4ED 208 897AUTHORTITLEINSTITUTIONSPONS AGENCYREPORT VOPUB DATECONTRACT - PDF document

1 4;ED' 208 '897AUTHORTITLE'INSTITUTIONSPO
4;ED' 208 '897AUTHORTITLE'INSTITUTIONSPONS AGENCYREPORT VOPUB DATE.'CONTRACT',NOTEAVAILABLE FROftDOCUMENT RESUMEIS 009 811Dlazeno Erica; Metzger *Jane:ilethods forEvaluating Costof Automated Oospital.Infbrmation Systems. Research.Summary Series..Little (Arthur D.),.Inc., Cambridge, sass.National Center for Heal* Services Research(DHHS/PHS),- Hyattsville, Md.DHHSPHS-81-3283; NCHSR-8I-17 .Jul 81*kB0A-TC-1-2.33-79-50b0i..52p.CAOH,t,NCHtR,* Publications and Information Branta, 37004'East-West Highway, Room, 7-44, Hyattsville, aD 20782(enclose a self-addressed envelope with request).EDRU PRICEMF01/PC03 Plus'Postage..Comparative Analysis; *Cost Effectiveness;*Evaluation Methods; *Health Serticies; Hospitals;*Information Systems; Labor; Methods; MultiplefRegression Analysis; Nursing; *Online Systems; Work-Sample Tests'IDENTIFIERS*Impact Studies; *Patient CareABSTRACT.----.6 .provide a compendium of methodologies on costimpac'tS of automated hospital 4iformation systems (AHIS), this reportsponsored by the National Center for Servres Research identifies,.reviews,- and sumiarizes ten studie; on inforsation systems whichmanage patient care data.The* studiesWere identified.by a litVcatureseardh and those4that included a description of- cost impacti wereselected for detailed analysis..-Most studies considered labor savingsthroggh Ircrkload reduction and/or time saved And labor- forcereduction. The saving of nUisin4 labor was the largest single costreduction-iUntified. Four' different techniques used for estimatingworkload changei

2 'were task analysis, job content analysi
'were task analysis, job content analysis, worksampling, and trend analysis. Othamethodological,issues indicate'that turning time. savings into labor force'reductiOns requires adeliberate effort to reorganize services and cost, and that benefitsmust be .evaluated.and docuaented in Selecting an AHIS. Also,regression analysis can, be used ,to study impacts of'AHIS4iiplementation retrospectively. kcomparison of three AHIS costlapacts, predictions, conclusions, and recommendations for an.approath to AHIS cost evaluation, and directions for future researchare included. Nine tables, 11references, and anextensive list of.current" National Center for Health Services Research publicat.i.ons areDESCRIPTORSptovided.----(RBF)-',IV._.:-....,..,.**************i*4******************************************************41.,ifeprauctions supplied- by EDRS, are the best that can be madev **from the original dodument..-.,*******************4****t****,1****************i******************11**** N:CNCOCOCDCVwl,..t o .RESEARCH SUMMARY SUES(\Methods forEvaluating-Costs ofAutomated.Hospital, 4Information,.SystemsUS. DEPARTMENT OF EDUCATIONNATIONAL INSTITUTE OF EOUCATIONEDUCATIONAL RESOURCES INFORMATION..CENTER (ERIC/The document has been reproduced asfeCerved from the person or organdabonong.natmg q%nor changes ha4 been made to .mproyereproduction qualityPoints of Yes.) or optmens stated al this document do not necessarily represent offickli NIEPOsaion or pokyErica DrazenJane Metzger,I,July 1981a r't?U S. DEPARTMENT OF HEALTH AND HUMAN SERVfCE

3 SPublic Health ServicesOffice of Health
SPublic Health ServicesOffice of Health Research, Statistics, and TechnologyNational Center for Health Services Research.,'DHHS Publication No (PHS) 81-3283f1 e.,This NCHSR Research Summary was,prepared by EricaDrazenand Jane Metzger, Arthur D. Little, Inc., Acorn Park,Cam -brige, MaSsachusetts. The projectwas supported by Basic Or-dering-Agreemen/ Task Order 1 233-79-3000 from the NationalCenter for Health Services Research. Copiesof the complete re-. port of the study are available froin the National Technical In-formatipn Service, Springfield, VA 22161 (tel.: 703/487-4650),and may be ordered as PB 80 178593 in eitherpaper 'copy ormicrofiche. Additional copies of thissummary may be obtainedon request from the NCHSR Publications sand InformationBranch, Room 7-.44, 3700 East-West.Highway, Hyattsville, MD20782 (301/436-8970). Current NCHSR. publicationsare-listedin the back of this publication.Office aliealthReirch,Statistics, and T hnology4National Center for Health Service3learchrGerald Rosenthal, Ph D.. DirectorRobert A Fordharn, Associate Deputy DirectorDonald E. Goldsto"ne. M D.. Associate Deputy Director.Medical and Scientific AffairsDivision of Extranlural ResearchArcher C Copley,rector (Acting)Gerald S Cohen. Md. Medical Information Systeriis ClusterRichard M DuBois. Php .Project Officer-The views expresiesi in this publication are those of the authors, and nonofficialendorse-ment by the National Center for Health Services Research is intended or should be in-ferred.U3 Foreword4The growing interest in the appl

4 ication of computertechniquesfor institu
ication of computertechniquesfor institutional and patient care m nagement has spawned avatiety of different system approach s. The cost implications ofthese approaches range from hundreds of t ousarids ofdollars to. severalmillion dollars for thelirtrvidual hospital. It is widely ac-cepted that other than facilities improvement (new construction,additions, major remodeling, etc.), automated data processingrepresenm.the largest single,capital expenditure facinghospitalsin the future.To use this expensive technology effectively, it is imperativethat the hospital community be able to not only projectthe costof a proposed system, but also be able to assesspotential sa#ingsso that the net cost (orsavings) resulting from the application ofsuch systems can be estimated: This study providesan.overview ofthe methods available for estimating and measuring the costim-pact..of automated hospital information systems. It willbe ob-'vious to the reader that no standard validated methods areavail-able for predicting the cost impact orthis technology. Thefind-ings in the report, however, provide the foundation for future re-search which shoulthultimately lead to the demonstrationandvalidation of methods which individual hospitals can use withconfidence.Gerald Rosenthal, Ph.D.Director47iiiL./7' OContents41 Introduction3 Studies, concerning,AHIS costs and costimpacts1L1MethodolCgical issues in evaluating AHIS costs31 Comiison of three predictions ofAFfIS cost impacts35 Analyzing revenue implications37 Conclusions and recommendations41

5 References4"V (IIIntroduction4Backgroun
References4"V (IIIntroduction4BackgroundI ,\,.- ,-Computerization offers the hospital a means of managing an in-creasing volume . of clinical and administrative information.,Since computerized hospital information systems were firSk usedin the 1960's, -numerous products offering varying scope, 'func-tions, and system configurations have been marketed; sOme hos-pitals have even developed their own systems. Consequently deci-sionmakers considering automated information handling arefaced with numerous options, each offering different benefits.Computerized information handling can be very costly. Acomprehensive hospital-wide systemone that integrates patient'N.records and reports for all hospital departments fora moderatesize hospital may have a purchase price of several million dollars.A smaller system processing information for one hospital depart-ment, though less expensive, will have more limitedimpacts onhospital operations. Thus, decisionmakers need to consider the'costs of the various forms of computerizedinformation handlingand weigh these against the benefits of each option.',.,The cost impacts of,Automated Hospital Information Systems(AHIS) are the changes resulting frOm automation that can beexpressed in terms of dollars. These impacts have been examined'in several studies. Some have been sponsored by the federal got,-ernment and others by vendors of commercial systems and indi-Nvidual hospitals. Since no compendium of the methodologiesused in these studies was available, the National Center forHealth Service

6 s Research sponsored a study to identify
s Research sponsored a study to identify and re-\ view these methodologies and to prepare a report summarizingand critiquing them. This report summarizes the findings of thestudy.*.,,iCopies of the complete report of;he study are available from the National Technical Information Service. The document number is PB 80 -178 -593.f6 .Objectives and Scope of the StudyThe study, had two basic objectives: to evaluate current methodsfor estimating cost impacts of AHIS and to recommend areas forfuture research Existing methodologies for studying cost impactsWere assembled, reviewed, apd summarized ina,format useful to.hospital administrators, health planners, and other tlecisionmak-.ers. These methodologies were analyzed and the usefulness ofmethodolOgies from other applications was also considered.The cost evaluation studies considered involved informationsystems that managed data used in patient carethose systemsdedicated to adNinistrative functions were excluded, as were sys-tems designed primarily for patient diagnosis and treatment(e.g., computed tomography scanners, automated radiotherapyplanning systems, etc.).ApproachAHIS cost evaluation studies' werewidentified by a literaturesearch, which included use of computerized abstracting services;a survey of vendors of systems per_forrning clinical informationhandling functions, a review of available informatn concerningapplications for Certificates-of-Need, and cotntacwithre-searchers in the field. Studies dealing explicitly with cost and.costimpacts were reviewed, and thos

7 e that included a description ofan evalu
e that included a description ofan evaluation of cost impacts were selected for detailed analysis.These studies include most of the published work in the field.2J,7.._..J btStudies conceroingAHIS costs andcost impacts.Ten studies.reprernt the currentState of the art of assessing costimpacts bf AHIS implementation and nkiSt of the.publishedwork. in the field (1. 10).. The studies art listed in Table 1,whichalso presents an overview of the pUrpose and approachof eachstudy.,Basic approachesTwo basic techniques have been.used in analyzingthe cost inrpacts of automated hospital information systems.In one method,the costs before system implementation andafter system imple-mentation are determined and compared in order toisolate thecost impact of the system. Inthe other type of methodology,_ costimpacts arepredicted before the systr is implementedbased on.baseline data concerning hospital operationS,-knowledge of po-tential,system impacts, and the experience of other users.Thistype of 'study is actually ndt a trueevaluation in that it is donepros ectively in anticipation of theimp,:icts expected in thefur re. The predictive methods areUseful for decisionmakerscoidering acquiring an automated kospitalinformation-*-ptem, but the retrospective type ofstudy that examines actual hos-pital.experience is necessary to provide data on actualimpactsfor use in developing or validating predictivemodels.711Before and after studies rely on measurements ofthe procest ofinformation-handling and on comparison of costs of certain as-pects of

