Bacteremia is common in critically ill neonatal foals,

Transcription

1 J Vet Intern Med 2008;22: Bacteremia in Equine Neonatal Diarrhea: A Retrospective Study ( ) A.R. Hollis, P.A. Wilkins, J.E. Palmer, and R.C. Boston Background: Bacteremia in sick foals is associated with survival, but the association of bacteremia and diarrhea is not reported. Hypothesis: Neonatal foals with diarrhea will commonly be bacteremic. Animals: One hundred and thirty-three neonatal foals. Methods: Records of all foals o30 days of age presenting with diarrhea between January 1990 and September 2007 were reviewed. Results: Sixty-six of 133 foals (50%) were bacteremic at admission, with 75 isolates from the 66 samples. The blood culture from a further 18 foals (13.5%) grew coryneform bacteria. Nine foals (6.8%) had 2 or more organisms grown on blood culture. One foal had 5 different organisms, interpreted as contamination. Forty-eight foals (36%) had no growth on admission blood cultures. No cultures isolated fungal organisms. Excluding coryneform bacteria, 43 isolates (57%) were Gram-negative organisms and 32 isolates (43%) were Gram-positive organisms. The most common isolate was Enterococcus spp. (22 isolates, 29%), followed by Pantoea agglomerans (13 isolates, 17%). IgG concentration at admission was not associated with blood culture status. Blood culture status was not associated with survival to hospital discharge. Conclusions and Clinical Importance: Bacteremia is common in neonatal foals with diarrhea. Decisions regarding antimicrobial selection should be made with these differences in mind. Key words: Colitis; Foal; Gram-negative bacteria; Gram-positive bacteria; Septicemia. Bacteremia is common in critically ill neonatal foals, with 29 36% of foals presenting to a referral hospital being blood culture positive. 1,2 Because septicemia is the major cause of death in foals aged o7 days, 3 it is important to identify and appropriately treat these animals. Several studies have investigated bacteremia in the overall population of neonatal foals presenting to equine hospitals, 1,2,4 13 but no studies have investigated foals with diarrhea as a separate population. A previous study noted that foals with Actinobacillus spp. bacteremia were more likely to have diarrhea than were those with other isolates, 12 a result that was not replicated by a study of all bacteremic foals, 4 and therefore requires further investigation. It is important to identify the commonly isolated bacteria and susceptibility patterns in these foals, because the administration of effective antimicrobial drugs within a short time period is associated with increased survival in septic humans, 14,15 and the same could be true for bacteremic neonatal foals. 4 The aims of this study were to investigate the prevalence of bacteremia in neonatal foals with diarrhea presenting to a referral hospital, to investigate common isolates and their susceptibility patterns, to investigate the association of adequate transfer of passive immunity with bacteremia, and to investigate the association of bacteremia with survival. From the New Bolton Center, Kennett Square, PA. Corresponding author: P.A. Wilkins, New Bolton Center, 382 West Street Road, Kennett Square PA 19348; pwilkins@ vet.upenn.edu. Submitted February 15, 2008; Revised March 24, 2008; Accepted May 21, Copyright r 2008 by the American College of Veterinary Internal Medicine /j x Materials and Methods Study Animals Foals examined at New Bolton Center s George D. Widener Hospital for Large Animals between January 1990 and September 2007 were eligible for this study if they met the inclusion criteria. These criteria were presenting with a primary complaint of diarrhea or development of diarrhea within 12 hours of hospitalization, and aged 30 days or under at admission. Exclusion criteria were foals 430 days of age, those developing diarrhea 412 hours after hospitalization, and those where the clinical record was not available for evaluation. Diarrhea was defined as the presence of feces of abnormally liquid consistency on more than one occasion in 24 hours. Critically ill foals were defined as those presenting with systemic diseases. Collection of Laboratory Samples During the study period, it was the standard care to collect a blood sample for bacteriologic culture and immunoglobulin G (IgG) concentration, among other diagnostic samples, at admission of all critically ill neonatal foals. For blood culture collection, the area over over the left or right jugular vein was clipped and aseptically prepared with a 5-minute chlorhexidine scrub followed by wiping with 70% alcohol. Sterile latex gloves a and standard masks b were worn and a 16 g, over-the-wire IV catheter c was placed by the Seldinger technique. Seven milliliters of blood was aseptically collected through the IV catheter into a sterile syringe, a sterile needle was placed on the syringe, and the blood was aseptically introduced into 70 ml of trypticase soy broth d enriched with carbon dioxide and containing sodium polyanethol sulfonate as an anticoagulant. The broth was then incubated at 351C for up to 10 days. After 1992, samples were routinely incubated under conditions that favored aerobic growth but also permitted anaerobic growth, as previously described. 