Whether the CRO makes the kits is a decision treated as logistics in most CRO operating models. It is presented as a question of cost — is it cheaper to make kits in-house or to outsource them — and the answer is almost always that outsourcing is cheaper. The decision gets made on that basis. The CRO outsources the kits to the sponsor, to a third-party packaging vendor, or in the worst case asks the site to assemble the kits themselves from raw supplies shipped to them.
We have made the opposite decision. We make our own kits, on every IVD study, as a core operational discipline. It is not the cheapest option. It is the option that controls the variables that determine whether the study runs cleanly — and those variables are not visible until the kit decision has already been outsourced and the consequences appear three months into enrollment.
The framework below explains the three reasons. None of them is "we can charge more for it." Each one is a piece of operational control that disappears the moment kits are outsourced. Together, they are why kit-making is, in our view, not an option for a serious specialty CRO. It is a must-do.
Kits control enrollment timing.
A site cannot enroll without kits. The team that decides when kits ship decides when each site can scale — and the corollary, when to throttle.
The first reason is the most direct. Sites enroll when they have kits and stop enrolling when they don't. The team that controls the kit shipment schedule controls the enrollment schedule. When the CRO makes the kits, the kit resupply cadence is a project-management lever — the PM can ship more kits to the site that is producing well, slow shipment to the site that is over-enrolling beyond its share of the cohort, and pause shipment entirely to a site that is failing the 30-day enrollment check.
When kits are outsourced, this lever moves. The sponsor's procurement schedule controls when kits are made. The third-party vendor's batch cycle controls when they ship. The site's own kit-assembly capacity controls when they have kits to use. None of those parties is in the position to make the operational call about which sites need more kits this week and which need fewer. The result is that the study's enrollment trajectory falls out of CRO operational control, and any divergence from the forecast becomes harder to correct.
The corollary matters as much as the lever itself. The ability to throttle a site — to shorten its kit pipeline when it is enrolling subjects that don't fit the bucket, or when it is consuming kit inventory faster than the cohort needs — is the operational counterpart to scaling. Studies that hit value-bucket cohorts are studies where the PM redirected kits in real time; studies that miss buckets are often studies where the kit pipeline kept feeding the wrong site because no one was empowered to redirect.
On a recent multi-site Hep B seromarker study, the kit shipment plan was sized to each site's projected enrollment rate: an initial shipment of 30 kits to one site (projected at 10 subjects per week) and 60 kits to another (projected at 20 per week), with explicit contingency that the second site's count would be adjusted based on the first's actuals. The kit pipeline was the enrollment lever, not a downstream consequence of it. Across studies, this approach has consistently produced cleaner enrollment trajectories than studies where kits arrive on the sponsor's procurement calendar.
Initial kit shipment per site is sized against projected weekly enrollment with a deliberate cushion. Resupply triggers are tied to the enrollment forecast — when actuals lag or lead, the next shipment adjusts. The PM, not procurement, makes the call.
The kit is the operational support.
Pre-labeled tubes, pre-printed forms, pre-bagged supplies — every piece of work the kit does in advance is one fewer decision the site has to make under time pressure at the bedside.
The second reason is that the kit, made well, is the most direct operational support a CRO can give a site. A coordinator running consents on a busy clinic morning has a finite cognitive budget for non-clinical tasks. Every label they have to print, every tube they have to verify, every supply they have to retrieve from a separate cabinet, every form they have to find on a shared drive is a task that competes with their attention to the patient and the consent.
A well-made kit removes those tasks. Pre-labeled tubes with the subject ID range already printed. Pre-bagged biohazard supplies. Pre-printed CRFs. The reflex collection card pre-positioned in the kit so the coordinator does not have to look up which arm gets which collection on which visit. When the site opens the kit at the bedside, the kit has done the upstream cognitive work, and the coordinator can focus on the patient.
