Well, that was a whirlwind.
The first year of residency is known to be a gauntlet. I thought I had dodged the bullet by picking a pathology residency instead of internal medicine. Nope! Sure, I bet they have it worse, but I have still experienced the absurd working hours and learning cliff of a starting resident. I have still been humiliated by my ignorance and deathly afraid of hurting patients with my incompetence. I've worked hard, burned out, burned back on, and struggled to find equilibrium.
And maybe, now, I've sort of found it? After a year, I don't think my inexperience has really harmed a patient. I've had layers of protection, from the Pathology Assistants (PAs) at the grossing bench, to the attending physicians at sign out. I've kept my wits about me and recognized my errors, and corrected them. I've had my senior residents, who are the perhaps the most understanding what what we are going through, close at hand with encouragement and wisdom. Perhaps most importantly, my lovely wife has nourished me, both with good cooking and emotional support. No one gets through this without help.
To be sure, the hours were unreasonably brutal. They remain unreasonably brutal, and will continue to be unreasonably brutal. Part of this is an artifact of my training path: college, 2 years in a lab, then an 8 year MD/PhD program. I am older than most used to be at this phase of training, but no more experienced in the specific area of practice. I have a lot of OTHER experiences that inform and educate my ability to do my job. I know some advanced math, computer programming, genomics and lab skills that could pay off someday. Right now it's all the time honored aspects of anatomy and histology, and I sure didn't come in knowing a whole lot of either of those.
(As a parenthetical: I continue to maintain that the way physicians are trained in this country is absurd. My friend from Ireland who had the benefit of entering medicine out of high-school, as is standard in much of Europe, and is arguably one of the most capable residents in the program. 4 year college before medical school is an extremely expensive luxury that might be totally pointless.)
I also have a wonderful family that demands time and attention. It's painful every morning to leave my son, and even more painful on days when I come home to find him already asleep. There's a different kind of pain reserved for when he wakes up at 4:30 AM on days that I have worked past midnight. I sometimes envy the younger residents who can devote themselves more fully to learning the craft.
But, as the new first years arrive, I realize how much has been learned while running this gauntlet. Sure, I could have learned more, but in one year I have gone from understanding extremely little about the discipline to knowing at least an outline of every area of pathology. Sure, I wouldn't in a million years trust myself with signing out cases at this point, but I now see the map of the territory I need to explore, and come to know. Its boundaries are mostly drawn, and some of the major features in the terrain are already sketched in.
In short, I am hopeful. Not that I will be working any less hard, but that my work is bearing, and will continue to bear fruit. I am learning more and more HOW to learn, and that ratchet will carry me up the cliff in safety. I look forward to the view from the top.
Saturday, July 9, 2016
An Ode to the Electronic Medical Record
Dearest electronic medical record,
Thank you for permitting me to read all the notes for my patients, going back for years, if not decades, in a typed, readable font. Thank you for storing all the lab values, also going back for decades, with a clearly available range of normal values for the laboratory that generated those values. Imaging results, from PET scans to chest X-rays, are similarly available at the click of a button. And thanks, also, for helping with billing. I don't really know what you're doing with all that, but I understand the the hospital administrators really like particular feature. It's really hard to underestimate how important these three things have been in improving care in hospital settings.
I just want to encourage you, electronic medical record, going forward, to learn from your mistakes and maybe offer some constructive criticisms.
You're starting to get a little big around the midsection. I call this "Note bloat". Doctors today are very busy, serving more and more complex patients, and often at a fairly high level of specialization. That means its more important than ever that they be able to communicate succinctly. But when I try to use you, sometimes I have to scroll through pages of auto-generated ... I hate to say it ... garbage, before I get to a decently written paragraph written by a human who has seen the patient recently. That paragraph can be located just about anywhere in the text, and is sometimes, bizarrely, not anywhere to be found.
You're also slowing down a bit. Maybe because of the bloat, you have some trouble with load times. Google is great in no small part because it is so damn fast. You go to a simple page that loads quickly, type a search, and get relevant results from the entire internet quickly. Medical records are big, sure, but why on earth does it take orders of magnitude more time to search the medical record than it takes to search the entire internet? Time is money, but it's also user-sanity. I hear that there are methods like indexing which can improve this kind of thing.
