Serial Disinfection with Heat and Chlorine To Reduce Microorganism Populations on Poultry Transport Containers

79 Journal of Food Protection, Vol., No., 00, Pages 79 797 Copyright q, International Association for Food Protection Serial Disinfection with Heat and Chlorine To Reduce Microorganism Populations on Poultry Transport Containers NIRAJA RAMESH, 1 SAM W. JOSEPH, LEWIS E. CARR, 1 LARRY W. DOUGLASS, A FREDRICK W. WHEATON 1 * 1 Department of Biological Resources Engineering, Department of Cell Biology and Molecular Genetics, and Department of Animal and Avian Sciences, University of Maryland, College Park, Maryland 07, USA MS 0-0: Received 19 September 00/Accepted 19 November 00 ABSTRACT A prototype system for the cleaning and decontamination of poultry transport containers was previously developed and evaluated as a means of eliminating foodborne pathogens entering poultry processing plants. While decontamination of the containers once with the use of either hot water (up to 708C) or sodium hypochlorite (up to 1,000 ppm) resulted in signi cant reductions in the numbers of coliforms and the elimination of small numbers of Salmonella, complete removal of pathogens was not attained. Therefore, the present study was conducted to determine whether repeated decontamination of the same containers could eliminate coliforms and Salmonella consistently. Individual ve-tier containers consisting of galvanized steel frames and berglass oors were identi ed (n ) and decontaminated once per day for ve consecutive days after being used to haul broilers from farms to the processing plant. Two types of containers were tested in this study: one had previously been used for broiler transportation, and the other had new oors. After each transport, the containers were rst precleaned with a cleaning agent using a high-pressure jet (,09 kpa) to remove debris and to loosen bio lms from surfaces. The containers were then immersed in an aqueous solution of 1,000 ppm of sodium hypochlorite at 708C for min. Samples obtained from the container surfaces before and after each cleaning and decontamination were analyzed to obtain coliform and Salmonella counts. Coliforms were completely eliminated from both types of containers following one decontamination treatment. Because no Salmonella were detected on the containers, the effect of decontamination in the elimination of Salmonella was not determined. Similar treatments on ve successive days also resulted in poultry transport containers that were essentially free of Salmonella and coliforms. This decontamination system involving a combination of heat and sodium hypochlorite can be used as a standard method for cleaning poultry transport containers in the poultry industry. It is recommended that such containers be cleaned after each use to avoid the potential risk of a buildup of signi cantly higher loads of pathogenic microorganisms and their bio lms. Microorganisms such as Salmonella and Escherichia coli can enter the poultry processing chain, leading to contamination of nal poultry products and compromising food safety. Critical points at which microbial contamination potential is high in the poultry farm to market chain have been identi ed (). Poultry processing plants have been found to be one of these potential contamination sources (9). Broiler birds transported from farm to processing plant in ve-tiered galvanized-frame containers with berglass oors were reported to shed Salmonella sporadically, thus contaminating container surfaces (). Poultry transport containers are not typically cleaned between uses and thus can act as a vector in spreading Salmonella and other pathogenic microorganisms from one batch of birds to the next during transportation. Since transport containers have been identi ed as one of the potential sources of contamination, several methods to clean these containers have been investigated, and most of the existing methods for container cleaning have been found to be ineffective in eliminating pathogens (). Existing cleaning and decontamination methods were inadequate because of either poorly designed equipment, improper disinfectant use, improperly set cycle * Author for correspondence. Tel: 01-0-; Fax: 01-1-90; E-mail: fw@umail.umd.edu. times, equipment malfunction, or recontamination of poultry transport containers after proper cleaning and decontamination. Therefore, considerable research is needed to develop an effective method for the cleaning and decontamination of poultry transport containers. In an attempt to develop a commercially viable decontamination system for the cleaning and decontamination of poultry transport containers, a prototype system was previously developed (). This system consisted of the washing of the containers with a cleaning agent with the use of a high-pressure jet (,09 kpa) followed by immersion in a tank containing hot water (at, 0, 0, or 708C) or an aqueous sodium hypochlorite solution (0, 00, 70, or 1,000 ppm). Although ph values were not measured on a regular basis, they were known to be.7. Thus, controlling ph might optimize chlorine use but would not change the results of this research. Both heat and chemical treatments signi cantly (P, 0.0) reduced the containers coliform counts. The disinfection of the precleaned containers in hot water (at 0 or 708C) produced a.