Food contamination alert system is possible

Firstly, a correction to my previous post: The authorities took not 2 days, but 3 days, to shut down the stall. It was revealed in TODAY on Thursday that the first reports of food poisoning from that stall appeared on a Wed 1 Apr. The stall was not ordered closed by NEA until Sat 4 Apr morning.

I’ve received some useful feedback from friends regarding my previous post, where I questioned why NEA officers took so long to shut down the Geylang Serai rojak stall that was allegedly responsible for three deaths and over 150 cases of food poisoning.

Some felt that I was being unfair by expecting NEA to react faster than it did, and that I appeared to be pinning the blame on NEA for the food poisoning.

While I don’t think NEA is completely blameless, I never said that they are entirely to blame. The NEA, Ministry of Health (MOH), the stall holder, doctors and even some of stall patrons could have played a part to avert this tragedy, or at least prevent it from ballooning into this nightmare involving over 150 people. Most of all, I feel it is “the system” which is to blame, and not any individual person or agency. I am not interested in playing any finger-pointing games at this point, but to suggest how the system can be improved to avert future mass outbreaks of food poisoning.

According to a Straits Times report (“Hospitals and GPs have system to track outbreaks”, ST, 8 Apr), it was only at 10pm that Friday night (3 Apr), after doctors at the hospital’s emergency department noticed a surge of patients with similar but more severe symptoms, that MOH was informed.

A Changi General Hospital (CGH) spokesman said: “These patients had started to come in after 5pm. Once we noticed a pattern, we reviewed the past 24 hours’ records for patients with the same symptoms”.

MOH’s spokesman said there was an established system of surveillance and reporting of mass food poisoning cases with all health-care institutions and general practitioners (GPs).

“This allows for epidemiological investigations to kick in immediately upon notification,” she said.

Singular and isolated food poisoning cases are not routinely notified to MOH, but when doctors have cause to suspect a common link between patients with food poisoning symptoms, MOH will be notified.”

The National University Hospital does not routinely report single and isolated cases of food poisoning to MOH.

But its spokesman said: “A doctor might flag such cases to MOH when he notices a sudden increase in the number of patients with gastroenteritis from a particular point source, especially when extended family members of the patient also have the symptoms.”

The KK Women’s and Children’s Hospital has its own internal system. When a discernible pattern is picked up at the emergency department, the relevant heads will be notified and the necessary action taken.

Family physicians approached by The Straits Times said they would inform the authorities when they see a similar pattern of symptoms coming from a cluster of between five and 10 patients.

Based on this report, several facts can be established:

1. There is a “system” in place for doctors to inform MOH of food poisoning outbreaks.

2. This system is not a clearly defined, uniform system. Every hospital and doctor seems to have its own variant of the system.

3. It relies on individual doctors or hospitals to notice trends in illnesses before reporting it to MOH.

4. All reports are made to just one agency — MOH.

What is a “discernible pattern”, “sudden increase” and “similar pattern”? Is similar defined as 5 patients or 10? NUH said a doctor “might” flag cases to MOH. “Might” but not “is required to”.

The most glaring gap in this “system” is that it relies on individuals to spot trends before reporting to MOH. If 10 different patients go to 10 different doctors or hospitals, how can we expect any trends to be spotted? We all know that doctors are severely overworked, especially in public hospitals. Given the short consultation times, would the doctor be more interested in treating the patient in front of her, or carrying out an investigation regarding exactly which stall the patient patronised and where it is located? Will she have time to share notes with her fellow doctors, even if they are working in the same hospital? I doubt so.

Even after MOH is alerted, how long will it take to piece together all the bits of information to form a discernible pattern? MOH then needs to inform NEA, which needs time to activate its officers to go down on the ground to investigate. If the source of contamination or poisoning is from food sourced from overseas (recall the recent melamine scare), the food will continue to be imported and consumed, perhaps at different establishments, if the Agri-Food and Veterinary Authority (AVA) does not block imports and order wholesalers and retailers to stop selling it.

All this time, the clock is ticking away and more people will continue to patronise the stall selling contaminated food. In the case of the Geylang Serai rojak stall, most of the victims ate at the stall on Fri 3 Apr — two days after the first poisoning cases were diagnosed. Lives could have been saved if the authorities were able to act faster.

