"GOVERNMENT, LAW, AND PUBLIC HEALTH PRACTICE F u t u r e N e e d s o f H e a l t h D e p a r t m e n t s A R id e in t h e l i m e M a c h in e : In fo r m a tio n M a n a g e m e n tC a p a b ilitie s H e a lth D e p a r tm e n ts W ill N e e d Seth Foldy, MD, MPH, Shaun Grannis, MD, MS, David Ross, PhD, and Torney Smith, MSHE provider incentives) may lead help meet these challenges, butW e have proposed needed WHAT INFORMATION MANAGE- information management ca­ health departments to stop deliv­ only if health departments de­ ment capabilities will be neededpabilities for future US healthering individual services, such as velop needed capabilities. by tomorrow’s US health depart­ departm ents predicated onimmunizations.2 Meanwhile, somements? The Public Health Ac­ trends in health care reform andHEALTH DEPARTMENTpredict that digitization and thecreditation Board establisheshealth information technology. ROLES AND INFORMATIONexchange of health care data willstandards and provides accredita­ Regardless of whether healthproduce “distributed access to in­ MANAGEMENT tion to US local, tribal, and statedepartments provide direct clin­ health departments. Because we formation without exposing theical services (and many will),Aspects of the ACA encourageare experienced in local, state, and details of the underlying data ...they will manage unprece­ private health care providers tofederal public health informatics collect[ing] only summarized data...dented quantities of sensitiveperform preventive care. Reduc­ (the systematic application of in­ or key results.”3 Taken together,information for the public healthtions in the number of uninsuredformation and computer science these trends suggest that healthcore functions of assurance andindividuals, improved financial in­ assessment, including population- and technology to improve public departments may eliminate directlevel health surveillance and centives for preventive services,health practice, research, or edu­ services and consume rather thanmetrics. Absent improved ca­ and business models such as ac­ cation),1a Public Health Accredi­ create health information products,pabilities, health departmentscountable care organizations andtation Board think tank asked us to thus minimizing their managementrisk vestigial status, with conse­ primary care medical homes,predict the effects of emerging of sensitive health information. quences for vulnerable popula­ could reduce the need for healthtrends on requirements for future We predict, rather, that healthtions. Developments in electronicdepartments to provide personaldepartments will struggle withaccreditation standards. health records, interoperabilityservices, such as immunizationsThe Patient Protection and managing more information toand information exchange, pub­ and tuberculosis management. Affordable Care Act (ACA) may improve service cost efficiency,lic information sharing, decisionHowever, this reduction mayradically change the functions of collaborate on prevention withsupport, and cloud technologiesbe offset by a continuing obliga­ US health departments. New de­ a leaner health care system, andcan support information man­ tion to ensure prevention for thoseagement if health departments velopments in health information meet demands for unbiased pop­ have appropriate capabilities. ulation health statistics. Competi­ remaining unprotected and an in­ technology (technology standards,The need for national en­ tive pressures will emerge rapidly creased demand for community-applications, and hardware forgagement in and consensusover the next 5 years. Health based services. Assumptions thathealth data) will profoundlyon these capabilities and theirdepartments that cannot manage ACA reforms will optimize popu­ change how information is man­ importance to health depart­ information appropriately may lation deliveiy of preventiveaged and exchanged. Differentment sustainability make themservices rest on continuous carepredictions about these changes become vestigial— to the detrimentappropriate for considerationof their communities (and espe­ access and accountability. Millionsmay produce markedly differentin the context of accreditation.cially to vulnerable underserved will still lack health insurancepredictions of required health de­ (Am J Public Health. 2014;104:populations). The pace of change (especially in states decliningpartment capabilities. 