𝗟𝗲𝘁'𝘀 𝗕𝘂𝗶𝗹𝗱 𝗣𝗮𝘁𝗶𝗲𝗻𝘁 𝗠𝗼𝗻𝗶𝘁𝗼𝗿𝘀
As many of you know, I was hospitalised for Covid, and spent about ten days in the ICU. If you listen to this video, but tune out whatever Manish is saying, you will hear in the background a cacophony of all kinds of medical devices beeping.
If you listen to these sounds for a long time, for instance, 240 contiguous hours interspersed by a few precious hours of sleep, you start finding melody and patterns. You start visualising the code that emits these infernal sounds. And, most importantly, as an engineer, you start imagining a utopia of open source hardware and software based medical monitoring devices with better UIs that let the nurses turn off these meltdown inducing sounds. Or better, the code is smarter and just doesn't emit noise when unnecessary.
A patient monitor is a device that costs about ₹40K/$500 in India. It continually monitors your SpO2, pulse rate, and some other vitals (they stick electric probes all over your torso). Every half hour or so, it inflates a blood pressure monitor permanently velcroed to your non catheterised arm. I was hooked to a monitor made by Philips. Every time my SpO2 fell below 90, it started beeping loudly. This typically happened every time I talked, ate, or took a piss (In an ICU you can only pee into a can while lying on your bed. Walking to the loo is heavily discouraged, and the first time you ignore the nurse, stand up, then swoon and fall back on the bed, you learn why. Peeing into a can as an adult was quite embarrassing and made my SpO2 and other vitals drop due to stress.)
Now, when my SpO2 went back above 90, the bloody thing didn't stop. It kept beeping. Loudly. Kept on beeping perpetually till someone was kind enough to come press its yellow button. Now, here's the crazy thing. I didn't have a call bell at my bed. (A lot of corona ICU wards were created in a hurry; maintenance is also an absolutely hassle.) So I had to shout out for the nurse. Which, guess what, would dip my SpO2. See how delicate a darling I am. So it was like crying wolf. You called the nurse or ward girl to turn off the beep. She came and saw a large red 88 and said the damn thing is just doing its job.
Anyway, they eventually understood what's going on, and also understood how annoying the sound was to me. I had to go back to the basics. My first resume, written in 2003, carried the tagline "I like to make technology user friendly". This wasn't user friendly. The beeps had all become comfort noise to them, they were tuned out, but it annoyed me at two levels - the noise annoyed me, and the realisation that some dolt of an engineer and product manager thought such persistent noises were a good idea in 2020, annoyed me even more. They appreciated all that and then tried to keep my bed quiet at all time. The other beds kept howling 🙄
Note also that manipulating these devices using the buttons and dials on them is hard. They are mounted high, higher than six feet, so they can be seen from a distance, over the head of anyone standing between them and the viewer. They are mounted behind the patient bed, so there's stuff like ventilators or bedside tables that need be moved out of the way to access them. And then the nurse needs to often clamber over something to get to the height of the buttons. Terrible, terrible design.
Perhaps India is behind the rest of the world. Perhaps the world has moved on to wrist mounted buzzers that alert nurses. But I have a reason to suspect they haven't. My hospital had very good quality ventilators. They had a fancy name, so they were likely European. They beeped continuously when working, and the nurses liked it this way, because they could tell the machine is working. So, in an ICU, each nurse was supposedly listening to the continuous beeps from three different ventilators, and telling it apart from the beeps of the nine other ventilators in the same room 🙄.
So basically, every contemporary medical device with a microprocessor in it beeps. It gives out comfort noise that keeps the nurses happy. The machines need to change. And hospital culture needs to adapt to the fact that we engineers can assure you that your machines are working, your patient is normal, and alert you when man or machine breaks, in better and more reliable ways than only sound.
I'm interested in building such devices, and the ecosystem - you need network based monitoring, reports, handheld devices that can manipulate the UI of a nearby monitor, wearable alerting devices, and perhaps even wearable wireless sensors for patients that can eliminate those cumbersome cables. Please leave a comment here if you'd like to get involved.
PS: I'm not interested in building a ventilator. That thing howls like wild the moment the person it is serving skips a breath, hiccups, does anything with their breathing. For good reason. A single missed breath is an indication of distress. Within seconds, you might lose the patient. The devices are adaptive, and they listen to every breath you take and tweak themselves. Even a basic BiPAP is adaptive. A misconfigured ventilator gives you lung fibrosis for life - imagine what a malfunctioning one can do (umm, inflate your lungs till you explode). Anyone who claims they can design a reliable electro-mechanical device with multiple feedback loops and continuous life-critical fraction-of-a-second monitoring in the middle of a pandemic, is lying. If the world didn't invest in ventilators before the pandemic, too bad. The tried and tested ventilator manufacturers, who perfected their product over decades, will extract their pound of flesh, and there is no way around that now.
