Imported hospital equipment rarely fails because it was badly made. It fails for a short list of local reasons, most of which are decided at specification and commissioning rather than during use.
Walk through the store room of any large Bangladeshi hospital and you will find equipment that was working three years ago and is not now. Very little of it was defective. Almost all of it failed for one of the reasons below, and almost all of those reasons were decided before the item was ever used.
Every hospital has a generator, so the outage is handled. What is not handled is the voltage between outages: sags when a large load starts, swells afterwards, the transient at the moment the generator picks up, and neutral to earth voltage on a distribution board carrying more than it was designed for. Imported equipment designed for a stable supply fails on its power supply board, repeatedly, and the fault looks random because nobody is measuring the supply.
What to do about it is decided at design: specify a wide input voltage range on anything that stays plugged in, put critical equipment on a UPS that carries through the generator transfer rather than for a token few minutes, and measure earthing at the outlet during commissioning instead of assuming it. An unearthed socket in a ward is both a safety problem and an equipment killer.
Sterilisers scale, dialysis membranes foul, and both fail years early because the plant was sized from a catalogue rather than from a laboratory analysis of the actual supply. The two most common local causes are hardness and iron, and the two most common local failures are a softener that stopped regenerating because nobody bought salt, and a carbon vessel that was never changed because there was no sample port to test at.
Monsoon humidity gets into unsealed enclosures and corrodes circuit boards slowly enough that the failure looks like bad luck. It also defeats a CSSD that cannot dry a load, producing wet packs that are not sterile. Dust from construction, which in most Bangladeshi hospitals is continuous somewhere on the site, blocks filters and cooling paths and finds its way into optics and platens.
Storage is where this does most damage. Equipment delivered early into a building under construction sits for months in dust and damp, and it arrives at commissioning already aged. That is entirely a scheduling decision.
An installation qualification and an operational qualification are documents, and when they are not produced nobody notices for a year. The consequences are specific: a steriliser never leak tested, a pendant whose earth continuity through the arm joints was never measured, an RO plant that made acceptable water on the day and was never tested at full demand, a bed lot delivered and stacked without anyone confirming the rail latches worked. Each of these becomes a failure that looks like a product defect.
This one is quiet and expensive. A washer disinfectant that the manufacturer never approved attacks the chamber and the instruments. A cheaper filter that fits but flows differently changes the pump duty. Oil put into a pump specified as oil free destroys it. Thermal printer paper that fades leaves a sterilisation record that cannot be read at audit. A chemical indicator bought on price that is not the class the cycle requires proves nothing.
The pattern is always the same: the consumable is bought by a different person, on a different budget, against a description rather than a specification. Fix it by putting the approved consumable list in the maintenance file and in the store's ordering system, with the reason written next to it.
A handful of daily habits account for a surprising share of failures. Suction pulled through a full jar puts secretions into the pump, which is a replacement rather than a repair, and the overflow protection device that prevents it is often present and unmentioned in training. Mopping solution run into unsealed castor bearings corrodes them within a season, which is why trolleys and beds end up with one seized wheel. A handset wiped with a wet cloth admits fluid past the membrane. A disinfectant that the furniture finish was never rated for crazes the surface and turns a cleanable panel into a porous one.
None of that is carelessness. It is what happens when equipment arrives without training, or with training given once to staff who have since been transferred.
The last reason is the most ordinary. A part with a long import lead time turns a two hour repair into a season out of service, and equipment that sits waiting for a part gets cannibalised for another machine. Meanwhile the one biomedical technician who was trained on it moves to another posting, and the knowledge goes with them.
Both are procurement decisions, not maintenance decisions. Ask at bid stage what is stocked in Bangladesh, write the response time into the contract, and insist that training covers two people with a written procedure left behind, so the department is not one transfer order away from losing the ability to run its own equipment.