1 Susceptibility patterns of isolated organisms were determined by use of a Kirby-Baurer technique from 1990 to 1994, after which susceptibility patterns were determined with a breakpoint inhibitory concentration system. e IgG concentrations were determined with blood collected into heparin-plasma tubes and 2 different commercially available ELISA tests ( f ; g ), and then by a commercially available turbidimetric

2 1204 Hollis et al immunoassay. h In addition, all foals had feces collected for culture to detect Salmonella spp. by routine techniques, and 93 foals also had feces submitted for the detection of fecal Clostridium perfringens enterotoxin and Clostridium difficile A and B toxins via ELISA. i Data Recorded Data recorded in the medical record at admission were retrieved and entered into a computer data base. j Not all data were available for all foals. Historical Information The following information was recorded: signalment, gestational age, age at admission, any available information regarding parturition (if attended, whether normal, dystocia and degree of dystocia, if recorded), cesarean section, premature placental separation, whether a primiparous or multiparous mare, and if any problems were noted with the mare before or after parturition (eg, if the mare was reported to be unwell before or after parturition, running milk before parturition, placentitis, poor lactation, retained placenta). In addition, any noted problems other than diarrhea (colic, contracted tendons, slow to suck, recumbency, seizure-like activity), the duration of diarrhea, and treatments administered before referral were collected. Laboratory Data Data were collected on the following: blood culture status, organism identified, and sensitivity profile; IgG at admission; fecal culture results; and the identification of C. perfringens or C. difficile toxins or both. Statistical Analyses Chi-squared testing was used to investigate the association between IgG concentration at admission and preadmission antimicrobial use and blood culture status. The association of the blood culture status with survival was investigated using logistic regression. These analyses were performed by commercially available statistical software, k with significance set at P.05. Results One hundred and fifty-three foals met the inclusion criteria, of which 133 had blood culture results available for evaluation and constituted the study population. There were 74 colts and 59 fillies; median age at presentation was 4 days (range 0 30 days; n 5 133). There were 79 Thoroughbreds, 16 Standardbreds, 12 Quarterhorse/ Paint horses, 9 crossbreeds, 8 Arabs, 6 Warmbloods, 1 Appaloosa, 1 Hafflinger, and 1 Welsh pony. Historical Information Information about parturition was available for 85 foals. Ten mares were recorded as multiparous, and 8 were recorded as primiparous. Fifty-nine mares had a normal parturition, 9 had a mild dystocia, 3 had dystocia, 2 had prolonged deliveries but no dystocia, 3 had premature placental separation, and 7 parturitions were not observed. No foals in the study were noted to be born via cesarean section. Information on the appearance of the placenta was available for 52 foals, of which 47 were normal placentae, 2 were retained, 3 had premature placental separation, and 1 placenta was described as not normal with no further information recorded. In addition, 5 mares were noted to be dripping milk before parturition and 1 mare waxed 6 weeks before parturition, 2 foals were orphans being raised on milk replacer, and 2 were orphans with nurse mares. Three foals were born at the hospital in the high-risk pregnancy program, all of which were discharged and later presented with diarrhea. Five foals had been hospitalized previously (2 with diarrhea), and 1 foal had accompanied previously a mare to the hospital. One foal had been treated at a different hospital for meconium impaction before presentation with diarrhea. Six foals were in contact with other foals that had diarrhea. Other historical features noted in the foal clinical record were seizure-like activity (2), entropion (1), weakness (2), dull mentation (3), slow to start (3), never right (2), contracted