Forum

The Meaningful Use Regulations in Information Technology:What Do They Mean for Quality Improvement in Hospitals?

B y passing rhe Health Information Technology for Economicand Clinical Health Act (HITECH), the U.S. Congress

sought to catalyze the use of information technology to improvethe quality and efficiency of health care in the United States. Thelegislation not only introduced new automated electronic report-ing requirements and defined "meaningful use" criteria for elec-tronic health records (EHRs) but also created financialincentives—and eventually penalties—for providers. In July2010, the Centers for Medicare & Medicaid Services (CMS) andthe Office of the National Coordinator released the final mean-ingful use regulations for 2011 -1 In light of the important rolethat CMS plays in health care delivery in the United States andthe size of the financial incentives (approximately $27 billion),providers and institutions are now scrambling to attempt to meetthese new requirements.3 This begs an important question:How should information systems be modified to achieve federalmandates and at the same time have the greatest impact onhealth care quality and efficiency?

Where to Set the Bar?Federal requirements arc developed through a iong consensusprocess. The aim is to set the regulatory bar high enough to meetbasic policy goals but low enough to ensure that most of the in-tended audience can meet the mandates. In the case of the mean-ingful use regulations, individual providers and providerorganizations are questioning whether the regulatory bar shouldhave been set lower, while advocacy groups and payers arc argu-ing that the bat should have been set higher^

Brigham and Women's Hospital-x\BostonV is adding a newclinical documentation module in its inpaticnt FHR, and qual-ity improvement (QI) activities tied to this effort have readilyunderscored the importance of taking an approach that strives toexceed Stage 1 meaningful use regulations from the start. To aidin the design of new applications within die EHR, we invento-ried measures that hospitals must report under current inpatientregulatory and accreditation requirements. We also conductedinterviews with our institution's clinical leaders to identify meas-

ures and QI efforts for which they deemed electronic measure-ment, decision support, or intervention (such as follow-up re-minders and critical alerts) to be important. It quickly becameclear that if we simply meet the current meaningful use regula-tions we would not have the comprehensive functionality thatwe need to improve quality and efficiency.

A comparison of meaningful use and current reporting re-quirements highlights the need for automated reporting that ex-tends far beyond federal regulations (Table 1, page 334). Inaddition, the findings from our interviews underscore the impor-tance of going beyond structured data collection and reportingfrom EHRs (Table 2, page 335). The final Stage 1 regulations doinclude some requirements that extend beyond reporting, such asthe use of some clinical decision support (CDS) and real-timemedication reconciliation. However, the committees making therecommendations have been reluctant to be too prescriptive in[his area, and much has been left to the discretion of provider or-ganizations.

Toward a Broader Set of FunctionalitiesFor EHRs to support sustainable and effective improvement ef-forts in hospitals, they will need a much broader set of function-alities than diose that currently exist in most EHRs.' Such a setwill need to cover areas including speeding the communicationof critical test results, enhancing transitions in care, improvingtest result tracking, and providing robust and complex real-timedecision support to providers.* Such functionalities should playa major role in enabling clinica! performance.

More specifically, HIT systems should be designed to improvesafety and efficiency by containing automatic reminder func-tionality for necessary follow-up in a wide range of clinical situ-ations.5 For example, when an inpatient radiological rest revealsan incidental finding, such as a solitary pulmonary nodule/thatrequires nonurgent but critical follow-up after hospital discharge,many current systems still rely on a complex process. The radi-ologist reports the finding, but the follow-up planning is left tothe inpatient physician to communicate to the outpatient

July 2011 Volume 37 Number 7

Table 1. Examples of Quality Measures Thai an Electronic Health Record Could Measure Automatically

Agency Reporting Requirements Measure ExamplesCenters-torMedicare &Medicaid Services

.-• NatfonaiHospital.. .;;;V? ;inpatient Quality Measures

• Aspirin -prescribed at.discharge" for acute, myocardiaj infarction {AMI) ..• patients"' " - - "• tjpid-:ov,'3ring.therapyatd(scharge:for:AMi."pat}ents*-:: .;/• Smoking cessation counseling offered to AMU heart failing and pnf- 'SimorNa pa-

tients* -;.-;--- - . - : . .. .• Perioperaiiue feeta blockade for patients taking a beta blocker prior to surgery"• Prompt; postoperative-.urinary.carterremoval* . " - - " ,•postoperative blood glucose confroHn:cardiac surgery, patients* '

The Joint j • National Hospital . : . = -- .Commission j tnpa'ient Quaiity Measures- • :- =

