THE POISON GARDEN website      Arum maculatum berries on a Cannabis leaf 

Search thepoisongarden.co.uk:

This free script provided by JavaScript Kit

Pontifications on Poison

Being some ramblings on events associated with poisonous plants.

Friday 23rd December 2011 

The American Association of Poison Control Centers has published the 2010 Annual Report of the National Poison Data System (NPDS). It ought to be fairly obvious that a report published on 21st December dealing with the previous year must have taken a lot of work to compile so it’s no surprise to find that the NPDS 2010 report runs to 203 pages.

There is such a lot of information in the report that I may get several days’ blog entries out of it. Certainly, all I intend to do today is look at some of the broader issues that leap out on a first pass.

One thing I’m struck by is that the report has lots of information at the highest level and lots at the bottom but in the middle it leaves a great deal unanswered.

I’ve mentioned before that the total number of calls to Poison Control Centers (PCCs) gives a misleading view of the extent of actual harms and 2010 is no different. The report says that there were 53,526 mentions of plants in calls to PCCs.

There is bit of confusion in my mind because the 53,526 is mentioned in the text but the detailed breakdown in Table 22 gives 53,295 as the total number of ‘case mentions’. But whether the difference is an error or just a result of differences between calls to PCCs and ‘case mentions’ needn’t concern us as the two figures are quite close to each other. The report separates information only calls from human exposure calls so the 53,526 figure is for potential poisoning incidents.

And the important part about that number is that, when the breakdown is given in Table 22, only 2 deaths resulted and 61 other cases had a ‘major’ outcome. In total, only 4,782 cases are recorded as ‘Treated in Health Care Facility’. It rather suggests that the majority of calls involved no potential for any symptoms of poisoning to appear. You can imagine that a parent catching its child putting a berry in its mouth might call a PCC even if the child was made to spit it out without chewing it.

Brugmansia suaveolens (syn. Datura suaveolens)

Brugmansia suaveolens, angel's trumpet
(syn. Datura suaveolens)

As I said, I want to look at one of the problems with the report by looking at the detail at the bottom and the top of the data. Before doing that it is worth mentioning that total human exposure calls were 2,384,825 so plants represent just over 2% of total calls and the 2 plant deaths detailed are from a total of 1,366 described in detail. (Of these, 1,146 were cases involving contact with a PCC and the other 220 were identified from news stories and the like.) So, plants accounted for only 0.14% of all deaths recorded as being due to poisoning.

What I’m referring to as the ‘bottom’ of the data is this detail on the fatal cases. The two plants involved were Brugmansia suaveolens (syn. Datura suaveolens), angel’s trumpet, and Allium sativum, garlic. That second wouldn’t normally be thought of as a poisonous plant so I’ll have to do some research and return to these individual cases another day.

The ‘top’ of the data is the analyses made of the 53,526 total ‘case mentions’. The report breaks this number down into age groups so that we can see that the overwhelming majority, 34,363, were related to children under 5. And it gives an analysis of the top 25 most mentioned ‘plant exposures’. Unfortunately, the categorisation includes terms like ‘Plants - cardiac glycosides’ along with some botanical names so it is not possible to compile a top 25 plants list. The ‘top 25’ covers 21,471 case mentions of the 53,526 total and, given that No.25 accounts for 375 mentions then there must be, at least, 85 other plants mentioned in the total.

Euphorbia pulcherrima, poinsettia

But, where the problem lies is between the two extremes. That is to say, there is no analysis of what plants caused the 4,782 cases that required the involvement of a health care facility. The overall category ‘Plants’ is subdivided but the subdivisions are a mixture of plant specific items, like colchicine, and symptomatic terms, like gastrointestinal irritants.  Nor are those 4,782 case ‘mentions’ broken down by age to enable one to see just how many of the 34,363 under 5s ‘case mentions’ produced symptoms of poisoning that required treatment.

Just to take one example of the difficulty this can cause, the report lists Euphorbia pulcherrima, poinsettia, at No.9 in its top 25 plant exposures but there is no way of determining how many of the 750 case mentions, if any, resulted in actual harm either by ingestion or skin contact.

For future entries, I want to look in more detail at the top 25 list, even though its usefulness is limited by the absence of any indication of the actual harm resulting. I also want to see if any useful data can be extracted for total mentions versus major outcomes for the different substance categories. The 53,526 figure is for plant exposures in plant form. The report gives separate data for pharmaceuticals derived from plants so things like morphine and atropine are dealt with a different heading and I’ll try and look at what that data tells us. Plus, as above, I’ll try and learn more about the two fatalities.

As it’s taken the AAPCC nearly twelve months to compile and publish the data I hope I can be forgiven for taking a few days to assimilate it.