1592-1600. doi:10.2105/AJPH.Medicaid expansion),4 move fre­ will challenge many health de­ Some have asserted that ACA2014.301956) quently,5 and experience providerpartments. Emerging advances inreforms (more people insured,turnover because patients and health information technology can preventive services covered, and American Journal of Public Health | September 2014, Vol 104, No. 9 1592 | Government, Law, and Public Health Practice | Peer Reviewed | Foldy et al.GOVERNMENT, LAW, AND PUBLIC HEALTH PRACTICE insurers will shop for value. TheHealth departments will need to • Which prenatal care factorsaccountable for the completeness,ACA reduces hospital reimburse­ accept electronic referrals as othercontribute to local infant mor­ accuracy, and transparency ofment, which may disrupt uncom­ providers do. Whether serving in­ tality disparities? whole-community health informa­ pensated care, although newsured or uninsured patients, in• Which infants lack follow-uption, ensuring public access to un­ nonprofit requirements may offsetpartnerships or separately, health for abnorm al newbornbiased information about thethis. Thus, discontinuities in insur­ departments will be expected to screens? health of the community andance, care, and information (pend­ deliver and document services • W here do concentrations ofsometimes providing impartialing improved health informationmore efficiently than before. people live with chronic diseaseassessment of the performanceexchange) will perpetuate preven­ Thus, departments will need to risk factors who are not receiv­ of health care providers and net­ tive service gaps. The ACA followsassess and satisfy patients’ needs ing preventive services? works. This accountability will re­ earlier reforms (e.g., health main­ in a timely, coordinated way, notquire health departments to useSuch information is required fortenance organizations, managedat the convenience of fragmentedand steward individuals’ healtheffective public health assessmentcompetition) that, despite greatprograms. Patients’ longitudinalinformation. and assurance and is supported bypromises, yielded mixed and modestand cross-program information The demand for person-linkedunique reporting mandates andchanges in preventive care.6 will be needed at the point of information across diverse sourcesHealth Insurance Portability andDespite recommendations to shiftservice and for performance im­ and systems will also grow be­ Accountability Act consider­ from service delivery to assurance,provement. This requires chang­ cause of the recognition of “syn-ations.15 Although accountablefrom 1997 to 2008 22% of localing the program-based culture of demics,” for example, interactionscare organizations may performhealth departments maintained orbetween HIV, tuberculosis, andmany health departments withwithin-network measurement andincreased the types of dinical ser­ syphilis and other sexually trans­ silos of disconnected informa­ may report similar metrics, thisvices they offered.7 In 2013 moremitted infections; relationshipstion.13 Cross-program business in­ will not assess people and prob­ than 90% offered immunizations,between obesity, diabetes, andtelligence is becoming business aslems falling between networks.more than 80% offered some in­ hypertension and other cardio­ usual in the private sector and isPrivate providers are also reluc­ fectious disease screening or treat­ vascular diseases; relationshipsbecoming critical to creatingtant to share information thatment, and more than 60% offeredbetween social factors, infant“learning health systems” thatmight affect competition forsome chronic disease screening.8mortality, adverse childhood ex­ continuously improve services.14 patients or payers. This conflictUnder Massachusetts’s ACA-like re­ periences, and chronic diseaseHealth departments that do notbetween competition and trans­ form, public health programs expe­ morbidity; and the relationshipsdeliver direct services will continueparency may create demand forrienced varying changes in demand,of these syndemics to health dis­ to receive and manage sensitivehealth departments to create im­ and an incapacity to analyze clientparities, whose elimination arehealth information for disease andpartial metrics of quality anddata impeded planning.9 Manya national priority.18 Segregatedinjury surveillance, outbreak andsafety, as they do for vital events,departments will continue servinginformation silos are inadequateemergency management (e.g., vac­ uninsured and hard-to-reach diseases, and injuries.16 Legallyfor studying and managing suchcine or drug countermeasure ad­ individuals,10 and expectations that authorized, neutral public healthsyndemics, which require a syner­ ministration), and maternal, child,they bill insurers, when possible, registries (e.g., immunization reg­ gistic (personcentric and needsand