8 hospital operations. Sincethe baseline
hospital operations. Sincethe baseline data required toevaluateimpacts are .usually not available from.hospitalrecords,two periods of data collection arerequired one before imple-mentation (to establish the baseline) and one after implementa-tion. Depending upon the scope of theevatation, this data col-lection effort can be substantial.Actual measurement of cost impacts hasusually taken place instudies in which the number of impacts -is small,the major im-pacts are easily measured andlied to AHIS functions, and the VSourceType of AHI4Arenson,-R L, and,Radiology OperationsJ W London, "RadiManagement Corn-ology Operations Man-puter System perform-agement Computera ing registration andSystem, Hospital ofscheduling, file roomthe University ofmanagement, patientPennsylvania," in Pro-tracking, reporting, ac-ceedingq or the Amen-counting and stet's-can College of Rath.ticsology Sixth Opnler--IV on Comptireiph anoncRath-- ology March 1979Table 19 Overview of AHIS cost studiesInstitutionspttingUniversity teachinghospital radiologydepartment perform-ing 130,000 diagnosticexaminations per y43arIPurpose_ofevaluationCompare annual costsof autoation and an-nual cost savingsachieved (focused ondirectly measureablecosts and cosrsavings)V,Study design. Cost factors/variables1Calculate annual costsfor system hardware,personnel, supplies.maintenance, and instal-lation2 Calculate saAgsachieved throughpersonnel reductions(workload reduced orfunctions eliminatedsuch as card filing,film library mainte-nance, and billing of-fice clerk)reco

9 vered pillinglosses (2.5%o lostcharges)r
vered pillinglosses (2.5%o lostcharges)rebuctiorrin papercosts (forms eliminat-ed)3Subtract savings fromcosts to yield net costs1Cost Factors'annual system hardware, software, main-Ptenarace and installa-tion over 8-year system lifetimeannual salaries (in-cluding benefits)annual supply costsRevenue recovery, as-sumed to have facevalue oLlost charges ?-Barrett, J P,P LHersch, and R J 'Cas-well Evaluation of themplementatton of theTechnicon Medical Infoilhation System atEl Camino HospitalPart II EconomicTrend Arialysis Bat-telle, Columbus, Ohio,May 14, 1979'Technicon Medical In-formation Systmhospitalwide comput'erized system proc-essing a broad rangeof medical and admin-istrative d5taEl Camino HospitalMountain View, CA464-bed general acutecare, non profit, short-term community hos-pital compared NinthGood SamaritanHospital, SantaClara. 315bed short-' term general corn-munity hospitalSequoia HospitalDistrict 4132bedshortterm, general,tax district hospital-Peninsula Hospitaland Med Ctr 343,bee, shortterm,community hospitalMills Memorial Hospital 301-bed shortterm, general com-munity hospitalValidate internal eco-nomic analysis at ECHNy a trend analysis'among ECH and four 'other California hospi-tals1Collect monthly hospitalkist and service dataover six yeaFs(1870-1975).2Create a uniform database by aggregationinto ntIrsing functions,therapeutic and diag-nqstcc ancillaries, sup.part services, all departments, and total hospi-tal, using direct expenses3 Expand sample of obser-vations by use of pooledregression'and estimat

10 eeconomic impact ot.TMIS4 Compare estima
eeconomic impact ot.TMIS4 Compare estimated im` pact with actual expertence at other hospitals72 variables tested relatingto administrative measures(occupancy, case flow rate,*personnel mix and turnover,units of service, laborhours per case and per patient day), patient characteristics (mix by age, rembursement type, LOS for 10selected diagnostic cate-gories, surgical proce-dures), and costs (per case.per day by service, perdiem rates, payroll ex-penses, prevailing wagerates) SosurceBrian, E.l's/, et alCost Benefit Analysisof Installation of aSpectra system atSanta Monica Hospi-tal. Unizersity ofSouthern CaliforniaCenter for HealthServrees ResearchMarch 21, 1975, Re-vised January 1976iType of AHISsHospital-wide mini-.computer-based sys-tem Sp4ctra 2000rTable 1. Overview of Atil'S cost studies,(continued)InstitutionsettingSanta Monica Hosptal, a 350-bed acutecare community hos-pitalPurpose ofevaluationDehirmine if comput-er-based informationhandling Is cost-effec-tive when costs arecompared with costsavmgs in charge cap-ture, forms replace-ment, and personnelreduction11Study designCosRactors/variables1Determine underchargeerror rate in patient bill-ings by random sam-pling of patient chartsand comparison withbills: project total revenue loss2 Calculate cost o manualforms to be repl ced1(could not be c mparedwith cos s foomputersystem supplies)3.a. Survey departmentsand select those whorepersonnel changesmight be significantb. By time-motionstudies estimate work-load reduction fromautomation.c. Estimate personnelcost s

11 avings (laborforce reductions) byp ?asin
avings (laborforce reductions) byp ?asing reductions tobe achieved from re-duced workload over 5years (30%, 50%, 66%,83%, 100% in the re,spective years)4 .Calculate cost to costsaving% ratio for eact.h.offirst 5 system years.Assumptions1Increased charge cqpture will 'result ineguivale'nt increases inhospital revenue2Predicted FTE workload .reductions acrossunits/departments andshine will result inlabor force reductions Cost-benefit Analyis:"of Data-Care System(unpublished). FpirfaxHospital System, FallsChurch, VA.DataCare, Inc hosprtal-wide computerizedsystem, processing abroad range of medi-cal and administrativedata for inpatients andoutpatiesntsFairfax Hospitala o'Determtnipresent1. Compute system costsfel? each year of system'lifea capitalize costs/feesfor hardware and soft-wareb phase costs forstart-up, upgrading,and internal systemsupport2Compute cost swingsfor each year of systtmlife,-a_ identify by functionanalysis, non-nursinglobs to be eliminatedand compute payrollsavingsb- compute savings forforms and current dataprocessiuc compule one-timecash flow beheflt forbilling office (reducedlag)3. Compute net benetit foreach year.4 Computwhnd sum pres-ent val.des (at discountrats of 6%, 9%,:and'teyield present- value 6f lifetimenetbenefitvalue of system costs,and savings over 8-year life cycle of auto-mated hosp,Itat infor-mation system1Prbsentvalues'of annual?costs, savings and netbenefits04'x. a.SourceGall. J. E. et ajDemonstration andEvaluation of a TotalHospital InformationSystem NatIonapeo-.1er for Health Ser

12 vicesResearch NTIS PB 282106. 1975tiType
vicesResearch NTIS PB 282106. 1975tiType of AHISTechnicoh Medical In.formation Systemhospital-wide comput-erized system proc-b essing a broad ratioof medical and admin-istrative dataIsSA.A4Table 1. Overview of AHJS a:4i studies(continued)Purpose of-evaluationInstitutionsettingEl Camino HospitalMountain View, CA464bed general acutecare, nonprofit, short-term comniunity hos-pital compared withsix similar nearby hos-pitalsCStudy designEvaluation of cost im;1. Predict labor savings bypact of ANISuse of Methods TimeMeasurement Modelcomparing manual vsAHIS pertormance ofclerical tasks2. Compare trends in nurs-ing labor and ancillarylabor haws actually ex:pended after 2 years ofAHIS with labor hoursexpected to be expend.ed under the manualmethod3. Compare trends in nurs-ing and ancillary labor.13at ECH and sixsimilarnearby.hospitals duringtwo 18month Bbspitalsbefore and after AHISimplementation and pre-dict difference at ECI-t.,.ICost factors/variables1Labor hours per month2. Nursing labor hours perpatient-day and ancillary& labor hours per admis.sJon3 Nursing labor hours perpatient-day and per atmission and labor hoursper ,dmission for se-lected ancillaries4 rHuff, W S and E. JBond Demonstrationof a Shared HospitalInformation System.The Sisters of theThird Order of StFrancis. Peoria, ILCustom-built systemfor three hospitalswith a shared com-puter facilityThree affiliated hospi-tals'evaluation car-ried out at only two.sites1'1Test five independent1hypotheses relating toquality, efficiency andcost of patient care bycomparing

13 selectedopeoting characteris-tics befor
selectedopeoting characteris-tics before and aftersystem implementa-2.tionControl other variablesbY eliminating changesin unit/department loca-tion, procedures, staff-ing, etc. during beforeand after experimentalperiodCost hypothesis. systemimplementation can beachieved 'without thgnifi-cantly affecting the costof patient careMeasure productiv0.of nursing staff(standard work hoursas a percentage of to-tal clock hours forRN, LPN, and trained'attendants) and of admilling staff by worksamplingMeasure re'duction incost of routine services by labchsplacement (standardMmes and volumesamplings) per pa-tient-day in nursingstations and admis-sionsMeasure reduction incost of radiology and.laboratory servicesby labor displacmentPercent idle time oftotal clock hours#hrs saved per patient-day timessalary/wagehrs saved per proce-dure timessalary/wage ..tTable 1. Overview of AHIS cost studiese(continued)...SourceType of AHISKim. S H. and H. H.(not stated)Sghtitz. 'A Cost-Benefit Analysis of aModular Hospital in-formation System," inM. W. Hopfe [ad IAmerican Institute forDecision SciencesProceedings. Boston,MA Nov. 14-16.1973.gaInstitutionsetting(not.stated))Purpose of.evaluation'Compare actual costf modular AHIS withpresent value of totalstream of net cashflow (resultly fromrecovery of lostcharges),1..1_ t.e1,Study design.Cost factors/variablea1Obtain revenue in).Assumptions.creases for first threecosts and revenue saving'syears iron) actual finanwill increase at 5%ciai data; project next 5years by use of 5% an-nual increase2 Obta

