndrome characteri zed by rapid progress ion from watery to bl oody diarrhea and marked by severe morbidity . 3 , 1 5 , 1 9 S igns of entero invas ive infection , such as fever or the presence o f fecal polymorphonuclear l eukocytes , are usual ly l acking or are not prominent . A particular serotype o f ï¿½ . QQ!i , 0 1 5 7 : 87 , is espec ial ly associated with
108 the syndrome and has been found to produce cytotoxins s imilar to the shiga toxin of Shigella dysenteriae type 1 . outbreaks of hemorrhagic col itis due to enterohemorrhagic ï¿½ . 2211 have occurred in persons of a l l ages , but have been prominent i n elderly res idents o f nurs ing homes . Deaths were frequent in nurs ing home outbreaks . Several outbreaks have been traced to the consumption o f beef and dairy products , and the organisms have been isolated from cattle . Thus , cattle are suspected of being a maj or reservoir . There is no comprehensive information on the epidemiology of the syndrome , and national surve i l lance data do not yet exist . Antibiotic res istance of enterohemorrhagic ï¿½ . QQli has not been an i ssue . Indeed , the role of antibiotics is paradoxical : their use appears to be a risk factor for development of hemorrhagic col itis if they are given during exposure : but they appear to have l ittle therapeutic value in the disease , in that the pathogenesis is toxin-mediated , rather than enteroinvas ive infection . 1 5 Infection caused by Yersinia enterocolitica , although moderately common in some European countries , is rarely recogni zed in the United States . There might be serious underdiagnos i s of infection caused by this species , but far more extens ive data would be needed to establ ish yers inios is as an important cl inical problem in this country . Most isolates are susceptible to tetracycl ine , although ampici l l in res istance is common . The committee knows o f no nationwide database that permits estimation of the incidence of infect ion with x . enterocolitica . Results of recent surveys in several areas of the United States suggest that Campylobacter spp . might cause at least as much illness and death as salmonellae . However , nationwide data on infections caused by campylobacters are not ava ilable yet , and antimicrobial resistance is not a recognized issue in the treatment of infections with them . 2 , 6 , 2 0 , 2 1 In summa ry , studies have i ndicated spread into farm workers and the ir famil ies of ï¿½ QQli originating in farm animals and poultry . I f a drug-resistant enteric flora is selected in farm animals or poultry by the use of ant ibiot ic supplemented feeds , the drug-resistant enteric flora might spread into the farm workers , their famil ies , and ult imately , to some extent into the community at large . There is no evidence , however , to suggest that drug-resistant ï¿½ Â£211 of farm origin are more infective or more virulent than drugÂ susceptible ï¿½ QQli of non- farm origin . Farm workers and their famil ies have not been found in l imited studies to have an increase in serious infections with diarrheal diseases , as compared to the population at large . There is evidence , derived from the study o f food-borne outbreaks of salmonel losis , that the causative salmonellae ï¿½-- - ï¿½ - Â·
109 were o f farm oriqin , and entered the human food cha in . In a number of outbreaks of multiple druq-resistant salmonellos i s , an animal or poultry source was impl icated , and the multiple druq-res istance was bel ieved to be due to the use of antibiotics in animal feeds . I n only one such outbreak was the evidence compell inq , with ful l documentation of the entire chain o f transmiss ion from infected cattle to infected humans . BEFEBENCES 1. Anderson , E . s . Druq res istance in Salmonella Typbimurium and its impl ications . Brit . Med . J 2 : 3 3 3-339 , 1968 . 2. Blaser , M . J . , J . G . Wel l s , R . A . Feldman , R . A . Pol lard , and J . R . Al len . The col laborative diarrheal disease study qroup : Campylobacter enteritis in the United States . A multicenter study . Ann . I nt . Med . 9 8 : 3 6 0-3 65 , 1 9 8 3 . 3. Carter , A . o . , A . A . Borczyk , J . A . K . Carl son , et al . A severe outbreak of Escberichia QQli 0 1 5 7 : H7 - associated hemorrhaqic col itis i n a nurs inq home . N . Enql . J . Med . 3 17 : 1 4 9 6 - 1 5 0 0 , 1 9 8 7 . 4. Centers for Disease Control . Salmonel losis-Kentucky . MMWR 2 6 : 2 3 9 , 1 9 7 7 . 5. DuPont , H . L . , and J . H . Steele . Use o f antimicrobial aqents in animal feeds : Impl ications for human health . J . Infect . Dis . 9 : 4 4 7 - 4 6 0 , 1 9 8 7 . 6. F inch , M . J . , and L . w . Riley . Campylobacter infections in the United States . Arch . Int . Med . 1 4 4 : 1 6 1 0 - 1 6 1 2 , 1984 . 7. Fish , N . A . , M . c . Finlayson , and R . P . Carere . Salmonel los i s : Report of a human case fol lowinq direct contact with infected cattle . can . Med . Assoc . J . 96 : 1163 -1165 , 1967 . 8. Habte-Gabr , E . , I . M . Smith , F . w . Gutman , et al . Effect of Antibiotics in Livestock Feed on Human and Animal ï¿½ . QQli Carriaqe and Human Diseases . Unpubl i shed manuscript . Department of I nternal Medicine , Univers ity o f Iowa ( by permission of I . M . Smith ) .