environmental health. Thewill compound information man­ istries and disease registries) oftenbased, not program based) para­ capability to match identifiableagement needs.1 1 have established community-widedigm for public health servicesrecords over time and across allBeyond clinical services, depart­ information sharing even as otherand the information systems thathealth care providers is needed toments may be invited to partner in health information exchange ar­ serve them. track important community-levelcase management and community-rangements fall prey to competi­ Thus, future health depart­ questions, such as, based services (foreshadowed intive pressures.17 ments must maintain and improveACA initiatives, e.g., community• What proportion of persons withTherefore, provider-controlledtheir capability to receive, secure,health teams and home visit pro­ HTV lacks antiviral treatment or aggregate reports may augment,manage, link, analyze, and usegrams).12 Success in such partner­ screening for tuberculosis andbut cannot replace, the mandatedindividuals’ personal health infor­ ships cannot be assumed becausesyphilis? universal reporting of identifiablemation for many purposes. Pri­ there is cost competition in the• Who received a first but notindividual reports to health de­ vacy and security will remainresource-constrained accountablesecond immunization for a newpartments. We believe health de­ critical concerns, requiring ongo­ care organization environment. pandemic influenza strain? partments will continue to be held ing capability building to stay September 2014, Vol 104, No. 9 | American Journal of Public Health Foldy et at. | Peer Reviewed | Government, Law, and Public Health Practice | 1593G O V E R N M E N T , LA W , A N D P U B L IC HEALTH P R A C T IC E DECISION SUPPORT laboratory results, and cancer clinical— not public health— use,ahead of emerging securityrisking misreporting or misinter­ threats.19 In many jurisdictions, diagnoses and care. In stage 3pretation. Data elements and vo­ Health care professionals relyprivacy concerns have led to pro­ (beginning 2015), EHRs may beon timely public health informa­ required to display patient vacci­ cabularies used in EHRs for pub­ hibitions against sharing informa­ tion, but delivering actionable in­ nation histories and incomplete lic health reporting must betion across public health programs.formation in the context of carethoughtfully defined (nationally)Unfortunately, such barriers limit immunization alerts from publicto improve medical decisions isdepartments’ ability to meet clients’ health immunization information and consistently used. a challenge. EHR clinical decisionneeds, improve programs, and systems.21 Previously, information • Manage larger information vol­ support systems can monitor careprotect the public health, when, exchange relationships between umes. For example, electronicand trigger alerts to improve di­ ironically, private companies now laboratory results have in­ health care providers and healthagnosis, treatment, and diseaseroutinely link identified data for creased report volumes overdepartments were established idi-prevention at the point of service. marketing and other goals of lesser manual methods.24 Processosyncratically using a variety ofThe EHR incentive program ispublic importance. standards and methods, and they automation can help staff man­ accelerating the adoption of stan­ Fortunately, technical advances age the increase. often failed to scale up to includedardized clinical decision supportcan facilitate the management, • Respond in real time to urgentlarge proportions of the popula­ systems. In addition to helpingsecurity, and use of tomorrow’s information. Health depart­ tion or to achieve operationalclinicians adhere to staticgrowing information challenges ments should leverage elec­ efficiency. For example, many im­ evidence-based practice rules,and may obviate the need formunization information and elec­ tronic reporting for fasterclinical decision support systemsevery health department, state anddata-driven suppression oftronic laboratory results systemshave been used by health depart­ local, large and small, to maintainemergencies, such as hepatitisreceive reports from only a mod­ ments to signal when a patient isall needed systems locally. A and meningococcal out­ est proportion of providers despiteat particular risk from a local out­ supporting multiple information breaks.25 break or a recent drug recall (situ­ ELECTRONIC HEALTH• Protect privacy and security.exchange formats.22 The newational clinical decision support).29RECORDS Electronic information mustcombination