The Regenstrief medical record: 1991 a campus-wide system

The Regenstrief Medical Record: 1991 A Campus-wide System Clement J. McDonald, M.D.*+, William M. Tierney, M.D., Douglas K. Martin, M.D.*+ J. Marc Overhage, M.D., Ph.D.*+, Zed Day With the technical assistance of: P. Adams, L Blevins, J. Clifford, P. Cassidy, M. Gaddis, T. Glazener, M. Hager, M. Lee, L Lemmon, B. Mamlin, S. Massengale, J. Meeks-Johnson, J. Metcalfe, B. Porterfield, JA Warvel, JS. Warvel, J. Waisburd.

*Regenstrief Institute for Health Care, Indianapolis, Indiana 'Richard L Roudebush Veterans Administration, Indianapolis, Indiana +Indiana University School of Medicine, Indianapolis, Indiana We will demonstrate a medical information management hospitals 1/4 to 1/2 mile away thmugh fiberoptic links system that has been in continuous operation at one (see Figur 1 for schematic). Each hospital has it's own hospital (WLshard Memorial) since 1973 [1]. Over the VAX computer and disk drive and runs a separate copy last few years we have expanded the scope of the of the Regenstrief Medical Record System. All are part operation to two other campus hospitals, have begun the of a larger VAX cluster. Consequently, the individual collection of "complete" doctor collected data in one computers can switch over automatically, with little user specialty (obstetrics) and have completed the one year of disruption. Moreover, physicians at any hospital, if they a study of physician order entry on our inpatient service. have the appropriate privileges, can switch among the These efforts have taught us many lessons about politics, medical records of the other 3 hospitals from any active about the difficulty of extrapolations, and about the terminal. This feature has obvious advantages to the importance of small environmental differences. We will Indiana University residents and attending staff, who staff demonstrate our system and describe some of these all three hospitals. lessons in the demonstration. Currently the Wishard patient file carries more than 35 million observations for over 380,000 patients. The Veterans file contains more than 4 million obsertions for more than 100,000 Veterans Hospital patients, and the Indiana University file contains 15 million results for about 150,000 patients. At Wishard, the Regenstrief-developed registration, laboratory, and pharmacy systems provide registration data, laboratory results, and pharmacy resultso the medical record system, and a bedside Vitalnet system provides inpatient vital signs, weight, and intake and output data. At the Veterans Hospital, the DHCP laboratory, pharmacy, and registration systems provide these data, via an ASTM format, while at Indiana University Hospital these data are provided by the CHC Figure 1: Campus Ethernet Schemadc laboratory system and the Phamis hospital information system through a Simborg System HL7 linlc The medical record content from all three hospitals is supplemented At Wishard and the Veterans hospitals, users access by manual coding and entry of diagnostic umpressions patient information thugh RS232 terminals linked to the from Xrays, echocardiograms, scintigrams, CT scan, computers through terminal servers and ethernet. More EKGs, endoscopy, and other narrative reports. Inpatient than 20 outreach and mental healtli clinics, scattered and outpatient diagnoses, as well as selected historical through Indianapolis, link to the Wishard computer via and physical findings information, are also obtained for Codex multiplexers At IU Hospital, access to the some patients and entered into the medical record by Regenstrief Medical Record and to the IU Hospital various means. information system (Phamis) is either thugh PCs or visual terminals located on an Ungermann-Bass network. The medical record system for all three hospitals runs on There, users can hot key between the Indiana University a cluster of Digital Equipment VAX computers located Hospital Information System, the Regenstrief Medical at Wishard Memorial Hospital. The VAX computers at Record System, and an IBM-based padent accounting Wishard are linked to diplay devices at each of the other system (Baxter Travenol). A collaborative of 15 city-wide 0195-4210/91/$5.00 © 1992 AMIA, Inc.