I • National.Parent Safety C5oa!s.I • National Quality Core MeasuresI • Joint Commission Intensive

Care UnitMeasureSet.

stroken 2 hours of an fschemic:

-Administration cvr Hn^i/sts? sterniris to patients at risk.of preterm delivery*l ijse::0f appropriate precsiiTions for Gicsiriti!um_ d/ffic//e patients* : ; ------'- . •! Useofappioprieriy dteioks ;o ensure biood typetotood- pcH^patibifity* . " - ; - - .l Assessment of risk for suicide* ". . . - -

Agencyfor..Healthcare

Quality and Safety

Leapfrog Group

State .{Massachusetts}

National Quality

Foium (NQF)".

• inpatient Qualify Indicators•.-Patient Safety Indicators (PSFs) - :--

B Pediatric Quality indicators

• Leapfrog Hospital Survey

• Medicaid ReportirkjMeasureS- "• NQF Ssrioiis ReportableEvents . : .= •

• NQF-Ereiorsed StandaEtis ::

..• Serious Reportaole. Events ::-

••(SR&K /.:. : :-

•-• : 1

• latrbgeriicpneurnotriorax rates in non-ncoiiates up to M years of age• Oecurrence of hirtft icaunia or injury to neoriate" . - . ;" ' : . _ LjT

• Occu'rrenee of postoperatWe hemorrhage or.;hematoma requiring^ rcopcration* .

• Newborn, screening for hyperfajlirubinemia prior to-distterge*-- V .; " : ":-".-.-::".-" " ;;

• Groap & Strep (GB3) prophylaxis prior la delivery for GBS+ patients* :': :• injury associated with a felt*- . - :

••Occurrence of stage 3 or 4 pressure ulcer*- \ Harm associafed v;!lh use of oontamirialed drug, device, or biologic*

• "lo'fant-discJiarged to wrong person* - . , -

• Retained .foreign body foliov/ing procedure' - ..: . . . . . - . ; " - . • . ": : ::: :: ; ::,• Disciosure conversation foitav.-ing an adverse? event™ ;:; . ; =-..; - . ; . : . -

"• Documentation of iraGrmed consent pnor to a procedure*" " ".." """ " •-":".-.--• Rates of deep sternaj wound .infection within 30 days following corgnary/anery -

.-bypass graft (CABG)'

• Completion of dysphagra screeh in stroke patients prior to oral -intake* . .

' Not included in final Stage 1 Weanrngfui Use regular" included in final Stays 1 fvteanisigrjl Lisa regulations.

provider (frequently the primary care physician), who must then

arrange the necessary foliow up. This process can be unreliable

at rimes, especially if the incidental folding is far removed from

the cause of admission or if the patient was admitted to a hospi-

tal not directly affiliated with the patient's outpatient provider.

Instead, in such cases, an EHR should prompt the radioiogist co

suggest a follow-up exam recommendation as soon as the iffigg-

jngis repotted as abnormal. The results should also be commu-

nicated directly to the patient and to the primary care provider.

CDS is another area in which EHRs have tremendous poten-

tial to improve quality and safety.6 However, while the meaning-

fill use regulations require the use of some real-time CDS,

citnical systems must rapidly add much deeper CDS if care is to

be improved for patients with multiple chronic diseases. For a

patient with heart failure and a low ejection fraction, basic CDS

can fulfill meaningful use by simply recommending an an-

giotensin-convertmg enzyme inhibitor (ACE-1). However, if the

patient has previously developed a cough with the use of an

ACE-I, a sophisticated EHR should also warn of the previous

side effect and simultaneously prompt a provider to consider an

angiotensin receptor blocker and ask for and record the physi-

cian's rationale in the case of nonprcscriptiprtT ' <TY\-*.4^C C

In addition, broad integration between systems is necessary so

that information including laboratory, pharmacy, billing, and

ordering information can be transferred between systems with-

out additional manual entry. The potential benefits arc great:

Responses to abnormal laboratory tests will be less likely to be

missed or delayed because they can be displayed or trigger re-

minders m muJtipic applications; electronic billing for specific

conditions can automatically become integrated into patient

problem lists to Improve both decision support and efficiency of

physician work flow; medication dosing can be guided and the

July 2011 Volume 37 Number 7

Table 2. Essential Electronic HeaJth Record Functionality to Enable Quality Improvement

Necessary—,Functignaliitv

Required Information Systeminfrastructure Elements

Functionality and : :;iSinfrastructure ElementsSpecifically Mandated in : ,Core Set of Meaningful UseRegulations' -__'^'<: Clinical Examples