-log reduction in coliforms, while the disinfection of the precleaned containers in 1,000 ppm of sodium hypochlorite produced a.-log coliform reduction. Although a -log reduction indicates signi cant inactivation of bacteria, it did not achieve our

79 RAMESH ET AL. J. Food Prot., Vol., No. FIGURE 1. Decontamination unit for the cleaning of poultry transport containers. (A) Rectangular berglass tank (. by. by. m) with steel reinforcement. (B) Sloped bottom to remove solids. ultimate objective of reducing the bacterial count to,0 CFU for each area examined (17 cm ). Therefore, the present study was undertaken to further improve the effectiveness of this system in cleaning and decontaminating poultry transport containers contaminated at levels of 10 to 10 7 CFU/17 cm. The tank solution included both heat and chlorine treatments. Previous studies have shown that the synergistic effect of heat and sodium hypochlorite increases bactericidal activity (). Therefore, the highest tested temperature (708C) and sodium hypochlorite concentration (1,000 ppm) were combined in the improved decontamination system, and the maximum exposure time was min. Because the previous intermittentdecontamination protocol did not completely eliminate coliforms, we surveyed the effects of cleaning the containers after each use for ve consecutive uses to assess the effectiveness of repeated decontamination. MATERIALS A METHODS Prototype decontamination unit. A berglass tank (. by. by. m) with the capacity to hold 1,000 liters of water was built for the cleaning and decontamination of the poultry transport containers (). The tank had a sloped bottom and a berglass mesh to remove the solids washed from the containers (Fig. 1). High temperatures were maintained in the tank solution by steam supplied through a thermostatically controlled valve via several nozzles located along the inside tank wall. The berglass decontamination tank was installed at a poultry processing plant near the beginning of the processing chain, at the point at which the birds were delivered onto a conveyor for further processing. Immediately after the birds had been delivered, the containers were available for decontamination procedures. Decontamination of poultry transport containers. Poultry transport containers used in this study (. by 1. by 1. m) weighed 0 kg (Fig. ). They consisted of galvanized steel frames with ve tiers and solid berglass oors. Six containers were used in this study. Three containers had been in regular use prior to the tests, and three were refurbished with new oors (as controls) prior to the tests. Microbial contamination levels for the control containers were compared with those for the test containers. The six containers were numbered, and a color-coded plate FIGURE. Poultry transport container (. by 1. by 1. m) consisting of a galvanized steel frame and berglass oors. was attached to each container to distinguish individual containers belonging to the two different treatment groups. Each container was lled in the eld with approximately 180 to 00 birds. Before the containers were used to transport the birds for the rst time, the populations of coliforms and Salmonella were determined by analyzing swab samples obtained from container surfaces. All six containers were used once daily to transport birds from the farm for ve consecutive days. Travel time varied from 1 to h. All six containers were tested each day on arrival at the processing plant before and after decontamination. This procedure was used to determine the effectiveness of the decontamination process in eliminating coliforms and Salmonella. After the removal of the birds at the poultry processing plant, the containers were decontaminated. Before the containers were decontaminated, the container surfaces were swabbed to determine the initial bacterial loads. To obtain a good representation of the entire container surface, bottom and middle tiers were swabbed at each of three locations, namely, the two ends and the center of the tier. A sterile metal template (1 by 1 cm with a 1.7-cm-diameter circular opening) was placed on the tier oor at the desired location, and the cutout area was swabbed. Three swabs collected from the same tier level (80 cm ) were pooled in the laboratory for bacterial analysis. To reduce