925

size and the cost of maintaining it [2]. At Wshard, the Veterans Hospital, and the outreach clinics, 4 reports are produced for each visit a patient abstract (Figure 3), a cumulative patient flowsheet, an encounter form (Figure 4), and a reminder report (Figure 5). The flowsheet is inserted in the outpatient chart at the time of the visit, and eliminates the need to file most individual diagnostic reports. The electronic record is also available on-line through video terminals located thrughout the institutions. Online access greatly reduces the need to pull the paper record.

prenatal clinics also use tfie medical record system to maintain computer prenatal and early pediatric records for all prepancies in the inner city. The prenatal clinics enter almost all of their clinical notes trugh multiple choice questionnaires (see Figure 2 for an emple). This data is entered into the Regenstrief System to provide availability of patient data to care givers at any participating site. w n" Hotai lot Wena Tenth __ _ ___d___p i.b_f IN 46202

nE

AA

_nr_A

DATO_____

OIL WC

For. Talaog H;o*y (Signed) IF PATIENT IS A TRANSFER PATIENT: Mendt Prenatal Care Began: Name of Tetrefuong Ageilcy REPRODUCTIVE HISTORY:

SAMPLE,PATIENT

Esdmared Due Dar Age od Menaiche (Yen) LB__ G_ T_ FT_ AB Ma al Cydl. (dcd. al dia apply) 4 m 1) noma a 6) days 2) beavn mem 7) menes days ) eidar se 3)11ght mene 4) 25 day c)de 9) d,sieione 2) 32 day cycle 31rth Courd Medad (citdi debem used) 1) B-C pU. 4) IUD 2) diaphragim-adoam 2) named Omilyph-mg 3) qermadldes 6) acne N1a1__ Gym Himwy (Cre all dti apply) 1) mdnmereosin S) abnoal papm 2) pocsd. nree 6) hfeefi 3) ntne leiomymam 7) odier_ 4) cie dymbee

l

PREGNANCT VARIAZSL

Ewdarifa/MetabLc Disna NO" 1) byperdiyridism 2) no-isuli dependent D 2) b6ypodyroidb! 6) cnage reular dnease 3) geadonal DM 7) odier

PRIfACY VARALEtS 04-$FS90 PAit TY 2 0 are * Of =OTtom (a Of) I

Pas P0fC6T 60CIG(T

N.m

6) 7) 0) 9)

blladnesr denness cerebri palsy

4) peivic lalammain dimare 5) onhebe 6) r)dilph__ 7) gaite hapcs 8) epatitis

ar Dinea Nae 1) byperedns 2) adisrnsb 2) diesaaic baet d e 6) dirnmbophlbks 3) cogeniel beart drne 7) pulmmoaqy embonb 4) mitWl vale prlapre ) other. N Riay D I) asdma 4) acne edie osid 2) btndhids 2) odir. _ 3) plIlnenia

C

11.6cm. us BP SITTING WASB SI TTING1

nag

CMV AMS ARC

0-4 42.0 0a

(10-A18. (10-AUll.

136

us1 A6t 16T11RnL SERUM AFP tEO AFPWS 110T MN VAL6E 811011000SStST 100-

goup B Bo nip

13) odter

C16CI01

11 181 00 79 1.1

10.0

/01. W0/t MGDL

P1OT101( TOTAL AL60Il1

4.7 81111. 7.1t MG/SI. 194 neni 6.210 G101. 3.6 1/I.

611ISA1

012 o5

TOTAL

ALI PlOS

SGOr CAST )

20'

LIVING( ARR6158R0 wth chit6r:n only P0OPLE 1 01SEWX0 4 MIwr noae TRSAPISTAtlIIN ItCE tb

babys father 30 T0? s7sT0 0111S16001.11810608 0 Y /S ATT1D SC1OOL 10 quit schoot STUDENT STATUS AGCt IAIYI fAtlERt 33 Y8SSCM1 0 C BAB FAT0E0 S 12

UNITS

UNtI TS

0S(1 SFA1H1R0T6 IC:TY blcic AIT' tFATHtER RISK NX SiCkLo ctll trait PCCIIGAICY EXPOSURE passive smite IlUEOEUATE StfSAJCtS

ERUITOLOGY 00LL POF 31-0U190 Wc a 00 1/CU 0 RIC S.O lMILL/CU M

nm 8111 011111010 001

17.? 1k

MrEN LA CUICOSC LUA

e

(0 0

39 0 Om f l 20 PC II 34 31000.0 NOteAjnsi

106Tt(E U6I8ALYSIS 02- V-90

-90

OCLAS5

000

WIC food sk4*t~t5 Aid to COodnt ChiPdren

8DIL

URll nCID C1811.16t01L

SCT Kws

n

1N. 161KW1

10

90.3 11/1 37.6 ST I 2.4

P10. (T01 STStSS TCST

n" n

erni y clothes

craiLd csre no.