tendons (2), incomplete ossification (2), flexural deformity (1), never nursed (1), and marked tendon laxity (1). Of the 133 study foals, 128 presented with diarrhea and 5 developed diarrhea within 12 hours of hospitalization. There was insufficient historical information to analyze whether abnormal parturition or periparturient problems were associated with increased risk of bacteremia at hospital admission. Medications Administered before Admission Ninety foals had information regarding medication administered before admission, of which 50 (55.6%) had received antimicrobials. Antimicrobials administered were ceftiofur (18 foals, 36%), penicillin (16 foals, 32%), gentamicin (11 foals, 22%), amikacin (7 foals, 14%), metronidazole (8 foals, 16%), trimethoprim sulfonamide (6 foals, 12%), chloramphenicol (1 foal, 2%), ticarcillin (1 foal, 2%), enrofloxacin (1 foal, 2%), ampicillin (1 foal, 2%), and cefuroxime (1 foal, 2%). There was insufficient historical information to evaluate if antimicrobials were administered before or after development of diarrhea and, therefore, it was not possible to analyze if any foals could have had antimicrobial-associated enterocolitis. Of the 50 foals that had been treated with antimicrobial drugs before admission, 15 (30%) had no growth on admission blood culture, 11 (22%) grew coryneform bacteria, and 24 (48%) had positive blood cultures, with 29 isolates. Of the 20 isolates for which susceptibility results were available, 12 (60%) were sensitive to the antimicrobials administered before admission and blood culture collection and 8 (40%) were resistant to the previously administered antimicrobials. There was no association between the administration of antimicrobials before admission and the blood culture status of these foals at admission (P 5.81). Blood Culture Status Sixty-six foals (49.6%) were bacteremic at admission, with 75 isolates. A single bacterial organism was isolated from 57 samples (86%), and multiple organisms from 9 samples (14%). The blood culture from a further 18

3 Foal Diarrhea 1205 Table 1. Bacteriologic isolates and their sensitivity patterns. Pantoea agglomerans Escherichia coli Salmonella spp. Klebsiella spp. Actinobacillus All Gram- Negative Isolates Enterococcus All Gram- Positive Isolates All Isolates # Isolates Amikacin (22/22) 100 (6/6) 100 (28/28) Amoxicillin (20/22) (10/11) 91 (30/33) clavulanate Cefoxitin (19/24) (2/11) 60 (21/35) Cefpodoxime (12/15) 0 9 (1/11) 50 (13/26) Ceftiofur (32/38) (7/24) 63 (39/62) Chloramphenicol (31/36) (19/26) 81 (50/62) Enrofloxacin (20/27) (7/11) 71 (27/38) Gentamicin (38/40) 86 (6/7) 94 (44/47) Imipenem (32/32) (7/9) 95 (39/41) Penicillin (22/26) 85 (22/26) Tetracycline (31/38) (19/26) 78 (50/64) Ticarcillin (30/38) (19/24) 79 (49/62) clavulanate Trimethoprim sulfonamide (30/38) 100 (4/4) 81 (34/42) Sensitivity to each antimicrobial is reported as the percentage of isolates tested that were sensitive to that antimicrobial, with numbers in parentheses indicating actual number of isolates sensitive compared with the number of isolates tested. foals (13.5%) grew coryneform bacteria (aerobic, Gram positive, catalase-positive rods morphology similar to Corynebacteria), and 1 foal grew 5 different organisms, interpreted as contamination. Excluding coryneform bacteria, 43 isolates (57%) were Gram negative and 32 isolates (43%) were Gram positive. Including coryneform bacteria, 84 foals (63%) were bacteremic with 93 isolates, of which 50 (54%) of isolates were Gram positive and 43 (46%) were Gram negative. The significance of the isolated coryneform bacteria is unclear; however, if members of this group of organisms are pathogenic, the predominant isolates of pathogenic bacteria were Gram positive. Despite the slight predominance of Gram-negative bacteria, the most common isolate was Enterococcus spp. (22 isolates, 29.3%) (Table 1). Other isolated Grampositive organisms were a-hemolytic Streptococcus spp. (3 isolates, 4%), Bacillus spp. (2 isolates, 2.6%), Listeria monocytogenes (2 isolates, 2.6%), with coagulase negative Staphylococcus spp., Streptococcus bovis, and group D Streptococcus having single isolates. Gram-negative organisms isolated were Pantoea agglomerans (13 isolates, 17.3%), Escherichia coli (8 isolates, 10.7%), Salmonella spp. (8 isolates, 10.7%), Klebsiella spp. (4 isolates, 5.3%), Actinobacillus spp. (3 isolates, 4%), CDC Enteric group 76 (2 isolates, 2.6%), with CDC VE type 1, Acinetobacter baumanii, Citrobacter freundii, Enterobacter spp., and Shewanella putrefaciens each having single isolates. Of the 9 foals with multiple organism bacteremia, 5 (56%) had Enterococcus isolated with another bacterium. Resistance Patterns All Enterococcus spp. isolates