As many of you know, I was hospitalised for Covid, and spent about ten days in the ICU. If you listen to this video, but tune out whatever Manish is saying, you will hear in the background a cacophony of all kinds of medical devices beeping.
If you listen to these sounds for a long time, for instance, 240 contiguous hours interspersed by a few precious hours of sleep, you start finding melody and patterns. You start visualising the code that emits these infernal sounds. And, most importantly, as an engineer, you start imagining a utopia of open source hardware and software based medical monitoring devices with better UIs that let the nurses turn off these meltdown inducing sounds. Or better, the code is smarter and just doesn't emit noise when unnecessary.
A patient monitor is a device that costs about ₹40K/$500 in India. It continually monitors your SpO2, pulse rate, and some other vitals (they stick electric probes all over your torso). Every half hour or so, it inflates a blood pressure monitor permanently velcroed to your non catheterised arm. I was hooked to a monitor made by Philips. Every time my SpO2 fell below 90, it started beeping loudly. This typically happened every time I talked, ate, or took a piss (In an ICU you can only pee into a can while lying on your bed. Walking to the loo is heavily discouraged, and the first time you ignore the nurse, stand up, then swoon and fall back on the bed, you learn why. Peeing into a can as an adult was quite embarrassing and made my SpO2 and other vitals drop due to stress.)
Now, when my SpO2 went back above 90, the bloody thing didn't stop. It kept beeping. Loudly. Kept on beeping perpetually till someone was kind enough to come press its yellow button. Now, here's the crazy thing. I didn't have a call bell at my bed. (A lot of corona ICU wards were created in a hurry; maintenance is also an absolutely hassle.) So I had to shout out for the nurse. Which, guess what, would dip my SpO2. See how delicate a darling I am. So it was like crying wolf. You called the nurse or ward girl to turn off the beep. She came and saw a large red 88 and said the damn thing is just doing its job.
Anyway, they eventually understood what's going on, and also understood how annoying the sound was to me. I had to go back to the basics. My first resume, written in 2003, carried the tagline "I like to make technology user friendly". This wasn't user friendly. The beeps had all become comfort noise to them, they were tuned out, but it annoyed me at two levels - the noise annoyed me, and the realisation that some dolt of an engineer and product manager thought such persistent noises were a good idea in 2020, annoyed me even more. They appreciated all that and then tried to keep my bed quiet at all time. The other beds kept howling 🙄
Note also that manipulating these devices using the buttons and dials on them is hard. They are mounted high, higher than six feet, so they can be seen from a distance, over the head of anyone standing between them and the viewer. They are mounted behind the patient bed, so there's stuff like ventilators or bedside tables that need be moved out of the way to access them. And then the nurse needs to often clamber over something to get to the height of the buttons. Terrible, terrible design.
Perhaps India is behind the rest of the world. Perhaps the world has moved on to wrist mounted buzzers that alert nurses. But I have a reason to suspect they haven't. My hospital had very good quality ventilators. They had a fancy name, so they were likely European. They beeped continuously when working, and the nurses liked it this way, because they could tell the machine is working. So, in an ICU, each nurse was supposedly listening to the continuous beeps from three different ventilators, and telling it apart from the beeps of the nine other ventilators in the same room 🙄.
So basically, every contemporary medical device with a microprocessor in it beeps. It gives out comfort noise that keeps the nurses happy. The machines need to change. And hospital culture needs to adapt to the fact that we engineers can assure you that your machines are working, your patient is normal, and alert you when man or machine breaks, in better and more reliable ways than only sound.
I'm interested in building such devices, and the ecosystem - you need network based monitoring, reports, handheld devices that can manipulate the UI of a nearby monitor, wearable alerting devices, and perhaps even wearable wireless sensors for patients that can eliminate those cumbersome cables. Please leave a comment here if you'd like to get involved.
PS: I'm not interested in building a ventilator. That thing howls like wild the moment the person it is serving skips a breath, hiccups, does anything with their breathing. For good reason. A single missed breath is an indication of distress. Within seconds, you might lose the patient. The devices are adaptive, and they listen to every breath you take and tweak themselves. Even a basic BiPAP is adaptive. A misconfigured ventilator gives you lung fibrosis for life - imagine what a malfunctioning one can do (umm, inflate your lungs till you explode). Anyone who claims they can design a reliable electro-mechanical device with multiple feedback loops and continuous life-critical fraction-of-a-second monitoring in the middle of a pandemic, is lying. If the world didn't invest in ventilators before the pandemic, too bad. The tried and tested ventilator manufacturers, who perfected their product over decades, will extract their pound of flesh, and there is no way around that now.