With one eye on the summer barbecue season the HPA focused on adequate cooking of meat - ensure it's piping hot, whatever that means - and avoiding cross contamination. Their press release also highlighted research published in 2005 estimating the number of cases of foodborne illness in England and Wales each year at 1.7 million. This figure is about 20 times the number of confirmed cases picked up by the surveillance system, showing not only the degree of under-reporting, but also giving some idea of the burden of foodborne disease on a population of about 54 million.

The FSA chose to advise consumers how to use food leftovers safely, after discovering that many households are trying to save money in these uncertain times by cutting down on food waste. Throwing away less food is clearly a positive trend, and is Government policy, but it does carry with it some food safety risks, which the FSA is keen to warn people about. Part of that warning concerns 'use by' and 'best before' dates, which many people ignore, but which assume a greater significance if we are intent on eating more of the food we buy. Despite repeated attempts to educate the public, many people still don't understand what these labels mean and don't take them seriously.

Perhaps the solution is to go back to square one and come up with something better, rather than try to educate consumers. It seems to me that applying 'use by' to more or less everything perishable dilutes the food safety message. For example, 'use by' is much more critical for cooked sliced meats than for yoghurt or orange juice. I suspect that 'best before' could be applied to a lot more food products without compromising food safety. That leaves the problem of how to label the rest - the foods that carry a real risk. 'Use by' doesn't really grab the attention. How about 'safe until' instead? That would certainly make more sense to consumers, but somehow I don't see food manufacturers being too keen to have it on their labels.

This kind of complexity is a big problem for the investigation of food poisoning outbreaks. Where do you start looking for the source of a foodborne pathogen and how do you trace it back to its origin? Witness last year's E. coli O104 outbreak in Germany. Initially the investigating authorities were confident that the source was most likely Spanish cucumbers and/or tomatoes. But further detective work later pinned the blame on fenugreek seeds from Egypt. Unfortunately, delays in identifying an outbreak's source can lead to additional cases of illness - sometimes even deaths - that might otherwise have been prevented.

The problem of complex supply chains is well known, but a newly published research study led by the Institute of Food Research in the UK has managed to quantify that complexity and map the networks that supply our food. The researchers used some sophisticated methodology to analyse food trade data from the UN. The maths are a bit beyond me, but the findings are clear. Supply networks are incredibly convoluted and complex, but there are a relatively small number of countries that act as processing and distribution hubs in international food trade - the Netherlands fills this role in Europe. The study's authors concluded that the global food supply network is great for speedy distribution of food, but not much use for finding out where it originated.

I think this supports the view that the food industry needs to pay much more attention to the traceability of its' products and ingredients. Simply meeting the 'one step forward, one step back' requirement of the law is no longer enough. Tougher legislation may be needed, along with industry cooperation to identify best practice. Without action, the problem can only get worse.

The issue was highlighted this week by the "Stericycle ExpertRECALL Index" of all the food recalls logged by the FDA in America for the first quarter of 2012. According to the Index, there were 142 recalls during the three month period, which sounds a lot, although it was apparently 19% fewer than in the previous quarter. Those recalls involved around seven million individual food units, which also sounds a lot, but is actually 90% less than in the previous quarter. The real interest for me lies not in the number of recalls, but in the reasons for them. By far the commonest cause was undeclared allergens, which accounted for 40%, followed by Listeria contamination (30%).

The high number of Listeria-related recalls can be viewed as peculiar to the USA, arising from the 'zero tolerance' policy towards Listeria. Only this week a large nationwide recall of potentially contaminated bagged salads was announced. In the EU that recall may not have been necessary, since food safety rules say that low numbers of Listeria are acceptable, provided they do not rise to dangerous levels during shelf life. Whether the American approach is better is difficult to say, but it makes life a lot more difficult for food manufacturers.

Recalls prompted by undeclared allergens are a different matter. The picture in Europe is much the same as that in the USA, with allergens accounting for a large proportion of food recalls. Are all those recalls necessary? Again, it's hard to say, but what it illustrates is that a lot of food manufacturers really haven't got their act together when it comes to allergen control. Many allergen related recalls are caused by packaging and labelling errors. That should be easily fixable because it just means always matching the product with the correct packaging. Cross contamination incidents are much less common, and adopting sound allergen control plans could prevent many of those. Food manufacturers have it within their capability to do far more to prevent allergen related recalls, which would not only benefit consumers, but would also make sound business sense.

The CAST report, titled 'The Direct Relationship Between Animal Health and Food Safety Outcomes', looks at how the health of food animals produced under different conditions can have effects on food safety at distant steps in the supply chain. For example, a subclinical infection with Salmonella might not be apparent prior to slaughter, but may have a considerable effect on contamination rates at the processing stage. The authors make some thought provoking observations and point out that 'humane' animal production systems, like organic and free-range, can actually lead to higher rates of infection with pathogens than intensive indoor systems.

Reading between the lines, the report seems to be saying that we will need to take a pragmatic approach to meat production if we are to meet the demands of our growing population. We might think that intensive production systems are bad for animal health, but some of the science says otherwise, and this has consequences for food safety. Whether you agree with that or not, it is a point worthy of discussion. So far so good, but the report also contains the following statement on the subject of antibiotics in agriculture, "The use of antibiotics in food-animal production, however, raises some concerns about antibiotic resistance in bacteria that could affect the efficacy of antibiotics in the treatment of human infections. Concern about antibiotic resistance is not equivalent to actual risk. Resistant bacteria were present long before antibiotics were discovered and found in many places without livestock exposure."

This is at odds with much of the scientific literature the authors must have consulted before reaching their conclusions. It is certainly true that antibiotic resistant bacteria have been around for a long time, but isn't that because many antibiotics are derived from compounds produced naturally by microbes in the environment? They can be seen as weapons in an arms race that has been going on for millions of years. Microorganisms compete for resources, and antibiotics provide an advantage if they kill off competitors. Equally, evolving antibiotic resistance as a defence confers a similar advantage. Introducing vast amounts of antibiotics into the environment through livestock means a shift in the balance of power, which is seldom a good thing in ecological terms. It may be true that "Concern about antibiotic resistance is not equivalent to actual risk", but the CAST report does nothing to allay that concern.