with susceptibility patterns available were resistant to more than 5 antimicrobial drugs. Two Enterococcus spp. isolated were resistant to penicillin (Table 1). Multidrug resistance was also noted with the Salmonella spp. isolates. The coryneform bacteria isolates were not tested for their sensitivity pattern, because the slow growth of most of the isolates precluded the usual testing techniques and for those few that will grow well the media to be used and the interpretation of breakpoints have not been standardized. IgG at Admission Eighty-eight foals had available plasma IgG concentration. Of these, 59 had IgG mg/dl, 16 had IgG mg/dl, and 12 had IgG o 400 mg/dl. There was no difference in the IgG concentration related to blood culture status (P 5.66). Of the 41 foals with a positive blood culture at admission where IgG was measured, 7 foals (17%) had IgG o 400 mg/dl, 7 foals (17%) had IgG mg/dl, and 27 foals (66%) had IgG mg/dl. Of the 12 foals that grew coryneform bacteria where IgG was measured, 1 foal (8.4%) had IgG o 400 mg/dl, 2 foals (16.6%) had IgG mg/dl, and 9 foals (75%) had IgG mg/dl. Of the 34 foals with negative blood culture at admission, 3 (8.8%) had IgG o 400 mg/dl, 7 (20.6%) had IgG mg/dl, and 24 (70.6%) had IgG mg/dl. There was insufficient information available to investigate any association of the IgG at admission with particular bacteriologic isolates. Of the 14 foals with isolated Enterococci for which IgG results were available, 3 (21%) had IgG o 800 mg/dl. Fecal Culture and Fecal Clostridial Toxins All 133 foals had fecal Salmonella spp. culture results available; 111 (83.5%) were negative, and 22 (16.5%)

4 1206 Hollis et al positive. Of the 22 foals with positive fecal cultures, 14 foals (64%) were blood culture positive (19 [86%] if coryneform bacteria are included), with a variety of isolates identified. One did not have blood culture results available. All 8 foals with Salmonella isolated on blood culture had Salmonella spp.-positive fecal cultures during the period of hospitalization. Ninety-three foals had results available for fecal ELI- SA for C. perfringens enterotoxin, with 75 (80.6%) negative and 18 (19.4%) positive. Ninety-three foals had results available for fecal ELISA for C. difficile A or B toxins or both, with 72 (77.4%) negative and 21 (22.6%) positive. Seven foals (7.4%) had both C. perfringens and C. difficile toxins present in their feces at admission. The overall prevalence of Clostridium spp. toxins was 34.6% in foals where feces were tested. Survival One hundred and twenty-seven foals had outcome recorded, with 78.7% (100/127) surviving to hospital discharge. However, there was no difference in survival related to blood culture status (P 5.24; OR 0.66, 95% CI ). Of nonsurvivors, 8 (29.6%) were blood culture negative, 2 (7.4%) grew coryneform bacteria, and 17 (63%) were blood culture positive, with 3 (17.6%) having more than 1 organism. Of survivors, 38 foals (38%) were blood culture negative, 15 foals (15%) grew coryneform bacteria, and 47 foals (47%) were blood culture positive, with 7 (14.9%) having more than 1 organism. When coryneform bacteria isolates were included in the analysis as positive blood cultures, there was still no difference in survival related to blood culture status (P 5.20). Excluding Enterococcus spp. and coryneform bacteria from the analysis, there was still no difference in survival related to blood culture status (P 5.74). Of the 21 foals with Enterococcus spp., the most frequently identified isolated organism, 18 survived (85.7%). Discussion Bacteremia was common in this population of neonatal foals with diarrhea, with 50% of foals having positive blood cultures at the time of admission. This proportion increases to 63% if coryneform bacteria isolates are included. This high number of bacteremic foals is in contrast to data from neonatal calves with diarrhea, where 19 31% are bacteremic, and critically ill foals, of which 29 36% are bacteremic. 1,2 It is possible to compare the foals in the present study with a study of all neonatal foals admitted to the same institution conducted over a portion of the same time period ( ), 1 because there are no differences in the age group of animals studied, handling of blood samples, or blood culture techniques. In addition, 26 foals in the present study were also included in the previous study (those presenting to the hospital from 1991 to 1998). In that study, 29% of foals were bacteremic at some point during their hospitalization, the large majority at admission. 