110 9. Holmberg , s . D . , M . T . Osterholm , K . A . Senger , and M . L . Cohen . Drug-resistant Salmonel l a from animals fed antimicrobials . N . Engl . J . Med . 3 1 1 : 6 17 - 62 2 , 1 9 8 4 . 10 . Hummel , R . , H . Tschape , and w . Witte . Spread of pl asmid-mediated nourseothricin res istance due to antibiotic use in animal husbandry . J . Bas ic Microbiol . 2 6 : 4 6 1-4 6 6 , 19 8 6 . lOa . Johnson , J . R . , et al . Aerobactin and other virulence factor genes among stra ins of ï¿½ 2211 caus ing urosepsis : Association with patient characteristics . Infect . Immun . 5 6 : 4 05 -4 2 , 1 9 8 8 . 11 . Levy , s . B . , G . B . FitzGerald , and A . B . Macone . Changes in intestinal flora of farm personnel a fter introduction of a tetracycl ine-suppl emented feed on a farm . N . Engl . J . Med . 2 9 5 : 58 3 - 5 8 8 , 19 8 6 . l la . Levy , s . B . and B . M . Marshal l . Genetic trans fer in the natural environment , pp . 6 1 -7 6 . I n M . Sussman , c . H . Col l ins , s . K . Skinner , and D . E . Stewart Tul l , Eds . The Release of Genetica l ly Engineered Microorganisms . London , Engl and : Academic Press , 1 9 8 8 . 12 . Lyons , R . w . , c . L . Samples , H . N . DeS i lva , K . A . Ross , E . M . Jul ian , and P . J . Checko . An epidemic of res istant Salmonella in a nursery : Animal -to human spread . J . Amer . Med . Assoc . 2 4 3 : 54 6-54 7 , 19 8 0 . 13 . Marx , M . B . The effect of interspecies contact upon diarrhea morbidity and salmonel l osis in children . J . Infect . Dis . 1 2 0 : 2 0 2 -2 09 , 19 69 . 14 . Parsonnet , K . c . , and E . H . Kass . Does prolonged exposure to antibiotic-resistant bacteria increase the rate of antibiot ic-resi stant infection? Antimicrob . Agents Chemother . 3 1 : 9 1 1-9 14 , 19 8 7 . 15 . Ri ley , L . w . The epidemiologic , cl inical , and microbiological features of hemorrhagic col itis . Ann . Rev . Microbial . 4 1 : 3 8 3 - 4 0 7 , 1 9 8 7 . 16 . Ri l ey , L . w . , M . L . Cohen , J . E . Seal s , M . J . Blaser , K . A . Birkness , N . T . Hargrett , s . M . Martin , and R . A . Feldman . Importance o f host factors in human salmone l l os i s caused by mul tires istant stra ins o f Salmonella . J . Infect . Dis . 1 4 9 : 8 7 8 - 8 8 3 , 1 9 8 4 .