Real-time decision •-.-•<'-'.support for managementof disease .-.; -

• Structured problem list• Structured medication list with aiiergfes andcontraindications - " - .• User interface to capture clinician action orresponse : v-V

Structured data (problem,medication, and allergy lisfe)

• Prompt to prescribe angiotensin-converting enzyme (ACE) or angiotensinreceptor blocker (ARB) at discharge in.heart failure patients with left ventricularsystolic dysfunction.• Prompt to provide venous throm-boeinbcJism (VTE) prophylaxis for =>;.appropriate inpatients -^:^.-

Reai-time decisionsupport based onlaboratory results

• Integration with Faboratory results: data• Integration with procedure schedufing sys-tem• User interface to capture clinician action orresponse ;::-; ~

Capability to exchange limitedclinical information only

• Prompt to consider canceling bariatrtcsurgical procedure in patient withpositive pregnancy test :v• Prompt to decrease warfarin dose inpatient with sup rathe rapeutic interna-tional normalized ratio (INR)

Real-time decisionsupport for diagnosticdecision making

• Demographic information• Structured problem list =:".- .V -• Structured medication list with allergies andcontraindications• Integration with laboratory, radiology,pathology, and endosccpy results data andcardiology and pulmonary testing data• Integration with procedure schedulingsystem -. -'-._ v ; ?V- :• User interface to capture clinician action orresponse ... :~ _ - ";L":-:-~:-

Structured data only (demo- -graphic data, problem, medica-tion, and allergy lists) :;;V :

• Prompt to order fbliow-up forpatient with mamrrtography findings• Support on what type of cardiactesting (if any) is optimal "-- :-

Gbmmunication andfoElow-up of critical testresults

• Ordering provider• Primary provider to foilowfup"• integration with laboratory, radiology,pathology, and endoscopy results data andcardiology and pulmonary testing data

Capability to exchange limitedclinical information only ; --:--"

• Prompt to refer patient to surgeonwhen colonic biopsy pathology revealsdysplasia . . . -» .• Prompt to schedule folloi ^p for . ; "patient with concerning pulmonary. ;.:-nodule on chest x-ray . ;

Preventive healthmaintenance

I Demographic informationI Structured problem fist

Structured data only (demo-graphic.dafa, problem list)

• Prompt to administer pneumococeatvaccine for patient over 65 years of age• Prompt for mammography or ;sv .colonosccpy :\^\. ;..y..--

population identificationfor population-based .:-management

i Demographic informationi Structured problem listI tnpatient documentation

Structured data only (demo-graphic data, problem list)

• Identification of ail patients with heartfailure to ensure they receive appropri-ate discharge education or enrollment indisease management programs

Documentation of criticalprocesses

• Structured medication list with allergies andcontraindications .. ^f:-~.• Templates for care processes /

Structured data only (medica-tion arid attergy lists) -"i- -

• Central line checklist documentation• Medication administration with bar-:code scanning to confirm right patient,right medication, right dose, right time,right route . ;/.

Assistance withtransitions in care

• Demographic information• Inpatientdocumentation .. $ \..• Structured problem list , .-':\ -'.• Structured medication list with allergies andcontra indications• Integration with laboratory results data• Discharge template• Ordering provider• Primary provider to followfUp"

Structured data only (demo-graphic data, problem, medica-tion, arid allergy lists)

• Electronically compiled sign-outbetween providers :• Electronically compifed discharge ;form that includes pending results andcare plan ;. . . ;::..:.:.

"The core set of Sage 1 meaningful use regulations require the use of only one decision support rule, which may ar may not enable one oFthe above functionalities.Additionally, the menu set of meaningful use regulations allov.' the choice of five objectives, some of which may enabie some of the above functionalities.

July 2011 Volume 37 Number 7

default dosing set based on the patient's age: .kidney function,and other factors; checklists can become a part of care deliveryat critical times, such as in the ICU or operating room or ac hos-pital discharge, and can easily be updated, when che evidence-

base changes. None of these features of F.HRs are specificallymandated in die initial federal criteria, but they are all needed forthe delivery of better care at lower cost.

Going Beyond the Federal Regulations

tions will not be sufficient to transform quality and deliver im-proved efficiency, even though more will be required in 2013 and

2015- To have an impact, EHR systems will need strong CDS,robust registry functions, and integrated tools chat support team

care, as well as broader reporting capabilities and additional func-tionalities that fit seamlessly into existing provider work flow.

