Currently, socket technology appears to have a technical competitor, at least in theory, and that is "osseointegration".

There, the prosthetic terminal device - hand, hook - is mounted on a bolt that sticks through the stump's skin and resides inside the stump's bone. With sockets, a liner, embracing type of socket, straps or vacuum type principle is used to keep a socket in place, that is attached to the remaining components of the prosthetic limb.

Some drawbacks of osseointegration include that the bolt has to be implanted surgically over a series of several operations, that the skin will provide a chronic wound that carries the permanent risk of infection, that the bolt or stump's bone are relatively brittle and thus can and will break as they broke in the past, that many repairs then involve surgery rather than just a prosthetic technician, as well as temperature sensitivity particularly to cold temperatures. Also, most health insurances might not cover the procedure nor its complications or after care, so one requires some 80'000 USD for the operation, then more money for all the rest, essentially pointing to a privately paid 150'000 to 300'000 USD expense with no guarantee of function.

The advertising promotes bolt implants or "osseointegration" by suggesting that sockets are a main problem of discomfort and hygienic problems. Comparing between both systems, it would be honest to say that one needs to clean equipment and skin frequently anyway. With the efforts one must provide to clean an osseointegration wound many times a day using special disinfecting solutions, one can also keep a prosthetic socket and liner clean. In other words, who cannot keep a socket clean also might be a candidate for implant infections.

That aside, I cannot see how my normal life (outside wearing a prosthetic arm) could survive with a hard bolt sticking out of my otherwise (mostly) soft body surface (other people actually complain about any hard metal type material anywhere around, so, no bolts due to intimacy reasons to begin with).

Furthermore, I have relatively bad circulation on my stump, in such a way that it needs compression all the time anyway. And with that, I am (in fact) better off with wearing tight sockets rather than no sockets - for simple reasons of circulation and associated stump and phantom pain.

When one considers the metric performances that one would wish to achieve with an arm, or more interestingly here, a prosthetic arm, one might want to take actual facts into account. Such as: what type of forces (in terms of numbers and figures) might we be looking at.

Upper extremities and prosthetic options are getting it from left and right these days. So what are we going to believe? As I am currently suffering from post traumatic instability in my left wrist and elbow (the "healthy" arm), working the hard parts of life with my prosthetic arm become a real very day issue. Now, negotiating single grams is not an option any more.

For the most part, osseointegrated bolts may be nice to wave a prosthetic arm around the office. But it appears that the fun ends there. Since a considerable while now, some outspoken members of the osseointegrated community had been invited several times by me to provide videos demonstrating how they lift some usefully heavy 25 kg bags or more or such, but no such demonstration so far has been provided so far. In fact, we even lack a side by side comparison in how osseointegrating an arm stump will in any way improve ADL (activities of daily living). So the ratio of C (cost) over V (number of videos that provide proof  of superior force and ADL function) C/V is indefinitely high for the osseointegration of arms application.

Because, it is about overuse. Suspecting also that the world's most proverbially notorious osseointegrated patient admits to suffering from overuse of his other extremity himself already points to a serious flaw in just how much (read: how little) an osseointegrated bolt will accept in terms of (ab)use. So the C/V ratio is estimated to be so high for good reason: there is no actual function that osseointegration gives "back". Furthermore, even pulling a regular Becker hand's cables might just be too much for these bolts which is an assumption from other backchatter, pointing at failure risk levels around 5 kg just from that. So much for the anecdotal bits.

Real life has numbers and costs attached to it.

After all and in the long run, there is just one thing that we could keep in mind for being the real reason why we would like to wear prosthetic arms (if not for additional reasons): to prevent overuse of the other arm by any means conceivable. Not to dance for the public, to expose one's armpit or wave around passive weights with the goal of wearing some arm without real function. And to work towards overuse prevention, extensive mechanical vibrations, serious blows and hits, strong pulls and torques, high repetitions and high loads, simple but reliable function, and tight control options are what a prosthetic arm needs to be primarily built for. Not shee shee froo froo pantsy boy demonstrations like "I can hold a water bottle" or "I can hold a tooth brush" but lifting a table, moving a shelf. moving 20 filled up box moving containers, scrubbing a toilet or shower, wood working, actually getting some real stuff done.

To provide minimal symmetry for posture demanding tasks such as typing, choosing a bionic or other myoelectric arm is an error! The correct answer is to wear a light weight and vibration insensitive arm with either a hook or a hand, that may be equipped with a pencil. My favorites here are my passive arm or the hook.

To provide reliable support for garden tools, high pressure cleaning or vacuum cleaning, ironing or other manually tasking work, the Becker hand or a hook proved to be best.

Personally, I have ripped out wrists from their prosthetic socket mount, I have ripped off 1,5 mm steel cables due to poor mount points, and I have irreversibly damaged prosthetic wrists or prosthetic hands built by lesser able companies. Now, I wear my own customized body powered arm, I use a Puppchen wrist, I wear Alpha gel liners, and I  routinely pull and lift weights of up to 25-30 kg, occasionally up to 40 kg. And that is not considering sports.

