Suction, oxygen and medical gas at the bedside: piped, portable and what actually keeps working
A piped gas system is only as useful as the jars, tubing and probes at the end of it. How to size suction provision per bed, decide between piped and portable, and run the maintenance that keeps a gas system safe in a humid climate.
A district hospital commissioned a piped vacuum system, used it for about a week, then went back to mobile units on every ward. The pipework was sound, the plant ran, the outlets held. What had not been bought was a second set of collection jars, so a jar could not be turned round between patients and the ward reverted to the machines it trusted. The installation was defeated by a consumable worth a fraction of a percent of its cost.
That is the normal failure pattern for medical gas here. The capital item is specified carefully and the things that make it usable are left to somebody else's budget line.
Piped and portable are not alternatives
They fail in opposite directions, which is why every piped area still needs mobile units. A piped system fails for a whole ward at once; a portable unit fails one bed at a time. The ratio between them is a decision to be made deliberately, not discovered after handover.
Piped
Portable
Availability
Always on, nothing to charge, nothing to carry
Limited by battery or by a mains socket being free
Failure mode
Whole zone at once, and usually without warning
One machine at a time, which is survivable
Capital cost
High, and only sensible at construction or major refurbishment
Low per unit, incremental, easy to add later
Maintenance
Planned, needs a competent contractor and records
Ward level: filters, jars, seals, batteries
Transport use
None; the patient is tied to the wall
The only option in a lift or an ambulance
Best fit
Theatre, ICU, resuscitation, neonatal, dialysis
Wards, outpatients, transport, and backup everywhere
What defeats it
A crossed pipe, a failed float valve flooding the line, an untested alarm
A missing jar, a perished lid seal, a flat battery
Size suction on simultaneous demand, not on bed count
Vacuum plant sized on bed count is either over-specified and expensive or quietly inadequate at the far end of the run. The number that governs it is peak simultaneous draw, and that is calculable from your own activity.
For each clinical area, list the procedures that use suction and how long each occupies a source. Theatre and ICU are near continuous consumers, a labour ward is bursty, an outpatient area may need almost nothing.
Establish the maximum number of sources in use at the same time in each area, then across the building, allowing for the fact that peaks in casualty and theatre correlate rather than cancel.
Add the reserve your consultant specifies and confirm the plant meets it with one pump out of service, because that is the condition in which you will need it.
Then check the far end. Vacuum at the plant tells you about the plant. Vacuum at the most distant outlet tells you about the system.
Record the design figures in the handover pack. Five years later, when a ward is added, this is the document that says whether the plant can carry it.
An ICU bed often needs two vacuum points at the head end, one for airway and one for drainage. Discovering that after first fix is expensive.
Neonatal is a different requirement, not a smaller one. It needs low, controllable, accurately indicated vacuum. High flow capability is irrelevant and low end gauge accuracy is everything.
Theatre needs its own reserve. A theatre that loses vacuum mid-procedure has no fallback except a mobile unit, so that unit belongs in the theatre, not in a store.
Count dialysis, endoscopy and delivery separately. These areas are routinely left off the calculation because they were commissioned later.
The consumable that strands the machine
Ask what a suction unit costs and you get a figure. Ask where its jar seal comes from and you learn whether it will be working in three years. The consumable chain decides service availability in Bangladesh and is almost never evaluated.
Is the jar a standard fitting or proprietary? A proprietary jar with a proprietary thread ties you to one importer indefinitely. This is the single most consequential question on the list.
Liner system or reusable jar? Disposable liners are cleaner and faster in principle, and only work if the liner is continuously available locally. If it has to be imported, a reusable jar with a locally obtainable seal is the more reliable choice, whatever the brochure argues.
Where do lid seals and float valves come from? These are the parts that fail. Get part numbers in writing at order stage and buy a holding stock with the machine.
Can tubing be bought by the roll in Dhaka? If not, the department cannibalises tubing between machines, which is how one broken unit disables three.
Is the jar autoclavable, and does your CSSD know? A jar that is not autoclave rated will craze and then crack, usually after somebody has put it through a cycle to be helpful.
Standardise the model across the hospital. One jar, one seal, one tubing size, one filter. Mixed fleets accumulated over five tenders are why store rooms hold consumables that fit nothing.
A working rule: order two complete jar sets per machine, so one can be cleaned while the other is in service, and hold seals and float assemblies. None of this is expensive and all of it is invisible in a tender that only prices the machine.