On a similar note, you're loosing agility. Record systems like you used to be bespoke amalgamations that individual hospitals and practices assembled from parts scavenged from medium size software companies and their own in-house ingenuity. Today, it's pretty much EPIC. While EPIC is often customized at time of purchase, the resulting software is hard to change.
Here's an example of the above points. I often work in a frozen section lab. We offer rapid tissue diagnoses, mostly telling surgeons whether a specimen they send is cancer vs. not cancer, in 20 minutes or less, while the patient is on the operating table. It's critical that we know which patients are coming in advance, and be able to look up their history. Looking up the histories the night before, I read notes on dozens of patients, often in the realm of about 30-40 or so. I get that list by culling a list of hundreds that are planned for the OR that day. I need to efficiently load each patient's record, find the notes that are relevant to the operation being performed, identify key elements of the imaging history and past medical history, and possibly identify prior operations that may alter the anatomy of the specimen we are going to process. This process can take hours, but of course it is only possible because we have an electronic medical record in the first place.
In the first step, culling a list of hundreds of operations down to those that actually are likely to require a frozen section diagnosis, currently makes use of custom computer programs written in the pathology department. Ideally, these programs could be integrated into a specialized "pathology view" of the OR schedule.
The second step, loading the patient record, requires 5 clicks, each of which can have a slow load time. There is also an issue where the program brings up multiple patient records from what should be a patient unique medical record number. This introduces a possibility of error (it has not, yet, but it shouldn't even be possible). Ideally the entire pre-filtered list of patients in the OR that day could be preloaded with one click, allowing access to each patient quickly and unambiguously.
The third step, reading through the records, requires me to identify important notes for the particular surgery of the day from a half dozen to hundreds of other recent notes. Many doctor's notes aren't that important for my specific purpose, e.g. the patient's physical therapy notes. Others are extremely important, such as the note from the surgeon seeing the patient before the operation. There is no way to generate customized algorithms to search, filter, and sort these notes based on what the operation is, who is doing it, and what sorts of notes we (as pathologists) generally find most useful. Sure, there is a search function, and a filter function, but these are clunky, slow, and very limited in their customization options. Better than nothing, but less than ideal.
The fourth step, actually reading those notes, frequently runs into the problem of note bloat. Thank goodness I don't have to read doctors' individual handwriting, but I do have to scroll through a lot of auto-generated content. That content is often there, sad to say, for the purposes of billing. Sometimes you get to the end of a note and realize that the entire thing is a billing document. This takes minutes, per patient, which adds up to hours over the course of a week. I can never get that time back, and I'm not really learning anything but what font sizes and formats the auto-generated stuff normally uses so I can scroll quicker.
We can do better! It's not even that complicated! Notes are really just formatted text stored in databases. A physician with a little programming know-how could write very simple programs that would save hours per week for doctors in a their specialty. But there is no avenue to perform such customization. Moreover, as the marketplace becomes increasingly dominated by very few corporate providers, if not a single dominant provider, will there be any motivation to improve the product?
Electronic medical record, you have so much potential, yet untapped! I dearly hope that you see it. If not, perhaps you could attend some motivational seminars. Get some help from some computer literate medical colleagues! With a little ingenuity, I think you can shed the weigh, speed up, and improve your agility. It'd be great for your health, ours, and our patients,
Sincerely,
A Faithful User
Thank you for permitting me to read all the notes for my patients, going back for years, if not decades, in a typed, readable font. Thank you for storing all the lab values, also going back for decades, with a clearly available range of normal values for the laboratory that generated those values. Imaging results, from PET scans to chest X-rays, are similarly available at the click of a button. And thanks, also, for helping with billing. I don't really know what you're doing with all that, but I understand the the hospital administrators really like particular feature. It's really hard to underestimate how important these three things have been in improving care in hospital settings.
I just want to encourage you, electronic medical record, going forward, to learn from your mistakes and maybe offer some constructive criticisms.
