Storage temperature records for lyophilized lots

A storage record describes where a lot was kept and what the air around it was doing. It is short, repetitive, and easy to fill in badly. Filled in properly, it is the document that decides what happens next when a unit fails overnight and someone has to say whether the lot was affected.

What an entry has to contain

The minimum set is small, and every element earns its place. Missing any one of them turns a fact into an opinion, which is the state most storage logs are in by the time anyone needs them.

• Equipment identifier, plus the position inside the unit, such as shelf or drawer
• Date, time, and the interval since the previous entry
• The value read, with its unit and the device that produced it
• The person recording, by initials, and the person who reviewed the sheet
• A field for a note whenever the value falls outside the expected band

Position matters more than it looks. A top shelf and a rear corner of the same unit do not warm or cool at the same rate, and a single probe placed for convenience describes one position, not the unit.

Units repay a second of attention as well. A sheet printed in degrees Celsius and filled in with a value read from a device set to Fahrenheit looks like an ordinary entry and is not one. Where two units appear in the same folder, print the unit on each line instead of leaving it in the column heading.

Logger or paper, and what each is good for

A data logger records at a fixed interval and produces a trace with no manual gaps, which is exactly what is wanted after an event. Its weakness is that the trace stops during a power interruption unless the device has its own supply, and its export is only as good as the file naming behind it. A paper sheet read at set times records a sampled picture and a person who actually looked. Its weakness is the opposite one: the hours between readings are invisible, and a value can be copied from the previous line without anyone noticing.

Running both, with the paper entry citing the logger channel, covers each weakness. Where only one exists, state in the record which one it is and how often it samples, so a later reader knows what the gaps mean.

Whichever device produced the value, its own status belongs beside it: when the probe was last checked against a reference, and by whom. An unchecked probe still records something, but not necessarily the temperature of the room. Put the check interval where a reader can find it, not in the memory of the person who set it.

How often to record

Frequency should follow two things: how fast the unit can move away from its band, and how long a departure could go unnoticed. A daily reading suits a stable unit in a quiet room; a per-shift reading suits an active study with frequent door openings. Adding entries does not improve accuracy, but it does shorten the blind interval, which is the number that matters when an event is reconstructed later. An interval chosen because it was convenient is a habit rather than a frequency.

After a departure, work in this order

The sequence matters, because each step narrows what the next one has to explain. Skipping to the conclusion is the usual error.

• Confirm the reading is real, with a second probe and a check for a door-open or defrost event
• Establish duration and extremes: highest and lowest value, and how long each held
• Check the positions that matter, since the front of a shelf and the back of it differ
• Compare against the storage statement issued for that particular lot
• Record the decision, the reasoning, and the name of the person who made it

The record itself does not decide anything. It supplies the four facts a decision needs: what the condition was, how long it lasted, where inside the unit it applied, and who judged it. A log that yields only the first of those four will be argued about.

A departure noticed an hour after it began is a different problem from one noticed three days later, because the second has an unknown middle. The record should state plainly which case applies, and how the gap was bounded.

How long to keep it

Keep the storage record at least as long as the study records it supports, and for a defined interval after the last vial from that lot is used. State that interval in the same document rather than leaving it to habit. Keep the logger export under the same identifier as the paper sheet, and hold a second copy somewhere that is not the laptop in the cold room. A record that cannot be produced is indistinguishable from one that was never made.

The standard to hold while the pen is still in hand is simple: could a reader who was never in that room answer the questions you would ask? Where the sheet cannot support that, it is a form that was completed rather than a record that was kept.