Oxygen in a country with load-shedding
There are three practical routes to bedside oxygen and each fails differently. The right answer for most Bangladeshi hospitals is a combination, chosen so that the failure modes do not overlap.
Route
Strength
What it needs to be safe
Cylinders and a manifold
No electrical dependency at all, which is a serious advantage here
A ventilated manifold room, full and empty stock separated and labelled, cylinders restrained, no oil near valves, and a delivery contract that survives a public holiday
On-site generation plant
Removes the logistics chain and the recurring cylinder cost
Continuous power, meaning a generator backed supply and a fuel plan, plus a cylinder reserve covering both the changeover gap and planned servicing
Concentrators at the bedside
Cheap, quick to deploy, useful on lower dependency wards
A backed socket per device, a filter cleaning routine that happens, and cylinder backup, because a concentrator with nothing behind it is a single point of failure on a patient
Two Bangladesh specific points belong in any oxygen specification. Monsoon humidity loads the drying stage of a generation plant harder than service intervals published for a temperate climate assume, so ask what those intervals become at sustained high humidity and get the answer into the maintenance contract. And size on measured consumption: units that size a plant on bed count are regularly caught out when a handful of high flow patients outrun it and the reserve bank empties in a day.
The maintenance that keeps a gas system safe
Medical gas is the one hospital service where a maintenance lapse presents as patient harm rather than inconvenience. The routine below is unglamorous and it is the whole job.
Every shift, check the float shut-off on each collection jar. A failed float lets fluid into the pipework, and decontaminating a flooded vacuum line is a zone shutdown, not a task.
Change bacterial filters on schedule and after any known contamination, and record it. An undated filter is an unchanged filter.
Weekly, read the vacuum at the furthest outlet in each area and trend it. A slow decline over months is a leak or a failing pump, visible long before anyone complains.
Monthly, check probes and sockets. A worn socket that no longer grips the probe is the commonest reason a nurse reports that bedside suction does not work, and it is a five minute fix.
Test the gas alarms and sign for it. Alarms are usually installed, often wired and rarely tested. Put them on the maintenance schedule with a signature line.
Exercise zone valves and audit their labelling annually. A zone valve nobody can locate during an incident is not an isolation device.
Have identity and purity verified at the outlet by an independent party, annually. At the outlet, not at the manifold. Testing at the manifold cannot detect a crossed pipe downstream of it.
Re-test identity in the affected zone after any pipework intervention, however small. Not negotiable, and the item most often skipped when the work is described as minor.
If your tender needs a reference to point at, ISO 7396-1 covers pipeline systems for compressed medical gases and vacuum. Naming a standard forces a bidder to state which parts they comply with, which is more informative than a general assurance of quality.
Commissioning tests to insist on before you sign
Anti-confusion test on every single outlet. Every one, not a sample. This is the test that catches a crossed pipe and there is no other opportunity to catch it.
Flow and pressure drop at the furthest outlet under simulated peak demand, with the results recorded per zone.
Alarm function demonstrated at the panel and the remote repeater, including low pressure and reserve-in-use conditions, not only the mains healthy light.
As-built drawings with every zone valve marked, on paper in the plant room as well as a file on somebody's laptop.
Training with a signed attendance list, covering ward staff and porters, not only biomedical engineering.
A handover pack containing part numbers for every consumable: probes, seals, filters, jars, tubing. This is the document that gets skipped and the one you will want two years later.
What to specify in a portable unit
Mobile units such as the SK-EX122 and SK-EX123 are the backstop for everything above, and the portable SU-510 class covers transport and the lift. The specification that matters is short.
Adjustable vacuum with a gauge readable and steady at the low end, so one fleet can serve paediatric and neonatal use.
A mechanical overflow shut-off, tested at handover by deliberately overfilling a jar with water. Ask for that demonstration.
A bacterial filter as standard, with its replacement part number on the quotation.
Battery runtime stated at a defined flow and verified at handover, then re-measured every six months against the serial number, so ageing batteries are replaced on evidence rather than after an incident.
A jar you can buy in Dhaka. Repeated because it is the point most often ignored.
Castors that survive real ward floors, and a handle that lets one person carry the unit up a flight of stairs when the lift is out, which in most Bangladeshi hospitals is routine rather than exceptional.