Boil Alert Fails to Serve Community

My own assessment of this week's boil alert is that it was pro forma, and that if this had been a real public health issue we'd all be screwed because Portland Water Bureau's process sucks (it is too slow, and lacking the information people really need).

A risk assessment from this week:

By the time the boil water order was given Friday morning, we'd already been drinking the questionable water for 3 days.  County health department officials have yet to report any discernible uptick in disease reporting (diarrhea) for the Water Bureau service area, so, how questionable was the water?  I think we can assume one of the following: 1) the bacteria were all dead, present but DEAD, and therefore harmless -- our tests don’t distinguish live-and-harmful from dead-and-harmless, and since PWB does not wear gloves (they use chemical hand-sanitizer instead) when they collect samples, dead-but-still-present bacteria can be transferred from hands to sample, 2) the E.coli was a friendly variety regularly found in our gut, 3) the bacteria count was really low and within a range most immune systems could handle, 4) the contamination wasn't widespread, or 5) the tests were misleading, possibly even wrong (after all, the immediate follow-up tests ALL reversed the results).  

Willamette Week tried to put the risk in perspective when the West-side-only boil water order went out one weekend in 2012:  “Probable number of live coliform bacteria that triggered the boil water notice: 1; number of that same bacteria U.S. Department of Agriculture regulations allow in a 4-ounce stick of butter: 1,100.”  (click here for that article) 

Here is how this week unfolded:

On Tuesday, May 20, a water sample at the outlet of a Tabor reservoir tested positive for a problem indicator known as Total Coliform, which was then identified specifically as E.coli.  A follow-up sample was taken as soon as that E.coli result came in, to verify the result and rule out error.  That follow-up sample came back clean.  On Wednesday, May 21, a water sample at a site down by the waterfront came back positive for Total Coliform, which was then identified as E.coli.  Again, a follow-up sample was taken as soon as that E.coli result came in, and that follow-up sample came back clean.  On Thursday, May 22, a water sample at the outlet of a Tabor reservoir tested positive for Total Coliform, identified as E.coli.  That result prompted a follow-up test on Friday, and while the City was waiting on the results to that test the Oregon Health Authority ordered Portland Water Bureau to issue a city-wide boil water notice.  Two of the three sites had already retested as clean.  The community had already been drinking the water for 3 days.  In my opinion, this boil order was largely pro forma.  

Yes, given the rarity of three days of E.coli positives, this case warranted special caution and OHA took it.  The pattern seemed to suggest that the last test would come back clean (and it did), and that the "event" might not really even be one of contamination, but possibly a flaw in testing protocols.  (Should we really count dead bacteria the same way we count live bacteria?  Shouldn’t quantity factor in this trigger system at some point?  Are these protocols really protecting public health by raising the alarm at the right time?)  The people at OHA didn't feel comfortable waiting another 24 hours (on top of the 72 already on the clock) to make that call and I sympathize with their decision.  Although, I don’t see what good it did as it was three-days late for the people who really needed it (the immunocompromised) and the risk was overstated for the rest of us.

A government process that fails the community:

A boil water order sounds like an emergency, doesn’t it?  It might surprise you to learn that a boil-water order in Portland can come after the risk has passed, when the water is already clean again (as it did in this case).  It might surprise you to learn that a boil water order can be given in Portland after even a single, dead, harmless bacteria is found in a water sample -- recalls on other products only happen when a threatening number of live bacteria are found in the product. 

A boil water order only protects public health if it is timely AND if it is actually necessary.  Late orders leave people vulnerable to real-time outbreaks.  Overstated risks cause panic, and panic causes decisions that are themselves public health risks.

Businesses, residents and schools full of kids flew into a panic on Friday, investing energy and resources into coping with the order because they thought they were in the midst of an emergency.  I was in our local elementary school on Friday when the kids started panicking because that morning they’d brushed their teeth, washed some fruit, or simply had a big glass of water.  It was a sea of little faces staring at certain but as-yet-unquantifiable doom.  When I explained they’d all been drinking the questionable water for 3 days and yet they were all still fine and could go about their normal day without worrying about what would happen, they were normal, safe kids again.  This extra bit of information helped frame the real risk of the situation and empowered them to make good decisions, including: focusing on their schoolwork instead of every little bubble their gut was making.

Our public officials did not clearly communicate enough facts so that individuals could make solid risk assessments on Friday.  Information leads to accurate risk assessment, the kind that individuals need to be empowered to make for themselves.  Water Bureau said, “boil your water.”  They didn’t make it clear that you’d been drinking the water for 3 days already.  Or disclose virility of the E.coli (alive vs dead, dangerous vs friendly species, quantity present).  Or attempt in any way to help us assess the risk they rather vaguely announced.  Because I already know a little about this boil-alert process, I was able to make some risk assessments for my friends and family.  For instance, in this particular case, since we’d all come through Tuesday, Wednesday and Thursday unscathed, my family kept using the water for teeth brushing, hand washing, dishwashing as we normally would and we substituted in some bottled water for Friday just to reduce our potential bacterial load.  This seemed reasonable, given the circumstances of this case.  Yet, I see people online so uninformed about the real risks facing them this week, that they are now trying to figure out how to clean their hot water tanks.

As a community, we should ask the following:

1) If this kind of detect is a real public health threat, what can PWB do to shorten the time between samples and results — there is now an 18 hour gap (or more if it is a holiday) between samples and final results of the tests?  If something dangerous shows up on Tuesday, let’s don’t wait until Friday to hear about it.

2) If this kind of detect is NOT actually a public health threat, as would appear to be the case given that there was no discernible uptick in illness, what can PWB do to better communicate the pro forma nature of this alert, so businesses and citizens aren't investing resources in a non-emergency?  Can we institute a tiered risk alert system?  Can we design a tiered boil-notice, one that can go out immediately so the immunocompromised have information on say, Tuesday (for this case)?  Can the new system come with tiered packages of actions to be taken by the community?