AskTog: Interaction Design Solutions for the Real World
Interaction Design Section   Living Section   About Bruce Tognazzini
NN/g Home > AskTog > ReaderMail > The Pump, November, 1999 Ask Tog, November, 1999 Special Edition

Ask Tog Reader Mail

 A Doctor Responds
 On Addictions to Pumps and Games
 Consider the People Systems as well as the Computer Systems
 Macho = Mucho Paino

A Doctor Responds


I’m sorry to hear that you had a painful surgical experience. Your description of the function of the pump, however, was somewhat lacking. Your readers might be interested to know the internal workings of the morphine pump.

Additionally, you ignored an important user interface component of the pump -- should it beep whenever you push the button? Should that beep be different when you receive medicine versus when you don’t?

Most Patient-Controlled Analgesia (PCA) pumps have four major settings. The first is a “basal rate” of pain medication infusion. This is a low rate of narcotic infusion (say, 1 mg per hour of morphine) that flows continuously. The second is a “bolus dose,” or the amount of medication that will be infused when the patient presses the button (eg, 0.1 mg morphone). Setting three is a “delay,” the amount of time after a bolus dose is given during which the button is disabled (e. g., 5 minutes). Setting four is a “lockout,” which is the maximum number of times per hour a patient can give a bolus (e. g., 10).

My understanding of the concept is that you should receive enough basal pain medication to control your pain while you are sleeping or resting. When you wake up, move around, or have more pain, the button will give you additional medicine to treat your additional pain. The delay prevents you from stacking up too many doses and causing yourself to stop breathing. The lockout controls how much total narcotic you receive.

A very interesting issue is what the pump should do when you push the button. Most pumps that I have seen make a generic beep whenever the button is pressed, _regardless of whether you received medication or not_. I think this is very important.

As you noted, the pain medication doesn’t kick in very quickly after you push the button. Maybe it would be nice to make a distinctive sound when medicine is being administered, and a different sound when the button is pressed during the delay period or after the lockout.

There is a substantial placebo effect in pain control. Giving a sugar pill to people complaining of pain can cause 30-50% reductions in their rated level of pain. So maybe the pump should always beep the same, whether you receive pain medicine or not.

The primary problem here, however, is the doctors. Doctors are inappropriately concerned about addicting patients to narcotics. There is a large body of research that shows that narcotics given to treat acute pain (sudden in onset, likely to improve in a few days to weeks) almost _never_ cause addiction. When they do cause addiction, it is in patients with a history of prior addiction, especially to drugs, alcohol, or cigarettes. I always ask my patients, “Have you ever been addicted to anything?” If the answer is no, and the pain is acute, I’m very liberal with narcotics....


James Cook, MD

So my description of the function of the pump was somewhat lacking, was it? Harumph! I did leave out the base rate medication, but purely for the sake of simplification. (My base rate medication actually was injected directly into my spinal column via an epidural.) I must confess that the anesthesiologist I consulted didn’t mention the one-hour lock-out.

The stupid pump I had also beeped, but I considered that fairly unimportant since it didn’t indicate anything beyond the button having been pressed. The button provided pretty fair feedback on its own, being a real hold-in-the-hand, long-travel button, rather than one of these dippy little molded dimples so popular on microwave ovens and calculators these days.

A beep that occurs regardless of whether the action has been successfully communicated is less than useless. Such a beep is about equivalent to those silly seatbell warning sirens that fire off regardless of whether you have put on your seatbelt. Stupid, stupid, stupid.

Your point in regard the placebo effect sounds good, but I would challenge how well placebos would work if you announced that 75% of the pills in the bottle were going to have absolutely no effect on me whatsoever. I don’t know about you, but I would take four pills at a time.

In the case of the pump, I, being a glass-half-empty kind of guy, assumed each button press had had no effect, rather than assuming I was receiving medication. I suspect, because of the enforced ignorance, that I actually received less pain relief than I might otherwise have had. I’m curious as to what percentage of the population experiences this same “anti-placebo” effect on these infernal machines.

The one place we seriously part ways, Doctor, is on the subject of treating drug addicts. My wife, The Doctor, is a board-certified addictionologist, in addition to being an internist, with a subspecialty in prescription drug addiction and signficant experience in running drug and alcohol treatment centers.

