Koopman RJ, Steege LM, Moore JL, Clarke MA, Canfield SM, Kim M, Belden JL. Physician information needs and electronic health records (EHRs): Time to reengineer the clinic note. J Am Board

ORIGINAL RESEARCH

Physician Information Needs and Electronic Health Records (EHRs): Time to Reengineer the Clinic Note Richelle J. Koopman, MD, MS, Linsey M. Barker Steege, PhD, Joi L. Moore, PhD, Martina A. Clarke, MS, Shannon M. Canfield, MPH, Min S. Kim, PhD, and Jeffery L. Belden, MD

Background: Primary care physicians face cognitive overload daily, perhaps exacerbated by the form of electronic health record documentation. We examined physician information needs to prepare for clinic visits, focusing on past clinic progress notes.

Methods: This study used cognitive task analysis with 16 primary care physicians in the scenario of preparing for office visits. Physicians reviewed simulated acute and chronic care visit notes. We col- lected field notes and document highlighting and review, and we audio-recorded cognitive interview while on task, with subsequent thematic qualitative analysis. Member checks included the presentation of findings to the interviewed physicians and their faculty peers.

Results: The Assessment and Plan section was most important and usually reviewed first. The History of the Present Illness section could provide supporting information, especially if in narrative form. Phy- sicians expressed frustration with the Review of Systems section, lamenting that the forces driving note construction did not match their information needs. Repetition of information contained in other parts of the chart (eg, medication lists) was identified as a source of note clutter. A workflow that included a patient summary dashboard made some elements of past notes redundant and therefore a source of clutter.

Conclusions: Current ambulatory progress notes present more information to the physician than neces- sary and in an antiquated format. It is time to reengineer the clinic progress note to match the workflow and information needs of its primary consumer. (J Am Board Fam Med 2015;28:316–323.)

Keywords: Decision Theory, Electronic Health Records, Information Systems, Medical Informatics, Qualitative Re- search

Electronic health record (EHR) use is increasing in primary care practices, partially driven in the United States by the Health Information Technol-

ogy for Economic and Clinical Health Act. In 2011, 55% of all physicians and 68% of family physicians were using an EHR system.1,2 In 2013, 78% of office-based physicians reported adopting an EHR system.3

EHRs can, however, be a source of frustration for physicians. A 2012 survey of family physicians revealed that only 38% were highly satisfied with their EHR.4 Among the barriers to EHR adoption and satisfaction are issues with usability, readabil- ity, loss of efficiency and productivity, and diver-

This article was externally peer reviewed. Submitted 27 August 2014; revised 31 October 2014; ac-

cepted 12 November 2014. From the Department of Family and Community Medi-

cine (RJK, JLB), the Department of Health Management and Informatics (MAC, MSK), and the Center for Health Policy (SMC), University of Missouri School of Medicine, Columbia; the School of Nursing (LMBS) and the Center for Quality and Productivity Improvement (LMBS), Uni- versity of Wisconsin, Madison; the School of Information Science and Learning Technologies, College of Education (JLM), and the Informatics Institute (JLM, MAC, MSK), University of Missouri, Columbia.

Funding: This work was supported by a grant from Miz- zou Advantage, University of Missouri.

Conflict of interest: none declared.

Corresponding author: Richelle J. Koopman, MD, MS, Department of Family and Community Medicine, Univer- sity of Missouri, MA 306-N Medical Sciences Building, DC032.00, Columbia, MO 65212 (E-mail: [email protected] health.missouri.edu).

316 JABFM May–June 2015 Vol. 28 No. 3 http://www.jabfm.org

on 4 S eptem

ber 2022 by guest. P rotected by copyright.

http://w w

w .jabfm

.org/ J A

m B

oard F am

M ed: first published as 10.3122/jabfm

.2015.03.140244 on 8 M ay 2015. D

ow nloaded from

gent stakeholder information needs, which all are crammed into 1 form factor.5

Physicians often come to an EHR with precise information needs that depend on the clinical sit- uation. One important function of the clinic note is to allow physicians to prepare for an upcoming visit by understanding what happened during a previous visit. For example, a physician may be preparing to see the next scheduled patient and notices that the patient was seen by another physician last week. The natural question is, What happened? The phy- sician looks at the note title and sees that it says “strep throat.” The physician now has a good idea of what happened during the visit last week without extensively reviewing the visit. However, if the pa- tient is now presenting with a rash that the physi- cian suspects may be an allergic reaction, the con- siderations are more complex and the physician may want to know a little more, including what treatment was prescribed. Nonetheless, the physi- cian’s information needs are goal-directed. Exces- sive information cluttering a visit note can impede information retrieval by increasing nonessential cognitive processing.6

