Diagnostic Concordance by Berger

I don’t know why I’m in the hospital!Patient-doctor diagnostic

concordance

Zackary Berger, MD, PhDAssistant Professor

Johns Hopkins General Internal Medicine

QuickTime™ and aTIFF (Uncompressed) decompressor

are needed to see this picture.

Doctors and patients think

differently.

Why am I in the hospital? Z. Berger, Hopkins GIM, ICCH

3

Doctor-patient diagnostic concordance

• Mr. Gonzalez says he’s more tired now because he’s getting old. Dr. Patel says Gonzalez has unstable angina.

• Are they talking about the same thing?


4

Ripped from the Bellevue Hospital casefiles:

The real story of Ms. D Ms. D’Angelo is a 61 year old English-

speaking woman from New York. She has a twelfth-grade education.

Why did her doctors admit her?

They say: atrial tachycardia, right-sided hemiparesis, and Broca's aphasia.

She says: Because I can't speak.


5

Doctor-patient communication in the hospital

• Why is it important?

• Why isn’t it important?

• What are the goals?

• How do we know we’ve done a good job?


6

Diagnostic concordance: a previous study (Tisnado 2006) Diagnosis Percent agreement

MI 93% Cancer 92% Diabetic retinopathy 79% Depressed mood 73% Arthritis 69% Angina 65%


7

Diagnostic discordance among medicine inpatients:

specific aims What proportion of medicine inpatients don’t

know the physicians’ reason for admitting them?

What proportion give a doctors’ reason for admitting which doesn’t concord with the reason in the EMR?

What are the covariates associated with concordance? Discordance?


8

Diagnostic discordance among medicine inpatients:

hypothesesIgnorance/discordance associated

with Not speaking English Less education (fewer completed

years of schooling)


9

Study design, setting, and population

• Design: Cross-sectional• Setting: Bellevue Hospital (New York) • Population: Internal Medicine Inpatients

• Excluding those whoo Did not speak English or Spanisho Were cognitively unable to participateo Were incarceratedo Were severely ill

o Study measures: via patient interview; medical record abstraction

• IRB approval: oral consent from all patients


10

Patient interviewoDo you know the reason why your

doctors say you’re in the hospital?

oUd. sabe el razón porque los doctores dicen que Ud. está en hospital?

oWhat was that reason?

oAlso collected information about age, gender, race/ethnicity, years of education


11

Medical record abstractionHierarchy of notes to abstract doctors’ reason for admission (most to least preferred):

oDaily progress note (housestaff or PA)

oDaily progress note (attending)

oFace sheet


12

How do we define EMR-patient diagnostic concordance?

• Agreement ono Organ systemo Pathophysiology

• Exampleso “Lung infection” and “colitis”

Agree neither on organ system nor on pathophysiology

o “Stomach pain” and “dysphagia” Agree on organ system but not on

pathophysiologyo “Blood clot” and “deep venous thrombosis”

Agree on organ system and pathophysiology


13

Patient characteristics (N=46)

• 51% women

• 50% older than 60

• 41% interviewed in Spanish

• 9 mean years completed schooling


14

“No sé porque estoy en hospital”

• Eleven percent of patients can’t give any reason why the doctors admitted them to the hospital

• Sixty percent of these are Spanish-speaking


15

Diagnostic discordance The diagnosis the patient reports as the

doctors’ reason for admission doesn’t match the diagnosis in the computer chart…54% of the time.

• 15% disagree on organ system and pathophysiology

• 39% agree on organ system but not pathophysiology


16

Discordance is greater among older patients

Concordant(N, %)

Discordant(N, %)

Age >= 60

Age < 60

5 (17%) 18 (83%)

11 (48%) 12 (52%)

30 (65%)Total 16 (35%)


17

Discordant patients are less educated

Concordant DiscordantMean years of schooling (SD)

10.6 (4.4) 8.9 (4.7)


18

Summary• There is significant doctor-inpatient

diagnostic discordance on the medicine service of a large public city hospital

• About ten percent of patients can’t give any reason why they were admitted to the hospital


19

Limitations• Small sample size doesn’t allow us to

examine patient characteristics associated with concordance/discordance• Future research could expand study

population• Single institution in New York City

• Future research could replicate study in other settings

Future questions• “Has the doctor told you?”• Controlling for physicians’ diagnostic

uncertainty (“rule-out” admissions)• Diagnostic knowledge or

concordance at admission vs. discharge

• Concordance ↔ better outcomes?


20


21


22

How do we address the problem of diagnostic discordance?

•Discussion of admission diagnosis between doctors and patients

•Auditing medical records to improve diagnostic cordance with patient

Make way for Diagnostic

Reconciliation!

Diagnostic Concordance by Berger

Documents

hopkins gim

esten hospital

bellevue hospital casefiles

diagnostic discordance

icch concordant n

s porque estoy en hospital

abstract doctors reason

patient reports