Physician shared decisionmaking

Physicians and patients make decisions together .. right? 

A sample of today's shared decisions:

Mammogram at 36? (no)
Antibiotics for acute otitis media? (no)
Aricept for Alzheimer Disease? (yes)
Biopsy to confirm diagnosis of condyloma? (yes)
Incision & drainage of sebaceous cyst vs. trial of antibiotics? (yes)
PSA at 54?  (no)
Colonoscopy at 56 (yes)
Counseling and/or family meeting for gambling addiction (no)
Antidepressant medication for depression/anxiety (yes – maybe)
Choesterol testing at 30 (yes)
Medication for smoking cessation (yes)
Physical therapy for back pain (yes)
Annual Pap smear at 35 (yes)

Some decisions, of course, are not final.  The prescription for antidepressants is written, but not necessarily filled.  Is it bad if the patient doesn't fill this prescription?   No.  I make it clear that it's not.  Indeed, despite the paternalisitic nomenclature that we use .. patients are not necessarily noncompliant .. they just decide not to take the medications that we prescribe.  Today I made it clear to my patient that the prescription was my method of putting the power into his hands.    If he decides to try the medication after reading about it and thinking and perhaps talking with his wife … that's OK.  If he decides NOT to take the medication, that's OK too. I won't be mad. If he has more questions about it … he should feel free to call me. 

I met a couple this afternoon who were looking for a new physician for their 3 year old daughter, and we started talking about shared decisionmaking.  I heard myself saying that I see myself as a resource for them.  I won't dictate what to do .. I'll just help guide them.  Yes . I may know a few things about medicine .. and yes .. I'll have opinions about what may be best for their daughter .. but they may know some things too  .. and if we are all open to each others' opinions … we can make good prevention and treatment decisions together.

So in this context, every decision is shared.  There are some issues that I'll feel strongly about.  Do I sometimes lobby for a given approach?  Certainly I do.  Do I mandate a given approach?  No.  Will I refuse to do something that I don't think is appropriate?  Yes. (I often refuse to treat colds with antibiotics … but I do so in the context of education .. ) 

Is an intervention really going to be effective if I impose it?  Not likely.  So we shouldn't bother imposing interventions .. right?

How many physicians does it take to change a lightbulb? (answer)  Get it?  If patients don't make the decisions with us .. then any decision we make for them is really irrelevant.

yes yes .. there are exceptions.  Self-destructive behavior may be one of them .. but even that's a grey issue.  Where do we draw the line?  Certainly a suicidal patient needs a decision made for them (but only after we fail to make a decision with them).   Is eating a Big Mac self destructive?  How about unsafe sex?  Smoking?

So I'm searching google for links on shared decisionmaking and I find what looks like a question from a final exam in a medical ethics class:

In a survey performed by the Presidential Commission for Study of Ethical Problems in Medicine, doctors were asked to consider 3 issues: (1) the issue of whether a pregnant woman over 35 should have amniocentesis; (2) the issue of which antibiotic to use for strep throat; and (3) the issue of whether to continue aggressive treatment for a cancer patient in whom such treatment had already failed. The doctors regarded (1) as a patient decision, (2) as a physician decision, and (3) as joint. Do you agree? What are the relevant differences among these issues?

What's the answer?

Ionized bracelets – no better than placebo – what you didn’t see on TV

A picture named Q-Ray_Combo.jpg

Today was the 1st time I've had a patient tell me that they actually bought a q-ray bracelet.  Yes .. as-seen-on-TV.  "These things are REMARKABLE" according to the very happy people on the infomercials.  My patient wasn't so pleased.  His shoulder still hurt .. and he's out the $90 for the bowling league since I told him to give it a rest for 2 weeks.

In this month's Mayo Clinic Proceedings is a paper that will likely (let's hope!) spread the work about these things.  In a randomized placebo controlled trial, the ionized bracelet was just as ineffective as the placebo bracelet.  