the organic loads and to loosen any bio lms on the surfaces, a high-pressure jet (,09 kpa) was used to prewash the containers with a disinfectant solution containing sodium chlorite and a detergent (Compound C) that proved effective under laboratory conditions (7). The high-pressure washing procedure was carried out manually and continued until all visible debris had been cleared. This process usually required to min per container. The container surfaces were swabbed at six locations with sterile templates (17 cm per location) along two tier levels (with the previously swabbed areas being avoided) to determine coliform and Salmonella loads on the containers after high-pressure washing. The containers were then immersed for min in the berglass decontamination tank containing a 708C aqueous 1,000-ppm sodium hypochlorite solution (Tilley Chemical, Baltimore, Md.). The temperature and chemical concentration of the tank solution were monitored and adjusted to desired values if necessary. The sodium hypochlorite concentration in the tank solution was determined by measuring the amount of total available chlorine in a sample solution by the iodometric method (1). If the existing concentration was,1,000 ppm, additional sodium hypochlorite was added. The added chemical was well mixed by recirculating the tank solution at 8 kpa through several chlorinated polyvinyl chloride (CPVC) nozzles located on the inside tank wall. The concentration was checked again after thorough mixing and was

J. Food Prot., Vol., No. DISINFECTION OF POULTRY TRANSPORT CONTAINERS 79 adjusted for maintenance at 1,000 1 ppm of available chlorine. Once the desired conditions were established, the tank solution was sampled to determine whether it contained bacteria prior to the immersion of the transport container. Results were analyzed only in those instances in which sample cultures showed that the tank water was not contaminated. The precleaned transport container was lifted with a,77- kg-capacity forklift (ACP 0, Allis Chalmers Manufacturing Co., Milwaukee, Wis.) tted with a special lift arm attachment () to pick up the container from the top and immerse it in the tank. The treatment time was calculated from the time the top of the container was immersed into the tank solution. After min, the container was removed from the tank and swabbed at six different locations (17 cm each) along the two tier levels near the container door, with the previously swabbed spots being avoided. Bacteriological analysis. Swab samples collected from the container surfaces before pressure washing, after pressure washing, and after immersion treatment were analyzed in the laboratory to determine coliform and Salmonella populations. Each swab was aseptically cut into two pieces of approximately equal sizes. One piece was used for quantitative assessment of the bacterial load (coliforms and/or Salmonella), and the second piece was used for Salmonella enrichment. The rst piece was touched to a xylose lysine tergitol (XLT-) agar (Difco Laboratories, Sparks, Md.) plate and streaked for isolation. The growth of black colonies in h indicated the presence of Salmonella, a nding that was later con rmed on the basis of biochemical reactions on triple sugar iron agar (Difco) slants and on lysine iron agar (Difco) slants. If Salmonella were not detected at this step, only the coliform population was determined. The same swab piece was placed in a tube containing 10 ml of 0.8% saline solution and vortexed to obtain a bacterial suspension. The bacterial suspensions for the three swab samples collected from the same tier level were pooled, and the bacterial population was determined by traditional serialdilution and spread-plating techniques. Diluted bacterial suspensions were plated on XLT- agar plates to determine Salmonella counts and on MacConkey agar (Difco) plates to determine coliform counts, and these plates were incubated at 78C. By counting the typical colonies (black for Salmonella and red for coliforms) on the plates after h of incubation, the initial Salmonella and coliform populations were calculated. The average bacterial counts for two tiers of the same poultry transport container were used to represent the bacterial load for that container at the time of swabbing. The second swab piece was placed in 10 ml of sterile m-tetrathionate broth for the recovery of Salmonella through primary enrichment and incubated at 78C for h. If Salmonella were not detected through direct streaking on XLT- agar plates, a loopful of the primary enrichment broth was plated on an XLT- agar plate. If there were no black colonies on the agar plate, another attempt to isolate Salmonella was made through secondary enrichment by incubating the primary enrichment culture for 7 days at room temperature. On the eighth day, a