The Van, the Letterbox and the Assay: Validating UK Home-Sampling Kits for Real Transit
Photo: British Library, No restrictions, via Wikimedia Commons
Home-sampling kits are no longer a niche pilot activity. They are being used for sexual health screening, renal monitoring, therapeutic drug monitoring, endocrine testing and a growing range of remote consultations. Yet many validation packs still treat the kit as a laboratory product that happens to be posted. In decentralised testing, that assumption is difficult to defend. The courier, the letterbox and the patient instruction leaflet are part of the analytical system, not an afterthought bolted onto the end of it.
A stability study that uses one ambient condition and a single 24-hour courier run is not a transport study. It is a bench study with a courier receipt. The purpose of transport validation is to establish whether the sample still produces a clinically defensible result after the journey the kit will actually take, including the journeys that go wrong.
Define the analytical system before writing the protocol
The first task is to agree the system boundary. For a home-sampling kit, that boundary should include the collection device, the patient at home, the instructions for use, the packaging, the postal or courier network, the receiving laboratory and the result-reporting pathway. A risk assessment should identify where the sample can change: time at room temperature, exposure to heat or cold, agitation, delayed collection, delayed posting, weekend sorting, failed delivery and return to sender.
This boundary matters because it changes what counts as validation. If the kit is validated only with spiked venous blood held at controlled room temperature, the study has not demonstrated performance in capillary blood collected by a patient, dried or stabilised in a device, and posted through a network that does not control temperature. The protocol should state which matrices, devices and transit routes are in scope, and which are explicitly out of scope.
Build a transport-stability matrix, not a single study
A useful transport-stability study is a matrix. At minimum, it should cover each analyte or analyte group, each sample matrix, each collection device and each seasonal condition that the kit is expected to encounter in the UK. For blood kits, that may mean capillary whole blood, dried blood spots and stabilised venous blood in a collection tube. For saliva kits, it may mean neat saliva, stabilised saliva and saliva collected into a device with a preservative.
The matrix should include temperature excursions that are plausible rather than convenient. A winter profile may include a cold letterbox, an unheated van and a sorting office at low temperature. A summer profile may include a hot delivery van, a south-facing porch and a delayed collection. The study should include a nominal control condition, but it should not rely on that control to carry the conclusion. Data loggers placed inside representative kits can document the actual temperature profile and support the interpretation of any failure.
Transit time should also be treated as a variable. Royal Mail 24 and Royal Mail 48 are not the same product, and neither is a tracked service that requires a signature. The protocol should include the intended service, the expected proportion of samples arriving within the return window, and a deliberate late-arrival arm. A kit that performs well at 24 hours but fails at 72 hours needs an instruction that tells the patient what to do when the 72-hour case occurs.
Simulate the postal network, then verify with it
Simulation is useful for controlled experiments, but it is not a substitute for the real network. A laboratory can use environmental chambers and shaker tables to test temperature and agitation, but it should also run end-to-end transit trials through the actual postal service. Those trials should include different regions, different seasons, weekend and bank-holiday periods, and both letterbox-compatible and tracked-return formats.
End-to-end trials should record despatch time, receipt time, temperature and sample condition on arrival. The receiving laboratory should assess haemolysis, clotting, volume, device integrity and any visible leakage. For saliva kits, the assessment may include viscosity, preservative volume, microbial overgrowth and device fit. The trial should also capture patient-facing failures: unclear instructions, missing labels, incorrect packaging and kits that are too large for a standard letterbox.
Acceptance criteria that commissioners can defend
Acceptance criteria should be defined before the study begins and should be analyte-specific. A generic allowance of plus or minus ten per cent is rarely sufficient. The laboratory should state the clinically allowable change for each analyte, the total analytical error that the method can tolerate, and the stability budget available for transport. Where a change is not clinically material but is analytically detectable, that should be documented rather than hidden.
For blood kits, acceptance criteria may include recovery of the analyte within a predefined interval, absence of clinically significant haemolysis, adequate sample volume and successful amplification or immunoassay performance. For saliva kits, criteria may include recovery, preservative effectiveness, device integrity and absence of interference. The study report should include the number of samples, the number of replicates, the conditions tested, the failures observed and the rationale for any deviation.
Instructions for use are part of the validation
Patient instructions are not a compliance appendix. They are a control that determines whether the sample arrives in a condition the laboratory can accept. The instructions should state when to collect, how much to collect, how to label the device, how to pack it, when to post it and what to do if the sample cannot be posted within the return window. They should also explain what happens if the kit arrives at the laboratory outside the stability window.
Comprehension testing is valuable here. A kit can be analytically stable and still fail because the patient posts it on Friday afternoon, stores it near a radiator or removes the desiccant. The validation pack should include evidence that the instructions have been tested with representative users, that the reading age is appropriate and that the packaging is robust enough for the intended postal service.
The documentation pack commissioners will ask for
Commissioners and receiving laboratories will want more than a certificate. A defensible pack should include the validation plan, the risk assessment, the study report, the temperature and transit records, the deviation log, the instructions for use, the training materials and the complaints and incident procedure. It should also show how the kit will be monitored after launch: sample rejection rates, arrival times, temperature excursions, failed deliveries and any analyte-specific drift.
Where the kit carries a regulatory marking, the validation pack should align with the manufacturer's intended use and the laboratory's own accreditation arrangements. The receiving laboratory should be able to explain how it will handle a sample that arrives outside the validated conditions, and how that decision will be recorded. The courier is part of the assay. The validation pack should read as though the laboratory knows it.
