The work program at FRPERC on poultry decontamination has developed over the past 10 years or so and has looked at many different techniques.  Studies have looked at a range of novel intervention systems and, in more detail, at the application of the more conventional techniques thermal techniques using hot water and steam.  In the more recent studies, a number of the interventions have been trialled under industrial conditions.  Much of this work has been carried out in collaboration with our colleague Janet Corry at DFAS.

Novel methods of treating chicken meat using UV light (Purnell & James, 2000), microwaves (Göksoy et al., 1999; 2000) or hot air (Corry et al., 2003) have been considered.  Initial trials using UV on chicken portions showed that reductions of 1.9 log10 cfu cm-2 in APCs could be achieved for skin-on chicken breasts exposed to a 3.4 – 3.7 mW per cm2 treatment for 10 s.  However, since UV is a “line of sight” process it was considered that there would be too many shadowing problems with whole chicken carcasses for this technology to be easily added to existing commercial processing plants.  Although there are published studies that suggest that microwaves could be used to surface pasteurise meats, our detailed studies have found microwave heating to be far too uneven and un-reproducible to reduce surface bacteria whilst avoid surface cooking.  A limited study has also been carried out on the use of an experimental, pilot-scale system to evaluate controlled, hot-air heating of food surfaces.  Results showed large reductions in counts of inoculated Campylobacter jejuni and Escherichia coli from chicken skin subjected to 15-min treatments with high velocity (ca 15 ms-1) warm air at 10, 40, 50 and 60°C.  Reductions of 1 – 2 log units at 20 and 40°C indicated that there was some form of non-thermal drying effect.  However, these data have not been fully validated.

Although some of these novel processes showed some promise more conventional heat treatments, namely steam and hot water, have been found to be more readily adapted to commercial processing plants.

Studies employing steam have looked at steam at reduced pressure, at high pressure or at atmospheric pressure.  Each offers different advantages.  At atmospheric pressure, steam will be created initially at 100°C.  At pressures below atmospheric (sub-atmospheric), the generation temperature will be lower than 100°C, while, at pressures higher than atmospheric, it will be above 100°C.  Generation at temperatures other than 100°C does not substantially change the heat capacity of the steam.  Treatment temperatures below 100°C are less likely to cause damage to the surface of the carcass, but will require longer treatment times than treatments at 100°C and above.  One key disadvantage of both low- and high-pressure treatments is that they are batch systems.

Small scale low pressure and high pressure treatment vessels have been developed and evaluated.  In the low pressure system steam at 75 or 85°C for 40 s was shown to reduce levels of Salmonella enteritidis inoculated onto chicken portions by 3 – 4 log units (James et al., 1998; Evans, 1999).  However, some degree of cooking was apparent on exposed muscle surfaces, although the skin was barely affected.  In the high pressure system (James et al., 1998) there were initial problems with extensive mechanical damage caused by the introduction of steam into the cavity (as has been reported by other studies) and practical problems were found in achieving rapid treatments that were sufficient to kill bacteria but prevent cooking of the chicken.

Modern poultry processing is very rapid and continuous, with birds being processed at speeds up to 12000 carcasses per h.  Such speeds make the implementation of batch pressurised steam systems particularly difficult.  An atmospheric steam system, similar to those used in scalding, offers greater potential for on-line operations.  The basic principle behind the use of atmospheric steam for decontamination is to utilise a vessel with an open base, permanently filled with steam.  The steam remains in the vessel because it is less dense than the surrounding air.  The poultry carcasses are simply raised into the steam for a set amount of time.  The effects of various atmospheric steam treatments on the appearance, shelf-life and microbiological quality of chicken portions have been investigated (James et al., 2000).  Initial experiments showed that a 10 s treatment on naturally-contaminated chicken breast portions resulted in a 1.65 log10 cfu cm-2 reduction in APCs.  However, in comparison with untreated controls, this treatment did not extend the shelf-life.  Overall, results indicated that significant reductions in microbial counts could be achieved for chicken meat using steam.

