Cost of Defensive medicine

There are many unexpected consequences for everything we do. For every action, there is an opposite reaction…. but in life (unlike physics), it isn’t always an equal and opposite reaction.

When I hear about healthcare costs, rarely does the discussion ever goes into some of the truer (hidden) costs or issues. But these issues run deep. I have an anecdote about Baxter, that happened to me, that I like to share.

 One huge eye-opener was from first hand experience while I was working for Baxter, making  medical instruments, and the issues around medical liability reared its ugly head. It can literally make or break entire companies.

I don’t think the public fully realizes the impacts of this on their healthcare costs, the negative impacts on quality they get because of it. Companies have to do gymnastics to avoid liability and "cover their asses". (And that’s before looking into the even worse behaviors by medical staff and providers). This was driven home when I first started at Baxter Edwards.

I basically was hired into pandemonium.

Infusion Confusion

The division I joined made infusion pumps; intelligent devices that infused drugs into a patient at particular rate, with all sorts of features about delays, mixtures, ramp-up and ramp-down speeds, and so on; basically an intelligent IV line.

Well, if you know software, you can guess what happened. One day, on leap year, whether the stars aligned just wrong, and a drug had a certain rate, and a name was a certain size, they overflowed a buffer, wrapped something to zero; and the pump decided to dump its load. This killed the patient. Now THAT is a software bug.

The company had gone through multiple auditing processes, done months or years of testing, and the machine had been in use for 10 years, all before these certain stars aligned incorrectly, and all hell broke loose. So the company had done everything reasonable to protect human life; but does any of that matter? Not when lawyers and money are involved.

An honest mistake meant the resulting lawsuit could have put the entire organization, out of business. Our ridiculous attitudes towards liability awards are such that a single event can wipe out a decade of profits or an entire organization. A device that helped saved 10’s of thousands of lives, could be taken off the market, because of a "good faith" mistake and a .01% failure rate. Thousands of people losing their jobs, all over something that everyone involved took amazing pains to avoid.

Think about those implications to society.

When I got there, the company was running around pulling every device off the market, and retrofitting them because of this newly discovered bug. No one factors in the costs of patients NOT having those devices. Many only see an opportunity to win the dead-relative lotto. 

Ironically, Baxter was saved, not because of their efforts in all this; that doesn’t matter in the real world of liability roulette. What saved Baxter was that the Hospital had ONCE had this particular infusion pump serviced by a non-Baxter authorized support person (his authorization had expired), so all liability was transferred to the Hospital because of a non-authorized servicing. Hurray! A company was saved… for all the wrong reasons.

The Problem

So a year or so goes by. There I was, at Baxter, working on a new medical instrument as one of their UI/Human Factors and Graphics lead.

My new instrument measured many parameters of the heart (SaO2: Arterial Oxygen Saturation, SvO2 : Venous Oxygen Saturation , CCO : Continuous Cardiac Output), and did most of this through a catheter inserted into the heart itself. These parameters are important for diagnosing what is going on; before, during and after surgery.

I noticed a problem.  The catheter was actually pretty smart, it had a small computer module, as well as a removable and disposable part that you inserted into patients; and then threw away when done.

NOTE: The money in the business was in selling these catheters (disposables). The instruments were given away for free, or loaned indefinitely to Hospitals that used the catheters. 

However, each time you plugged one of these devices in, you need to calibrate it, and enter many different numbers (5 different 10-digit numbers). This was because the light going through the catheter varied in manufacturing, or other issues (variances and tolerance corrections). Whenever you moved the catheter, you had to manually reenter the calibration values for the device, and if you entered any of it wrong, then it could cause mistakes in information, that could lead to misdiagnosis and harm or death for a patient. Which as I’ve expressed, is a bad thing.

To aggravate the problem, you aren’t just plugging in a patient once, and leaving them. There could be many stages:

  1. You set things up in a prep-room (one instrument)
  2.  Then move them to the OR (Operating Room) and do the procedure
  3. Then move them intensive care
  4. and then to a final recovery

Each of these places might have their own instrument, and require moving the catheter. And each of them required re-entering the calibration numbers.

The Solution

I looked at that processes and thought, "there has to be a better way". And being a geek, I thought of a solution. 

I crunched over a weekend, and created a working prototype; and demonstrated it to some higher ups, who immediately wanted me to demo it for an approval board. This was big.

My solution was simple:

  1. the operator just pushed a button "transfer patient", which downloaded all the calibration numbers to the catheter, with a time-stamp and some checksums.
  2. When you plugged the catheter into the new device, it imported in the calibration numbers, did some fancy data validation, and even did a little recalibration test (to verify that things were correct)
  3. It had a safety timeout, so transfers had to be within a few minutes (couldn’t be stale data)

I put in many fail-safes that made it much, much, much safer for the patient than the old way.

Management was impressed; I’d taken the initiative on my own and single-handedly reduced the steps and complexity involved, and created a process that would save lives. They let me stay in the room for the discussion (which was rare). 
It ended something like this;

"Thank you for your initiative, this would increase productivity and save lives – but we can never use it".
The reason? Well, when the nurses were forced to enter that information, if anything went wrong, then it was the Hospital (nurses) fault. With my solution, while it was much, much less likely to ever have a mistake in the process (the saving lives part), and it was much faster and easier to use, the problem was that we’d shifted liability from the Hospital to the Manufacturer. If anything went wrong, they couldn’t say "they did it" (and blame the Nurse for entering the information wrong), which was a free out for any lawsuits they got named in. But by nature of the process, they’d get named in many more lawsuits, and have less of an escape path. 

 So we couldn’t use the far better solution that would save lives and make a better product, because of product liability.


Before you get snippy with Baxter for decisions that would cost lives, remember, they had an obligation to the stockholders and organization not to take that risk and risk the company, for a potential accident that someone else might commit. Recent events had made them extremely sensitive to such things. And every company is forced to make decisions that cost lives, if it means saving their own asses (liability). Think of THAT next time you hear of a big liability award. I know I do. 

In the end, they stayed with the old process. There are probably very few lives ever lost because of that; but most importantly, there is far less liability on the company, they had to make a bad decision, because of Americas bad liability laws (and democrats/progressives especially that fight against all tort/liability reform — because trial and liability lawyers contribute well to them). It is why Obamacare refused to put in any tort reform, even with multiple studies showing you could save up to 10% of all medical costs by doing so. 

Economics 101 is not just thinking about visible gains, but invisible costs. The seen versus unseen: called the Broken Window Fallacy. So let’s do that. Think about: 

  • all the companies that have gone out of business
  • the lives lost because of drugs/instruments they don’t make.
  • the delays in bringing life saving drugs/instruments to market because of liability
  • how many drugs/treatments/devices the FDA keeps things off the market (up to a decade more than necessary) for the sake of bureaucracy
  • all the drugs that would save lives, but have too small a market (and too high of risks) to recoup their costs, so are never validated/completed
  • how many die waiting for things that are well within our technological reach, but well outside the financial/liability risk/reward balance

When you compare the lives lost, or suffering while waiting, versus the very few saved due to blocking quackery (or rewarding family members for accidents), then you have an idea of what the real problems are.

Written: 2003.05.05
Edited: 2016.01.01