OBSTACLES TO CAItt chitd care finPncialt

08 DIETFOM 09 oIET FUIfi 0-FEI-90 MILK GI" DIET ItX 2 MEAT GC2 DIET NX 3 FRUIT/VEG CRaW DIET MX *6

F1TAL ULT601 CSt. WY-US

Otber MedlW Disn

0 IeDICAL S rSTORT 0 68108610T OAA604-9F-90

90

23F

22

GM

18ST1JAL CYC1 NX naomI I11T0 CT1 16CO1 tural fmity plai txOOCSINt MItTAOOLIC SX diabeis met inalin

N

blood osf_sin 1) anemia

2) bemoglobinopadiy 3) innimmuizration 4) gall bladder dineoe

PRENA0TAL A6O IN I15W1190

2) congenital nomaieas 6) DES eqonre 7) canrer (apeciy) 8) odier_

pox

Sb

mmman SCRTNN !Em

OS ECAT

SlCCiL. CELL SCS£Elt pm UILLA 1SCM pe

90

IS

-

ANTIGEN8

restIve

pM

PAO StA 10M 12-JUN-90 a goaerrinme

4)

Os

)m oy 3) tonsiLmy D&C dcokgcyeMy 10)9) odler. S) ampoinire 6) cemiage FAMILY HITRY.*s peanssa (cude al di 9aa9y) 1) Endo SI R&W as

INFECT10NaX

SNEG TESTS1-ATSU-90

CC C 0l0.t0A1111 no .rLou"

00-AL-90

CUlTrURES 02 18V 90 USttNE ULtTURE URXm ONG. Cs noxt cotup-)

(OS-OCT)

(00-6Wl1

dep

Pla:

ALLERGY NXU1str TO TEST VAR nq IlWATITIS sCa xX nes NIV TEST XX nq

StCt01C1ELL s sxX nanot SU LLA SCI NX 010E.LA IN" EVE01n8

no

brrier

8

SEAVtNGS/O3AY SERVINGS/CAT

SCRV NS/DAr

RCA OL C2 IOET ioX 6 SERVINGW/AY OTHERS GA"I DIEt ItX 4 SEAVINGcS/DAY DIET PRC81E1 mx dI ental problem WEIGHtT PROBLEM oozsity DIET KCAL/DY 2200 CAL DIET 7PXULMEALMA 3 OltT PUU SMACKS/DAY 3 MGt SPECIAL DIET disactic diet Y" OlGT COSSELlNC PREtC¢r WEIGHT GAlIN excessive pr" weight

gin

_w

2 1TOII c1001un.60o0 * days-en 1TON 601T/6AY 1 0inks/day 0 c001s/d AMi0 TDAY /

C1018ETT0

U1STA1C18 A6USI0X _mrine 11-m8 03 SocrAL SISToRT

01(61.A8 ISTOR8 Foot 04-F01-90 I se next coza)

This abstract contains recent medical data and does not reflect the complete medical history of the patient. Record last modified 23-DEC-90

6) Hean Arct 7) High Blood Preme 1) Mulple BEt16

4) BlE Defem

P161D01111

MAt. WE.LNIG

y HIls

2) CVA

(04-Fl No

e NG

CUOC C £ATDIllft

PArEAMAI Aft' SUX leC-F-9

1) no pioLs sgety 6) coniadon a" Updm 2)apdmy

3) Dla Media

C118012 S-6J6-90

L"s L..1

110

70

0O SOCCAJt StCR1 FORM C4-EIfE90 (fr prwvious cetumn) MARITAL StATUS di.oreds

viru not isol 109-01T7

ASS ISTANCE PtOGRANS

(13-inc

CL68t COENT 10-A8.90 300 TSUtESTC0R LASS 0E0T VISIT CWCOLA NOT VISIT CJtTENSIVC SOMA MIT VlISITr

-

S

170

ftT1LN6.880T

.-- 4-

-GLHSrOlty *Dt2

,so

W001180 LIt

EOC6 01110 91111 1IG0T

odier Noao

9) 10) 11) 12)