1 This is suggestive, although not definitive, of a difference between the subset of foals presenting with diarrhea and all critically ill neonatal foals. The apparently increased incidence of bacteremia in the subset of foals with diarrhea could be because of a compromised intestinal barrier increasing the risk for bacterial translocation and subsequent bacteremia. Integrity of the intestinal mucosa is one of the principal factors in gastrointestinal protection, 19 with bacteremia occuring after lymphatic enteric drainage of viable bacteria, rather than via the enteric venous system, after both transcellular and paracellular translocation. 19,20 Damage to the intestinal mucosa secondary to enterocolitis may therefore increase the risk of secondary bacterial translocation and subsequent bacteremia. Repeat blood culture is not routinely performed in the study hospital on foals with diarrhea, and blood cultures taken after admission were not included in the analysis, in part because of their relative infrequency and in part because of our specific interest in the prevalence of bacteremia at the time of admission. The collection of only a single blood culture at admission could reduce the likelihood of correctly classifying certain foals as bacteremic or nonbacteremic during the entire hospitalization period, 2 because bacteremia can be transient. Although multiple samples might have identified additional foals as bacteremic, the study still demonstrated a high prevalence of bacteremia at the time of admission in these foals presenting to this hospital. In addition, the administration of antimicrobials before admission can reduce the likelihood of isolating bacteria from a bacteremic patient, even if the bacterium is not sensitive to the antimicrobial administered. Despite the relatively large proportion of foals in this study receiving antimicrobial treatment before admission, there was no association of antimicrobial administration with the blood culture status at admission found in this study, in agreement with a previous study where antimicrobial use before blood culture collection was not significantly associated with culture outcome. 6 Sixty percent of blood culture isolates collected after administration of antimicrobials were susceptible to the antimicrobial agents administered, suggesting that bacteremia is not merely transient in these animals and that bacterial translocation may be ongoing. The bacteriologic isolates found in this group of animals are especially interesting when compared with those with a variety of other complaints. In several studies, E. coli was the most common isolate, 1,4,6 8,10 13 although Actinobacillus was more common in some older reports. 5,9 Enterococcus spp. accounted for 0 19% of isolates in these prior studies. 1,4,6,12 Differences in bacterial isolates between 1 study 1 and the present study cannot be accounted for by climatic, geographic, or temporal differences, and suggest that isolates differ because of intrinsic differences between the disease conditions of the study populations. This could be because of a compromised gastrointestinal tract leading to overgrowth of certain bacteria with subsequent translocation of those bacteria. It is interesting to note that in 1 study, foals with Actinobacillus spp. isolates were 6 times more likely to have diarrhea than foals with other isolates. 12 This was not confirmed by the present study, where 4% of isolates

5 Foal Diarrhea 1207 from foals with diarrhea were Actinobacillus spp. The lack of association of Actinobacillus spp. with diarrhea in the present study is in agreement with another recent study of a more general population of bacteremic foals, and the potential role of Actinobacillus spp. in foal diarrhea continues to deserve further study. 4 It is perhaps not surprising that Enterococcus is a common isolate in foals with enteritis as this organism has emerged as one of the top 3 causes of bacteremia in children during the same period as our study, 21,22 and is widely thought to be a low virulence pathogen of immuno-incompetent hosts, in whom translocation from the intestinal lumen to the blood stream is thought to occur readily. 22 Seventy-nine percent of foals with Enterococcus bacteremia had adequate transfer of passive immunity, suggesting that failure of transfer of passive immunity is not associated with colonization and subsequent bacteremia. Enterococcus bacteremia is not often fatal in humans, 25 reflected in the lack of difference between cases where Enterococcus was isolated compared with those with negative blood cultures in the present study. All Enterococcus spp. isolates were resistant to more than 5 antimicrobial drugs and considered multidrug resistant. Of these isolates, only two were resistant to penicillin. Enterococci are reported to have intrinsic resistance to a number of commonly used antimicrobials Their relative resistance to penicillin and other b-lactams, including cephalosporins, appears to be related to unique enterococcal penicillin binding proteins that allow synthesis of the cell wall in the presence of these antimicrobials, 23 which may result in in vitro stasis when susceptibility tests are performed, but not in vivo killing, leading to a loss of correlation between the susceptibility test results and clinical efficacy. 