loopful of the secondary-enrichment culture was plated on XLT- agar and incubated at 78C. The growth of black colonies in h indicated the presence of Salmonella, a nding that was con rmed on the basis of characteristic biochemical reactions on triple sugar iron agar and lysine iron agar slants. The XLT- agar plates from both primary- and secondary-enrichment cultures were held for up to 7 h before they were considered negative for Salmonella (, 8). Statistical analysis. Coliform and Salmonella counts were converted to log 10 values and subjected to analysis of variance with Statistical Analysis Software (8). Least-squares means of the TABLE 1. Log 10 coliform counts for poultry transport containers immediately after the transport of broilers (before treatment [BT]), after manual pressure washing (AP), and after immersion decontamination treatment (AT) Day no. Cage no. a 1 1 1 1 1 1 Coliform count (log 10 CFU/ml) b BT AP AT.19.97..8.8.1.87.70.18.9.0.9..9 7...90 7.0.9 7.81.8. 7.8.8.9.78..8..10.01.89.1.7.9.0.7.8.0.7.1.09.1.79 7.1.. 7.11.0 7.1.7.0.8.8.1...9.0.00 a Cages 1,, and were in regular use; cages,, and were refurbished with new oors. b Mean count obtained by culturing samples swabbed from the middle and bottom tiers of the containers at six different locations. Each swab was taken from a standard 17-cm area and suspended in 10 ml of saline solution., not detectable;, not applicable (containers were not disinfected owing to practical limitations). coliform and Salmonella counts obtained from the two types of containers before and after decontamination were compared. RESULTS A DISCUSSION Coliform counts for each of the poultry transport containers before the application of any cleaning treatment, after pressure washing, and after immersion treatment are presented in Table 1. Salmonella were not detected on any of the six poultry transport containers tested even before treatment. The method used for the detection of Salmonella was adequately sensitive and had previously proved to be effective in detecting the pathogen (). In the absence of Salmonella on the containers, the effect of the decontamination process was evaluated on the basis of the elimination of coliforms. Since Salmonella is more susceptible to heat and

79 RAMESH ET AL. J. Food Prot., Vol., No. TABLE. Mean log 10 coliform counts for poultry transport containers immediately after the transport of broilers, after manual pressure washing, and after immersion decontaminationtreatment Sampling point Immediately after transport of broilers After pressure washing After immersion treatment Mean coliform count (log 10 CFU/ml) for containers a : In use.97 1.00 A.80 0.97 A B With new oors.17 0.7 A.9 0.7 A B a, not detectable. Means with different letters in the same column are signi cantly different (P # 0.0). was assigned a value of zero for calculation purposes. chlorine than coliforms are (, 10), the decontamination process could be considered effective against both organisms if it could eliminate coliforms. In both of the treatment groups (containers in use and containers refurbished with new oors), the decontamination process proved to be capable of reducing coliforms to undetectable levels. There were two occasions on which coliforms were present in small numbers on containers after decontamination treatments. These two data were considered outliers in the statistical data analysis, and thus there were no detectable coliforms on the containers after decontamination with the immersion treatment. Mean coliform counts for the two treatment groups before and after decontamination are presented in Table. A high level of coliforms, 1.1 10 CFU/cm (.17 log cycles 1 10 CFU/ml 1. 10 CFU/ml 10 ml of saline solution per 17 cm of swab area 1.1 10 CFU/cm ), was detected before treatment on containers each day immediately after the delivery of birds to the processing plant. This nding suggests that during each transport, containers can become contaminated with a large load of organisms that could potentially infect the next ock of birds transported in these containers. While visible debris was removed through washing with a high-pressure jet, pressure washing did not signi cantly (P. 0.0) reduce coliform counts (data not shown). The extents of debris removal and bio lm loosening were not estimated, and therefore the in uence of pressure washing on these factors is not completely known. In the present study, the decontamination of poultry transport containers with 1,000 ppm of sodium hypochlorite maintained at 708C was found to be effective on the rst day of cleaning, suggesting that a one-time decontamination treatment with the use of our experimental parameters could essentially reduce coliforms on the containers to nondetectable levels. In our previous study (), however, we observed that either a