Everything that makes a laboratory result worth believing is found outside the analyzer: Controls at defined intervals, on defined material, against defined limits. An operator whose competency has been assessed and reassessed. An instrument in a room held at a known temperature. A director who reads the exceptions and signs their name.
The analyzer is the one part of that you can put in a bag. And that is roughly what the country is now trying to do.
For example, Rite Aid went from having 2,451 stores to none in less than four years, closing the last of them in October 2025. According to a 2025 study published in JAMA Network Open, roughly 17.7% of Americans live in a pharmacy desert. Similarly, after the sole pharmacy in Snow Shoe, Pennsylvania, closed, followed by the local federally qualified health center, residents were reported to be taking blood pressure medication every other day, cutting insulin, and sharing medication with neighbors. No one in a town like that is driving 90 minutes for an eGFR, so the test comes to them instead—in a van, on a bike, in a bag—run by someone who has never worked a bench.
That much is already underway, and I would not argue against it. What is missing is any serious accounting of what gets left behind.
Laboratory conditions are already eroding
Carry the analyzer out of the laboratory, and the systems and controls that make its results reliable, accurate, and reproducible may disappear. The more uncomfortable point is that most are going to places that still have a laboratory. Vacancy rates run as high as 28.5% in some laboratory departments, while ASCLS reports small rural labs often run mostly on short-contract travelers. Meanwhile, CLIA sets no education requirement for personnel running waived tests. CDC guidance makes clear that waived tests are not error-proof, with errors clustering where instructions are not followed or operators are unfamiliar with the system.
None of this is the fault of the operator. It describes what happens when the quality system goes and the test stays.
Decentralization aims to make testing available to people who have none, which is a good reason. But what makes a test result meaningful is not simply the test itself—it’s the quality system around it. Transport the test without the quality system, and you have not extended the laboratory, you have only extended the appearance of one.
A quality culture does not travel well
I have spent close to two decades building diagnostics for places with no laboratory, no grid, and no realistic prospect of a second visit. The lesson that took longest to learn is that a quality culture cannot be exported. A device travels. A procedure travels. What does not travel is what makes people run controls in a hospital, i.e., a director who will ask, an inspector who will review protocols and records, and experienced laboratory staff who have worked at the bench for years.
At a screening camp in a village, the device may be there, but the quality culture is not. The workflow has to carry it instead. In the field systems I have worked on, the service is not delivered unless the quality control check has cleared. Not flagged for review. Not logged for later. Not delivered. It sounds severe until you consider the alternative: a result that will never be verified going to a patient with no standing to question it.
The same logic applies to the instrument. Fifteen minutes at 42°C and 83% humidity can push a point-of-care glucose reading 33 mg/dL higher (a 30% bias), while other research has found meters unreliable within the manufacturer's own stated range. Indoors, the laboratory controls the environment an instrument operates in. Outdoors, a device and its workflow must be able to account for changes in environment, because an instrument operating beyond its limits does not announce its results are no longer reliable. The failure is silent, and it simply keeps producing numbers.
The real deliverable is access to care
The second lesson took longer to learn: capturing a result is not the same as closure.
The gap between reporting an abnormal test result and acting on it is well documented. A systematic review found that 6.8% to 62% of abnormal laboratory results went unfollowed in ambulatory care. The low end came from a fully computerized system that fired mandatory alerts at named providers, yet 6.8% still received no follow-up at 30 days. If the loop leaks inside a building with an EMR, a paper form in a van is not a plan.
In screening work I have been involved with, we routinely find women walking around with a hemoglobin of 7 or 8. Any laboratory professional would know that number is clinically significant. But these women do not. And these types of failures to follow up on abnormal results are not unique to remote settings: a 2026 analysis of 84,259 patient records in the US found 85% of patients with low hemoglobin never received follow-up testing for iron, ferritin, or B12.
In the same analysis, pathologists reviewing Pap smear cytology found cervical cancer in 270 rural women who did not yet know they had cancer. Each woman was referred to a nearby hospital. Finding the cancer was only the first step. The real deliverable was getting these women care.
The profession should set the terms
ECRI, a nonprofit focused on healthcare safety and quality, reports its top patient safety concern for 2026 is AI diagnostic tools deployed without adequate safeguards. Reduced access to rural health care is second. In actuality, these are not two separate concerns but one described twice: a diagnostic output arriving without the machinery that would justify believing it. It would be poor progress to answer the second with the first.
Testing will continue to be decentralized, whether laboratories lead the shift or not. Testing moved into pharmacies and urgent care while the laboratory watched and other people set the terms. What matters now is not whether testing is available, but whether the quality system comes along with it. From the patient's perspective, testing with that system and testing without it may look identical until the result needs to be verified, acted on, or followed up with access to care.
Rural health keeps asking how to save the hospital. But the real question—the one laboratorians are qualified to answer—is what needs to be in place to test patients and refer them to the hospital in the first place. A hospital cannot treat a disease that is never identified.
