CrossRoads

CrossroadsTransforming Health Care Delivery with

Informatics: A perspective from the Massachusetts General Hospital (MGH)

Henry C. Chueh, MD, MS

Coming up...

Introduce the problem

Suggest a theme to solutions to the problem

Provide a brief description of MGH

Describe our path through stories about tools

Revisit the theme

One Day in the Life of a PCP

10.6

7.4

6.0

Hours needed to manage 2,500 patients...

Chronic disease (10 Dx) Prevention FreeYarnall KS, et al. Primary care: is there enough time for prevention? Am J Public Health 2003; 93:635

Ostbye T, et al. Is there time for management of patients with chronic diseases in primary care? Ann Fam Med 2005; 3:209

“Computerized clinical information systems will help physicians close this quality gap

by performing many of the repetitive, protocol-driven tasks.” (McDonald, 1976)"the burden of harm conveyed by the

collective impact of all of our health care quality problems is

staggering" (Chassen et al., 1998)“The current care systems cannot do the job...trying harder will not work...changing

care systems will.” (IOM, 2001) "Our recovery plan will invest in electronic health records and new technology that will

reduce errors, bring down costs, ensure privacy, and save lives." (Obama, Feb 2009)

Lack of time Flawed process

demand

supply

CDSS

MDs = 178 • FTEs = 101 • Practices = 15 • Patients = 155,590

MGH Primary Care

(at MGH)IMA

WHABMGMWI

(near MGH)DowntownBeacon Hill

MGMGSenior Health

Back BayNECHC

Charlestown

Everett

Chelsea (2)

MGH WestRevere (2)

IT infrastructure at MGH

Online registration and scheduling

Outpatient electronic health records and e-prescribing

Clinical data repository for results and reports

Inpatient provider order entry

Patient portal

Information “push”

patient

nurse care manager

primary care provider

1

2

Decision support in diabetes(A1C, SBP, LDL) Visit

Modest improvements in process outcomes only

info

info

reco

mm

end

Patients don’t always come in for clinic visits.

Clinic visits are busy.

Providers are good at deciding, but bad at doing.

Getting there from here

Loyalty cohort: Connectedness

FastTrack: Enhanced CDSS/SSCD outside visits

ACCORD: Involving patients

Who are my patients?

Visit to registered PCP

Repeated visits to a specific practice

Age and home address as variables

Connection Status

59%

34%

6%

PCP connected Practice connected Not connected

n = 155,590

MGH Preventive Metrics

0

17

34

51

68

85

Mammography (n=35,865)

Pap Test(n=65,860)

CRC Screening(n=37,605)

Per

cent

age

(%)

P<0.0001 P<0.0001 P<0.0001

PCP Linked Practice Linked

MGH Disease Metrics

0

18

36

54

72

90

Diabetes A1C(n=9,632)

Diabetes LDL(n=9,632)

CAD LDL(n=6,612)

Per

cent

age

(%)

PCP Linked Practice Linked

P<0.0001 P<0.0001 P<0.0001

A patient-tailored information letter

A hard copy Rx, to be signed

Automatic electronic documentation

A 6-week reminder timer

Results

64%

15%

9%

11%

PCP/screened Practice/screenedPractice/overdue PCP/overdue

Breast cancer screening

patients

care delegate

primary care physician

Decision support in breastcancer prevention

Outside the Visit

population info

Mammography FastTrack

action

0

7.5

15.0

22.5

30.0

All Patients MD-Linked Practice-Linked

Intervention Control

P = 0.01

P = 0.01

P = 0.02

Mammography FastTrack: 6-month results

Com

plet

ion

rate

(%)

Awareness of Issue

Risk assessment

Plan for care

Follow-up

Complete care

?

A Fragile Loop

1/3 have no system, <1/3 satisfied with system

Patients want communication

Lack of documentation

Lack of patient understanding

Follow-up

ACCORDAmbulatory Care Compact to Organize Risk and Decision-making

ACCORD characteristics

Patient-Provider preferences

Explicit agreement with documentation

Fail-safe monitoring

High visibility

CDSS

CDSS

Population-based

Continuous, not visit-based

Patient-provider preferences

challenges. tools.