14 in additional costs(of system for first
in additional costs(of system for first year,project for next 7 yearsled on 5% annual in-creases3 Compute net caste flowchanges (gains in reve-nue minus additionalcosts) for each of 8years of system lifetime/4 Compute present valuesat 6%, 8%, and 10%discount rates and .sum-NPV for 8year systemPlife for each rate5. Calculate profitability in-dexes by dividing NPVby cost of AHISlif.. Mishelevich, D J ,W. G. Gipe, J R. Rob.arts, C Denny, A DStern and M W Day"Cost Benefit Ana lyists in a Computer-Based Hospital Infor-mation System," inProceedings of theThird Annual- Sym.posium on ComputerApplications in Medi-cal Care, 1979Radiology departmentordering, work management and reportingsystem developed byParkland HospitalParkland On-Line In.formation System(POIS)ls.University TeachingHospital (Dallas Coun-ty District Hospital)Dqtermine total de-partmental and unitcosts with and with-out the system for astandard patient load1By use of industrial engineering techniquesand compartmentalanalysis, derive costs ofan hypot4etical,averageproceduredefine average proce44dure.measure direct timeand labor for allsteps in proceduremeasure indirect timeand personnel timeby work samplingdistribute overheadover transactionVolume for standardpatient load. adjust-.ing post implementation volume to reflectavoided duplicateprocedures (3%)4Annual total cost of producing transactionvolume for standard patient toad'Assumptions1. Departmental overheadand x-ray equipmentcosts same before andafter2Clerical salaries increasedue to requirement for,data I

15 nput3. Physician costs exclud-ed (fee fo
nput3. Physician costs exclud-ed (fee for service)house staff costs in-cluded-r7. Souice..Schmitz, H. H:, AnEvaluation of the im-mddiate Financial Im-pact, orthe HospitalInformation System atDeaconess Hospital,p 265-282 in IR, CKozo led.) Health In-formation Systems%Colorado AssociatedUniversity Press flout.,der, Colorado, 1874Type of AHIS(not stated)..,r.'Table 1.a,Overview of AHIS cost studies;continued).histitutio%-'4."Isetting ``Deaconess HospitalSc-kpuis, MO ,N.4Purpose ofevaluationDetermine if any sig-nificant changes inrevenue and volumeindicators fpr the hos-pital changed as a re-suit of AHISr-acStudy designCost factors/variables-1Calculate direct expenseper patient-day beforeand after.compute non-roomrevenue per patient-dayeliml'nate effect of in-flation2 Calculate number oftests performed per Oa.tient-day before and al. "ter to test for changesother than recovery of"lost" charges3Evaluate possible ex-planations for changes-171Annual system hard-wate, software, mainte-nance and installationover 8-year system life-\ time2. Annual salaries (includ-ing.benefits)3 Annual supply costs 71F4titwTechnicon Cost-Bene-fit Analysis of theTechrucomMedical In-formation System(MIS) For "X" hospital..December 1975.Hospital-widez(Methodology may beapplied to any setting)'13-Project savings in#labor, forms, equip-ment Over first 5 pperational yeirs of PAIS.1Calculate labor time sav-ings by Job title per unitof work2 Multiply traffic d;ta(dilly volume) times unitsavings to yield FTElabor reduction for allbut nurstn?Jabor3 Obta

16 in `realizable workforce reduction byrou
in `realizable workforce reduction byrounding down for ancil-lary departments-andconsidering rob shifting -and other measures forother departments.4 Calculate labor savingsperyear based on incre-mental changes in staff-ing patterns, attrition,.etc including expectedsalary increases,5 Calculate cost avoidancefor nursing labbr by re-gression based ontrends in nursing.laborand assume trend to in-crtases will be arrested(based on experience atEl Camino).6 Estimate cost sayingsfor eliminated forms.1., Minutes per task2. Minutes pef day3, FTEFltE increase avoidedbased on nursing hoursper patient-clayefi.Sp rday impacts will be realized in a short time (so that the influence ofother variables is minimal) Therefore, most of these studies thatmeasured cost savings havebeen conducted for departmental sys-tems. The study by Arenson and London (3) is an example of thistype of approach.The two, studies at El Camino Hospital (1, 2)-and, the evalua-tion at the hospitals of the Sisters\f the Third Order of St. Fran-cis (6) are examples of attempts to determine impacts of a com-prehensive AHIS by measurement and comparison of a widerange of information handling actisities and costs before and af-ter system implementation.Methodologies fdr predicting the potential cost impacts ofautomated hospital information systems have been developed toexamine both comprehensise hospital information systems anddepartmental systems. Most of the predictive methods examinethe personnel and supplies (mainly paper forms) impacts andcompare these with ti

17 n system costs. Examples of this type of
n system costs. Examples of this type ofstudy are the Technicon methodology (10) and the study byMishelesich et al (8). Most $redictive studies consider one-yearsystem costs and cost sasings.Three of the studies listed in Table1 include,a multi year cost analysis..the Techilicon methodology(10) includes the first 5 years ofsystem operation and the FairfaxHospital (5) and Kim and Schmitz (7) studies Aver life cycle (8-year),,c8st impactsPurposes and perspectivesVaried purposes in implementing an AHIS and in undertaking acost study.and varied expectations regarding cost and other im-pacts are reflectedin the ten studies identified. Examination ofcost impacts was thy focus of studies at Santa Monica Hospital(4), and Fairfax Hospital (5)amonVotherswhereas quality ofcare benefits (most of Which are difficult to expres,k in dollars), aswell as (lost impacts have been evaluated in the El Camino (2)and the Sisters of the Third Order to St. Francis study (6).40.Characterrs of AHIS and settingsSystems for radiology departments are examined in the studies byArenson and London (3) and Mishelevich et al(8). The otherstudies included in Table 1 consider hospital-wide automated in-formation handling. In the case of the systems at Santa Monica1419 Hospital (4) and Deaconess Hospital, St. Louis (7,.9), the systemconfiguratton is a modular or distributed one. Commercial sys-tems -are involved in Ihe.Technicon (10), Fairfax Hospital (5),and Santa Monica Hospital (4) evaluations; whereas the systemsstudied at El Camino Hospital (1, 2) a

18 nd the hospitals of Sistersof the Third
nd the hospitals of Sistersof the Third Order of.St. Francis (6) were early versions of sys-tems.that are now commercially available.In addition to system characteristics, the work to date reflectsindividual dePartment (3, 8), community hospital (1, 2, 5, 6),and university teaching hospital settings (4, 7, 9). Except to thedegree that the institutional setting influences the informationhandling needs, the type of cost evaluation conductvikand thespecific methodologies employed are determined more by the ob-jectives for implementing automation and the purpose for andperspective of the evaluation than. by the characteristics of thespecific system or hospital setting. Therefore, this review of the _studies has been oriented toward specific methodological issuesand the generic approaches to analyzing cost impkts found inone or mare of the evaluation studies cited.Practical considerationsAll work on AHIS cost, iwpacts is limited by the resources avail-'able for the study. Ther?fore, most studies are focused on themajor impacts for which sufficient data can be obtained.4.-2015 . Methodological issues in evaluating AHIS costsEstimating cost impacts of labor savingtHospital information handling is very labor-intensive. Clericalfunctions have been estimated to consume 25% of hospital stafftime and as much as 18% of nursing staff time. Therefore, laborsavings, achieved through the reduction of clerical workload, area major economic rationale for automating` information 'han-dling. A hospital realizes direct cost Savings, howev

19 er, only when,*the savings in time devot
er, only when,*the savings in time devoted to clerical work are reflected in labbr,force reduttions. Therefore there are two, aspects to the cost im-pacts of labor savings:(1) Workload reduction and/or redistribution(timelsaved) and V;(2) Labor force reduttion (personnel eliminated) and conse-quent payroll reductfilo..Most of the studies done to date Piave considered labOr savings,and savings of nursing labor have been.theIargest single cost re-ductions identified (or predicted) in most of these.Four different techniques have been ennployeefo'r estimatingworkload changes:(l) task analysis (2, 4, )3,which considers time savings inindividual tasks by use of industrial engineering techniques,such as Methods Time Measurement;(2) job content analysis (5) that estiyiates time savings for.jobssubstantially changed;(3) work sampling (6) which involves obserNion of actual'time devoted to tasks; and(4) trend analySis (10)which is based on historical trends inproductivity and anticipated changes.In task analysis, individual work activities that will be affectedby automation are identified, e.g., filling out an X-ray requisi-tion, telephoning for a laboratory result. Then the time spent inthe activity in the manual system Is measured. The time spent in16-21. 'itthe activity with automation is measuredstudy) or predicted (in a prospective evanon is combined with information onthe salary of personnel performinglabop saviNs Table 2 illustratesprovides a complete inventory ofing time savings in specific tasks tocult. as will be disc

20 ussed later.(in A retrospectiveanon). Th
ussed later.(in A retrospectiveanon). This informa- ''.----.e volume of activity ande task in order to CalculatehiS teOnique. Task analysistentill impacts but tra slat-eductions in payroll isiffi-.0_.-Job content analysis overcomes some og the problems of tasiWanalysis since it focuses on the emplpyee, rather than the task.The steps in this process include identifying jobs that will be per-formed differently, w.ith AHIS, estimating (or measuring) thetime that- with be saved, and redefining (combinin,g) jobs torealize the savings, Table 3 shows the results of a job contentanalysis.Job content analysis is usually performed in collaboration withthe supervisor of a department, sincethorougleknowledge of thearea is required. and since supervisors must agree toKnplementany changes Job content analysis is less time-consuming to per-form than task analysis and results are easier to translate intorealistic staff reductions. It may also be less accurate since judge-ments of time savings are used and there may be incentives toin-flate or underestimate savings..Work sampling (also referred to as activity sampling) can alsobe used to estimate labor savings. In this case the work performedby individuals is documented by measuring periodically the ac-tivity of personnel in a unit over an extended period. The data'are used to describe how individuals orecategoriesof seaff spendtheir time: Then the activities affected by AHIS can be identified,And fhe potential labor savings can be estimated fin predictivestudies) of measured, Oh ret

