Friday, October 5, 2007

Accounting for Demand

"Insanity: doing the same thing over and over again and expecting different results." Albert Einstein

Most of our efforts so far have been directed toward reducing backlog. We've had some success in other areas, particularly when we asked family physicians to consider referring to "Urology Associates" rather than a specific urologist. This let us distribute referrals to the urologist with the shortest wait list. As our backlog shrinks (wishful thinking?), we need to think about making changes to our office workflow. Otherwise, the backlog may creep back up.

"Managing demand" is the jargon for many system changes that improve workflow. We've been reducing our internal demand (patients who we recall for follow-up appointments) by carefully considering which patients truly need to be followed by a specialist and which ones should be followed by their family physician. Managing external demand (patients newly referred by family physicians) is more difficult.

Occasionally, I may receive a consultation request that can be managed without needing the patient to visit me. In those cases (especially when the patient may have to make a long trip in to see me), I'll write a letter to the family doctor giving advice on further management. Most situations, however, will require an appointment with me. How can I reduce that demand? After all, that's how I make a living!

Maybe we need to change the way we perceive external demand. At present, we consider each consultation request as being one "unit" of demand on the physician's time. What if we looked at demand units in the same way a lawyer would look at "billable hours"? Lawyers charge for each time they handle your file. This may include making phone calls to you or on your behalf, or even just picking up your dossier and thinking about the contents.

Let's think about external demand as including every time I handle a patient file: the actual visit, filling out x-ray or lab requisitions, reviewing results afterward, calling the patient with those results, dictating a letter to the referring physician, and even reading the initial referral letter before the appointment was made. Each patient encounter may generate multiple actions (demand units). Each of these actions is "billable" toward external demand. This opens up opportunities to manage demand.

Here's how the work might flow for a common urologic problem. The family doctor's referral letter initiates the process. I review the letter to decide the urgency of the situation. Let's say the problem is asymptomatic microhematuria, a trace of blood in the urine, found on urinalysis, but without any pain on urination. This is a frequent cause for urologic consultation and has a fairly standard workup. I'll see the patient in my office for discussion and examination. At the end of that visit, I'll arrange some lab tests and an ultrasound of the kidneys. I'll also have my staff arrange a bladder examination (cystoscopy). The ultrasound and cystoscopy will be carried out on separate days. I'll perform the cystoscopy and (if the results are available) discuss the lab tests and kidney ultrasound. If the results aren't back yet, I'll have to phone or write the patient with them.

Here's the tally on this process:

Demand units (minutes)
Initial review of referral letter
Office visit
Complete requisitions
Review chart before cystoscopy
Review of lab and ultrasound results and phone call to patient (assuming no telephone tag!)

Here's the tally I'd like to see:

Demand units (minutes)
Initial review of referral letter, including (already completed) necessary lab and ultrasound results
Review chart before cystoscopy
Meet patient at cystoscopy clinic, Cystoscopy, review of all results

What's the trick? Communication between referring and consulting physicians! If the referral letter clearly states the patient's problem, then I can anticipate which necessary steps can be completed before the visit with me. The referring physician can arrange the lab tests and kidney ultrasound and the results will be available when I first meet the patient. Further, I can meet the patient in the morning and then arrange the cystoscopy later the same day.

Even better, I can cut out the office visit entirely and consolidate everything into a meeting/cystoscopy. The purpose of the initial office visit is to take a history and perform necessary examination. If the patient's past medical history, list of medications, etc. are already in the referral letter, then the office visit is redundant and I can assign that appointment time to another patient (begone, backlog!).

It's tempting to point out inadequacies in referral letters, but the problem is just as often an inadequate flow of information from the consultant to the referring physician. Many conditions have a standard set of investigations that can be made into a checklist or algorithm. If I provide this checklist to the referring physician, all necessary testing can be completed before I see the patient. This looks like a job for the Internet! (Speaking of which, our Urology Associates website and my own practice website are up and running, albeit bare-bones at the moment.)

So, the answer to managing demand is to look at patient encounters from an accounting point of view. (I may be reinventing the "Lean" wheel here, but it was an exciting breakthrough for me nonetheless.)

Then I looked at this from a patient's point of view. And I realized... It's not the demand, it's the muda, stupid!

Patients aren't "demanding" (to use the fist-pounding, foot-stamping connotation) my time; I'm wasting theirs! Look at the microhematuria encounter from the patient's point of view:
  • Multiple trips into Saskatoon.
  • Messages on my answering machine when the doctor called while I was out.
  • Worrying about the ultrasound results.

Why can't they just coordinate everything on one trip and get it over with?

If I look for ways to improve my patient's experience, it will satisfy my need to reduce external demand while keeping my attention on providing exceptional service.

Looking at our work through "Patient-view lenses"... That's a powerful prescription.

P.S. Crazy thought: Let the patient know about the checklist/algorithm for their condition. After all, who's more motivated to make sure that all the necessary information is available to the consultant at the appropriate time?

P.P.S. Check out this chart!

Early signs of a significant change in our system? Stay tuned!


  1. Originally posted by Steven Lewis (Access Consulting) 10/05/07 11:25 AM

    Another great entry. What's fascinating about this journey is the questions that appear once you will have dealt with the muda and the operations research issues. The next two: appropriateness and perverse incentives. We don't talk much about appropriateness in Canadian health care, and even the term "demand" is loaded, implying some sort of natural and by definition legitimate market force. If we talk about need instead of demand - an essential concept in a public system, I would suggest - we don't just accept that everything that presents needs to be there. As has been shown by a lot of research, supply creates its own utilizaton, be it beds or consultations or procedures. Ironically, if you solve the backlogs and make appointments highly accessible, the tendency is to increase throughput and lower the threshold for service, which then creates another apparent mismatch of supply and "demand",.....

    And then there are the perverse incentives. As you point out, implicitly, you make more money from failure than from success. If a GP successfully manages a potential referral, you lose income. (Of course, you won't really, because your slate will still be full.) In many cases, muda = moolah for all types of practitioners. No doubt, in blog #35 or so, you'll put forward your solution to the payment conundrum, drive it through the SMA/Govt. negotiations process, and happily accept the Nobel Prize - not for economics, but peace!

  2. Originally posted by Richard Walker 10/09/07 4:20 PM

    I find this fascinating as I am an IT consultant trying to implement things in health care (in MB) that other industries did years ago. Why? I often ask myself. Part of the problem is 'being special' (I think). While health care is clearly a special 'business', this should not be used to excuse poor practice - and fundamentally, I think it is used this way too often. Also, the God-like status afforded physicians further enforces this. Don't get me wrong, docs often have to play in a God-like manner, but the patient could care less if the hang up is the doctor's or their minions fault. It just causes them delay. Keep it up! This is great stuff.

  3. Originally posted by Michael Farrell (Philanthropy Coach and Counsel) 10/24/2007 9:00 AM

    What a concept...actually telling the patient what lies ahead. Nice work. I heard a physician from the Boston area speak about his practice (G.P.) that is heavily dependant on "on line" communication. However I made the assumption listening to him that it is dependant on the U.S. system where billing is not so tightly other words he could get compensated for on line consultation. The billable hours analogy/model is a good one. You certainly present a powerful case. I will stay in touch.