She has informed me that (1) most “enlightened” doctors are liberal with narcotics for those who have not been previously addicted and stingy with those who have been hooked, and (2) these doctors don’t know what the hell they are doing.

(Of course, she said no such thing. If you think police officers stick together, you should check out doctors sometime. However, that’s exactly what she said, only translated from doctor-talk into English.)

Here’s the problem: Drug addicts, particularly those addicted to narcotics, have a vanishingly low threshold of pain. That’s why withdrawal from heroin, which is apparently equivalent to a bad flu, is so painful to the addict. It is also why the addict typically needs far more pain medication than average, rather than less.

What about the danger of getting people re-hooked? If the medication level, no matter how high, is kept below the threshold of euphoria, the chances for re-hooking the patient are low. And if the patient does become hooked, there are programs and facilities that can handle that problem.

The worst part about asking the patient if he or she has previously been hooked is that patients not in recovery, who are likely to con you out of all the drug they can, are going to lie, while patients in recovery, who will use the pump responsibly, are going to tell the truth. Therefore, those addicts who are most likely to get in trouble with high doses will receive high doses, while the rest are left in significant pain.

In my own case, I wrote down on my drug history that I had once had a single methadone tablet. This was prescribed by my oncologist while I was recovering from cancer some 20 years ago. I took one tablet and concluded, based on the side effects, that I’d rather die than take another. Nonetheless, one of the seven anesthesiologists that took turns deciding, a couple months ago, how much narcotic the pump would supply me “today,” took this fleeting mention of methodone to indicate that I was a dangerous heroin addict who had been on methadone maintainance. So he did the only thing reasonable. He reduced my pain medication to almost nothing.

My wife had a little chat with the boy. His head is now floating in a glass jar in her office.

On Addiction to Pumps and Games

Dear Tog,

15 years ago I was in the hospital for emergency surgery after discovering precisely how wide a bridge railing was not. There was no pump on which to push a button; I had to wait until the nurse managed to get around to me to give me an old-fashioned shot. Fortunately, I was in an Intensive Care Unit and the nurses were very good.

I was amazed at how quickly I became dependent on the morphine. When they started taking it away, I was quick to over-emphasize the pain in the hopes that they would give me just a bit more. I wasn’t plotting—it wasn’t really a conscious decision. I just WANTED it. Something deep within my being WANTED the feeling that morphine gave me.

Suddenly, without any kind of conscious decision or choice, I was addicted.

There’s no crime in manipulating the doctors or nurses--I’m in pain here! I’m the victim! I’m in pain! I DESERVE to feel good because I’m in SO MUCH PAIN! These people have an OBLIGATION to make me FEEL GOOD.

That’s the scary part about alot of these addictive medications. They don’t just make the pain go away, they make you feel good. And before you know it—because you don’t know it—you’ll do anything to get that feeling back. And alot of the pain isn’t even real—it’s just that you felt so good before and you want to feel that good again.

It is an interesting interface problem. Consider the example of the drug machine that tells you how many minutes it will be before your next dose. Wonderful! This way you don’t spend time pressing the button “for nothing.” But what do you think the patient’s reaction will be when the machine is set to 20 minutes instead of 15 minutes. I know I’d’ve been buzzing that nurse and doctor and letting them know that I am still in incredible mind-numbing pain and how DARE they set that machine to 20 minutes! Again, these people have an OBLIGATION to make me FEEL GOOD.

Personally, I think that this is one of those RARE cases where providing this information is not beneficial to the user, even though the user might believe otherwise.

On a related note, though, how do you design an interface for delivering addictive content? Consider the so-called addictive games. Obviously, these diversions provide entertainment and make us feel good, like a drug. But there is a risk of becoming addicted. So do we allow only 2 hours of game-play during a 24 hour period? Do we start with unlimited game play but if the user plays more than 12 hours in a 24 hour period, start cutting down their play-time until it is cut off completely for a week and then they can start again with unlimited play time? Do we just stick in a Preference and let the user decide what their dosage will be? Or do we insist that our product is not addictive?

It’s a neat thought, especially if you compare/contrast it next to things like cigarettes or marijuana.

-Peter Merchant

Actually, Peter, the scenario you outline with the pump is far less likely to occur than you might expect.