21 rospective studies). Since work sam-'pii
rospective studies). Since work sam-'piing provides- information on how time is allocated to work ac-tivities, it can be used to determine .how activities have changed(use of freed-up time) and howrduch "downtime" exists in man-ual and automated systems, etc. Table 4 provides results of worksampling of information handlingctivis in a surgical inten-sive care unit.Work sampling is especially useful in analg labor impactsin nursing units. Nursing staff perform many information han-dling tasks in conjunction with other patient care activities butmay not devote large blocks of timesolely to information han-dling. Sometimes information handling and patient care tasksare performed simultaneously. Therefore,changes in nursingstaff activities can be best captured and quantified by direct ob-2217 aTable 2. Example of task anIs for admitting departMentFunctionJobtitleTimesavingsVolume(min(unitsper unit)per day)FTEreductionBasis for time savingsProcess PreAdmissionAdmitting Clerk#8.026t,0.6Bed Aisignment forDept. Head (60%14.0310.4'Daily AdmissionsAssislant Dept. Head4,0210.2.2dmitting Clerk1.0510 2Sub-Totalj0.8Process EmergencyAdmitting Clerk,50,17.0.3AdmissionNProcess SOtleduledand Urgent Admissiondmitting Clerk.5.0520.8g)9,Preparation,of Admission Record and typ-ing and filing of admission form,replacedby direct, type-in of reservation/pre-admis-sion data on VMT from phone call ormailedin Preadmit form. Case Card Fileeliminated. Simplifies entry and release ofpreadmit orders and screening...liminates Maintenance of Bed Avail

22 abil-ity records and preparation of Bed
abil-ity records and preparation of Bed Cards.(Replaced by Scheduled admit list and bedassignment sheetr--,Eliminates transcriptionofAdnirsIondata .io Admission Form. Faster bed as-signments. Eliminates phone calls fromEmergenciand checks of Case Card File.Automatic assignment of new controlnumbers. MIS provIdeti on-lin&case # file.Eliminates distribution of Admit Records.Eliminates distribution of patienOinforma-tion to Information Desk and.Switcbboard1,.and other patient lists't. /rlt-.0PProcess fiansferlDept. Head (60%)3.0100.17---.1Assistaryt Dept. Head3.070.1SubTotal\-0.2Prpcess DischargeAdmitting Clerk3.0.0t2d.4Preperation of AdmissionDept. Head or Asst.15.01"'0.1Lists of Daily Admissions,Transfers, DischargesDEPARTMENT TOTALDept. Head0.5IAsst. Dept. HeadAdmitting Clerk----------1,--.0.42.3t3.2Eliminates preparation and distribution ofTransfer Slip. Eliminates making new bed. assignment on boardrEliminates preparation of Discharge No-.t ice and distribution of same,Preparation eliminated. Computer automatically prints'list.Source Tecbrocon Mbdical Information Systems Corp . Cost-Benefit Analysis of the Technicon Medical In formahoh System (MIS) for Unnamed Hospital) December 1975a,.NeAil 1..., '..1-.1."6.-..-._ aTable 3. Example of job content analysis for a radiology departmentAl lected,Saved hour'JobtasksPer dayTechnician 1 ..perform exams.0Technician 2...perform exams0Technician 3....scheduling, perform exams1Technician 4. ..perform examsletrieve files,3ClerkfClerk 2-Clerk 3..Clerk 4,Clerk 5,transcription of

23 reports7scheduling/phone.4phone/xFay fi
reports7scheduling/phone.4phone/xFay filing..)6 'phone/schedulineport distribution2phone/requisitiorocesbling/receptioni%Source Arthur 0 Little, Incservation (as in work sampling).Trend analysis involves examining trends incost variables(i.e., nursing cost patient day) priorto AHIS with hose afterAHIS to identify and quantify changers. The techniqe is usedmost often in retrospective studies but can also be usedrospec-tively if throe is inforAtion available to project trends. P form-ing a trend analysis to estimate labor savings involves fit-Itcing ,historical data on trends in labor productivity andextrapo-lating these trends into the future. In retrospective studies,pro-.ductivity after. AHIS would be plotted, in prospective studies ftture productivity must be estimated. Once -the trends with an d\414.without the AHIS have been plotted, then the,\rwo productivity\lines are compared in order to estimate labOructivity differ-ences at differpm points in time. Table 5 and Figure 1 show an\*example of trend analysis of'productivitypf nursing laboron; nin pa den tat ni tResults from trend analysis are difficult to interpret because,changes in labor cost or productivity can be the result of manycauses other than changes in information handling tasks. Majorinfluences on staff prodUctivity'may result from changes in thepatient mix, the occupancy rate,regulatory and paymentpolicies, or the addition of new services (especially intensivecareservices.Identifying and adjusting for these other variables toisolatthe impact of an A

24 HIS is a difficult undertaking. Further-
HIS is a difficult undertaking. Further-.the_large number of institutional variables make ifirnpos-sibleto generalize the results of trend analysis performed inoneinstitution to other hospitals.An overview of the steps involved in estimating workload andpersonnel reduction for each methodology is presented in Table6, along with the advantages and disadvantages of eachap-proach. At the present time, the most efficient approachappears,20 VTablet Results of nursing staff work sampling in aeurigical Intensive care unit4e,CadeActivity\Ilours per shirtHours'rpatientDayEvening!NightDayEveningNightTotal11Patient Measurements,t40,8'0.50,20.1'0.10.4.,,(Monitor Type)32Non-drug Ordering,0.80.41.1'0.10.10.1,0.333Drug Recording0.100.134Nondrug Recording4.93.04.40.60.30,645Review Patient fiecorqs3.9'3.43 30,440.40.452Communication with Nurses4.13.60.50.52.11"53Commymication With Doc,tors3. h1.63.10.4,0.4i .01.71Research--72Te4ching or Education4.250.30.50.10.6Total28.2.13.616.53.21.52.16.8.,Source Arfttur D Little, inc , A Revert 01 theMediab System in the Surgical Intensive Care Unit at the Massachusetts General Hospital, Appendix A Reportto the National Center for Health.-Services Researchend Development. Contrict HSM 110-70-406, March 1913,...4t(T Table 5. Example of prediction of avoided increases in nursing costsInpatient serviceYear endingDecDecDecDecDecMedical/surgical0 03 fetlIC.79777978197979807987Nursing hours per patient day:iv-without MIS,r7.697.908.108.318.51-with MIS .......... ................. 7.607.607.607.607.60--

25 increase aillrfance with MIS0.090.300.51
increase aillrfance with MIS0.090.300.510.710.91Med/Surg. patient days per month,9,6989,4859,2728,846Med/Surg. nursing hours increase avoidance with MIS8732,8464,7299,0598,060Full Time Equivalents5 0416.4227.286,432Average salary, including fringe.$1,432$1,53837.11$1,774$1,33311,652.-Meci/Surg. cost avoidance.$6,718$23,513$41,969$61,305$82,404Nursing hours per patient day:;.'-without MIS,8 809.72'10 6411.56.12.48. -with MIS,,8 428.428.428.428 42- increase avoidance with MISr0 381 302.22.,3.144.06Obstetrics patient days per month'730699669638608Obstetrics nursing hours Increase avoidance with MIS`2778271,4852,0032,469Full Time Equivalents..1 604.778.5711.5514.25Average salary, including fringe...$1,333$1,432$1,538.$1,652$1,774'Obstetrics cost avoidance$2,133$6,834$13,1,79$19,094$25,274..-----J..f NewbornDec1977Dec.1978Dec. z-1979'Dec1980Dec1981Nursing hours per patient day:-,-without MIS.9.309.73,10.4811.2311.98-with MIS.,8.988.988.988.98-irtdrease avoidance with MIS.321.071.822.573.32NewbOrn patient days per month,08578578547517Newborn nursing hours increase avoidance with MIS1956181,0521,4061,716Full Time Equivalents1.12'3 56-6.078.119.90Average salary, Including fringe$1,333$1,432$1,538,$1,652$1,774/Newborn cost avoidance$1,493$5,110.$9.336$13,401$17,5700Summary - all services4Full Time Equivalents- Medical /Surgical5 0416.4227.2837.1146.45-Obstetrics.,1 60".4.778.57.11.5514.25-Newborn .1 122.686.078.119.94-Total'7 7623.8741.9256.7770.60Cost Avoidance)._-Medical/Surgical$6,718$23,513.$41,969$61,305$82,404-Obstetric

26 s2,1336,83413,17919,09425,274-Newborn1,4
s2,1336,83413,17919,09425,274-Newborn1,4935,1109,3313,401-17,570-Total4...$10,344$35,457$64,484$93,800$125,248Source Techrucoe MedICSI Infomitten Systems Dory , Cost-Benefit Analysis'Ol the TechnIcon Medical In lormatiog System (MIS) for Unnamed Hosplta0 December 19753B to be using job,content analysis for non-nursing labor andworksaiepling for nursinglabor.Several questions must be answered before thepotential. work-load reductions can be translated into realisticestimates of laborforce reductions:What is the minimum staffing in eacharea that will be re-quired independent of the clerical workload?Can time savings within a unit,be realizedacross shifts?Can time savings within a unit be summedacross labor cate-gories?Can time savings be realizedacross nursing wards?Can savings of partial FTEs be realized?val'vON°°°Increaseavoided(Productivity savings)Predicted AHIS trendlineIII'.197519761977197819791980198119821983YearNote Trend lines are not based on actual data and are presented only to illustrate theuse of trendanalysis to estimate cost avoidanceFIGURE 1. Sample trend analysis of nursing productivityon medical/surgicalservice for use in study of laborcost avoidance2 9 flTable 6. Overview of methodologies for predicting cost impacts of labor savingsTask analysis1Identifyinformationhan-dling tasks2Measure time spent to per-form task by each labor category3Estimate time required toperform task with AHIS andthe labor category involved4Calculate time difference foreach task by labor category5Document volume of tasks6Mul

