An Inaction Level

An Important Detail

Listen up.  One of the key differences between a safety assessment and a risk assessment is this:  After a safety assessment has been concluded, the decision about what is acceptable has been made, and that is true even if you call the safety assessment a risk assessment.  After an actual risk assessment such as the one produced for an apple juice action level, it hasn’t.  You still need a risk manager to decide.

An Intervention Analysis for Arsenic in Apple Juice

The Apple Juice Risk Assessment was patterned after the EPA 2001 Drinking Water Rule.  It contained an intervention analysis that evaluated how much inorganic arsenic exposure and the increased risk of lung and bladder cancer from arsenic in apple juice would be changed by the application of different putative action levels.  A supporting document also estimated the amount of apple juice that would be eliminated from the food supply at each limit.  The results looked like this:

Arsenic Limit (ppb)	Average Arsenic (ppb)	Per Capita Risk (per million)		Apple Juice Eliminated
3	1.4		2.5 (0.0, 6.8)	69%
5	2.7		4.8 (0.0, 12.8)	46%
10 and above	4.4		8.0 (0.0, 21.3)	0%

Some technical details: This table only displays risk at 3 ppb, 5 ppb, and 10 ppb since none of the samples in the subset of 94 samples taken during October 2011 had inorganic arsenic levels above 10 ppb. Consequently, risk with all arsenic limits above 10 ppb would be identical to a limit at 10 ppb.  All estimates are for combined case rates for both lung and urinary tract cancer per million persons with the 5th and 95th percentiles are given as confidence intervals.

The long and the short of it is that while there probably is some apple juice out there with inorganic arsenic concentrations above 10 ppb, it isn’t very much.   So, setting a limit at 10, 20, or 50 ppb will all have about the same result; no reduction in exposure or risk.  If a limit is set lower than 10, then significant percentages of juice get eliminated from the market.  The information in the above table is not exactly a cost-benefit analysis, but it’s getting there – which is far more than what the agency would typically do for an action level that isn’t legally enforceable anyway.

The Proposed Action Level for Arsenic in Apple Juice

The guidance limit of 10 ppb was proposed in a supporting document (FDA, 2013); it was not specifically identified by the risk assessment.  Reasons given for the selection of 10 ppb were a) there is a significant risk associated with exposure to arsenic from apple juice (that is debatable), b) a limit of 10 ppb is feasible and can be met by the juice industry, and c) a limit of 10 ppb would be consistent with the limits for arsenic in both bottled water and tap water.  Unfortunately, effective risk reduction was not among the reasons given.

To be fair, the term “risk management” is probably ill-advised; “risk-policy” would be better.  First of all, the best result will often be partial risk mitigation, where the risk is not completely and entirely managed. Secondly, there is a strong tendency to assume that the “agency managers” are also the “risk managers”.  But personnel management skills may have little correlation with devising effective strategies for reducing risk.  Instead, that task may be best handled by non-managers with more practical experience than the agency managers, or by a distributed group people, some of whom may not be in the agency at all.

 

In any case, one might think that having a transparent decision making process would lead to better decisions.  At least, that’s what I thought would happen.  But, even with multiple layers of review (the risk assessment was controversial, the guidance document was not), it didn’t work out that way.  No one seemed to care whether the risk was actually reduced or not, and onward and out the door the proposal went.  However, the guidance isn’t finalized, and perhaps it never will be.  Does anyone really care?  Consumer Union thinks a limit lower than 10 would be a good idea, but I’m guessing that either they don’t like apple juice very much anyway, or they don’t care how much it costs.  But, even they seemed to think that taking a first step that goes nowhere is a step in the right direction (Bottemiller, 2013).

The main problem, in my estimation, is that it just didn’t occur to very many people at all to consider the issue rationally.  If the cost of monitoring were factored in, I would guess that no action level for a very minor risk factor would be supported by a formal cost-benefit analysis like the one done for the drinking water rule.  Yet most reviewers, including the industry critics, seemed to simply assume that the risk assessment justified the guidance, and that is surely because that is way the EPA typically operates.  The FDA gave them something different, and they didn’t know what to do with it.  That also may explain why the analysis for the EPA drinking water rule couldn’t be done in-house; the reviewers are too used to toying with precautionary assumptions.

A Potential Source Elimination

If someone were interested in reducing the risk, the apple juice risk assessment does contain a clue as to how some arsenic might be prevented from getting into apple juice.  Comparison of total arsenic concentrations in juice to projected values from juice concentrate after water is added indicates that at the low end of the distribution, arsenic concentrations are higher in the finished product.  Thus suggests that about one third of the arsenic in the juice may come from the water that is used to reconstitute the concentrate.  Since it is much easier to remove arsenic from water than juice, creating a standard for “food-grade” water used to reconstitute juice of 3 ppb or even 1 ppb might be a cost-effective intervention.  Yes, that would be inconsistent with the EPA standard, but what is wrong with that?

References

Bottemiller, H (2013).  FDA Proposes Limit for Arsenic in Apple Juice.  Food Safety News, July 13, 2013

Carrington CD, Murray C, and Tao, S. (2013). A Quantitative Assessment of Inorganic Arsenic in Apple Juice. 

U.S. Food and Drug Administration (2013).  Supporting Document for Action Level for Arsenic in Apple Juice.

Official Post Soundtrack

Santana (1971).  You Just Don’t Care.  Santana (first album), Track 7.

Post Notes

Post Thesis #37.  This concludes the apple juice saga, but the arsenic tale may continue with rice.  Even though I think the AJ action level was fiasco, I don't blame the Center for Food Safety.  The orders came from higher up, and when that happens only they can stop it.  They didn't.