3) cmadia

(29-AG1

MCSFtS CUILTURCe

00 SOCIAL OISTORY

(02-00

00 AsnltACr CL o0tA 02-aov-90 Inc. , IT

4) dnq siene 2) odi

lafednan

1) so bkrwn allergie 4) vee 2) pelnkI S) odier 3) sdfa

39.0aO

CESTATION WI1EKS

3) rend ran mt

1) ndyl.s1 2) cpitis

56A0Y90

LAP

NOW

-d Oh L epilepq 2) mgeaine 3) uipe clernais 4) tmal retbrdaden 2) smke

(fro pr4wien cot11a) 50.000-74,000/cc

LAST VtSrT DATA

Renal Dis..

1) glmeAdepb ns 2) pycboep ds

PRENATAL SCREKINO ET On Caea CULUE8 02i-i 90

GPtA10A0

4) imiln-dependeni DM Year DM diapered

MEDICAL HISTORY (dide all that )appy) Prepepm Weogl________

08 .U

#0099999-5 AGE:31 SEX:M RACE:B PHONE:535-4435 PRENATAL CLINICAL ABSTRACT

Figur 3: Prenatal Abstract

SUBSTANCEW8 ALCOHOL 0

PM

I2nddTri

in Tri

Day/Mo

Dda/DayI

II

ClGS/Day

MEDICATION S2N1 LMAW (edle a1 dist appIy - wte In oedr dneg name)

2w

p~~ose / SiffD

120 I sit

1) Danddo

10) Esnrso

2) Cad_o_

11) 12)

3) Udim

4) Vaic

Ad

13)

_

5) Teosycli 6) AcaMe 7)AA

14)

_

1) 16)

_

0) Tenequan 9) Dluredas

Feroilly Dnup

17) is__

__

_

_

__

_

_

1)

_

_

_

_

_

_

_

Figure 2: One of Many Prenatal Input Forms At Wishard, where the implementation is most mature,

the medical record cariies information about all diagnostic studies, treatments, and diagnoses, in a fully coded and computer understandable form. The system does not replace the paper chart, but it does reduce its

926

The encounter form is patient and clinic-specific and is the only worldng document for the visit. The reminder report is based on a large set of quality assumnce protocols, and used by the computer to review the patient's medical record. The system has two different queiy/retrieval systems. The CARE system pernits the specification of veiy complex conditions, including protocols of care [3]. When CARE is used as a queiy language, the computer retrieves data for statistical analysis [4-5]. When used for quality assurance, CARE remids physicians about clinical conditions that need attenion. These reminders have had significant influence on physician decisions [6]. The FAST query system, which retrieves data via special indices, has fewer capabilites than CARE but it is faster and easier to use. Queries that might take many hours searching trugh the 340,000 patients' records via the CARE system can take minutes with the FAST system.

SAHPLZ,EXOTNIZR

TUE AX I Dr: BLACKBURN B0Y3T *WINhVrd HsoriaLt ospital 2700 Cc Martin Luther KINg Jr Se EIVLaNapolNi IN 46202 CLINIC VIStT, 2 LINITO VISIT, S N U ILL. LAS ONLY, 5 NO CHAOG

24-sEP-SS 317-926-7400 .*- HRC L001 W TNTH ST 46202 IONOANAPOLIS 1

Zc1iadrr_ esi Xeariou tooh

and other chargeable activities. The billing diagnoses are entered as part of the maintenance of the problem list. In some clinics, the corputer automatily translates 90% of the physician problem names into standard ICD9 code. return to Bogenstrief Institute, do not chut.

please

TIDBTS FROM THE COMPUTER These suggestions are based on incosplete data; your judgunt should take precedence.