Next on the agenda… magnetized back brace.  Not much in the literature (that I can find) on this yet.  There was a pilot study a few years ago that showed no effect.  I've heard anecdotes from several patients (including one today) that this is effective.   We'll have to see.

Handsets and PDA Owners Done with Monochrome

PDA Cortex reports that handheld uders are moving away from monochrome devices.  I'll finally admit that I would agree that (finally) color is a reasonable option for a handheld purchaser to consider.  As is speed.  I now have a Palm Tungsten T  .. and the speed is significant, as are the visual cues that color makes possible.  Now if only HandheldMed would make that OS5 version of their reader software available … I could use my textbooks again.

How Perils Can Await the

How Perils Can Await the 'Worried Wealthy'. Hordes of consumers are flocking to centers around the country to have their bodies scanned by three-dimensional computerized X-rays. By Jane E. Brody. [New York Times: Health]

This is a compelling article that I'll print out for the office.  Too much testing CAN be a bad thing.  The article clarifies how false positives on screening tests can lead to serious problems .. and great expense and hardship. 

Today I saw a 53 year old man for a "annual physicial."  Unfortunately, the common misperception is that a physicial examination is the conerstone of this annual visit.  It's not.  As a primary care physician, my job is to assess risk, and prevent disease.  I always ask patients if they wear seatbelts, since automobile accidents are more likely to contribite to their death or illness than anything I hear when I listen to their lungs.   Testing always comes up too.  "Do the all of the bloodwork" I often hear …  "I want to be sure everything is OK"  Hmm … How do I respond?

Of course, I can't do "all of the bloodwork."  Doing so would be both impossible and irresponsible.  I need to help my patients make careful, informed decisions about which screening tests are appropriate and indicated.  I often make use of the USPSTF screening guidelines to guide the disucssion.  To some, "It's just a blood test" and the risk of a false positive or ambiguous result isn't appreciated.  I don't want to talk my patients OUT of getting a PSA.  ? or do I?


It's interesting to look at the access and referrer logs for the docnotes website.  Rererrer logs tell me who has linked to docnotes .. access logs tell me who has visited .. and often .. from where.

Where do people come from?
Most often … it's either "unknown" meaning they typed in the link or used a bookmark on in the browser … or from a search engine.
Other referrers are the medical weblogs .. or
Where are people logged in from?
It's usually possible to determine the domain from which a user is coming from.  So if a user is on AOL .. I can tell .. or AT&T .. or MSN.   More interesting is when the domain is from a health related webiste.  Today's interesing login is from Medfusion.  It's a company that offers several services to physician practices. They do their programming in ColdFusion, which is good .. and they seem to offer services to practices including online lab result reporting, online scheduling (for patient appointments) and online bill payment.  It's not clear, though, how they interact with the myriad practice management systems.  There is no mention of interfaces on the website.  This is important to know.  INdeed, the reason that we're moving out of relayhelath is because we don't want to have several separate data repositories.  We want out patient information to live in ONE place.  So we're building our own simple (yeh .. coldfusion, of course) replacement for RelayHealth.  It's going to run as a webservice on our external server.  The internal server will poll the exteral server every few minutes.  If there are new messages, The external server responds with the (encrypted) data, and the internal server will post the messages to our internal messaging system.

So it's not clear to me what the role of Medfusion will be.  No mention on their website of pricing.  Id be surprised if it's cheap.  Build vs buy?  I'll continue to build.



Information Technology Projects Evaluation Process

Evaluating technology projects is hard .. and many organizations don't (yet) have an infrastructure for doing so.

Using Reider's Rule (see below) is my method.  But it's more complicated that that.  Using technology in a business, an organization, a school .. or even a medical practice … It's important to make certain that there is a clear process for how such projects are identified, prioritized, and implemented.  In the absence of such a process, one can't be sure that the right projects are getting implemented.

