Tuesday, April 9, 2013

On Doctors And Data Entry

Managers are typically keen to know how hard the bees are working in the beehive.

When the bee is, say, a factory worker it's fairly easy to measure work output and quality: count the number of glass objects placed on a conveyor belt in an eight hour period and note the amount of breakage, if any.  Workers can be easily compared using this metric.

When the bee is a family practice physician or internist measuring work output and quality of work becomes problematic because the work elements that count defy straightforward analysis.

Undaunted by the complexity of this task managers sometimes plow ahead with a simple measurement, believing, it seems, that if something can be measured, it should be measured, regardless of how limited the return.

For example, I recently learned about a very good clinic that came up with the very bad idea of reviewing physician performance based on their skill and versatility at data entry.  

In this case, a computer review of physician records evaluated how often doctors documented that they, among other things, recommended colonoscopies and mammograms to their patients.  The information had to be entered in a specific place in the electronic record or it was assumed that the doctor did not make the recommendation, even if the information was entered elsewhere.

Getting to the designated documentation spot involved navigating the electronic record with a variable number of clicks depending on where the doctor was in the patient record.   The time necessary to click and enter the information likely ranged from a few to thirty seconds.

When the study period was over the computer analyzed all the data and the physicians were assigned grades (a report card).  Those who scored low were encouraged to shape up.

Using the computer, with its pixilated wisdom, to answer the above questions is actually a good idea.  The documentation of screening tests is important.  And placing the information in a standard place makes lots of sense.

But considering that the average fifteen minute appointment is chock a block with patient concerns, doctors have precious little time for data entry.

It's of great interest that two of the very best clinicians in this study got the lowest grades.


It appears that given the choice of talking with their patients or entering data, they chose the former. What really counts - empathy, availability, ethics, intelligence, thoughtfulness - cannot be measured in the usual sense and certainly not by a doctor's talent for data entry.

In the end, the measure of a doctor is not subject to formulas but by reputation, the child of character.