Goats and Apples

OK, one last post about our recall rates/internal demand. I’ve been fixated on this topic for many recent posts, and it’s probably time to move on… after I show you this chart:

Looking good!

In July, 6 out of the 8 docs who were working had patient recall rates in the single digits, and the clinic average recall rate was 6.8%. That’s the first time we’ve had a clinic average in the single digits. We need to maintain these gains, and I think we’ll be helped by a change coming to our office this fall.



We’ve used an electronic medical record (EMR) for several years, and finally converted all the physicians to it earlier this year. The biggest benefit of the system has been improving accessibility of our record, both in our office and remotely. However, our current system lacks the ability to use forms and templates to automate and streamline workflow.

For example, I may have removed someone’s kidney tumour and plan to return their long-term follow-up to their primary care provider (PCP). We’ve developed a form that lists the date and type of surgery, the pathology results, and an algorithm for postoperative testing. The PCP and the patient each receive a copy. The problem?  It’s all paper-based, which means fill-in-the-blanks by hand. Each form isn’t individually onerous to complete, but when you add up all the forms I fill out in a day… Anything that makes this task easier makes it more likely that I’ll do it consistently.

Our new EMR supports templates that can be completed by clicking on multiple-choice items. We’ll develop similar forms for common urologic problems that can be managed by PCPs. Templates will also make it simple to include educational material for PCPs.  Who knows what other Clinical Practice Redesign opportunities our new EMR will present?

---

Unfortunately, we won’t be reaping the benefits of reduced recall rates in the near future. Most of our recalls are 6 to 12 months from now, so we won’t see the increased capacity for a while. Until then, we’re still battling the backlog.  Our core team has been contemplating a new approach to this. It has to do with developing a “common currency” in our practice.

Before coins and bills were developed as a way to facilitate trading goods and services, people used the barter system:  I’ll give you these 5 fine watermelons for that mangy goat of yours. But, who’s to say that your goat is as valuable as my watermelons?

Our practice functions like a barter system. In our partnership, we pool our income and then each take an equal portion as a salary. Even though we may generate different revenue individually, we have agreed that our efforts are all equally valuable to the overall conduct of the practice. One of us may spend time providing a service that isn’t reimbursed very well under the fee-for-service system. However, we recognize that the service is important for our patients, and so that partner provides the poorly-compensated service as part of our overall practice profile. Of course, in practices that have a different financial structure, there would be an incentive to cherry-pick services that pay well, and ignore ones that don’t. Not that anyone would do that, of course - I’m just saying, theoretically.

In our current system, then, the common currency is time. As long as each partner is putting in the same number of hours at work, nights on call, etc., we’re satisfied that everyone’s contribution is equal. But, because we’re not measuring output/service provided, some people may be making disproportionate contributions. An extreme example would be 2 workers whose desks are side-by-side. Each is responsible for processing papers from their in-box to their out-box.  If their work is measured based solely on when they punch the time clock, the volume of paper that they each process may be very different.

At this point, it’s easy to jump to the conclusion that one worker is lazy or incompetent.  Instead, let’s presume that they are equally able and motivated, but that the one who processes less paper during his workday is lacking feedback. He doesn’t know how his output compares with his partner.  If he did, he would seek out ways to improve his output. Perhaps he doesn’t have the same equipment or is using inefficient work habits.  Also, the current system implicitly tells him that he adds value simply by being present, and not by processing a certain volume of papers. In order to change his behaviour, the system needs to change.

In a medical practice, perhaps each of the partners needs to know how their “output” compares to that of the other physicians. But, what output measure should we use?  Revenue generation is based on a fee schedule that is somewhat arbitrary, and weighted heavily in favor of more complicated (but not necessarily more valuable for the patient) procedures. That’s not what we want.

Some groups assign values to each procedure or service a physician provides. This works like a barter system. We agree that taking out a kidney stone has the same value as spending 3 hours seeing new patients in the office.  Each procedure/service must therefore be assessed a relative value. That’s too complicated and time-consuming to set up.

If we could agree on our basic unit of exchange, then we would have a common currency. It would have to be a service that all of us provide equally. Our core team has been exploring the idea that the common currency is the number of new consultations seen by each doc. We all see new consultations, even though we may have an area of sub-specialization. Most new consultations are assigned the same amount of time for assessment. The unpredictable part is what follows from the initial consultation. Some new consultations may lead to a patient having further testing, and even surgery. Other patients will be seen only once, and not generate further contact with the urologist. If new referrals are distributed evenly (as per our pooled referral system), these sequelae should be evenly spread among our group.

If it turned out that there was a disparity of new consultations seen (as per the paper-pusher example above), then the docs should be motivated to search for ways to improve their output. We’ve made some preliminary measurements over the last 2 months, and (surprise!) there is quite a difference in number of new consultations seen per physician. We’re not ready to communicate this information to the group yet, because summer holidays make it difficult to interpret the results. We’ll continue to measure into the fall.

When we do present the data, it will be essential to point out that anyone who is seeing fewer new consultations is not “a bad doctor” (learned my lesson in Tight Spot!). Rather, we would emphasize the opportunity to learn from other’s practices, and provide the appropriate tools to help docs optimize the number of new consultations seen.  For example, if a patient is being referred with blood in the urine, one doc may be in the habit of seeing that patient in the office to take the history, and then again at the hospital for a bladder examination. Another doc may have the habit of seeing the patient only once – at the hospital – and combine taking the history with the bladder examination. The second doc has opened up an appointment slot at the office and can see an additional new patient. The first doc isn’t doing anything wrong; he may just be unaware of a different work habit.

Other methods to increase the number of new patients seen include reducing internal demand/recall patients (see above).

If we begin to share this information as a way to change behaviour, we also need to consider what behaviour we’re promoting. If we record (and share) the number of new patients seen, then our docs will be encouraged to see more new patients. That’s great, right? In most cases, it would be great, but sometimes the best care for a patient is something other than a physical visit to our office.

The fee-for-service system pays for a face-to-face encounter with the physician. However, some problems can be resolved with a letter to the referring physician. Such letters can save patients a trip to Saskatoon, while still resolving their health concern. Also, it means that the appointment slot they would have used is now open for another patient. So, if we begin to track new patients seen per physician, we should measure how many new referrals are managed, rather than how many new patients are seen in the office. That is, docs should get credit for time spent reading referral letters and composing responses to referring physicians, even though the patient may never be seen in the office.

I think this system would let us compare apples with apples.