Initial experiments into the use of hot water (Göksoy et al., 2001) determined the maximum times that chicken-breast samples could be immersed in water at temperatures between 50 and 100°C, before the appearance of the samples changed irreversibly.  The data produced show that these times range from 120 s at 50°C to 1 s at 100°C.  All the samples in this study were packed in a polyethylene film.  While it is probable that the rate of temperature-rise at the surface of unwrapped samples would be higher, the exposure time before noticeable changes occur is likely to be lower than those reported.  Most of the changes in appearance caused by the heat treatments were to the cut edges and exposed muscle of the samples.  Temperature analysis showed that, at the lower temperatures and longer treatment times (50°C for 480 s and 60°C for 300 s), the temperature measured below the skin almost reached that of the surrounding water.  However, the short immersion times at higher temperatures were not sufficient to allow heat to penetrate through the skin.  This is not unexpected, since chicken skin has a relatively low thermal conductivity, 0.357 Wm-1K-1 (Morley, 1972).  Tests on E. coli serotype O80, added artificially, and utilising treatments that had no effect on visual quality were found to be ineffective in reducing microbial counts.

In subsequent work (Purnell et al., 2004), an experimental in-line processing unit for poultry carcasses using hot-water immersion was developed and evaluated in a commercial poultry plant.  Treatment at 75°C for 30 s significantly reduced APCs and counts of Enterobacteriaceae and campylobacters, but the skin tended to tear during trussing.  However, treating carcasses at 70°C for 40 s, followed by a 12 – 15°C spray-chill treatment for 13 s, did not detrimentally affect the skin.  Microbial counts remained significantly lower than the controls for eight days under typical chill-storage conditions.

Further work with our colleagues in DFAS (James et al., 2005; Corry et al., 2006; James et al., 2007) has compared both atmospheric steam and hot water treatments, under laboratory and commercial conditions, using pilot equipment designed and constructed by FRPERC.  In experimental studies whole chicken carcasses, inoculated with ca. log 6 Campylobacter jejuni and Escherichia coli K12, were treated with steam at atmospheric pressure for up to 20 s in a pilot scale cabinet, or with hot water in a pilot immersion system for 20-30 s at 75 and 80°C.  In steam, numbers of C. jejuni were reduced by ca.     1.8 log10 cfu cm-2 in 10 s and 3.3 log10 cfu cm-2 in 20 s.  Corresponding reductions in numbers of E. coli K12 were 1.7 and 2.8 log10 cfu cm-2.  However, the 20 s treatments caused the skin to shrink and change colour.  The optimum treatment for maximum effect on C. jejuni and E. coli, least skin shrinkage and change of colour was concluded to be 10-12 s.  In hot water, a reduction of 1.3 log10 cfu cm-2 in counts of E. coli was achieved using a 20 s, 80°C treatment.  A 1.66 log10 cfu cm-2 reduction in C. jejuni AR6, was achieved by a 30 s, 75°C treatment.

Trials in a commercial poultry plant using naturally contaminated carcasses, compared treatments for 10 s in steam with 20 s in hot water at 80°C.  The appearance of the treated carcasses was assessed visually at intervals until the end of shelf-life, and checks made for pseudomonads, Enterobacteriaceae and campylobacters on breast skin.  Initial levels of Campylobacter spp. were low (~1 log10 cfu cm-2) and variable, but reductions (similar for steam and hot water) of about 2 log cycles were obtained for the other two groups.  Numbers of campylobacters were reduced, but not eliminated.  Visual assessment indicated that the hot water treatment caused less change in appearance than the steam treatment.  Carcasses produced using either treatment could be used for production of ‘skin-off’ portions.  It was considered that changes to appearance of skin-on carcasses or portions would be acceptable to many consumers.

This work shows that either atmospheric steam or hot water interventions could be readily incorporated into existing commercial processing plants and in doing so reduce numbers of campylobacters and salmonellas on fresh poultry at retail.  This could potentially have a significant effect on reducing the incidence of human campylobacteriosis and salmonellosis.