Recently, one of the medical directors of one of the adult medicine practices at our sister institution sent me an email asking if I knew how to get certain data on different facets of our practice.
She was wondering if we regularly collected data on our no-show rate, as well as how and when appointments were cancelled or rescheduled, for individual resident providers, as well as some other pretty basic reporting. I was embarrassed to have to say that, unfortunately, the information she was interested in wasn’t readily at my fingertips, and that this type of data quite often proved to be a challenge to retrieve from our electronic medical record.
A few years ago, when we revamped how we scheduled our patients, in an effort to improve continuity of patients with their provider, and providers with their own patients, we asked the developers to create a series of reports for us on what we call the “continuity indexes”. These numbers show us what percentage of the patients a provider sees are their own assigned patients, and what percentage of a patient’s visits are with their own primary care provider (PCP). This provided a lot of really useful information, and helped us define and refine the rules we use for scheduling that we think really improved care overall.
Her practice has been struggling with patient empanelment, figuring out how to affix a particular provider as the PCP for a patient. When they converted from a different electronic system over to the system we all use now, all of their assignments of themselves as PCPs were apparently lost in the switchover.
As our practices have grown, there is more and more interest in improving the experience for patients as well as providers, and seeing your PCP as much as possible for your care is a key driver of these measures. As someone responsible for running a practice and for making sure that things move smoothly, she wanted to know several things about the nature of her practice, in the hopes of using that data to build a better system and to make it easier for providers to practice there.
Of course, our electronic medical record has extensive reporting capabilities, but it seems like a lot of the stuff wasn’t built by the people working on the front lines. True, there are some things that are pretty easy to retrieve, such as how many of my own patients with diabetes have a hemoglobin A1c over 9%. And there are countless administrative reports and billing reports that are churning away in the background. But those of us who are seeing patients find it incredibly challenging to try and figure out something we want to know, something specific about a patient or a provider, about a group of patients, or about some other characteristic of our practice.
If I, as a medical director of our practice, want to know how many left-handed white males over the age of 60 (that’s me, by the way) have had more than three visits to our practice in the last 6 months with a chief complaint of abdominal pain, have a cholesterol between 190 and 195, have had a colonoscopy less than 3 years ago, and live in Zip code 10023, I should be able to get that data right when I want it. But it turns out that it’s more complicated than that, and every one of these data demands requires submission of a report request, the assignment of an analyst, the development of a reporting tool, and more money being spent.
Last week, as part of an effort to improve vaccination rates for our patients, we wanted to create a letter from all the primary care providers in our practice to all of our patients who had not yet received the COVID-19 vaccine, sending them a personal endorsement from their PCP saying what a good idea we think getting vaccinated would be. The system even has a way that I can send a letter as an e-mail attachment for those that are enrolled in the portal.
I asked the developers if there was a way to send a different letter to patients based on their selected language preference. This item is a part of the core data on the front page of a patient’s chart, right there with their medical record number, date of birth, and telephone numbers. It should be easy to assign which version of the letter they get based on this data point in their health record, so that the Spanish version of the letter would go to someone who had selected that as their preferred language.
But it’s apparently too much to ask to have language preference that’s already noted in the patient’s chart help decide which version of a letter they get. We would have to do that manually, patient by patient, provider by provider.
We want running dashboards showing us just the information we want, when we want it, displayed how we want it. We want support from systems smart enough to spot trends, prevent errors, and help us think of new ways to do things better. Maybe someday the system will be running like HAL in “2001: A Space Odyssey,” suggesting that maybe we don’t need to check daily labs on this patient, because we took no new actions based on yesterday’s labs and there is nothing to suggest that today’s will be any different.
And we’d like to generate reports on who hasn’t been seen, who needs to be seen, or who is missing essential parts of their care. If a medical director wants a report, if they’ve decided that some statistic on the characteristics of their practice is important to them, don’t you think it’s important that they get that information as quickly, easily, and effortlessly as they possibly can?
Shouldn’t be too much to ask.
Fred N. Pelzman, MD, of Weill Cornell Internal Medicine Associates and weekly blogger for MedPage Today, follows what’s going on in the world of primary care medicine from the perspective of his own practice.
Last Updated March 08, 2021
Source: MedicalNewsToday.com