high water temperature (708C) or a high sodium hypochlorite concentration (1,000 ppm at ambient temperature) was not effective in the complete elimination of coliforms. It appears that both factors (temperature and sodium hypochlorite) exerted an apparent synergistic effect on coliforms. Our results are consistent with an earlier nding () that the bactericidal effect of sodium hypochlorite is enhanced at higher temperatures. Further studies might show that a combination of a lower concentration of sodium hypochlorite and heat may be adequate for complete decontamination for subsequent use of the containers if treatment is performed after each use. To test our hypothesis in this study, we used the highest concentration and the highest temperature that were found to be effective in our previous study (). Although a one-time decontamination treatment involving a 1,000-ppm sodium hypochlorite solution at 708C reduced coliform counts for poultry transport containers to undetectable levels, it is necessary to repeat the cleaning procedure after every transport trip because coliforms accumulate on the container surfaces during each trip, as demonstrated by the results of this study. In all but two instances, coliforms were not detectable on the containers after immersion treatment (Table 1). Without thorough treatment, E. coli levels on containers could be as high as 10 CFU/cm, a level suf cient to contaminate the next batch of birds transported in a container. After of the 0 cleaning trials, there were low levels of coliforms (8 to 0 CFU/cm [100 to 0 CFU/ml]) on container surfaces after immersion treatment. On one of these two occasions, a delay in applying the decontamination treatment caused the organic debris on the surface of the container involved to dry up, perhaps indicating that a dried surface offers more resistance to disinfection. Thus, it might be advantageous to clean containers soon after birds have been delivered to the processing plant (Table 1). In summary, our results suggest that the cleaning of containers once with a process involving a high temperature and sodium hypochlorite was effective in eliminating coliforms. However, we recommend that transport containers be cleaned every time immediately after use because of the potential risk of the buildup of signi cantly larger loads of coliform organisms and bio lms. Our decontamination method can be automated to clean each transport container soon after the delivery of birds to a poultry processing plant. Our method can be used as a standard method in the poultry industry for the cleaning of poultry transport containers to eliminate the risk of introducing pathogenic microorganisms into the food chain. ACKNOWLEDGMENTS This research was supported by a gift from Delmarva Power to the Department of Biological Resources Engineering, University of Maryland, College Park, Md., and by the Maryland Agricultural Experiment Station. We thank Gary Seibel for his technical assistance and Sean Walsh, James Adkins, Leslie Lorenz, Josh Hayes, and Shirong Wei for their help with data collection. REFERENCES 1. American Public Health Association. 199. Standard methods for the examination of water and wastewater, 19th ed., p.. American Public Health Association, Washington, D.C.. Bhatia, T. R. S., and G. D. McNabb. 1980. Dissemination of Salmonella in broiler chicken operations. Avian Dis. :1.. Carr, L. E., C. Rigakos, G. Carpenter, G. Berney, and S. W. Joseph. 1999. An assessment of livehaul poultry transport container decontamination. Dairy Food Environ. Sanit. 19:7 79.. Dychdala, G. R. 1977. Chlorine and chlorine compounds, p. 17

J. Food Prot., Vol., No. DISINFECTION OF POULTRY TRANSPORT CONTAINERS 797 19. In S. S. Block (ed.), Disinfection, sterilization and preservation, nd ed. Lea and Febiger, Philadelphia.. Mazzotta, S. A. 001. Thermal inactivation of stationary-phase and acid-adapted Escherichia coli O17:H7, Salmonella, and Listeria monocytogenes in fruit juices. J. Food Prot. :1 0.. Ramesh, N. 1999. Development and evaluation of a prototype poultry transport container decontamination system. Ph.D. dissertation. University of Maryland, College Park. 7. Ramesh, N., S. W. Joseph, L. E. Carr, L. W. Douglass, and F. W. Wheaton. 00. Evaluation of chemical disinfectants for the elimination of Salmonella bio lms from poultry transport containers. Poult. Sci. 81:90 910. 8. SAS Institute. SAS user s guide: statistics, version.1. SAS Institute, Inc., Cary, N.C. 9. Timoney, J., W. Kelly, J. Hannan, and D. Reeves. 1970. A study of Salmonella contamination in some Dublin poultry processing plants. Vet. Rec. 87:18 191. 10. U.S. Food and Drug Administration. 1998. Hazard analysis and critical control points (HACCP): procedures for the safe and sanitary processing and importing of juice. Fed. Regist. :00 08.