111 17 . Rowe , B . , E . J . Threl fal l , L . R . Ward , and A . s . Ashley . I nternational spread of multiresistant strains o f Salmone lla typhimurium phage types 2 0 4 and 1 9 3 from Britain to Europe . Vet . Rec . 10 5 : 4 68-4 69 , 19 7 9 . 18 . Ryan , c . A . , M . c . Nickels , N . T . Hargrett-Bean , et al . Mass ive outbreak of antimicrobial-res istant salmonel losis traced to pasteurized milk . J . Amer . Med . Assoc . 2 58 : 3 2 69 - 3 2 7 4 , 19 8 7 . 19 . Sack , R . B . Enterohemorrhagic Escherichia QQli . N. Engl . J . Med . 3 17 : 1 5 3 5 - 1 5 3 7 , 1 9 8 7 . 20 . Schmid , G . P . , R . E . Schae fer , B . D . Pilkaytis , J . R . Schae fer , J . H . Bryyner , L . A . Wintermeyer , and A . F . Kaufmann . A one-year study of endemic campylobacterios is in a midwestern city : Association with consumption of raw milk . J . Infect . Dis . 1 5 6 : 2 18 - 222 , 1987 . 21. Skirrow , M . B . Campylobacter enteritis : A " new" disease . Brit . Med . J . 2 : 9 - 1 1 , 1 9 7 7 . 22 . Spika , J . s . , s . H . Waterman , G . w . Soo Hoo , et al . Chloramphenicol-resistant Salmonella newport traced through hamburger to da iry farms . N . Engl . J . Med . 3 16 : 5 6 5 - 57 0 , 1 9 8 7 . 23 . Tackett , c . o . , L . B . Dominque z , H . J . Fisher , and M . L . Cohen . An outbreak o f multiple-drug res istant salmonella enteritis from raw milk . J . Amer . Med . Assoc . 2 5 3 : 2 0 58-2 0 6 0 , 1 9 8 5 . 24 . Threlfal l , E . J . , L . R . Ward , and B . Rowe . Epidemic spread of a chloramphenicol -resistant stra in of Salmonella typhimurium phage type 2 04 in bovine animals in Britain . Vet . Rec . 1 0 3 : 4 3 8 - 4 4 0 , 1 9 7 8 . 2 4 a . Vaisanen-Rhen , v . , et al . Plasmid clones among uropathogenic ï¿½ ï¿½ stra ins . Infect . Immun . 4 3 : 14 9 - 155 , 1984 . 25 . Will iams , E . Salmonella dubl in skin les ions in a veterinary surgeon . Lancet 2 : 7 3 7 -7 3 9 , 1 9 6 9 .
VI I THE RISK MODEL : OVERVIEW OF THE PROBLEM AND NEED FOR A MODEL The committee took as its princ ipal charge the quantitative assessment of hazards to human health from the subtherapeutic administration of penic i l l in/ampic il l in and the tetracycl ines to faï¿½ animals . The committee del iberately chose to consider the tetracycl ines and penicil l in G together , rather than separately , for several reasons : o Antimicrobial resistance in salmone l l ae and ï¿½ QQli to each of these drugs is predominantly plasmid-mediated . o S imul taneous resistance to both ampicil l in and the tetracycl ines is commonly found in the same individual animal isolates of salmonellae . Of 7 17 isolates of ï¿½ typhimuriun ( see Table V-2 ) , 5 2 % had tetracycl ines resistance , and 3 7 % had both tetracycl ine and ampic i l l in res istance . o Exposure to either penic i l l in G or a tetracycl ine of ï¿½ Â£Qli or salmonella stra ins bearing R plasmids that encode both tetracycl ine and ampici l l in res istance markers sel ects for such R-plasmid-conta ining stra ins in a mixed population . Thus , exposure to either of the two antibiot ics would enrich the population of microorganisms res istant to the other , as wel l as resi stant to it itsel f . o The tetracycl ines far exceed penici l l in G in use in l ivestock and poultry feeds . For example , in 1 9 8 5 , tetracycl ine accounted for 4 9 % of annual sales of antimicrobials for animal feeds , and penicil l in accounted for only 5% ( Table IV- 6 ) . Because penic i l l in use was only 1 0 % of that of the tetracycl ines , it did not seem to the committee that performing a separate risk analysis for pen ic i l l in G would provide use ful information . o In performing the risk assessment , the committee could not f ind evidence suffic ient to j usti fy the use of d i f ferent death rates for stra ins resistant to ampic i l l in ( penic il l in ) , as opposed to the tetracycl ines . This task requires the study of broad questions regarding the e f fects of drug res istance on the epidemiology o f various pathogens and diseases and the ef fects o f feeding subtherapeutic doses of antimicrobial agents on ( a ) the 112