27 tiply time difference fortask by volume.
tiply time difference fortask by volume.7Converttimesavingsfortask into FTE labor cats-gores8Sum FTEs for all tasksJobconlent analysisWork samplingTrend analysis1Identify job positions withworkloadtobe changedsubstantially2Estimate amount of time de-votedto the affected tasks3Estimatk time savings byjob categoryI1.Measure staff time spent inr3erform !Fig tasks by activitysampling2.PredicttimerequiredforInformation handling tasksafterimplementationofAHIS3Estimate time savings byjob category1Collecthistorical data ontrends in labor productivity2Plot trends and extrapolateinto future3.Estimateproductivitychangeafterimplementation of AHIS4. Plot trend line for post-AHISproductivity'5 --Calculate productivity differ.encebetweenhistoricaltrend line and AHIS trendlinb at specific times in thefuture9.Identify staff reductions10Calculatepayroll savings. foreliminated jobs.4Realign jobs and alter staff-4ing patternsfor adjustedworkload55Calculate payroll savipgs forteliminated JobsCorrelate time savings loathstaffing changes.Calculate payroll savings forelfrninated jobs.6Calculate payroll difference(cost avoidance) based onproductivity savings.'Complete inventory of potentialimpactsCostly and time consuming:-Less costlytand results easier totranslate Into realized labor savingsaLess accurateProvides "snapshot" of how timeis spent before and after automa-lionCostly and time consumingEasy to applyLack of productivity ?landards forvarious labor categthies make accurate laborsavingpredictingdifficult Some of the'published work to

28 date on AHIScost impacts hasbeen based o
date on AHIScost impacts hasbeen based on unrealistic assumptions aboutrealizing the poten-tal labor savings and this has leadto overstated cost impacts.To illustrate the importance of thesequestions, cobsider thedata presented in Table 7. Heresavings in minutes of timeperday have been convertedto FTEs by dividing by. production workminutes per day (adjusted for vacations,holidays and illness) andsummed for each nursing station, shift andjob category. A quiteconservative estimate-will be obtained ifone rounds savings downto the nearest whole FTE for each laborcategory within eachshift and unit. This wouldmean that in two wards no savingswould, be predicted, and onlyone or two staff reductions wouldbe counted in the remainingareas. A total of 17 saved FTEswould be predicted. Alternatively, ifone assumes that all partialETE savings can be realized, then.,theseprojections could becon-verted diDectly into labor foice savingsof 4.7.4 FTEs. It is easy to)nsee that the treatmt of partial FTEs is critical to the resultingestimate of.realizab e labor savings.-The real savings to be realized from theexample in Table.7probably lie somewhere between 17 and47.4 FTEs. For ex-1 Ymple, ifone were to take a conservative approachto convertingtime savings to FTE staff reductions,one might:'Assume savings can only be summed withinone department.Assume transfer of tasks from lowerto higher level staff torealizee savings: RN's taking on functions previouslyper-formey aides or clerks, technicians replacing clerks in ancil-lary departments.

29 For nursing, departmentsassume small ,am
For nursing, departmentsassume small ,amounts of time sav-ings (less than 0.2 FTEs)can be transferred between shifts.Assume that 50% of the time savings of 0.5FTE or more canbe realized by use of part-time staff. (In'some cases savings won'tbe..possible becausc ofcore staffing requirements, e.g., at least..one ward clerk must be on duty oranswer phone).,,Round remaining partial FTEs downto next lOwict wholeFTE.,.When these assumptions are appliedto the workload savingsestimates from El Camino Hospital,the resulting estimate ofrealizable labor saving for the nursing unitsin Table 7 is 3FIFTEs(Arthur D. Little, Inc., estimate).26v1 Table 7. Example of potential labor savings for nusin/servicesEstimbted labor savings IE TE)NightDay shitEvening sh tshirrWardRNILVNIWardStationClerkAideClerk1 SouthMaternity461 234---- 781.6723402 East.1.041 37,42Illp2 West1,001 3244Pediatrics8088333 West951.A7404 West1 01.1 31445 East.1 20'1 66'535 West921.20436 East.901 23386 West811.11.344NILVNIRN/L VN/AideAideTotal66.51'2.641 00 \1 665.9693.824 5885814.42.60.76,3 371 00844.66.91.884 55,%,,1 i5945.38764 0873:4.057662,371,L47 40Source Gall J E ri pi ora,,,naraaan and 1. ,aluazion ala 1 ,aal Ihopual 1.11arrnanon SwirlNational Center forHealth lama rs Itnreart h N 115 PH 262106 1975Realizing labor savingsThe automation of informationhandling and communication01through an AHIS offers many potential labor-Sing benefits tothe hospitEliminating entire functions (e.g., preparationofbilling ch rge slips) often means that easilyidentified jobs can

30 beeliminat(e.g., billing clerks). Howev
beeliminat(e.g., billing clerks). However, the bulk of theclericallabor savi gs are moreJikely to oteufpiecemeal across manyde-partments, shifts, and labor categories.Turning these time savings into labor force reductionsrequires.a deliberate effort toreorganize services by 'undertaking `workmethods improvement, redistributing jobtitles, and redefiningjob roles among various labor categories.Without an activebenefits realization effort, activity displacementmay.result in in-creased idle time for hospital staff rather thanin payroll savings.Also, labor benefits following implementationof anAHISareonly achieved slowly as the staff learns to use the systemand tikeadvantage of its labor-saving features. Actual savingS arerealizedin increments and maximum savings are notlikely to result forseveral years.Choosing an economic analysis methodology.,1In decisionmaking about acquiring an AHIS, costsand benefits(cost savings) should be evaluated as tkey would befpr any majorinvestment. Typically, a hospital first incurs a large initial capi-tal cost and -only realizes economic benefits, incretnentallyin later .ar3227 LyearsPresent value methodology applied,to a life-cycle costanalysis proides a useful analytical framework. for -valuingtheyanticipated stream of economic savings (net benefits)over theuseful lifetime or the system. It employs discountingproceduresto reflect the fact that.benefits-received in futureyears are worthless than the` same amount ofmoney today. Thus, it permits acomparison of expected future befits andpr

31 esent expendituresthat incorporates the
esent expendituresthat incorporates the time valueo money.Selecting the appropriate discountrate is one problem inapplying the present value technique. DisCountrates used rangefrom 2 5% to 120the larger the discount rate, the lower thepresent value.of future moneys. Because there isno consensusamong economists as to which rate to use, severalare often usedand the results compared. Life cycle analysis requiresan assump-Lion abOut the useful lifetime of the AHIS. Althoughcomputersystems such as AHISare continually maintained and enhanced,at some point the requirements on the system and the availabilityof more cost effective systems make them obsolete.A system lifecycle of eight yeaq has been used inmost AHIS cost evaluationsand seems -to strike a reasonable balance between the timephas-ing of new computer generations and the investmentrequired toimpletnent a new system.Calculating system,costDocumentation of system costsappears at first to be- a simple,straight forward task. Indeed most previous studieshave devotedlittle attention to this area, using onlysystem purchase price andinstallation cost as measures ofsystem costs. Actually total systemcost involves a large number of components:manpower required for selection and planning activities;initial hardware Cost;initial software cost;facilities renovation to accommodate thesystem;cost of money to finance the installation;installation costs (vender charges and personnel supplied byhospital);hardware and software modification's or additions (by vendororin- house) both d

32 uring implementation and subsequentlyove
uring implementation and subsequentlyoverthe system's lifetime;28r)(-1ki P0.0"supervision of system installation and start-up;training of staff in AHIS-Use andtime spent by users in learn-ing the sytem; and-support personnel to managethe system :and interfaCewithusers;cost of special supplies, .supportservices, overhead, and otheroperational expenses.Thus a careful accounting of system costsmust include expen-,..ditures for services and equipmentobtained from a system orservice vendor and the investment4n,in-house personnel requiredto install andutilize the system. When system costs arebeingcompared with labor savings, the costof a benefits realization ef-fort to bring about the necessaryoperational changes needs to beincluded as well.-Using regression analysisOne approach to studyingimpacts of AHIS implementationretrospectively is regression analysis. Inits simplesi form, the sta-tistical technique uses time and .trendlines for various cost or.productivity indicators (suct as nursinghours per patient day).These indicators are calculatedfrom historical data points,which are used to extrapolate whatthe indicator would havebeen without AHIS. The extrapolateddata are then comparedwith acues for the comparableindicators after the AHISisuse. Morecomplicated forms of regressions include com-parison of trends across comparablehospitals and multiple re-gression analysis, which attempts toinclude in the equation themany-independent variables, in addition toimplementation ofan AHIS, that mayinfluence the dependent variableof i

33 nterest.Simple trend analysis does not a
nterest.Simple trend analysis does not accountfor confounding factorssuch as implementation ofutilization review or change in casemix of the patient population;multiple regression, which canhandle these factors., is more difficult toapply.A number of factors makeregression analysis difficult to use:In multiple regressions, measuresof important independentvariables may not be available.41IP'.The results of regression analysis maybe biased by variablesthat are unknown or omittedfor practical reasons. Variablesmust be selectedcarefully and interactionsunderstood or the re-sults may be misleading.%-0 Aki "1 Seasonal effects, which can be important, requireseveral yearsof data both before and after AHIS implementation, in ordertocontrolfortheirinfluenceonthedependentvariable(s),,examined.Although comparisons across hospitalscan help control for theeffects of outside influences, inter hospital comparisOnsare ham-pered by the difficulties of easuring all the factorsand inherentdifferences between hospitals.Nevertheless, when applied-carefully, regression analysiscanbe useful in evaluating the impact,of AHISon hospital costs, andin validating the results of industrialsengineeringstudies. Regres-sion analysis is used in some previous studies (1, 2).However, re-gression analysis can be costly interms of its4pe'cial data collec-tion requirements, and because of the difficulty inunderstandingthe underlying relationships between -variables,itprobably.should not be the only analytical technique used.Including quality of care benef