You're starting to get a little big around the midsection. I call this "Note bloat". Doctors today are very busy, serving more and more complex patients, and often at a fairly high level of specialization. That means its more important than ever that they be able to communicate succinctly. But when I try to use you, sometimes I have to scroll through pages of auto-generated ... I hate to say it ... garbage, before I get to a decently written paragraph written by a human who has seen the patient recently. That paragraph can be located just about anywhere in the text, and is sometimes, bizarrely, not anywhere to be found.
You're also slowing down a bit. Maybe because of the bloat, you have some trouble with load times. Google is great in no small part because it is so damn fast. You go to a simple page that loads quickly, type a search, and get relevant results from the entire internet quickly. Medical records are big, sure, but why on earth does it take orders of magnitude more time to search the medical record than it takes to search the entire internet? Time is money, but it's also user-sanity. I hear that there are methods like indexing which can improve this kind of thing.
On a similar note, you're loosing agility. Record systems like you used to be bespoke amalgamations that individual hospitals and practices assembled from parts scavenged from medium size software companies and their own in-house ingenuity. Today, it's pretty much EPIC. While EPIC is often customized at time of purchase, the resulting software is hard to change.
Here's an example of the above points. I often work in a frozen section lab. We offer rapid tissue diagnoses, mostly telling surgeons whether a specimen they send is cancer vs. not cancer, in 20 minutes or less, while the patient is on the operating table. It's critical that we know which patients are coming in advance, and be able to look up their history. Looking up the histories the night before, I read notes on dozens of patients, often in the realm of about 30-40 or so. I get that list by culling a list of hundreds that are planned for the OR that day. I need to efficiently load each patient's record, find the notes that are relevant to the operation being performed, identify key elements of the imaging history and past medical history, and possibly identify prior operations that may alter the anatomy of the specimen we are going to process. This process can take hours, but of course it is only possible because we have an electronic medical record in the first place.
In the first step, culling a list of hundreds of operations down to those that actually are likely to require a frozen section diagnosis, currently makes use of custom computer programs written in the pathology department. Ideally, these programs could be integrated into a specialized "pathology view" of the OR schedule.
The second step, loading the patient record, requires 5 clicks, each of which can have a slow load time. There is also an issue where the program brings up multiple patient records from what should be a patient unique medical record number. This introduces a possibility of error (it has not, yet, but it shouldn't even be possible). Ideally the entire pre-filtered list of patients in the OR that day could be preloaded with one click, allowing access to each patient quickly and unambiguously.
The third step, reading through the records, requires me to identify important notes for the particular surgery of the day from a half dozen to hundreds of other recent notes. Many doctor's notes aren't that important for my specific purpose, e.g. the patient's physical therapy notes. Others are extremely important, such as the note from the surgeon seeing the patient before the operation. There is no way to generate customized algorithms to search, filter, and sort these notes based on what the operation is, who is doing it, and what sorts of notes we (as pathologists) generally find most useful. Sure, there is a search function, and a filter function, but these are clunky, slow, and very limited in their customization options. Better than nothing, but less than ideal.
The fourth step, actually reading those notes, frequently runs into the problem of note bloat. Thank goodness I don't have to read doctors' individual handwriting, but I do have to scroll through a lot of auto-generated content. That content is often there, sad to say, for the purposes of billing. Sometimes you get to the end of a note and realize that the entire thing is a billing document. This takes minutes, per patient, which adds up to hours over the course of a week. I can never get that time back, and I'm not really learning anything but what font sizes and formats the auto-generated stuff normally uses so I can scroll quicker.
We can do better! It's not even that complicated! Notes are really just formatted text stored in databases. A physician with a little programming know-how could write very simple programs that would save hours per week for doctors in a their specialty. But there is no avenue to perform such customization. Moreover, as the marketplace becomes increasingly dominated by very few corporate providers, if not a single dominant provider, will there be any motivation to improve the product?
Electronic medical record, you have so much potential, yet untapped! I dearly hope that you see it. If not, perhaps you could attend some motivational seminars. Get some help from some computer literate medical colleagues! With a little ingenuity, I think you can shed the weigh, speed up, and improve your agility. It'd be great for your health, ours, and our patients,
Sincerely,
A Faithful User
Subscribe to:
Posts (Atom)