.NeNgti.R List. on tos 2 IttCNT LOS _R Ls

SAMPLE,MOTHER #99999-5 AGE:35 SEX:F RACE:B PHONE:630-7400

37SP SITTINCN __

Scheduled to BLACIKURN OBGYN

* -

6

5

I--

IUN

5

T I

on 02 -APR-91 at 10:10 AN

I

B

11 *

GONORRHEA SCREEN

ar11I04

ta, Oi

9 CK0TATIO70 T

11 FETA NrU

rm_t

12

_a

. 1.

I

Z

lt

HERPES CULTURE

C. .. TA

1

la*oJ

ST I a S

(1)

CHLAHYDIAZYME

VP

CBC

FNr METNa

NIV (vith patient's consent)

13 f[TAL MVMN

PPD

RUBELAA SCREZN ON 06-SEP-89 WAS poe *

RH BLOOD TYPING ON 26-OCT-90 WAS pos

IV (.ith pstidnts connt)

IN Glu cose (Nistw of 6bmr0 etua *Urine C- tur (~ 16 wio gtation

Os Zict

I

r

26t

(~

tests)

ke "sain

This patient has a history of abnorsal glucose tests: Please consider having a 1HR GLUCOSE drawn today.

Staff

signBture:

010vider to

Ietur rovider

etw

Wt Jt Osto

~~~~~~~~~~~~~~~~~~~Printed:01-Apr-91tMut_:3

ENCOUNTR FOILM NR. HC ...

46202

t CLINIC

ll* all laRnk*

i

ACETAINPHRN H/COO (30 NO 011N723.2 I.I-IAY-9 5t-NA-SO (schedule 31 (05A requiLred)

2

TARS PO 045H Eps

PACKAGEI ASD 144421-MAR-91 RsCt21-NAR-01

1270124.2

10012237.2 '

(1

5ROHOCRIPTZRR J2. Rd.2io-HAR-9O ISes20-NAR-90 (325 NO) rtsousSuLrArs

TARS

PO

S__nature_Pr 020API -91 e ~eOt

written

1 DOXES

REALS

last

1

(A3)

Rfs.

Dimp

R23.90 90 T.AssQS $4.

enougn weignt

0 1 A 0

#99999-5

Figure 5:

..

0

0

()

Glucose.

gain)

lsa....

bottLes 04.70

SOXES

Prenatal Reminder Report

0.

DLp

l'ishard, physicians in the

1At 1

Re.fs

sp

..

medicine

wards,

enter

medcicine their

clinics, and orders

on

the

through

microcomputers linked to a central VAX computer and to

each other via an Ethernet network (see Figum 7).

May sobstituts /

/

SILA

0o

$5ice 4,9 e1A tOC /to 02-APR-91 at 10:10AM Pri.ted:01.Vr-91 OUt_opc.NI

ENCOUNTER FORM

Figure 4:

I

n_a)

-'Ide tO

SAMPLE,MOTHER

Disp

l00 TASS

1A1

1/2

Res

(AJI

24 HEALS

2/3 PAIR

Rxp.

#.270124,1

as

Hr

Please note on the encounter sheet, any fetal sovesent and all educational topics discussed with patient

04.10

TABS

12

1/2

WETH

PO 310 WNTH 1/1

LC0012237.1 EeIS-AAOLz9 ASt20-AR1-9Rap. 0 Eset2l-NAR-91 Rots21-AR-91

Dispens

(Not

(REDSCENV)

1/1

TAra

IssN2N-KAR-SL Reft28-NAR-91 Zxps ESUPEU 1400 NG I 2 TABS P0 QNX

1270124,3 6

1

Please consider referring this patient to Dietary for a consultation.

offosa

-1/1

(ER2) $S.34

0012237,3 ESa.20-NAR-90 RefC20-NAR-90 Xxpt S CLOcOSTEE ASO (HESCENE) $29.00 7

Patient over 28 vks gestation: Please consider ordering a

St

t

PR

Ep.

3

4

2700 Or SartLn Luthec Elng Jr EadLauupolLs IN 46202 3 RN EtSPILL. 4L* OUL, 5 N10C t

2 LINITO VISIT.

OLOIn at bottN

1

ALCOHOL PADS

VISIT

ORGY)

1

P R 2 s c R I P T I O N S ---------------

--------A C T I v It

2

consider collecting speciman

Patient over 28 vks gestation: Please consider ordering a CBC.

OPB-8

Dr: BLACKBURN WLshard NNoPIal RospLtal ANE

TUE

24-SP-SS 317-926-7400 1001 TSTUSH ST SlOZtUAAPOLES

Patient over 16 vks gestation: Please for the Urine Culture.

02-APR-91 at 10:1 0 AM

#99999-5 )

SAIPLE,XOTHER

I /@ACKtlLZCBGY 1,-ice A'.

mwnths

02~~~~~~~~~~~~k

fmeowvsTer D.t.

SAMPLE MO THER

OPB-W

Prenatal Encounter Forn

The FAST system can display distributions for any 2 patient cohorts. It can, for example, display a distribution of chest x-ray fndings in patients with a diagnoses of chronic lung disease and the comparable distribution from a random population, adjusting for the baseline prevalence of findings (Figure 6). At Wishard, the medical record system is part of a larger system that captures patient charges, schedules appointments, and helps manage the clinic. It collects all needed administrative, insurance and billing information. Its scheduler provides appointment scheduling with options for multi-scheduling, wait-listing, assignment of phiysicians to patients (to assist continuity of care), patient ViSlt reminders and load balancing. At check-out, it captures the diagnoses and records, procedures, supplies,

