The preservation medium is the least visible component in a collection kit and the one most likely to invalidate a result. It is also the decision most often made by copying whatever the last kit used.
Here is the decision framework, in the order the questions actually have to be answered.
Start with what the laboratory does to the sample
Not with the specimen. Not with the collection device. With the assay.
If the laboratory extracts nucleic acid and runs PCR or sequencing, the sample does not need to be alive. It needs to be intact and stable. That opens up inactivating chemistries and, with them, ambient shipping.
If the laboratory needs to culture the organism, or needs viable cells, an inactivating medium destroys the thing being measured. You are now in viable transport and cold chain, and the whole logistics profile of the program changes.
If the target is a protein, a metabolite, or a drug, the question is different again: what degrades it, and does the medium interfere with the detection chemistry.
Everything downstream follows from this answer. Get it from the laboratory in writing before you evaluate a single medium.
Then ask what the sample has to survive
A medium that performs well in a validation study run two floors below the collection point tells you very little about a kit that sits in a mailbox in Phoenix in July and then in a sorting facility over a holiday weekend.
The real transit profile is: how long, at what temperature range, with what handling. Ask for stability data that covers the worst case you will actually ship into, not the nominal case. If that data does not exist for your specimen type in that medium, you need to generate it, and that is a study, not a phone call.
Ambient stability is worth paying for. It removes cold chain, removes shipper qualification, removes the temperature excursion investigation, and removes the single largest recurring cost in a direct to consumer program.
The main categories, by what they do
Guanidine based nucleic acid preservation
Lyses cells, inactivates pathogens, stabilizes DNA and RNA at ambient temperature. The workhorse for molecular testing on almost any specimen type. The trade is that nothing survives it, so culture and viability testing are off the table, and the chemistry itself is an irritant that has to be handled correctly by a consumer.
Saliva collection media
Formulated for stabilizing DNA in oral fluid at ambient. Well understood, consumer friendly, and forgiving of the volumes people actually produce, which is less than the instructions request.
Viral transport media
Keeps virus viable for culture and antigen work. Requires cold chain, and viability degrades on a clock. Choose this only when the laboratory genuinely needs a live organism.
Molecular transport media
Inactivating and ambient stable, aimed at molecular workflows. Sits where viral transport media sits in the workflow but gives up viability to gain shipping simplicity.
Blood stabilization chemistries
Preserve cell free nucleic acid by preventing cell lysis, which is the opposite problem from the guanidine case. Highly specific to the analyte and generally not interchangeable.
Dry formats
Absorbent substrates and volumetric absorptive microsampling remove liquid from the kit entirely. No spill, no irritant, no leak in transit, and a simpler packaging problem. The trade is elution recovery and lower sample volume, which the laboratory has to accept.
Volume is a decision, not a detail
Fill volume relative to specimen volume sets your dilution factor, and dilution moves your limit of detection. A medium that works at a one to one ratio may not work when a user contributes half of what the instructions asked for, and users routinely do.
Ask the laboratory what the minimum acceptable specimen volume is, then design the fill so the assay still works at that minimum rather than at the target.
Compatibility is the failure nobody plans for
The medium has to be compatible with the extraction chemistry, not just with the specimen. Carryover of a lysis buffer into a downstream reaction is a real and common failure, and it does not show up until someone runs the full workflow end to end.
The only reliable test is the laboratory validating the actual medium, at the actual fill volume, in the actual vessel, from an actual production lot. Not a sample vial from the supplier. A unit from the lot that will ship.
The vessel and the human matter as much as the chemistry
A preservative is a chemical in the hands of an untrained person. Closure integrity, the force required to open and reseal, whether the vessel tips easily on a bathroom counter, whether the liquid is visible enough that a user knows not to drink it. These are not packaging concerns. They are the difference between a working kit and a support burden.
If your medium is an irritant, it drives label warnings, it drives the instruction sheet, and in some programs it drives the choice of a dry format instead.
Regulatory posture
Whether a medium inactivates infectious material can affect how the returned specimen is classified and shipped. That determination belongs to your regulatory lead and to the specific medium and specimen combination. Do not assume it carries over from another program, and do not let a supplier’s marketing language substitute for your own assessment.
Then somebody has to fill it
This is the part that gets treated as trivial and is not.
Once the medium is chosen, every unit needs the same volume of the same formulation, from a controlled lot, with the fill documented. Manual filling drifts. Drift changes the dilution factor. A changing dilution factor moves the limit of detection across your lots, and nobody finds out until the results look strange months later.
Automated liquid handling with a documented fill volume per lot is not a luxury on a program that intends to scale. It is the thing that makes lot to lot results comparable.
The short version
Ask the laboratory what it does to the sample. Ask what the sample has to survive in transit. Choose the chemistry that satisfies both, then validate it in the real vessel at the real fill volume from a real lot. Design the fill for the volume users actually give you, not the volume you asked for. Then make sure whoever fills it can prove they filled it the same way every time.
Related reading: custom formulation and aliquoting covers filling a medium to your own specification rather than a catalog format, and building a kit around a novel collection device shows how the media decision interacts with the rest of the build.