34 itsThe motivation for implementing an AH
itsThe motivation for implementing an AHIS is rarely only-costsav-ings. Improved information handling can result in benefitsto 13a-tiept 'care that are difficult to translate into economicterms.Benefits such as improved response times for communicationsand increased availability of nursing staff time for direct patientcare have been documented at all sites where they have beenstudied (2, 6). There are indications that AHISmay shortenlength of stay by shortening turnaround time for ancillaryserv-ices (1). The cosC savings imialkatiOns of changes such`PaS thesehave not yet been tied conclusively to automated ititormitIonhandling.30 ../0Comparison of three predictionsof AHIS costimpacts'Three predictive cost studies of comprehensive AHISare availa-ble for community hospitalsof similar size (400-600beds): thestudy by Gall et al. at El Camino(2), the Technicon study at anunidentified hospital (10), and thestudy for Fairfax Hospital (5).These were compared in order toexamine the consistency of theresults and to determine howdiffering assumptions affect thepredictions of cost impacts.A comparison of predictedtotal annual and per-patientday,savings is presented in Table 8.In order to make, the studyresults..more comparible,the figures shown do norinclude estimates ofsavings based upon uniqueand less accepted assumptions.Theestimated savitigs"dueto' "work ometliOdsimprovement" at El, .Camino were expected toresult from an intensive programtorealize nursing labor. savings.Because these savings are not,di-rectly atttibu.table t

35 o theAHIS alone, they are notincluded. T
o theAHIS alone, they are notincluded. The"nursing cost avoidance" estimatesin the Technicon studybased.upbn trend analysis andthe hypothesized stabilizationof nursingproductivity following implementationof the AHIS are also notshown. Note that the Fairfaxstudy did not includenursing sav-ings beciuse the systemCould be cosi justified on thebasis of say-'ings in non-n rsing laborand the authors believed thatrealizingnursing labor avings wouldbe difficult...,-..Table 9 pre ents the estimates costsavings in the ancillary andsupport areas,by'indiyidual departments for thethree hospitals.As indicated, not all threehospitals ex+ined savings inallareas.---)F...../,.o'3654, 1Table 8. Comparison of predicted annualand per patient day savings from three prediCtive studies of AHIScost ImpactsSavingsAnnual,Per patient dayTypeo/sawngsEl Camino(75 -'79)Technic on('77-'81)Fair lax'81)El CaminoTic/tit/conFairfaxNursing personnel...$ 807,000604,999 '$NA$ 6.564.84$ NANon-nursing personnel..209,616,..350,765533,1071.712.812.80Total 6bor.......$1,016,616$ 955,764$533,1078 277.65Equipment....$7,044$13,788$ 50,8480611.27Forms..'-57,73275,21620,800,'.47.61.11Materials and other.13,38092,1924..11.48Total nonlabor$78,156$89,004a $10,840.6472.86T QTAL$1,094,772$1,044,768$696,947 '$ 8 90$ 8 36$ 3.66°A'Cltie;not intlude nursing savings.---',Work Method Improvement not in eluded"Cost AvOidance" nainclugedgra_e 4' e$-%'Table 9. Estimated ancillary and support cost savings from three predictive studies of AHIS costs..-.....44Areat.Total annual sav

36 ingsEl Crimino%Tecenicon%Fairfax%Inpatie
ingsEl Crimino%Tecenicon%Fairfax%Inpatient registration /admission .0Laboratory .Medical recordsOutpatienUER/chnicsPharmacyX-RayUtilization review......BuslnesS off ice/aCCoirictir-4Central service ..,,,.eI,,$ 62,14237,80912,73453,83210,93035217316$ 46,47646,47630,984117,22317,34243,423155515.50,10396.14-$ 39,500-139,52496,530490,700199,14579,24837,900-620141329126TOTAL,$177,445100$301,924100$682,547100"....s4) Qsi LIC.___),ea After eliminating the savings basedon questionable assump-tions, the estimated cost impacts for the hospitals consideredinthe studies, are approximately $1 millionper year, equivalent toabout $9 pef patient day. These impactswould increase withtime, principally because of inflation(estimated at 7-8% peryear)Since personnel costs escalate .faster th,ansystem costs(which_tend to be more fixed by contract),the comparison of sav-ings versus AHIS costs would also tendto improve over time,These projected savings haveto be compared with system costs inOrder to estimate the netcost impacts of AHIS implementation.The majority of the predictedcost impacts (over 90%) come.from potential savings in personnel.Such savings amount to ap-pr.04imately $8 per patient day,.. Of thesesavings, the majoriy(60-80%) is attributed to reduction of nursingpersonnel require-ments As discussed previously in this report, achievementof thisreduction is dependent upona significant effort directed io'achieve the savings through reorganizationof nursing staff tasksand staff schedules, so that the personnelreductions do takeplac

37 e.The non-personnel savings are consider
e.The non-personnel savings are considerablymore modest, un-der $1 per patient day. Reduction of forms in thehilspitals cancontribute about $.50 of this reduction.The ancillary department with the consistentlygreatest savingsCvas pharmacy, 'which accounted for between 21-39% of estimated ancillary savings. The nextmost important departmentwas inpatient registration/admitting, which accounted for be-tween, 6% and 35% of ancillarysavings. These were followednext by laboratory with 15% and 20% of savings* and radiology,with 6% and 12% of the savings.AThe savings shown in Tableg 8 and 9 should beconsidered in-.dicative of order -of- magnitude estimates only, sincethey apply todifferent, time peri-mt-,--andwere based upon somewhat differentassumptions with regard to personnel and non-personnelinfla-tion rates, personnel fringe benefit rate,etc.F,A1ih..r Inwpit...het.' thew/MI.. ,. were quantified340' Analyzing revenue implications41.4Published analyses of AHIS 'ostimpa&thave varied greatly intheir treatment of the revenueimplications of art AHIS. In theFairfax Hospital study (5), reduction in receivables wastreated asa one-time cost saving.One study did not include the effects ofincreased revenue capture because of the-faci that payetsreim-bursing'the hospital on *a cost basis would adjustrates downwardto counteract any increase in thevolume of charges (1). At theother extreme, the analysis of the AHIS in theRadiology Depart-ment at the Hospital of the Universityof Pennsylvania counts as acost benefit the fullvalue of los

38 t charges captured as a result ofautomat
t charges captured as a result ofautomation (estimated at 2.5% of totalartment charges) (3).This same method was used in the sty aanta Monica Hoipi-ta14.4). Others, sap as Sg,hmitz ievaltattbn of the AHIS atDeaconess Hospit'al in Saint Louis (9), recognize thatthe evalua-tion of revenue implications is complex, and mustconsider reim-bursement patterns and patient mix.An AHIS can have two general types of impacts Qnhospitalre enue. Itapsed time between service delivery and billingis r-',the h spital may be Able to reduce the amountof aworking.capitoan4aad hence the interest owed) or receive aone-time increase in available cash'. Either way, reductionin re-ceivables is a tangible financial benefit to the hospitaland thehealth care system.The second type- of revenue impact is the captureof revenuepreviously lost to`the hospital because of inaccurate orincom-plete infOrmation about services delivered topatients. More ac-\curate billingsf r services, however, does not affect the costs ofcare. SinCe theosts are incurted when the services aredelivered,these costs will exist whether or not the services arebilled. Moreaccurate billing through 'automatedcharge capture may influ-ence how costs are paidfor and wko pays for them and itmayalso affect the hospital's revenue; but it does not reduce the costsof care delivery. Therefore, cost savings and revenue capture Aretwo separateimpacts of an AHIS. The inclusion of revenuecapture as a costsavings in studies done in the past has led to overstatedcostpacts. Revenue implications of

39 an AHIS maybe of interestto ahospital a
an AHIS maybe of interestto ahospital administrator. However, analysis ofrevenue impactsshould be condUcted as a separate study, notas part,of a cost im-pact analysis.Any evaluation of revenue implications must recognize thatthere are at least three sets of variables thatgovern the revenueItsimpacts at a particular h6spital:The type of system being ev,aluated and themeasure of reve-nue' capture used in the analysis. For example,revenues to de-partments witin hospitals are often analyzed differently thanrevenues to the hospital.The proportions of patients treated in the institution whopaycharges themselves or are covered by insu(ance plansthat payhospital charges, and the proportion of patients covered by insur-ance plans that reimburse the hospital on some cost-related basis., The formulas used to determine hospital payment by eachpayer using a cost related rate, and any overall limitations thatare imposed on the hospital's budget or charges by state or Fed-eral regulatory agencies.In order to understand any particular analysis of AHIS- relatedrevenue increases, it is necessary to understand each of these setsof variables and their effect on the calculati6n ofrevenues.a36 Conclusions and recommendationsConclusionsFew rigorous studies havebeen conducted of cost impactsofimplementation of automated hospitalinformation systems.Most of thes.work done in this areain the past has involved pre-dicting cost impacts. Somewhatoverstated cost savings (largelythe result of unrealisticlabor savings and inclusion of revenueimpact

40 s) and understated system costs mayhave
s) and understated system costs mayhave produced un-realistic estimates of the net cosiimpacts ofAHIS.Little work has been done onvalidating the results of predic-tive methodologies. Therefore,there is very little documentationof actual impacts of an AHIS onthe productivity of hospitalstaffor on overallchanges to guide the conduct of costassessments.Through improvements ininformation flow, an AHISoffersbenefits to the quality of servicedelivery, and these are usuallymajor motivations for ahospital to implement a system.Morework needs to.be done exploringthe cost implications ofirrrprov-ments in the qualityof service delivery, such asimproved turna-IN"round time for test. resultsreporting and decreased, loss of infor-mation.Revenue recovery fromcharge capture inot a measure ofchanges in the costs ofoperating a hospital and,therefore,should notbecounted amount the cost impacts.Cost studi6s have been conductedfor different hospital settingsand system configurations.The specific methodologiesemployedare generallyapplicable, however, to anysetting or system .type.The purpose in undertakingthe cost study determinesthe ap-proach that is used.-37 a.,Recommendations concerningapproachtoAHIS costevaluationUntil the results of predictive studiesare validated, no singlemethodology for predicting AHIScost impacts can be totally en-dorsed. The approach recommended,however, builds upon thework dope in the past and avoids majorpitfalls. The approachinvolves several steps.(1) Defining system goals and objectivesThis will establish th