DISTRIBUTION OF AN OBSERVATION Distribution: Union of all CXR for cohort LUNG-DISEASE Patents with parameter COHORT 483 out of 483 LUNG--DISEASE 303 outof 483 Random 483 (Use arrow keys to scroll, press ESC to quit) CHI-SOUARE = 472.4157, D.F. -

Findings

COPO overinfiation interstit fibrosi interstitial mark in bilateral

infiltrate heart normal right HOSP CP angles blunt left RUL pleural effusion worse

Figure 6:

927

LUNG-DISEA Random - 4 5.611 60.248 43.064 6.931 35.404 11.551

31.056 38.923 37.474 65.631 30.849 57.971 22.981 26.915 18.427 21.946 20.290

=

107

# ofLUNG--DISEA Random- 4 17 291 208 21 171 35 150 32 188 58 181 56 144 317 47 149 280 134 -

-% of

10.561 19.142 18.482 47.525 15.512 44.224 9.901

15.182 7.261 10.891 9.241

111

130 89 106 98

Distrbution of CXR Findings in COPD from FAST Retrieval

30

46 22

33 28

Orders

HOSPITAL.ORDERS 999999-6 TESTs, PATIENT 999999

ePR 1 91-10:

1CIt

_

5)

THEODUR

6) TRIAITEREI11 cIiF options options

CIDAM NCDWILD

C4

CXIIRN1ouPastHxof *l'' 011112 Hext routite tia

rx Orders choices 1) ACTIVE ORDER nEMU 2) POBLEII sPEciric nsu TESTG's active orders 4) IU FLUID ORDERS

149Ks 1.Ss 10:29AN

FUROSENIIDE 209

PO

Chruonic RX ($9.1l4db) 7) FUJDSENIDE 0) HYDROCHLORDIIIAE

9) NETOLAZNE

Ir anUt! A nurrr, hlost & ftasLirt RX 19) FUROSENIDE 111J (mrore choices)

Figure 7: Gopher Order Entry Screen The same will soon be true at some clinics in the Veterans HospitaL The network service is provided by Novell's Netware. An Intel 386 base PC and the VAX are fie servers on this network. The network 386 PC file server holds a cache of medical records for the few thousand most active patients, and all of the hospital inpatients. A patient's medical record is fetched from the file server to the microcomputer whenever a physician is writing orders for that patient. The record is updated as results are produced and verified by the lab or entered by data entry specialists through the VAX. To improve quality, orders are reviewed against the content of the patient's record as the orders are being written [7]. The computer is "smart" about what it presents on the menu of options. The content depends upon the specialty of the entering physician, the ward location of the workstation, the patient, and/or the problem for which the order is being written. Information, such as the price of the item being ordered and the date the test or treatment was last ordered, is presented when the order is written. Phyicians also have access to patient information, eg. flowsheets and past results, and teabook information (eg., American Hospital Formulary Service monographs) on demand, while they are writing orders. The provision of information about a patient's past results and calculated probability of an abnormal result to the ordering physician, have reduced test ordering significantly [8,9,10]. Some of the lessons learned Medical computing is difficult. We have provided encounter forms and flowsheet summaries to Gynecology, Medicine, Pediatrics, all of the medical subspecialties, and other clinics for years. The acceptance of these reports had been grateful and immediate. The installation in Obstetrics was different. In contrast to the first mentioned