Presenting users with information they want and need in a prioritized manner could improve EHR information review and cognitive processing, which could reduce cognitive load, error, and fatigue.6

The first step to better information presentation is determining what the physician user wants and needs from the patient’s record for ambulatory clinical care.7 Therefore, with the aim of informing better information display, we interviewed physi- cians viewing typical acute and chronic care visit notes in preparation for a patient visit and asked them to explain their approach. We sought to un- derstand how physicians reviewed notes, their per- ceptions of the most and least important parts of those notes, and how they thought the EHR dis- play could be improved.

Methods We used cognitive task analysis methods to help characterize the information processing and in- formation needs of physicians performing a typ- ical clinical task that required processing com- plex information, namely, reviewing notes to prepare for a visit.8 Cognitive task analysis in- volves defining a typical task and observing the user performing that task, often using supple-

mental data collection methods such as inter- views. Cognitive task analysis involves “charac- terizing the decision making and reasoning skills, and information processing needs of subjects as they perform activities and perform tasks involv- ing the processing of complex information.”8

Cognitive task analysis has especially been used to help understand whether technology systems are meeting user information needs, which made it an ideal method for our purpose.

An early step in cognitive task analysis is de- fining a scenario. We chose a representative ac- tivity of primary care physicians: preparing for a patient clinic visit by reviewing the patient’s re- cent clinic notes.8 We presented physicians with an acute and a chronic care visit note represen- tative of EHR documentation typically generated by physicians. The acute clinic visit note con- cerned a patient with a cough, and the chronic clinic visit note was a follow-up visit for a patient with type 2 diabetes, hypertension, hyperlipid- emia, and depression. We asked the physicians to highlight the sections of each note they would review in preparation for a return visit of a pa- tient in 3 different scenarios reflecting the pa- tient’s primary care physician and who last saw the patient: (1) you are the patient’s primary care physician and you last saw the patient; (2) your partner is the primary care physician and your partner last saw the patient; and (3) you are the primary care physician and your partner last saw the patient. We anticipated that these 3 sce- narios might produce different information needs for the physician participants.

The physicians were presented with tasks to highlight important parts of the note and then to highlight parts of the note they found unimportant. While highlighting, we asked the physicians to communicate their cognitive process using think- aloud prompts.8 After highlighting, we interviewed them using a semistructured interview guide, ask- ing about their preparation process, what they looked for in a note, what they considered extrane- ous, what they thought about the structure and function of current progress notes, and how current notes suited their information needs. A second in- terviewer was present and made field notes on the task analysis.

We studied 16 primary care physicians practic- ing in community clinics associated with a medical school department of family and community med-

doi: 10.3122/jabfm.2015.03.140244 Physician Information Needs and EHRs 317

on 4 S eptem

ber 2022 by guest. P rotected by copyright.

http://w w

w .jabfm

.org/ J A

m B

oard F am

M ed: first published as 10.3122/jabfm

.2015.03.140244 on 8 M ay 2015. D

ow nloaded from

icine and a division of general internal medicine. Physicians were sampled for maximum variation in sex, years in medical practice, and experience with EHRs. Attending family medicine physicians formed the sample majority; smaller numbers of internal medicine attending physicians and resident physicians were sampled for confirmatory purposes after reaching saturation.

Interviews were audio-recorded and transcribed and then analyzed by 3 qualitative coders: a family physician clinical researcher (RJK), a human factors engineer (LMS), and an information and interac- tion design scientist (JLM). Coders also referenced and coded highlighted documents and field notes. This blend of analysts brought an “insider” clini- cian perspective, which was balanced by “outsider” human factors engineering and information science perspectives. All 3 coders independently coded the first 3 transcripts; they met after each transcript to come to consensus on codes and meaning and to create the codebook. The remaining 13 transcripts were independently coded by dyads of 2 coders, again with coders meeting to come to consensus on codes and meaning using a thematic analysis ap- proach.9 Analysis was aided by NVivo 9 qualitative analysis software (QSR International, Doncaster, Australia). The study was reviewed and approved by the University of Missouri Health Sciences In- stitutional Review Board.