23 There is also an inoculum effect, where a small inoculum tested in vitro may appear to be susceptible to penicillin because of a low concentration of penicillinase produced, whereas with a larger in vivo inoculum the amount of penicillinase produced may be clinically important. 21 These factors combine to make Enterococcus spp. typically tolerant to all b-lactams. The minimum bacteriocidal concentration may therefore be many times higher than apparent from the susceptibility results. 22 Enterococcus also exhibits low to moderate intrinsic resistance to aminoglycosides, even in the absence of acquired resistance, which is thought to be because of an inability of this antimicrobial class to cross the cell membrane. 22 However, there is apparent synergism of penicillin and aminoglycosides on these pathogens; therefore, the combination of high levels of penicillin with an aminoglycoside is recommended for this pathogen. 21,22 In addition, susceptibility results to antifolates, such as trimethoprim-potentiated sulfas, may also be misleading because some isolates have the ability to use folic acid from exogenous sources. 23 When selecting antimicrobial agents for foals, the guidelines presented in the recent ACVIM consensus statement on antimicrobial use should be adhered to, 26 combined with the knowledge of likely susceptibility patterns of commonly isolated organisms. This knowledge allows goal-directed antimicrobial therapy for these animals. There was a high percentage of foals with positive fecal Salmonella cultures that were blood culture positive (64 or 86% if coryneform bacteria are included). All the foals with Salmonella spp. bacteremia also shed Salmonella spp. in their feces. The relatively small number of foals and potentially confounding factor of Salmonella bacteremia precluded statistical analysis of the number of bacteremic foals compared with bacteremia in Salmonella-negative foals. However, it is not surprising that foals with salmonellosis are more likely to become bacteremic, because they could have more significant gastrointestinal compromise and therefore be at increased risk of bacterial translocation and subsequent bacteremia. 19,20 The significance of the coryneform bacteria isolates remains unclear. This group of bacteria includes members of the genera Corynebacterium, Brevibacterium, Dermabacter, Microbacterium, Rothia, Turicella, Arthrobacter, and Oerskovia as well as several other genera. Coryneform bacteria are widely distributed in the environment as commensals colonizing the skin and mucous membranes of humans and other animals, and are commonly considered contaminants without clinical significance. 30,31 However, the high percentage of coryneform bacteria isolates found in this study, and the presence of coryneform bacteria with other pathogenic bacteria in some cases, may allow reasonable challenge of this assumption. In older studies, coryneform bacteria may have been reported as Corynebacterium spp. which included Corynebacterium equi (now Rhodococcus equi), making direct comparisons difficult. 10 In more recent studies, one considered these contaminants and classed foals with coryneform bacteria as blood culture negative 1 ;however, another considered these as pathogens, with 2 Corynebacterium spp. isolates reported in that study. 4 In humans, clinically important infections with certain coryneform bacteria have been identified in both adults and children, and, in 1 study, clinically significant coryneform bacteremia has been reported in 3.1% of positive blood cultures. 35 In many cases, coryneform infections are reported as opportunistic pathogens in immunocompromised patients. 33 Clearly, further work needs to be performed to investigate whether some members of this group are truly pathogenic and therefore important isolates. It may also be that these isolates are simply markers for loss of the gastrointestinal barrier or failure of protective mechanisms against bacteremia. There was no association of the IgG concentration at admission with the blood culture status at admission. Although there is some debate about the importance of the role of failure of transfer of passive immunity in the development of sepsis, it is generally accepted that this is permissive, rather than requisite, for the development of sepsis. 36 These results are different from those seen in neonatal calves with diarrhea, where blood culture positive animals had a significantly lower IgG than blood culture negative animals. 18 The reasons for this difference are unclear; however, the results from the present study suggest that adequate transfer of passive immunity, using the traditional cut-point of 800 mg/dl for adequacy, does not protect the neonatal foal from devel-