When shots are administered intramuscularly (i. e., in the butt) every four hours, patients don’t received four hours of equally effective relief. Instead, the effects of the drug die away in a smooth curve. As a result, patients typically are juiced up into euphoria for a while, then drop through discomfort into pain. This sort of cycle is ideal for generating addiction. (The faster the cycle, typically, the more powerful the addiction. Consider smoking: Take a puff and it is at your brain in three seconds. Then, the drug levels start dropping precipitously. Time for another hit. And so forth.)

The pump can flatten this curve almost out of existence. With a proper base rate, the drug flowing into the patient will be enough to neutralize most of the pain without causing either the euphoria or the toppy-turvy cycle that shots induce. The potential for post-surgical addition is drastically reduced.

The best way to cause post-surgical addiction is to undermedicate in the beginning. That way, the patient will be so starved for endorphines that, once the pain begins to drop away, they will lie, cheat, and steal to get up on and stay up on the plus side of the equation—the euphoric side that spells trouble.

As for computer games, you seem to have laid out pretty well the secondary game played by millions and millions of parents with their children, trying desperately to control how many hours they are hitting the PlayStation.

When our son went through his Nintendo phase, I found it tempting to “initiate the kind of controls that would lead to responsible game-playing.” Instead, we held to an agreement we’d had with the kids since they entered school: Maintain an A-average and you can pretty much do what you want. Our son played furiously, but he knew not to let his grades drop. Then, one day, he was through. We avoided all that screaming and yelling, the child did not end up behind some dumpster pressing buttons on a Game Boy with half-dead batteries and, I imagine, he spent exactly the same total hours playing as he would have otherwise.

As for getting adults to give up similar addictive behavior, that becomes even more problematic. Perhaps we can incarcerate adult Nintendo-nuts and web addicts, as we are so fond of doing with those choosing unapproved herbal supplements.

Consider the People Systems as well as the Computer Systems


...Not to throw rocks, and I do recognize the continuity, but your article does not criticise the design of The Pump per se. Rather, you rightly raise the issue of patient care as related to pain management. I haven’t given the matter any thought, as I was completely at ease with my pump, but it would be curious to explore ways to improve the design of The Pump. A design goal could be to improve feedback, so the patient’s sense of control is increased.

Like the rest of us, doctors have difficulty keeping up with all the advances in their profession. I’ve had the pleasure of being treated by brilliant, compassionate doctors. I’ve also had others that nearly killed me. The lesson learned is to guard your self-interest by educating yourself about your health. Here’s two helpful articles about pain management (aka “pain and toxicity”).

Controlling the Pain of Cancer, by Kathleen Foley
Scientific American, Sept 1996
/not available online/

The Tragedy of Needless Pain. by Ronald Melzack
Scientific American, Feb 1990
/not available online/

I encourage anyone about to undergo surgery to give copies of these articles to their surgeon, oncologist, etc. In my experience, most doctors are responsive to patients who speak with authority about their own care. In the event that the health care provider doesn’t respond favorably, I now get a new doctor.

Cheers, Jason Osgood

Good advice, Jason. You really can tell rather quickly whether you are having an adult-to-adult conversation with a doctor. And the article does focus more on the human interaction around the pump rather than the interaction with the pump.

This kind of interaction is often even more important for the interaction designer to consider than is the interface itself, because a whole team will focus on the interface. You may be the only one focusing on all the people systems that surround it.

In an earlier part of your letter, you had said:

Like any tool, The Pump can be misused, even counter productively. The intent of tallying bolus requests is to evaluate the effectiveness of the pain management strategy. Lots of clicks, the rational response is increase the max dossage. Fewer clicks, reduce the max dossage.

The rational response, due to the extreme fear of addiction and doctor’s ignorance in how to treat it should it arise, is to do the opposite of what you have outlined. The designers of the pump felt as you did; the average physician is continuing to act oppositely.

Macho = Mucho Paino

“(God help you if a male surgeon who has never had surgery himself is in
charge of pain managment. You're lucky to get a couple of extra-strength Tylenol.)”

When my previous dentist retired, I went to the dentist who’d taken over his practice for care of a painful abcessed tooth. As he was probing around in my mouth, he casually mentioned that he “didn’t really feel pain much.”

Needless to say, this turned out not to be one of the best dental care experiences one could hope for...

Carl Maniscalco
San Diego, CA

Don't miss the next action-packed column!
Receive a brief notice when new columns are posted by sending a blank email to

return to top

Contact Us:  Bruce Tognazzini
Copyright Bruce Tognazzini.  All Rights Reserved