41 e prioritygiven to cost savings relative
e prioritygiven to cost savings relativetoother motivations for automating and identifythe parameters ofthe cost study.(2) ldfntifying major cost impactsThis is necessary hot only becauseanalyzing all impacts isverytime-consuming and costly, but also because tiepartmentsper-forming thetgreatest yohime of clericalinformation handling arethe most likely to experience actualcost reductions.(3) Valuing labor impactsFor lab-or savings, the most realisticapproach i% to estimate work-load reductions by jobcontent analysis (for non-nursing labor)and mirk sampling (for nursing staff)and then to develop newstaffing plans incorporating workforce reductions in consulta-tion with shift supervisors and departmentmanagers.(4) Checking predictionsIdeal?), predictions should be comparedwith actual experiencewith similar installations at other hospitals.Until this type of in-formation- becomes available,predictions can be compared withthe results of previous predictive studies.(5) Identifying costs'Care has to be taken to identify allcomponents of costs and to de-velop realistic estimates,of the totalcosts of the system,(6) Establishing time frame forstream of costs and cost savingsLife cyde analysis andpresent value: methodology providearealistic analytical framework fora predictive AHIS cost study.(7) Documenting actualsystem costsBecause of the uncertainties regirding theaccuracy of predictedcost impacts, any predictive study should beaccompanied by acommitment to monitor actualcost impacts after the system isimplemented.4338s. 4Re

42 commended directions for futtire researc
commended directions for futtire researchMOie researchisneeded to help decision makershospitals,health planners, and regulatorsevaluate the cost impacts ofAHIS. Since decisionmakers are primarily concerned with pre-dicting saviags prior to an iniestment in AHIS, it is important tovalidate whether the techniques being used to predictsavings areaccurate:,Many different types of AHIS are available and they vary/greatly in terms of scope, design, and cost. Information aboutthe costs and benefits of indikidual system applications, such asmedication ordering and radiology scheduling, would be usefulfor comparing systems with different features and for designingmore cost-effective systems.Optimal use of an AHIS and realization of all potentiaLbene-fits requires a directed effort on the part of the user's. Althoughsome work has been done on developing a benefitsrealizationprogram for certainANL§ benefits (2), the expansion and trans-fer of this type of program from one institution to another hasnever been studied. Research onincentives is also needed to en-sure that the benefits of future AHISimplementations are maxi-mized. This could involve disseminating informatiOn about thecost impacts of AHIS and the factors thatdetermine savings, prp-siding technical assistance on conslucting benefits realization*programs, and considering changes in reimbursementpolicy thatwould offer hospitals more immediate financial incentives torealize AHIS benefits.Before hospitils' Can incorporate the experience at other sites,into their decjsion

43 making about AHIS, the transferability o
making about AHIS, the transferability of thatexperience has to be considered. Ideally, guidelines should beavailable concerning the range of cost impacts of different:auto-mated systems and the factors that determine the magnitude ofthese at each installation.Most AHIS, and all comprehensive AHIS, are implementedprimarily to improve patient care by improving the communica-tion of information. Therefore, techniques for examining thecost impacts of these improvements ininformation Sow would beextremely valuable in completely assessing the cost impacts ofautomation.39 References"1.Barrett, J. P., P. L. Hersch and\ R. J. Caswell, Evaluation ofthe Implementation of the Technicon MedicalInformation Sys-tem at El Camino Hospital, Part II,Economic Trend Analysis,Battelle, Colurribus, Ohio, May 14, 1979.`2.Gall, J. E., et al., Demonstration and Evaluatiqn of aTotalHospital Information' System, National Center forHealth Serv-ices Research, NTIS PB 262 106; 1975.3. Arenson, R. L., and J. W. London, "RadiologyOperations41.Management Computer Sygtem, Hospital. of theUniversity.. ofPennsylvania," in Proceedings of the American Collegeof Ra-diology Sixth Conference on Computer Applicationsin Ra-diology, March 1979.4. Brian, WI, M.D., et al., Cost Benefit Analysisof Installationof a Spectra System at Santa Monica Hospital,University ofSouthern California,Centerfor Health ServiceResearch,January 1176.5. Fairfax Hospital Association, Cost-IlenefitAnalysis of DataCare System (unpublished), Falls Church, Virginia, 1979.6. Huff, W. S. a

44 nd E. J. Bond, Demonstrationof a SharedH
nd E. J. Bond, Demonstrationof a SharedHospital Information System, The Sisters of the ThirdOrder ofSt. Francis, Peoii;, Illinois, 1970.7. Kim, S. H. and H. H. Schmitz, "A Cost:BenefitAdalysis ofa Modular Hospital InformationSystem," in M. W. Hopfe (ed.)American Institute for Decision StienceProceedings, Boston,Massachusetts, November 14-16, 1973.'8. Mishelevich,.D. J.,' W. G. Gipe, J. R. Roberts,C. Denny,A. p. Stern and M. W. Day, "Cost-Benefit Analysisof a Com-puter-Based Hospital Information System," in Proceedings of the.,Third Annual Symposium in Computer Applications inMedicalCare, 1979..,9. Schmitz, H. H. "An Evaluation of the ImmediateFinancialImpact of the Hospital Information Systems at DeaconessHospi-4145 tat:, pp 265-282, in R.C. Koza (ed.)Health Information Sys-teM;Colorado Associated University Press, Boulder, Colorado,1974.10. Technicon, Medical Infoimation Systems Corp.,Cost-Bene- -fit ,Inalysis of the Techniron Medical Information System (MIS)for (unnamed hospital),December 1975.1 1.Arthur D: Little, Inc.,A Review of the Medlab System in theSurgical Intensive Care Unit at the Massachusetts General Hospi-tal AppendixA Report to the National Center for Health Serv-ices Research and Development, Contract HSM "110-70-406,March 1973.42r4e, Current 4,CHSR PublicationsI ,National Center fur Health Services Research publications of interest tothe health community are available on request to NCHSR, Publicationsand Information Branch. 3700 East West Highway, Room 7 44, Hyattsyille, MD 20782telephone: 301,436 8

45 970). Mail requests will befacilipted by
970). Mail requests will befacilipted by enclo ure of a self addressed,, adhesive backed mailinglabel. These publications also are available for sale through the NationalTechnicalInformationService(NTIS),Springfield,VA 22161(telephone: 703 487 1650). PB and HRP numbers in parentheses areNTIS order numbers. Publications which arc out of stock in NCHSR areindicate,d-1ts available only from NTIS Prices may be obtained from'theNTIS order desk on request.Research DigestsThe Research Digest Series provides overviews of significant researchsupported by NCHSR..The series describes either ongoing or completedprojects directed toward high priority health services problems. Issues areprepared by the principal investigators performing the research, in collaboration with NCHSR staff. 'Digests are intendedfor an inter-,disciplin,aryaudienceof health,servicesplanners,administrators,legislators, and others who make decisions on research applications.(HRA) 76-3144Evaluation of a Medical Information System in aCommunit(Hospital 4PB 264 353, available NTIS only)(HRA) 76-3145Computer-Stored Ambulatory Record (COSTAR) (PB 268 342)(HRA) 77 -3160Program Analysis of Physiciap,,Extender Xlgorlthm Projects(PB 164 610, available NTIS only)i/(HRA) 77 3161Changes in The Costs of Treatrhent of Selected Illnesses,1951-1984-1071 (HRP 0014598, available NTIS only),(HRA) 77-3163ImpactofState Certificate-of-Need Laws on Health CareCosts and Utilization (PB 264 352, available NTIS only)(HRA) 77-3164AnEvaluationofPhysicianAssistantsin DiagnosticRadiolog

46 y (PB 266 507, available NTIS only)(HRA)
y (PB 266 507, available NTIS only)(HRA) 77-3166Foreign Medical Graduates A Comparative Study of StateLicensure Policies (PB 265 233, available NTIS only)(HRA) 77-3171Analysis of Physician Price and Output Decisions (PB 273312)(HRA) 70173Nurse Practitioner and Physician Assistant Training andDeployment (PB 271 000, available NTIS only)43 (HRA) 77 3177Automation of the Problem Oriented Medical Record (PB 266881 available NTIS only)(PHS) 78 3190Uncertainties of Federal Child Health Policies. IMpact inTwo States (PB 283 202)(PHS) 80 3229Responses orCanadian Physicians to the Introduction ofUniversal Medical Care Insurance The First Five Years inQuebec (PB 80-137 979)(PHS) 79Israel Study of Socialization for Medicine (PB 293 887)(PHS)3235AAMC Longitudinal Study of Medical School Gradates of1960 (PB 294 689)(P 5) 79 3238Some Effects of Quebec Heal* Insurance (PB097)(PHS) 79 3261Medical Education Financing Issues and Options (PB 80-134851(-(PHS) 80 3285A Social and Economic Analysis of Professional Regulationin OptometryResearch SummariesThe Research Summary Series provides rapid access to significant resultsof NC.fiSR supported research projects. The series presents executivesummaries prepared by theinsestigaturs.Specific findings are highlightedin'a more concise forMithan in the final report. The Research SummarySeries is intended for health services administrators, \pianners, and otherresearch users %silo require recent findings relevant to immediateprograms in health services.(HRA) 771162Recent Studies in Health Servic