6 1208 Hollis et al oping diarrhea and subsequent bacteremia. This could suggest that the traditional cut point of an IgG concentration of4800 mg/dl needs reexploration. C. difficile toxins A/B were detected in the feces of 22.6% of foals in the study, which by inspection is similar to the reported prevalence in mature horses with diarrhea (28%). 37 In addition, C. perfringens enterotoxin was detected in 19.4% of foals in the study. Only 7.4% of foals had both C. difficile and C. perfringens toxins in their feces, giving an overall prevalence of clostridial toxins of 34.6%. The importance of detection of these toxins remains somewhat uncertain. There was an overall survival rate of 78.7%, with no difference in survival between bacteremic and nonbacteremic foals. Owing to the retrospective nature of the study, it was not possible to distinguish between foals that died or were euthanized because of intractable disease, or those that were euthanized because of financial or other concerns. It is rare in our hospital for foals with diarrhea to be euthanized strictly because of financial concerns, because cost estimates are generally provided before admission, and the authors cannot recall such a case in recent years. This compares favorably with other reports of survival in critically ill neonatal foals (10 72% of bacteremic foals and % of all foals). 4,7,11 13,38 42 In contrast to 1 study of bacteremic foals, where fewer foals with multiple organism bacteremia survived, 4 there was no difference in survival between those with one or multiple organisms isolated in the present study, whether coryneform bacteria are considered as a pathogen or as a contaminant. This may be because of the relatively small number of foals with multiple organism bacteremia in the present study or because of underlying disease in the particular subset of foals with diarrhea at the study institution. These results are in contrast with those of neonatal calves with diarrhea, in which bacteremia was associated with decreased survival. 16,18 This could be a reflection of the prevalence of low virulence pathogens in the neonatal foals compared with those found in neonatal calves, where E. coli was the predominant isolate. 18 Acknowledgments The authors thank the clinicians, residents, nursing technicians, and students at New Bolton Center, who cared for the foals in this report and recorded most of the data present. Footnotes a Latex Micro Touch Surgical Gloves, Ansell Healthcare Inc, Dothan, AL b 3M Filtron High Performance Surgical Masks, 3M Health Care, St Paul, MN c ArrowCath, Arrow International, Reading, PA d Septi-Check, Becton-Dickenson Microbiology Systems, Sparks, MD e Sensititer Plate, Trek Diagnostics, West Lake, OH f CITE Foal IgG, Idexx Laboratories, Westbrook, ME g SNAP Foal IgG, Idexx Laboratories h DVM Rapid Test, Value Diagnostics, Immunosystems Inc, Spring Valley, WI i Tech Lab, Braxsburg, VA j Microsoft Excel, Microsoft Corporation, Redmond, WA k Stata 9.0 for Windows, StataCorp LP, College Station, TX References 1. Marsh P, Palmer J. Bacterial isolates from blood and their susceptibility patterns in critically ill foals: 543 cases ( ). J Am Vet Med Assoc 2001;218: Corley K, Donaldson L, Furr M. Arterial lactate concentration, hospital survival, sepsis and SIRS in critically ill neonatal foals. Equine Vet J 2005;37: Cohen ND. Causes of and farm management factors associated with disease and death in foals. J Am Vet Med Assoc 1994; 204: Corley K, Pearce G, Magdesian K, et al. Bacteraemia in neonatal foals: Clinicopathological differences between Gram-positive and Gram-negative infections, and single organism and mixed infections. Equine Vet J 2007;39: Dimcock W, Edwards P, Bruner D. Infections observed in equine fetuses and foals. Cornell Vet 1947;37: Henson S, Barton M.. Bacterial isolates and antibiotic sensitivity patterns from septicemic neonatal foals: A 15 year retrospective study ( ). In: Proceedings of the Dorothy Havemeyer Foundation Neonatal Septicemia Workshop 3, Talliores, France 2001; Koterba A, Brewer B, Tarplee F. Clinical and clinicopathological characteristics of the septicaemic neonatal foal: Review of 38 cases. Equine Vet J 1984;16: McKenzie H, Furr M. Equine neonatal sepsis: The pathophysiology of severe inflammation and infection. Comp Cont Educ Vet Prac 2001;23: Miller W. Foal disease Two years survey. Br Racehorse 1950;2: Platt H. Septicaemia in the foal. A review of 61 cases. Br Vet J 1973;129: Raisis A, Hodgson J, Hodgson D. Equine neonatal septicaemia: 24 cases. Aust Vet J 1996;73: Stewart A, Hinchcliff K, Saville W, et al. Actinobacillus sp. bacteremia in foals: Clinical signs and prognosis. J Vet Intern Med 2002;16: Wilson W, Madigan J. Comparison of bacteriologic culture of blood and necropsy specimens for determining the cause of foal septicemia: 47 cases ( ). J Am Vet Med Assoc 1989; 195: Harbarth S, Garbino J, Pugin J, et al. Inappropriate initial antimicrobial therapy and its effect on survival in a clinical trial of immunomodulating therapy for severe sepsis. Am J Med 2003; 115: Kang C, Kim S, Park W, et al. Bloodstream infections caused by antibiotic-resistant Gram-negative bacilli: Risk factors for mortality and impact of inappropriate initial antimicrobial therapy on outcome. Antimicrob Agents Chemother 2005;49: Lofstedt J, Dohoo I, Duzer G. Model to predict septicemia in diarrheic calves. J Vet Intern Med 1999;13: Fecteau G, Pare J, Van Metre D, et al. Use of a clinical sepsis score for predicting bacteremia in neonatal dairy calves on a calf rearing farm. Can Vet J 1997;38: Fecteau G, Van Metre D, Pare J, et al. Bacteriological culture of blood from critically ill neonatal calves. Can Vet J 1997; 38:

7 Foal Diarrhea Balzan S, de Almeida Quadros C, de Cleva R, et al. Bacterial translocation: Overview of mechanisms and clinical impact. J Gastroent Hepatol 2007;22: Gatt M, Reddy B, Macfie J. Review article: Bacterial translocation in the critically ill Evidence and methods of prevention. Ailment Pharmacol Ther 2007;25: Butler K. Enterococcal infection in children. Semin Pediatr Infect Dis 2006;17: Chatterjee I, Iredell J, Woods M, et al. The implications of enterococci for the intensive care unit. Crit Care Resus 2007;9: Moellering R. Emergence of Enterococcus as a significant pathogen. Clin Infect Dis 1992;14: Amyes S. Enterococci and streptococci. Int J Antimicrob Agents 2007;29(Suppl 3):S43 S Salgado C, Ison M. Should clinicians worry about vancomycin-resistant Enterococcus bloodstream infections? Bone Marrow Transpl 2006;38: Morley P, Apley M, Besser T, et al. Antimicrobial drug use in veterinary medicine. J Vet Intern Med 2005;19: Andersson A, Weiss N, Rainey F, et al. Dust-borne bacteria in animal sheds, schools and children s day care centres. J Appl Microbiol 1999;86: von Graevenitz A, Punter-Streit V, Riegel P, et al. Coryneform bacteria in throat cultures of healthy individuals. J Clin Microbiol 1998;36: Shapiro M, Smith K, James W, et al. Cutaneous microenvironment of human immunodeficiency virus (HIV)-seropositive and HIV-seronegative individuals, with special reference to Staphylococcus aureus colonization. J Clin Microbiol 2000;38: Stauffer F, Kittler H, Forstinger C, et al. The dermatoscope: A potential source of nosocomial infection? Melan Res 2001;11: Young V, Meyers W, Moody M, et al. The emergence of coryneform bacteria as a cause of nosocomial infections in compromised hosts. Am J Med 1981;70: Adderson E, Boudreaux J, Hayden R. Infections caused by coryneform bacteria in pediatric oncology patients. Pediatr Infect Dis 2008;27: Lagrou K, Verhaegen J, Janssens M, et al. Prospective study of catalase-positive coryneform organisms in clinical specimens: Identification, clinical relevance, and antibiotic susceptibility. Diagn Microbiol Infect Dis 1998;30: Meyer D, Reboli A. Other coryneform bacteria and rhodococcus. In: GL Mandek, JE Bennett, R Dolin, ed. Mandell, Douglas, and Bennett Principles and Practice of Infectious Diseases, 6th ed. Philadelphia, PA: Elsevier Inc; 2005: Balci I, Ekşi F, Bayram A. Coryneform bacteria isolated from blood cultures and their antibiotic susceptibilities. J Int Med Res 2002;30: Furr M. Systemic inflammatory response syndrome, sepsis, and antimicrobial therapy. Clin Tech Equine Pract 2003;2: Ba verud V, Gustafsson A, Franklin A, et al. Clostridium difficile: Prevalence in horses, in environment and antimicrobial susceptibility. Equine Vet J 2003;35: Brewer B, Koterba A. Bacterial isolates and susceptibility patterns in foals in a neonatal intensive care unit. Comp Cont Educ Pract Vet 1990;12: Gayle J, Cohen N, Chaffin M. Factors associated with survival in septicemic foals: 65 cases ( ). J Vet Intern Med 1998;12: Freeman L, Paradis M. Evaluating the effectiveness of equine neonatal care. Vet Med 1992;87: Hoffman A, Staempfli H, Willan A. Prognostic variables for survival of neonatal foals under intensive care. J Vet Intern Med 1992;6: Baker S, Drummond W, Lane T, et al. Follow up evaluation of horses after neonatal intensive care. J Am Vet Med Assoc 1986;189: