4.5 Site Facilities, Equipment Maintenance, Calibration & Trial Supplies
Key Takeaways
- ICH E6(R3) Annex 1 section 2.2 requires the investigator to have adequate resources — qualified staff, facilities and equipment — available for the duration of the trial, not only at selection.
- Calibration proves an instrument reads true against a traceable reference standard; maintenance keeps it working; qualification (IQ/OQ/PQ) proves it was fit for purpose before first use.
- Calibration and maintenance certificates are essential records because they are the evidence that a reported measurement is reliable.
- An out-of-tolerance calibration result triggers a retrospective impact assessment of every measurement taken since the last passing calibration, not just a repair.
- Temperature-controlled storage requires continuous monitoring with alarms, documented excursion limits, defined escalation, and a validated contingency plan.
Site Facilities, Equipment Maintenance, Calibration & Trial Supplies
Quick Reference: Every number in a clinical database begins as a physical measurement made by a physical instrument in a physical room. If the instrument cannot be shown to have been accurate on the day of the visit, the number is not data — it is an assertion. That is why the ACRP-CP Exam Content Outline lists "Maintenance and use of equipment and supplies" as a Domain 4 knowledge statement and includes "Perform and/or verify equipment calibration and maintenance" and "Verify appropriate staff, facility, supplies, and equipment availability throughout the study" among the task statements.
What the guideline requires
ICH E6(R3) Annex 1 section 2.2 (Investigator — Resources) obliges the investigator to have, and to keep having, the resources needed to conduct the trial properly. The phrase that carries the exam weight is throughout the trial: adequacy is a continuing condition, not a box ticked during feasibility. Principle 9.2 adds the data-side rationale — systems and processes that aid data capture must be fit for purpose.
Four words candidates conflate
| Term | Question it answers | Typical evidence |
|---|---|---|
| Qualification (IQ / OQ / PQ) | Was the instrument correctly installed, does it operate as specified, does it perform in our setting? | Installation, operational and performance qualification reports, signed before first trial use |
| Calibration | Does it read true against a standard traceable to a recognised reference? | Dated calibration certificate stating the standard used, the tolerance and pass/fail |
| Verification (calibration check) | Is it still within tolerance today? | Daily/weekly check log, e.g. a balance checked against a class weight |
| Maintenance | Will it keep working? | Preventive maintenance schedule and service records |
Exam Watchout: Passing preventive maintenance is not calibration. A centrifuge that was serviced last month but whose tachometer was never calibrated cannot support a claim that samples were spun at the protocol-specified relative centrifugal force.
Traceability
A calibration is only as good as the standard behind it. The certificate must identify the reference standard and show an unbroken chain to a recognised national or international measurement standard. A certificate that says only "calibrated, passed" without naming the standard, the tolerance or the technician is a monitoring finding.
Equipment commonly in scope at a site
| Equipment | Why it matters | Typical control |
|---|---|---|
| Centrifuge | Incorrect speed or time alters serum/plasma separation and invalidates analyte results | Tachometer and timer calibration; RCF conversion documented |
| Refrigerators, freezers (−20 °C, −70/−80 °C), controlled ambient rooms | Stability of investigational product and of biological samples | Continuous monitoring, min/max recording, alarms, mapped units |
| Thermometers and data loggers | They are the evidence for everything above | Calibrated against a traceable reference; certificate on file |
| Balances / scales | Weight drives dosing, eligibility and endpoints | Daily check with class weights; periodic calibration |
| ECG machines | Cardiac safety endpoints and QTc exclusion criteria | Annual calibration; correct lead placement documented; date/time set correctly |
| Blood pressure devices | Primary endpoints in whole therapeutic areas | Validated device list; cuff sizes; calibration schedule |
| Spirometers | Respiratory endpoints | Daily calibration/verification with a 3-litre syringe, ambient conditions recorded |
| Infusion pumps | Dose accuracy | Biomedical engineering service and calibration records |
Date and time are part of calibration
Instruments that print or transmit a timestamp — ECG machines, glucometers, data loggers, EDC-connected devices — must have the correct date and time, ideally synchronised to a single site clock. A contemporaneous record with the wrong timestamp fails the contemporaneous element of ALCOA-C and will be queried at audit.
When calibration fails: the retrospective impact assessment
This is the highest-yield process in the section, because the intuitive response — repair and move on — is wrong.
Out-of-tolerance calibration result
│
▼
1. QUARANTINE the instrument (label it, remove it from trial use)
│
▼
2. DEFINE the affected window
from the last PASSING calibration ──► to the failure date
│
▼
3. IDENTIFY every trial measurement taken in that window
(which participants, which visits, which data points)
│
▼
4. ASSESS the impact: does the magnitude of the error change
eligibility decisions, dosing, safety assessments or endpoints?
│
▼
5. NOTIFY the sponsor; document as a deviation if data are affected
│
▼
6. REMEDIATE — repeat assessments where clinically meaningful,
flag affected data, raise a CAPA
│
▼
7. RECALIBRATE / replace and return to service with documentation
Exam Watchout: The affected window starts at the last passing calibration, not at the date someone noticed the problem. If calibration is annual and the failure is found at month 11, eleven months of measurements are in scope.
Facilities
Beyond instruments, "adequate resources" covers the rooms themselves:
- Space for the visit — private examination and consent discussion areas, and a defined area for procedures the protocol requires.
- Secure, access-controlled storage for investigational product, source documents, signed consent forms and the subject identification code list.
- Emergency provision — for trials involving first administration, infusion or known hypersensitivity risk: resuscitation equipment, in-date emergency drugs, oxygen, and documented staff competence.
- Backup power for freezers and monitoring systems, and a written plan for what happens when it fails.
- Environmental controls where the protocol depends on them, for example a temperature- and lighting-stable room for a photosensitive product.
Trial supplies that are not investigational product
Non-IP supplies are a routine source of avoidable deviations because nobody owns them:
| Supply | Failure mode |
|---|---|
| Laboratory kits (tubes, cryovials, requisition forms) | Expired kits used at a visit; wrong tube type causing sample rejection |
| Dosing diaries, questionnaires, patient-reported outcome devices | Out-of-date version used; unvalidated translation |
| Consumables (needles, syringes, gauze, dry ice) | Stock-out on a visit day forces an out-of-window visit |
| Device accessories and calibration standards | Cuffs, electrodes, spirometer syringes past their own service dates |
Good practice: a single named custodian, a supply log with lot numbers and expiry dates, a reorder trigger set well before stock-out, and a check of expiry dates in the pre-visit preparation for every participant visit.
Realistic exam scenario
Scenario: At a monitoring visit in month 9 of a cardiovascular outcomes trial, the monitor asks for the ECG machine calibration certificate. The site produces a certificate dated 14 months earlier and a service record from three months ago showing "annual service — passed." Fourteen participants have had protocol-specified ECGs in the past four months, two of whom were excluded at screening for a prolonged QTc interval.
Evaluation:
- The service record is not a calibration record; the machine has been out of its calibration interval for two months.
- The affected window runs from the last passing calibration, so every ECG in the past two months is in scope, and the site must determine whether any earlier drift is plausible.
- The two screening exclusions are the serious part. If the machine were reading QTc high, two people may have been wrongly denied access to the trial — an ethical issue under Principle 1, not merely a data issue.
- Correct actions: quarantine the machine; calibrate it immediately and obtain a certificate stating the reference standard and tolerance; compare the as-found result against tolerance to judge whether the readings were biased and in which direction; list the affected participants and visits; notify the sponsor; document a deviation; repeat ECGs where clinically meaningful and re-evaluate the two excluded candidates if the as-found error is material; raise a CAPA that puts every instrument on a tracked calibration calendar with an owner and an advance reminder.
A site’s balance fails its scheduled annual calibration in November. The last passing calibration certificate is dated the previous November. Which participants’ weight measurements fall within the impact assessment?
Which record best demonstrates that a −80 °C freezer holding investigational product was suitable for use throughout the trial?
A coordinator discovers during pre-visit preparation that the central laboratory kits scheduled for tomorrow’s visit expired last week. What is the appropriate first action?