After analysis was completed, we engaged our physician sample and primary care colleagues in a focus group with an interactive presentation of our findings. This informant feedback served as a mem- ber check.10 We also presented prototypes of 3 different note displays that might improve presen- tation of information needed at the point of care; these displays were based on interview findings. Another source of feedback was presentation of these findings at an international primary care re- search conference.

Results Participants included 14 family medicine physicians and 2 general internal medicine physicians. Among these 16 physicians, 5 were women; 12 were attend- ing physicians and 4 were residents. Five had �5 years in practice, 3 had 5 to 10 years, and 8 had �15 years. All had used an EHR for �1 year, with 10 using an EHR for �5 years.

Emerging Themes Eight themes emerged from the analysis and are presented below. Themes, definitions, and sup- porting quotes are summarized in Table 1.11–15

Understanding Context Drives Visit Preparation To prepare for an outpatient clinic visit, physicians reported viewing typically the last 1 to 3 contacts with the health system, including visit notes, phone messages, and urgent care or emergency visits. Preparation was influenced by prior knowledge of the patient, the reason for the current visit, the complexity of the patient’s problems, the volume of transactions, and amount of time since the last visit with the primary care physician. Increasing com- plexity and time since last visit increased the effort needed to establish context for this visit. The rea- son for the current visit also influenced what sec- tions of the previous notes might be most perti- nent; for example, procedures history was generally not important but gained importance if the chief complaint was abdominal pain.

Forces Driving Note Content Physicians expressed that clinic visit notes have become more structured, lengthy, and complex over time in response to a multitude of evolving demands from diverse stakeholders. On top of this, they noted that the EHR has now added poor syntax and has lost the story narrative because of note construction templates and structured data elements. Physicians perceived the following as current drivers of note construction:

● Billing (checklists for each section, especially re- view of systems)

● Quality improvement measures (eg, diabetic foot examination)

● Avoiding malpractice ● Compliance (eg, documenting informed consent,

patient education) ● Relating what happened during the visit

Physicians lamented that most of these drivers do not support the physician’s information needs for the care of the patient. They expressed frustration that text responding to external drivers has come to dominate the clinic note.

318 JABFM May–June 2015 Vol. 28 No. 3 http://www.jabfm.org

on 4 S eptem

ber 2022 by guest. P rotected by copyright.

http://w w

w .jabfm

.org/ J A

m B

oard F am

M ed: first published as 10.3122/jabfm

.2015.03.140244 on 8 M ay 2015. D

ow nloaded from

Ta bl

e 1.

Th em

es ,D

efi ni

tio ns

,a nd

Su pp

or tin

g Q

uo te

s

T he

m es

D efi

ni tio

ns Su

pp or

tin g

Q uo

te s

U nd

er st

an di

ng co

nt ex

t dr

iv es

vi si

t pr

ep ar

at io

n

T he

m ai

n re

as on

to re

vi ew

pr io

r cl

in ic

no te

s w

as to

un de

rs ta

nd pa

st ev

en ts

an d

th at

m ig

ht he

lp th

e ph

ys ic

ia n

un de

rs ta

nd th

e pr

ob le

m be

in g

ad dr

es se

d in

th e

cu rr

en t

vi si

t.

“S o

it’ s

pr et

ty co

nt ex

t de

pe nd

en t.

I al

w ay

s re

ad m

y la

st no

te or

th e

la st

no te

,a nd

us ua

lly an

y ot

he r

re ce

nt no

te s

th at

an yb

od y

ha s

w ri

tt en

.A nd

th en

if th

er e’

s a

pa rt

ic ul

ar pr

ob le

m th

at I

id en

tif y

th en

I go

se ar

ch in

g fo

r ot

he r

de ta

ils .”

(P hy

si ci

an 1)

“I f

th er

e w

as ,l

ik e,

lu ng

ca nc

er in

th e

fa m

ily hi

st or

y I

m ig

ht w

an t

to kn

ow th

at .T

ha t

m ig

ht m

ak e

so m

eb od

y ki

nd of

�w or

ri ed

� ..