This is possible—Kaiser, for example, has already achieved im-pressive results m this regard.'-"5 Although meeting che Stage 1

meaningful use regulations will help providers begin the journeyThe transition to effective EHRs will be no smali task, especially j^ee-tho-sead to bercer care, meeting these criteria alone will not bebecause of their currently limited use. For example, accordingto a recent study, fewer than 3% of hospitals in the United States

are using "'comprehensive" £HRs.H" Adding to the challenge isthe need for organizations to design mformauon systems thatcan function in ways chat enhance work flow and do not disrupt

critical cognitive processes, making both usability and work-flowintegration critically important to adoption.:r' New approaches

are needed to ensure that HIT systems expand beyond basic re-porting and functionality, while maintaining ease of use and en-suring effectiveness. Neither regulation nor certification

addresses—and both are likely ilKuited co address—these Issues.

Instead, we believe that the mosc effective way to achieve these\J. characteristics will be through regular and frequent evaluations

of systems (perhaps akin to the evaluations reported In Con-). In addition, Information systems must undergo

periodic testing (like the tests of hospicals performed by theLeapfrog Group [http://www.leapfroggroup.org}) to assess

whether key decision support is In place.11 In che meantime,however, If organizations simply install systems without takingthe necessary and thorough examination of their work-flow and

care processes, they may not only waste resources in implemen-tation but also miss the opportunity to Improve care.

The design and im piemen cation of new HIT systems will also

require approaches focused on clinical outcomes and process im-provement rather than the collection of administrative data. Asthis Is done, systems developers, providers, and organizations

must recognize diat these new HIT systems should be designedwith the understanding that HIT systems can bring their own sec

of errors, counteracting the quality benefits that they can ereate.12 When errors in care delivery occur. It will be critical co as-

sess, track, and rooc out the contributions of new i IJT systems.Developers must be willing to continuously and rapidly refine

systems to correct any defects that cause problems.

enough to meet even our current clinical needs from HIT. If

provider organizations are serious about improving quality and

efficiency, they must take ownership of advancing their EHR ca-pabilities far beyond just meeting the federal regulations. I

Tfte authors thank Dr. Ashish jiia Sor his valuable thoughts and comments on anearlier version of this manuscript.

implement incentives and reguiessential to increase HIT adoption races. I lowcvcr, organizations

must recognize that simply meeting the current Stage 1 regiiia-

1. Blumeiitrtil D., "lavenner M_: The "meaningful use" regulation tor electronichealth records. VftWMWjfff-Sftl-'itU. AJig S. 7010

2. Blurncnthai D_[Stimulating the adoption of healB information technology.NEngljMed360:1477-1479, Apr. 9, 2009.3. DesRoches C.M., el al.: hlccrronic health records' limited successes suggestmore targeted uses- Heairk-Aff (Millwood) 29:639-646, Apr. 10, 2010.4. SchiffG-f)-, Rate? D-WjCin electronic clinical documentation hc3p preventdiagnostic errors? NEn&fMcd 362:1067-1069- Mar. 25, 2010.5. Roy C.L., et aL: Patient safetv- concerns arising from test results thar return

/n£-rnJMa/]43:121-128, Jul. 9, 2005.aL: A roadmap for national actionjjn dJnjalji£Cj(3Jiiri-Stip-^-

, Mar.—Apr. 2007- Erratum in:bn Meet Inform Assoc i 4:389, May-Jim. 2007.

7. Wright A., etaL: Governance for clinical decision support: Case studies andrecommended practices Irom leading institutions. J Am Med Inform Assoc

18:187-194. Mat 201 i.•SrJKaA.K., et aL|A progress report on electronic health records in U.S. hospi-tab. HakhAff(MiiiwweQ 29:1051-1057, OCL 2010-9- Jha A-K-: Meaningful use of electronic health records: 'I'he road ahead. JAMA304:1709-1710, Oct. 20, 2010.10. Han Y.Y., ct aL: Unexpected increased mortality after implementation of acommercially sold computerized physician order entry system. Pediatrics116:150( 1512,_Dec. 20G3&

.(• computerized010.

12_ Karsb B., et a!.: Health tnforniation technoSos;;: fallacies and sober rcvili-&s.JAm Med Inform Assoc 17:617-623, Nov. 1, 2010.13-IiangL. (ed.J: Connected for Health: Using Electronic Health Records to Trans-form Care Deliv&y. San Francisco: Josscy-Bass, 2010.

July 2011 Volume 37 Number?