47 es Research,.VolI (July 1974through Dece
es Research,.VolI (July 1974through December 1976) (PB 266 460)(HRA) 77 3176QualityofMedical Care Assessment Using OutcomeMeasures (PB 272 455)(HRA) 77 3183Re,cent Studies in Heaitp Services Research. Vol II (CY 1976)(i.E3 279 198).(PHS). 78 3187Criterion Measures of Nursing Care Quality (PB 287 449).(PHS) 77 3188Demonstration and Evgluation of a Total Hospital Informa-tioo System (PB 271 079)(PHS) 78 3192AssVatvg the Quality of Long-Term Care (PCB 293 473)(1)1S) 78 -3193Optimal Electrocardiography (PB 281 558)(PHS) 80a10Comparisons of Prepaid Health Care Plans in a CompetitiveMarket. The Seattle Prepaid Health Project(PHS).78 3201A National Profile of Catastrophic Illness (PB 287 291)(E'FIS)19 3223Medical Direction in Skilled Nursing FacilitreS (PB 300 845)(PHS) 79 3230Per-Case Reimbursement for Medical Care (PB-294 688)(PHS) 79-3236Nurse Practitioners and Physician Assistants A ResearchAgenda (PB 294 084)(PHS) 80-3244Quality of Ambulatory Care(PHS) 79.3247A Demonstration of PROMIS(PHS) 79 3248Effects of the 1974-75 Recession on Health Care for theDisadvantaged (PB 80 138 449)(PHS) 79 3250Effects and Costs of Day Care and Homemaker Services forthe Chronically III A Randomized Experiment (PB 301 171);(PHS) 81 3260The Rhode Island Catastrophic Health Insurance Program.The First Three Years44.Q,*.r (PHS) 80 3265How 4 Medical Information System Affects Hospital Costs9The El Camino 'Hospital Experience(PHS) 80 3266Developing a Quality Assurance Strategy for Prim,ry Care(PB 80.162 878)(PHS) 81 3277Comparison of Strategies

48 for the Development of ProcessMeasures
for the Development of ProcessMeasures in EMSiPHS) 81 3298Regulating Hospital Capital Investment The Experience inMassachusettsResearch ReportsThe Research Report Series provides significant research reportsin theirentiretyupon the completion of the project.Research Reports aredeveloped by the principal investigators who conducted theresearch, andare directed to setatecrtilsers of health services research as part of a conti-nuing NCHSR effort to expedite thedissemination of new knowledieresulting from its project support.(HRA) 77 3152How ILaWyers Handle Medical Malpractice Cases (HRPV0014313)(HRA) 77 3159 17-qp Analysis of the Southern California ArbitrationProject,-January 1966through June 1975 (HRP 0012466)(HRA) 77 3165Statutory ProvisionsforBinding. Arbitration ofMedicalMalpractice Cases (PB 264 409. available NTIS only)(HRA) 77-31841960 ond 1970 Spanish Heritage Population of the Southwestby County (PB 280 656, available NTIS only)(HRA) 77-3189Drug Coverage under National Health Insurance The PolicyOptions (PB 272 074)(PHS) 78.3204Experiments in Interviewing Techniques Field Experimentsin Health Reporting (PB 276 080, availableNTIS only)TelehealthHandbook. A 'GuidetoTelecommunicationsTechnology for Rural Health Care, (PB 292 557. available NTISonly)Emergency Medical Technician Performance Evaluation (PB285 961)Evaluation of Child Abuse and Neglect Demonstration Pro-jects, 1974.77, Vols1 and2(PB 278 438 and 278 439: vols1 12, PB 278 437, the set)Needed Research in the Assessment and Mcrrittortng of theQuality of

49 Medical Care4PB 288 826)A Cost Effective
Medical Care4PB 288 826)A Cost Effective Approach to Cervical Cancer Detection (PB295 515)Costs and Benefits of Electronic Fetal Monitgring AReviewof the Lite'rature,11P.B 294 690)ComputeApplications in Health Care (PB 300 838)Effects and Costs of Day Care and Homemaker Servicesforthe Chronically III k Randomized Experintent (PB80.138100)Health Status. Medical Care Utilization, and Outcome AnAn-notated Bit)liography of Empirical Studies, Parts I-1V (P13284-997, available NTIS only)E$tirnated Economic Costs of Selected Medical EventsKnown ctr Suspected to be Related toAdminiStration of Com-mon Vaccines...Health Care Stirveys Using DiariesA Social Sna EtoripmiC Analysis of Professional Regulationin Optorrfetry45(PHS) 79-3210(PHS) 78 3211)PHS) 79 3217 ;I(PHS) 78-3219(PHS) 79.3237(PHS) 79-3245(PHS) 79.3251(PBS) 79.3258(PHS) 8113263(131-1S) 81-3272(PHS) 80.3279(PHS) 81.3295 (PHS) 81 3300HispanicHealthServicesResearchA PreliminaryBibliography(PHS) 81, 3303Health Planning Review of Medical Information SystemsResearch ProceedingsThe Research Prdseedings Series extends theavailability pfnew researchannounnisl at kel, conferences. smposia and seminars sponsored orsupported In NCHSR In addition to papers jsented, publications in thisseries include discussions and responses Ahever possible. Theseries isintended to help meet the information needs of health services providersand others IA ho require direct access to concepts and ideas evolving from.the exchange of research results.(HRA) 77 3154Advances in Health Survey Research Me

50 thods (PB 262 230,available NTIS only)NC
thods (PB 262 230,available NTIS only)NCHSR Research Conference Report on Consumer Self-Carein Health (PB 273 811)International Conference on Drug and Pharmaceutical Ser-vices Reimbursement (PB 271 386)Emergency Medical Services Research Methodology (PB 279 63096)Health Survey Research Methods. Second Biennial Con-ference (PB 293 492)IHRA) 77 3181(HRA) 77 3186(PHS) 78 3195(PHS)78 3207(PHS) 78 3208Drug Coverage Under National Hedth Insurance (PB 293 468).(PHS) 79.3209 "Health Services ReWearch in Puerto Rico (PB 292 326)(PHS) 80 3215eost Accounting for Pharmaceutical Sprvices (PB 80-1.57 936)(PHS) 79 3216ykdical Ikhnology =TheCulpf it Behihd,Health Vare Costs'-(15B 299,408)-.,,,.(P79 3225-1 Emer eft 'y : MedicalServicesResearch Methodblogy,.'llibr13.p'l (PB 294 648),t,i:(PHS) 79 3225 2 Emer'gencyagdicalServicesResearch Methodology'_..Wockshop 2 (PB 80-147292)-r,Effects of the Payment Mechanism on the Health CareDelivery Systm (PB 291.2433)-A National ConferenciPon Health Policy, Planning, and Finan-cing the F -uture of Health Care for Blacks in Amery (PB 292-559)4Emergency Medical Services Systems as a-Health Servic4Rgsparch Setting (PB 297 102)Medical Technology (PB 80-149,11)Sharing Hpalth Care Costs.(PB 80'162 795)(PHS) 78-32274PHS) 79 3228(PHS) 79, 3233(PHS) 79-3254(PHS) 79-3256(PHS) 79'3257,(PHS),81.3268(PHS) 80-3288'41..Health Facility Reuse Retrofit, and Recinfiguration (PB80-142 383)...,*Health Survey Research Methods. 1979Hispanic Health Services ResearchtResearch Management"The Research Managem'entcle*ribes

51 programmatic rather thantechnical aspect
programmatic rather thantechnical aspects of the NCHSR resgarA.effort.Information is presentedon the NCHSR goals, research objective; and priorities; in addition, this46' sseries contains lists of grantsa ndcontracts, and adminis.trative inforina-non on funding. Public anon, in this series arcintended to bring bask in-formation on NCHSR and its programs to researchplanners, 'ad-ministrators, and others who are involved with tht allocationof researchresources..(PHS) 79.3220�(PHS 80 3271(PHS) 81 3287(PHS) 81 3304(PHS) 81 3307r-tiEmergency Medical Services S stems ResearchProjects,1978 (PB 292 558)Emergency Medical Services Systems ResearchProjectsAbstracts. 1979Computer Tapes Available from 20 Health Services ResearchProjectsEmergency Medical Services Research Project Abstracts,1980Summary of Grants and Contracts Active on September30.1978NHCESThe Aatumal Health Care Expenditures Study Series presentsinfopna-tion and anal)ses on critical national health policy issues.Basic data wereobtained from the National Medical Care ExpenditureSurcA, astatistical picture of how health services are used and paidfOr:. DataPreviews give preliminary estimates of key measures.(PHS) 80 3276Data Preview 1 Who are the Uninsured')(PHS) 80 3275Data Preview 2 Charges and Source's Qf Payment for DentalVisits with Separate C6'arges(PHS) 80 3,278Data Preview 3 Who Initiates Visits t,o Physicians' Of fices9(PHSJ 80-3290.DataPreview 4 Health Insurance Coverage of Veterans(PH5) 80.3291Data Preview-5 Ch4rges and Sources ofPaymentfor Visits t

52 oPhysician OfficesData Preview 6 Waiting
oPhysician OfficesData Preview 6 Waiting Times in Different MedicalSettingsAppointment and Office Waits(PHS) 80.3296PoliCy ResearchThe Polic3, Research Series describes findingfarom theresearch programthat have may-)r significance filitpolicy issues of4hemoment. Thesepapers are prepared l.'s members of thestaff of NCHSR or by indepen-dent investigators. The series is intended specifically toinform those in'the,public and private secuirs who must consider, design, andimplementpolicies affecting the delivery- of health services(HRAI 773182Controlling the Cost of Health Care.IPB 266 885)Program Solicitations(PHS) 81 3292Grants for Research on Health Promotion and DiseasePrevention47 (PHS) 81-3293(PHS) 813299.59Extramural Associates Program, 198Grants for pissertation Support, 1981NCHSR Fact SheetsNCHSR Fact Sheets provide concise information on NCHSR's organiza-tion and mission, as well as on key problems in health services andresearch diiected toward meeting health care needs.National Center for Health Services Research (English)National Center for Health Services Research (Spanish)National Center for Health Services Research (Chinese)NCHSR Research Focus. 19811982Catastrophic IllnessHealth Care and the DisadvantagedHospital Cost and Utilization ProjectLong-term CareManpowerMaternal and Child Health CareNational Health Care Expenditures StudyUnited Mine Workers of American Health Care PlanEmergency Medical Services SystemsNCHSR Research Grant-Application and Scientific/Technical Merit Review ProceduresComputer Applicat