This is what most outsourced kit programs miss. Vendor-assembled kits ship the supplies; they do not ship the operational design. The labels are generic, the supplies are bulk, the documentation is in a separate envelope, and the coordinator does the assembly work at the bedside that should have been done at kit assembly. The site's enrollment rate falls accordingly, not because the site is incapable but because the cognitive overhead of each consent is too high to sustain pace.
RDI kits for a recent IVD study contained, for each subject ID range: a biohazard bag with prekit supplies, two pre-labeled 8.5mL tiger-top SST tubes with subject ID labels, sixteen pre-labeled 2mL cryovials with sequential subject-numbered labels (subject ID-1 through subject ID-16), and a butterfly needle, gauze, bandage, alcohol wipes, and sterile transfer pipette. The kit's design absorbed the tube-labeling, supply-organization, and inventory-tracking work that would otherwise have fallen on the coordinator. Sites enrolling from these kits consistently reported faster per-subject collection times than sites running comparable studies from generic supplies.
For every kit the team designs, walk it through the consent and collection workflow as the coordinator will see it. Every decision the coordinator has to make at the bedside is a candidate for being moved into kit assembly. Move as many as the science permits.
Quality is a process, not a hope.
Every kit goes through a structured QC check by a second person before it ships. Caught at QC means it does not break a sample at the bedside.
The third reason is the one most easily lost in a cost analysis. Kits made by the CRO go through structured quality control before they ship. The person who assembled the kit signs and dates the assembly record. A second person — independent of the assembler — runs through a QC checklist against the same kit, item by item, and signs and dates the QC record. Lot numbers and expiry dates are recorded for each component. The QC record names what was checked, when, and by whom.
This is the discipline that catches the problems before they become protocol deviations. A wrong-sized cryovial. An expired alcohol wipe. A mislabeled tube where the subject ID label has the wrong cohort code. A kit short by one component. None of these problems is rare. All of them happen during kit assembly. Every one of them, caught at QC, is a problem that does not arrive at the bedside. The kit that arrives at the bedside without QC is a kit whose problems are discovered at the moment of consent — by a coordinator who cannot proceed until the problem is resolved, or worse, who proceeds and produces a sample that has to be discarded later.
The QC discipline is also what produces the audit trail. Every kit on every study has a chain of accountability — who made it, who checked it, when, with which lot numbers — that survives any subsequent question about sample integrity. Outsourced kits have a chain of accountability with the vendor; CRO-made kits have it with the CRO; sites that assemble their own kits have, in many cases, no chain at all.
RDI's kit QC checklist for a recent study tracked, per kit, the biohazard bag with prekit supplies, the tiger-top SST tube labeling, the cryovial set (sixteen vials, subject ID-1 through subject ID-16), with comments and date/checked-by columns. The checklist is a standing artifact, used on every study, with the kit components and SID ranges adapted per protocol. Issues identified at QC — wrong-format labels, missing components, lot expiration concerns — are resolved before the kit ships.
No kit ships without a completed QC record. The QC is performed by a person other than the assembler. The QC checklist names every component, lot, expiry, and label format. The records are maintained for the life of the study and through the audit window after lock.
What this framework rules out.
The three principles describe why kit-making is a CRO discipline, not a logistics decision. They also rule out a few conventions worth naming.
They rule out kit-making as a cost decision. The cost analysis treats kit assembly as a fungible service to be optimized. It is not fungible. It is one of the few operational levers that touches every site, every visit, and every sample.
They rule out vendor-supplied kits as operationally equivalent. Vendor kits ship supplies; they do not ship operational design. The kit's design is the support, and design is not what kit vendors are selling.
They rule out site-assembled kits as a sustainable approach for any study where sample quality matters. Sites assemble kits when no one else does, and the assembly happens between patients in a room not designed for it. Quality is correspondingly variable.
The framework is not closed. When the study outcome matters, you call RDI. Kit-making is not an option. It is the discipline that makes the rest of the operational framework executable, and outsourcing it is outsourcing the part of the study that determines whether the data lock cleanly.