Mimi si daktari, mimi ni mhandisi [I am not a doctor, I am an engineer]
I’m in Tanzania to use my engineering skills to use repairing hospital equipment, and I’ve already fixed my first bit of kit (a blood pressure cuff) and made a right mess with my second repair. Almost all hospital equipment in Tanzania is donated, the norm in LEDCs, which brings a whole host of problems for engineers to tackle: no service manuals, mismatched power supplied, no spare parts, and so on ad infinitum.
For the first month I attend technical lectures 4 days a week delivered by Larry Fryda, former dean at Pennsylvania College of Technology. The course roughly follows the textbook Medical Instrumentation in the Developing World, and covers everything from oxygen concentrators to electro-cardiograms to hospital beds. Larry is a very knowledgeable teacher who owns a bunch of guns and drives an 18mpg Chevy. We rub along just fine.
After lectures are labs, which have started simple, though it turns out I’ve been soldering wrong this whole time. We’ve built extension cables, familiar to me but useful as most of the American students have never worked with the grounded, fused, 220V plugs used in Tanzania and the UK. In fact, most of my fellow students are on biomedical engineering degrees: great for specific knowledge in western hospitals, but maybe less applicable to the general problem solving approaches needed here. We shall see.
On Fridays we visit Mt. Meru Hospital in Arusha and this week I fixed a pile of blood pressure cuffs. These are simple bits of kit: a pumps with a valve you open and close, a pressure gauge, a bladder you inflate with the pump and a material cuff holding it all together.
To use one you put the cuff on the patient’s upper arm and pump up the bladder until the pressure cuts off blood flow through an artery so you can’t feel a pulse. Then you slowly release air through the value to lower the pressure. At some point the pressure in the cuff drops below the systolic (peak) blood pressure and the pulse comes back. Feel for the pulse, record where the dial is when it comes back and tadaa: you have found the systolic blood pressure.
Typically the problem with the “broken” cuff was leaks somewhere in the tubing. Removing the tubes, trimming the leaky ends and refitting them fixed most of the cuffs in about 30 minutes. For the rest we bastardised bits from different systems to make frankencuffs: since the underlying tech is so simple this works beautifully.
The most frustrating was a “broken” paediatric cuff (henceforth “kiddie cuff”) where the only issue was torn stitching on the cuff holding the bladder in place. Kiddie cuffs are super hard to find, and many hospitals don’t have one. None of us had a needle and thread with us to repair it, so we brought it back and will return it next week.
The next fix I attempted was more substantial and a whole lot messier: an operating bed with a foot pump to raise and lower it, except you could only raise it, not lower it. Not ideal for surgery to be working up around your chest.
At Larry’s suggestion we flipped the bed on its side to get a good look at the underside. Hydraulic fluid poured out all over the floor.
Working around a growing puddle, we managed to unsieze the bolts (difficult), clean up the oil slick (more difficult), and remove the whole jack from the table and move it outside.
To operate the bed when it’s working you push the foot pedal to half way repeatedly to raise the bed; you push it all the way down and the bed lowers.
A look inside the case shows how it works.
With the little cylinder full of fluid, you push the foot pedal, move a piston into the cylinder and force the fluid through a one way valve into the big cylinder. The big cylinder has a bigger cross-sectional area than the little one, so a big linear movement in the little cylinder moves the bed up only a little bit. Combined with the long lever this gives a big mechanical advantage letting you lift the heavy bed relatively easily, exactly like a car jack. When you take your foot off the pedal, the little cylinder fills back up from the reservoir. When the lever is pushed all the way down, a second valve opens between the big cylinder and the reservoir, and the weight of the bed pushes the fluid back out into the reservoir. Rinse, repeat.
Unfortunately, no one had rinsed this thing for a looong time. Basically the problem was there was a load of congealed gunk in the big cylinder which blocked the release valve, stopping the big piston retracting. So in the end this is a clean out job: to fix it we’ll clean the gunk, replace the fluid (car brake fluid works great) and refit the jack. The oil slick occurred because, unusually, the reservoir on this model isn’t sealed to the air at all, it just has a cover: tip the bed and it pours right over the side. Lesson learned, we’ll return next week with some more fluid and get the bed going again.
Thanks for reading, here’s another poster from the TCDC library for your troubles.