  1. Idea Solicitation. It's important to develop a mechanism to ask for ideas.  If we implement projects from the top of the Ivory Tower .. we can't be sure we're really meeting the needs of the users.  Develop a method for proposals to be submitted.  A web form, a paper form, or even just a contact person.  The more detailed informatin that is gathered about the proposal — the better.
  2. Align Projects with Strategic Goals and Objectives.  If you don't know what the strategic goals are .. then you'll have to go back and do some strategic planning.   The Value of a project can me measured on many variables .. but alignment with strategic goals is probably the most important.   Other methods of assessing value will depend on the proposal.  Does the proposed project make it easier for workers? Will it enhance revenue?  Will it decrease cost? .. enhance communication? .. etc.
  3. Develop a method of determining the work that would be required to implement the project.  Who is going to do it?  How much is it going to cost?  How long will ti take?  Who will maintain it once it's "up."  How will it be kept up-to-date?

I'm off to the Airport .. more tomorrow …



Couple Remain Hospitalized With Bubonic

Couple Remain Hospitalized With Bubonic Plague. A man who is battling bubonic plague in a Manhattan hospital is a former top official of the New Mexico agency responsible for investigating criminal cases of insurance fraud. By Cecilia M. Vega and Tina Kelley. [New York Times: Health]

In addition the news reports … I've received three e-mails about this from the Health Department.  It's a great use of technology to communicate important information in a timely manner.

Reider’s Rule

P = V_W




We’ve been doing lots of work on prioritization recently. 

Choosing which projects to initiate, which products to choose, which initiatives to let fall.

Borrowing from the wonderful usefulness equation, we’ve been using a method of prioritization that seems to capture an element often missed.  Often, when people talk about prioritization, they are – without knowing it – assessing the subjective value of an initiative.  What’s often missing is how much effort is needed to do what is intended.  Indeed, the team members who know (or could know if asked) how hard something would be are not even involved in the conversation.

Let’s think through some examples … 

Should we build a new office, buy a new computer, or design a new website for the residency?  (or should we do all three?) 

Let’s prioritize.  If PRIORITY = VALUE / WORK .. 

Let’s first score value.  On a scale of 0-100 .. what’s the value of a new office building?  Well .. now it gets hard (all-of-a-sudden).  Value to whom?  How do we understand value?  Is value = revenue?  (I hope not) .. is value = better health for the people we serve?  (could be) .. is value = alignment with strategic goals?  (YES) .. so this is all of the above and much more.  So we’re going to need to canvas our teams, make sure that we score the “V” really well, yet understand that we’re probable guessing.  That’s ok.  Guess.  We can always change it.  I’m going to say that the value of a new office building is 70.  Why?  We have an office.  It’s suboptimal, and there are mice here (really!) but it works and most people are happy with it.  The rent is acceptable too.  A new one would, however, be fantastic, and the impact would be broad (we could really create an experience that is head-and-shoulders above what people have today – really change the way folks think about going to the doctor .. ) So .. shorthand:

Office = 70
Computer = 30 (serves small # of people, improves efficiency)
Website = 45 (attract new residents, explain the program without printing & mailing, show innovation).

Yes – it’s all pretty subjective / qualitative but that’s ok. 

Now: work. This one is easy but if we don’t incorporate it, we’ll just pick the highest value.  
Office: $100,000
Computer: $3000
Website: $2000
P=V/W  (we can multiply by 100 to get rid of leading zeros)
Office: 0.07
Computer: 1.0
Website: 2.25

So – the highest priority initiative is the website!

A few final thoughts:

a) Sometimes we use the square root of “work” to normalize work estimates
b) “Work” can be expressed as dollars or time or both.  To pull them together – we sometimes calculate cost of time (~ $100/hr) and then add that to the $$ so we are using the same denominator.

This really works – even if just to get a gestalt of what things look like.  If you go through this exercise and then look and say “that’s nuts!” then cool .. why do you think it’s nuts?  What needs to change?