.‘ I’

m w

or ri

ed th

at th

is m

ig ht

be lu

ng ca

nc er

an d

th er

e’ s,

lik e,

so m

eb od

y el

se in

th e

fa m

ily .’

A nd

th en

sm ok

in g

hi st

or y

w ou

ld pr

ob ab

ly be

im po

rt an

t, th

at th

ey ’r

e a

fo rm

er sm

ok er

,n ot

a cu

rr en

t sm

ok er

.” (P

hy si

ci an

11 ,

sp ea

ki ng

ab ou

t a

pa tie

nt pr

es en

tin g

w ith

a co

ug h)

Fo rc

es dr

iv in

g no

te co

nt en

t C

lin ic

al ,l

eg al

,a nd

bi lli

ng re

qu ir

em en

ts al

lc on

tr ib

ut e

to no

te co

nt en

t, of

te n

w ith

le ng

th y

re su

lts .

“Y ou

kn ow

,I ap

pr ec

ia te

th is

ef fo

rt bu

t w

e’ re

sw im

m in

g up

st re

am he

re ,a

nd th

e st

re am

is th

e to

rr en

t of

bi lli

ng -r

el at

ed re

gu la

tio ns

,r ig

ht ?

Fi rs

t an

d fo

re m

os t,

th is

se rv

es th

e go

d of

bi lli

ng an

d it

ha s

to co

nt in

ue to

do th

at ,a

nd I

w is

h it

di dn

’t, yo

u kn

ow ?”

(P hy

si ci

an 14

) T

he as

se ss

m en

t an

d pl

an is

th e

m ai

n in

fo rm

at io

n ne

ed ed

T he

A ss

es sm

en t/

P la

n w

as vi

ew ed

as th

e m

os t

im po

rt an

t an

d in

fo rm

at iv

e pa

rt of

th e

no te

.

“I op

en a

ne w

no te

I im

m ed

ia te

ly sc

ro ll

do w

n to

th e

as se

ss m

en t

an d

pl an

to ge

t a

br oa

d vi

ew of

w ha

t I’

m go

nn a

be lo

ok in

g at

.. ..

I m

ay be

di dn

’t re

al iz

e it

bu t

th at

is al

w ay

s th

e fir

st th

in g

I lo

ok at

.” (P

hy si

ci an

7)

A na

rr at

iv e

H P

I su

pp or

ts th

e as

se ss

m en

t/ pl

an

T he

H P

I w

as vi

ew ed

as a

se co

nd ar

y na

rr at

iv e

so ur

ce of

in fo

rm at

io n,

as an

ad ju

nc t

to th

e A

ss es

sm en

t/ P

la n.

“I w

ill ju

st re

ad th

e as

se ss

m en

t an

d pl

an an

d th

en if

so m

et hi

ng do

es n’

t jib

e w

ith w

ha t

I re

m em

be r

se nd

in g

th em

fo r

I’ ll

go ba

ck up

to th

e hi

st or

y of

pr es

en t

ill ne

ss an

d be

,l ik

e, if

th ey

go t

a di

ff er

en t

hi st

or y

th an

I di

d or

th ey

di dn

’t ge

t a

pi ec

e of

in fo

rm at

io n

th at

I kn

ew ,t

he n

I’ d

pr ob

ab ly

go ba

ck up

to th

at .”

(P hy

si ci

an 10

) R

O S

is “u

se le

ss ”

T he

ph ys

ic ia

ns di

d no

t re

ga rd

th e

R O

S as

pa rt

ic ul

ar ly

us ef

ul .M

an y

id en

tifi ed

th is

as a

so ur

ce of

no te

cl ut

te r.

“A nd

th e

re vi

ew of

sy st

em s

I sk

ip en

tir el

y be

ca us

e ho

pe fu

lly so

m eb

od y

to ld

m e

so m

et hi

ng in

th e

H P

I th

at w

as go

nn a

be im

po rt

an t

fo r

th at

vi si

t. O

th er

w is

e, I

ha te

th e

re vi

ew of

sy st

em s.

I th

in k

it’ s

us el

es s.

.. .I

do it

in m

y no

te s

to sa

tis fy

bi lli

ng re

qu ir

em en

ts be

ca us

e I

ca n&#x20