So, figures now.

Real torques of socket mounted prosthetic arms may range around 120-160 Nm:

With gravity at g = 9.80665 m/s² and an average lower arm length at 0,4 meters, torque t = r x F with r = radius (m) and F = force (N), and in this instance (lifting), force F=m x g, so F = 30 to 40 kg x 9.8 m/s² = 294 to 392 N, and thus at the elbow, t = 0.4 m x 294 to 392 N = 117 to 156 Nm - not considering friction, peaks or accidents.

A fall may cause forces up to a few thousand Nm or so.

Osseointegration may hold up to around 37 Nm before fracture (the ultimate failure):

According to a recent study [1], maximum torque for osseointegrated upper arm bones (humerus) was 36,7 Nm when fracture occurred under strict laboratory conditions. That is around 7-8 kg. That is what was to be somewhat expected - but, guys? That's not much.

Comment

That is not even a third of what I would do with my conventional body powered arm using customized components - without any fracture. If anything, my pin lock might get a bit wobbly after carrying some five to ten 20 kg boxes or so - then I use a screw driver to fix it back into place. Even if I entirely deglove my cable setup (which happened with commercially available kits of a certain large Germany based company) I still will not have a fracture that causes pain, hemorrhage, loss of limb length, surgery, costs, time and suffering. I just don't.

That means:

• If you want to stay out of trouble with regard to overuse or repetitive strain injury of your remaining arm and hand, and you very well may want to target such a situation, then you may want to consider staying away from osseointegration, and you also may want to consider staying away from "bionic" or other myoelectric technology. That is, if you are facing manual work at all, and if you are inclined to do it.
• If you try to improve your body feeling by incorporating Tai Chi type anatomical perfection in a way that allows you to move weightlessly and effortlessly, if you are after seamless perfection of your anatomical shape and outline, and if you can manage to stay away (far away) from any real load, real weight, real work and realistic effort to your prosthetic (and maybe other remaining) arm, then (who knows) going for osseointegration and "bionic" parts may be the right thing for you. Then also, not wearing any prosthetic might work just as well though - in terms of overall cost, waste of resources and time and energy, the "no prosthetic" option or a "passive arm" option are not really that bad.
• In terms of hygiene: if you use the same energy you would need to sufficiently clean an osseointegrated chronic wound site to keeping your stump and liners clean, you should be alright.

Notwithstanding other considerations (What about intimacy and having a stump with a protruding metal bolt? What about biofilms? What about uncovered costs for implantation, and uncovered costs for complications, uncovered costs for after care or prosthetic options?) the mechanical weakness of the osseointegration concept for prosthetic arms as such might be a relevant aspect to consider.

General considerations

There are a number of indicators that society moves arm amputees into the corner of people without strong hand or arm function by default; instead, we are targeted as "pretty to look at anatomically (and stay there, that is, within looking distance)" rather than "functioning" or "soft on touch".

These are recent trends indicating just that:

• Hand transplantations are on the rise. Transplanted hands' strength, however, is never as good as prosthetic hand or gripper strength can and will be. Rehabilitation efforts are ridiculous, though. And immune system suppression can have a very high price.
• "Bionic" hands are on the rise. They have a hard time opening pop corn bags and definitely are of no help for serious weights - and yet, arm amputees world wide seem to run after these and big and small media clapping hand(s) all along.
• Osseointegration is extremely restricting in many ways, but not in one particular aspect: it allows to easily clamp a prosthetic to the stump.
• All kinds of "activities" are shown to "depend" on either hand transplantations, or "bionic" hands - people brushing their teeth, walking through a foreign town, making instant coffee, having to actually switch their hand functions in order to use a simple coffee machine, or reading a newspaper.I do these with the stump tied to my back, mind you, as those are properly termed Activities of No Particular Prosthetic Need. But this is a current trend, and not even Otto Bock's own showman employed his Michelangelo hand to plug computer cables.

Despite these "solutions" to arm amputation being notoriously useless from where I am standing, the trend seems to be that arm amputees increasingly parade - or are being paraded - as "anatomically complete" doll-type beings that never mind just having an upper load limit of a few kilograms. So the most relevant aspect currently seems to be that "successful" arm amputees use their arms and hands as visible symbols, not as engineering solutions to actually transmitting any forces, with their stumps not being sensitive soft touchable parts of their bodies but discardable ergonomically useless interface wasteland - just like electric plugs never pretty up a wall or desk.

Is that what currently is considered "human"?

@article{welke2013stiffness,
  title={Stiffness and ultimate load of osseointegrated prosthesis fixations in the upper and lower extremity},
  author={Welke, Bastian and Hurschler, Christof and F{\"o}ller, Marie and Schwarze, Michael and Calliess, Tilman},
  journal={{BioMedical Engineering OnLine}},
  volume={12},
  number={1},
  pages={70},
  year={2013},
  publisher={BioMed Central Ltd}
}