Is such an alert system even possible? I believe it is. But it will take the coordinated effort of all stakeholders and an investment in technology. Here is how I think such a system could be designed.

Firstly, there needs to be an agreement in medical and official circles that food poisoning is a serious issue, not to be treated lightly. When lives are at stake and have been lost, I believe it warrants a sense of urgency and importance.

Secondly, a clearly defined, mandatory and easy-to-use reporting system needs to be put in place for doctors and consumers to report food poisoning to the authorities so they can take action.

Thirdly, the data from reports must be collated and analysed, and trends must be immediately detected and “red flagged”.

Lastly, the authorities must act immediately on any red flags.

Fortunately, we are in the technological age. An IT system can play a big part in this reporting system. Let’s call it the Integrated Food Contamination Reporting System (IFCRS) for convenient reference in this article.

IFCRS could be a web-based system that allows all GPs and hospital doctors to report cases of food poisoning that they encounter. The report could be as simple as stating who was taken ill, what the suspected contaminated food was, when the incident took place, and where the stall was located. The “who” is necessary in order for the authorities to get in contact with victims to conduct their investigations.

Similarly, consumers who fall sick after eating food, or detect that a certain food has a “funny smell” could also file a similar report into IFCRS. My wife and over 10 of her colleagues once fell ill after eating some catered food. She called up NEA to report that caterer, but was told that they needed proof before taking any concrete action. In all likelihood, the report was filed away and ignored. Other consumers might have fallen sick but either didn’t know how to report it or didn’t bother to do so.

Doctors should be required to log all diagnosed or suspected food poisoning cases. They should not wait for “discernible patterns” before doing so. As explained earlier, individuals cannot spot discernible patterns until it is too late. Consumers should be encouraged to file reports for the public good, although it is probably not reasonable to make it mandatory.

Once the reports are entered into the system, the IFCRS should take over to crunch the data and alert the authorities when a pattern is detected. What is defined as “a pattern” could be adjusted within the system. However, my view is that two or more cases from the same food establishment should be red flagged for investigation.

The alerts should be sent to MOH, NEA as well as AVA — MOH and AVA, because they would have the facilities to test the food for contamination and to determine its source, and NEA because they are the ones who have officers on the ground to check on the stalls and shut them down if necessary. Currently, doctors only notify MOH, and there is surely a lag time before NEA officers are activated.

However the authorities need not over-react by immediately shutting down a stall on the slightest suspicion. Procedures need to be put in place to quickly investigate cases to prevent more people from getting poisoned, but with minimal disruption to businesses which are not at fault. This could involve advising the stall holder to discard all his food on hand and do a thorough cleaning of his stall before selling more food. The current enforcement system appears flawed because the authorities require proof before taking action, rather than taking preventive action.

I hope the authorities would seriously consider implementing such a system. Isolated food poisoning cases may not constitute a national concern, but three deaths and 150 taken ill should surely be a wake-up call to improve the reporting system for food contamination or poisoning cases. The system could come in useful for not just food poisoning, but also help trace sources of contaminated food for the AVA to block imports and prevent further harm from being done.


To all readers: I would like to seek your feedback, comments, and criticisms of this proposal. I will use your feedback to refine this proposal before compiling and submitting it to MOH, NEA and AVA. Please send your feedback either through comments, Facebook (if you’re my friend) or email me directly.

Author: Gerald Giam

Gerald Giam is the Member of Parliament for Aljunied GRC. He is a member of the Workers' Party of Singapore. The opinions expressed on this page are his alone.

4 thoughts on “Food contamination alert system is possible”

  1. I think this is a good proposal. You’ve identified the gaps and come up with a practical solution. Yes, this should be made known to MOH/NEA quick. Like all good systems, it requires good inputs so ensuring the proper input of accurate data from all relevant sources will be a key success factor. That area requires the most attention, I feel.

  2. Thanks keenlen. A friend Facebooked me separately to suggest that this could be integrated with the Electronic Medical Records Exchange (EMRX) system, which was launched in 2004 but has yet to be deployed to private hospitals.

    Please keep the feedback/comments/suggestions coming.

  3. Instead of calling it IFCRS which is quite a mouthful and easy to forget, how about calling it “Integrated Contamination Alert & Reporting Exchange (I-CARE). Easy to remember and shows that you and I care.

Comments are closed.