departments, the Obstetrics medical staff wanted to capture almost all of the physician recorded data. They were already using a structured manual form to do tiis. Tlhus, our installation required a big change in the habitual process. So, success came much more slowly. It requird tailoring of the organization and content of the clinical reports and flowsheets, conversion to a laser printer (this had an immense effect on attitude), many changes in manual procedures, and the passage of time. Wehave been using computers to enter lab orders in thie clinic for almost 7 years. No problems. But, after two years, the ward order entr effort is still

uphilL Everyting is more difficult; tle number of orders written and their variety is greater than the outpatient service by 2 orders of magnitude: telephone orders, student orders (to be co-signed), and negotiable orders (transfer orders that take effect if the bed is needed by a sicker patient) add complicated wrinkles. User turnover is higher (a batch of almost 50 new physician/medical student users per 6 weeks). The inpatient study requires a 24 hour/seven day week operation. The average resident has finally reached acceptance, but not enthusiasm. To get to ftis stage has required redesign and re-redesign, as we discovered new operational realities, converted from monochrome 12 mgHz 286 PCs to flat tension color monitors on 20 mgHz 386 PC's, eliminated most of the special forms that house staff had to fill out. It takes the physicians longer to enter their orders with computers than without. However, some of this time is recovered by other time savings of the computer. The physicians favor keyboard over the mouse by at least ten to one. They favor color by even a greater margin. We continue the quest. The installation of systems in different organizations is always more difficult than we guess. Subtle influences and barriers stand in the way. We will discuss these at the demo.

928

This work was suepod a tmtheAgen for Health PRX#H 49 and HS-05626). Care Policy and Research

References 1. McDonald CJ: Protocol-based computer reminders, the quality of care and the non-perfectibility of man. NEJM 1976; 295:1351-1355. 2. McDonald CJ, Martin DM, Overhage JM: The Regenstief medical record system: 20 years' experience in hospital outpatient clinics and neighborhood health centers. Jnl of Ambulatory Care Management (in press). 3. McDonald CJ: Action-Oriented Decisions in Ambulatory Care. Yearibook Medical Publishers, Chicago, 1981. 4. Tierney WM, McDonald CJ, Martin DM, Hui SL, Rogers MP: Computerized display of past test results; Effect on outpatient testing. Annals of Internal Medicine

1987;107:569-574. 5. McDonald CJ, Mazzuca SA, McCabe GP: Diureticinduced laboratory abnonnalities that predict ventricular ectopy. J Chron Dis 1986; 2:127-135. 6. McDonald CJ, Hui SL, Smith DM, Tierney WM, Cohen SJ, Weinberger M: Reminders to physicians from an introspective computer medical record. Ann Intern Med 1984; 100:130-13 7. McDonald CJ, Tierney WM: The Medical Gopher -a microcomputer system to help find, organize and decide about patient data. West J Med 1986; 145:823-829. & Tiemney WM, McDonald CJ, Hui SL, Martin DK: Computer predictions of abnormal test results. Effects on out-patient testing. JAMA 1988; 259(8)1194-1198. 9. Tierney WM, McDonald CJ,Martin DK, Hui SL, Rogers MP: Computerized display of past test results. Effect on outpatient testing. Ann Intern Med 1987; 107:569-574. 10. Tierney WM, Miller ME, McDonald, CJ: Informing physicians of test charges reduces outpatient test ordering. N Engi J Med 1990 322:1499-1504.