0:00

Number seven, be curious.

0:04

We are physicians, we are scientists, we are individuals

0:08

that have a natural curiosity to us.

0:11

Be curious about a case.

0:13

I really enjoy opening up the patient's electronic medical

0:17

record, electronic health record,

0:19

and reading about the actual traumatic event.

0:22

Unfortunately, you know, a lot of the ED cases that I read,

0:25

it's, you know, royal trauma

0:27

or you know,

0:29

MVCI like reading about the motor vehicle collision.

0:32

Was it a head-on collision? Was it a rollover collision?

0:34

Those things have some import in the understanding

0:38

of whether there is a dangerous mechanism or not.

0:41

And similarly, in patients who have a new mass in the brain,

0:45

it's very helpful to know where they have a known primary,

0:49

even if that primary is 10, 15 years old

0:52

and they were thought to have been cured of

0:55

that primary tumor.

0:56

It's particularly true with breast cancer, unfortunately.

0:59

So explore the clinical history.

1:01

I I, I'm one of the fast readers in our division,

1:05

and yet I would believe

1:07

that I look up the patient's clinical history more

1:11

commonly than many of my colleagues.

1:12

It's worth my time, even after my blind review.

1:16

You've got a patient who has diffuse white matter lesions,

1:20

or you've got a patient who's got

1:21

multiple nodules in the lung.

1:22

Is this a coal miner? Is this a person who has known tb?

1:25

Is this a person who has, you know, erythema nodosum

1:29

and you know, is known to be at risk

1:31

for things like sarcoidosis?

1:33

I'd like reading the, the clinical histories.

1:36

So I think that there will come a time

1:41

where that is the standard of care that we are required

1:44

to not just look at what the clinicians put on our,

1:48

you know, indication,

1:50

but to actually look at the medical record.

1:53

It's there for us now

1:54

and I'm afraid that over the course

1:56

of time we will be expected to look at that.

1:59

We have at Hopkins what's called crisp boot, which is a

2:03

PAC system of multiple hospitals outside

2:06

of the Johns Hopkins Health system.

2:09

It ha it includes hospitals that are in other states than

2:14

Maryland where Johns Hopkins is.

2:16

And being able to look at old films from, you know,

2:21

the University of Virginia is very helpful when you're

2:24

looking at a tumor or something new on a case.

2:28

So benefit from the magnitude of

2:30

what is available in your electronic medical record

2:33

and understand that over the course of time it may be that

2:36

that looking things up is the standard of care,

2:40

not just relying on the clinical history that's

2:42

provided in the, uh, indication on the image.

2:48

So looking at perceptual

2:49

and interpretive error in diagnostic radiology causes them

2:53

potential solutions.

2:54

Again, this is Elizabeth Kapinsky. She will say that,

2:58

You know, one of the strategies

3:00

to reduce informational error is

3:02

to mine the electronic medical record

3:05

for relevant information to provide for each examination

3:09

and this evolving era of big data,

3:11

rapid computing machine learning

3:13

and robust software tools, efforts

3:15

to streamline quality improvement issues would be

3:18

undoubtedly welcomed by the radiology community with regard

3:21

to having easy access to the electronic medical record

3:25

and the relevant portions of that record.

3:29

So in addition,

3:32

I'd say be curious about technological

3:34

advancements in your field.

3:37

Be curious about where we are with artificial intelligence,

3:40

where we are with computer assisted diagnosis, where we are

3:43

with error detection in reports.

3:46

AI now can look at a report

3:48

and identify whether you've flipped right to and left,

3:51

or you haven't put in the impression

3:54

and important finding

3:55

that you had in the body of the report.

3:58

So you're not driving home

3:59

and going, oh my God, I forgot that kidney thing.

4:03

Exactly. Um, AI can help you with comparison studies.

4:08

It can actually identify new MS plaques when you're

4:10

comparing from a prior one

4:12

or a new lung nodule when a patient

4:15

who has multiple lung nodules undergoing chemotherapy.

4:18

That type of technology is really useful to radiology

4:22

and will reduce the errors.

4:23

Obviously peer review is another way

4:25

that we can reduce errors.