The temperature bands a blood bank has to hold, why a domestic refrigerator cannot hold them, how monitoring and alarms should be arranged, and what happens inside a stocked cabinet when the mains goes down.
A blood bank fails quietly. Nothing breaks, nothing alarms in a way anybody hears, and the units look exactly as they did the day before. The loss is discovered at issue, or it is not discovered at all. That is why almost every specification decision in a blood bank is really a decision about monitoring, and about what the equipment does while nobody is watching it.
Each component has its own storage condition, and they are not compatible with one another, which is why a blood bank needs several pieces of temperature controlled equipment rather than one large refrigerator.
Take the exact figures from the standard your service works to under the national safe blood transfusion programme rather than from a supplier brochure, and print them on the cabinet door. Staff should not have to remember which door holds which band.
This is not a question of build quality. A domestic refrigerator is designed to keep food acceptably cold on average. A blood bank refrigerator is designed to hold every point in its chamber inside a band at all times. Four differences do the work.
Storing components in a domestic refrigerator is the most common serious non compliance in small blood banks, and it survives because nothing visible goes wrong. The units still look like units.
Monitoring has three parts, and skipping any one of them makes the other two decorative.
The sensor should sit in a bottle of fluid that behaves thermally like a unit of blood, not in free air. Air temperature swings every time the door opens, and a probe in air produces a record full of alarms that staff learn to ignore. A probe in a simulated unit records what the blood actually experienced.
The record should be continuous and independent of the cabinet, whether that is a chart recorder or an electronic logger. A manual reading twice a day says nothing about the eleven hours in between, and the eleven hours in between is when the excursion happens.
The alarm has to reach a person who can act. An audible sounder inside a locked blood bank at midnight is not an alarm. Route it to a location that is staffed continuously, and test the route rather than the buzzer, on a schedule, with a signed record. Power the alarm circuit from its own battery as well, so that a mains failure cannot silence the alarm that should announce it.
A refrigerator holding a full load of units has a large thermal mass, and thermal mass is your ally. A full cabinet with the door shut stays in band far longer than an empty one. So the first two rules during load shedding are simple and rarely followed: keep the door closed, and keep the cabinet full, using sealed bottles of water as ballast where stock is low.
The second consideration is what happens when power returns. Compressors do not restart instantly against a head of pressure, and many have a deliberate delay before they will run again. A supply that flickers off and on repeatedly, which is exactly what an unstable evening supply does, can leave a compressor locked out although the lights are back on. That is why the blood bank belongs on a stabilised supply, and why the changeover arrangement matters as much as the generator itself.
Air conditioning belongs in this chain and is usually left off the essential circuit. A refrigeration cabinet in a room at 38 degrees works its compressor far harder than the same cabinet in a cooled room, and the failure that follows looks unrelated to the power cut two weeks earlier.
The cold chain does not end at the cabinet door. Units travel to theatre, to the ward and sometimes to another hospital, and each of those journeys either preserves the band or does not.
Use transport boxes with a documented packing arrangement, and validate them under your own conditions rather than accepting a figure generated in a temperate climate. A box that holds for several hours in a European van behaves differently on a Dhaka road in June. Validate with a data logger inside a simulated unit, on the actual route, in the hot season, and keep that record. It is the evidence that supports every subsequent journey.
At the ward end, the widely used rule is that a unit out of controlled storage beyond a short defined period cannot be returned to stock. Ward staff frequently do not know this, so units are sent up, not used, and returned hours later to sit alongside good stock. Print the time limit on the issue slip and give the ward a fast way to return a unit when a transfusion is deferred.
Around the storage sits the serology work: grouping, antibody screening and cross matching. That bench needs a serological centrifuge, a thermostatic water bath for incubation and for controlled plasma thawing, a cell washer where the workload justifies one, and proper lighting for reading agglutination. A bath such as the GEMMY YCW-01, with a stirrer moving water around the tank to even out temperature, is the right kind of item here, because an unstirred bath has a gradient across it that makes the setting on the dial close to meaningless.
Donation itself needs a reclining couch or therapy chair that can be put head down quickly, a collection mixer and weigher, a tube sealer, and a place where a donor who feels faint can lie down without being moved anywhere. A chair that cannot go head down quickly is a false economy the first time somebody reacts on it.
Records are the part that gets rebuilt in a hurry before an inspection and never convinces anybody. Keep the temperature charts, the alarm test log, the calibration certificates for every probe including the independent one, the maintenance history, the transport validation, and the excursion reports with the decisions attached. Calibrate the monitoring probes at a defined interval against a traceable reference. An uncalibrated probe reading inside the band tells you nothing at all, and it is the item most often missed.