of national standardsFor example, a provider caring forbe protected during transmis­ and provider incentives createsa preschool patient with diarrheaThe federal Electronic Healthsion, storage, and use toa compelling opportunity (if notcan be alerted to a current localRecord (EHR) Incentive Programavoid loss, corruption, andrequirement) for health depart­ daycare-associated dysentery out­ (often called “Meaningful Use”),diversion. ments to migrate toward morebreak, potentially improving diag­ which began in 2010, is acceler­ universal, rapid, and automatedNew organizations and tech­ nostic and therapeutic decisions.30 ating health professionals’ andelectronic communication withnologies may facilitate publicEHR systems can solicit infor­ hospitals’ adoption of EHR sys­ providers’ EHR systems.23 Thishealth access to EHR information.mation from public health immu­ tems.20 The requirement for EHRcould increase the ascertainment,Health information exchange or­ nization registries to alert providerssystems to be certified to newspeed, and efficiency of reportingganizations can facilitate reportto immunization deficiencies andinteroperability standards (whichbut requires health departmentdeliveiy and record access.26from prescription drug-monitoringenable machines to exchange andcapability to do the following: Distributed data-mining protocolsdatabases for evidence of sub­ use information with minimal hu­ increasingly allow health depart­ • Update public health systems to stance abuse.31 Clinicians can alsoman intervention), combined withments to actively query EHRs tonew interoperability standards be alerted to opportunities to ad­ incentives to achieve meaningfulaugment or replace passive sur­ for secure transmission (e.g., dress health disparities on the basisuse objectives of public healthveillance of provider-initiated re­ of elevated risks in a patient’s geo­ the Direct project protocol),reporting, is creating more stan­ ports (while concealing sensitive graphic and demographic cohort.32 formatting (e.g., HL7 versiondardized information exchangepersonal identifiers, if desired)27 Health departments are2.5.1), and vocabulary (e.g.,between health care providers andBecause health departments re­ uniquely able to provide local,SNOMED-CT). health departments. main accountable for surveillance timely, and population-based• Interpret and improve theThe rules of meaningful usedata quality and completeness, information; thus, they havequality of information derivedstages 1 and 2 specify messagea unique obligation to supportfrom EFIRs. Public health they must decide whether suchformats and vocabularies forsuch situational clinical decisionsystems increasingly rely on arrangements are to be used and,reporting immunizations, syn­ support in clinical EHRs as the information recorded for if so, how.28 dromic surveillance, electronic American Journal of Public Health | September 2014, Vol 104, No. 9 1594 | Government, Law, and Public Health Practice | Peer Reviewed | Foldy et al.G O V E R N M E N T , LA W , A N D P U B L IC HEALTH P R A C TIC E technical capability grows. As fa­ Crowdsourcing has alreadyfrom the increasing availability ofdisplay with minimal human in­ miliarity increases, health depart­ assisted disaster and outbreakmultiple overlapping granulartervention. Natural language pro­ ments should also deploy decisionmanagement37 and many peopledata sets). cessing and sophisticated algo­ support in their own systems tovalue opportunities to communi­ Technical inequalities, for ex­ rithms might reduce the need tomanage caseloads more efficientlycate about potential hazards andample in computer skills or high­ standardize data tomorrow, butand effectively. events.38 Online information fromspeed Internet access, are some­ meaningful use standards canoutbreak “cases” may replacetimes called “the digital divide.” Assubstantially advance automationENGAGING THE PUBLICmuch future public health inter­ the use of electronic tools fortoday.45 AND THE DIGITAL DIVIDE viewing (as it has replaced voicehealth become more widespread,The second is to enable infor­ interactions for many purchases,such inequalities must be identi­ mation, currently separated intoMembers of the public are in­ travel reservations, and appoint­ fied and managed to avoid rein­ program-oriented silos, to becreasingly using personal healthments). Information collectedforcing health disparities. Trendslinked on the basis of person,records (PHRs), EHR patient por­ electronically from affected indi­ sometimes confound expectations.specimen, location (e.g., address),tals, social media, and mobileviduals is sometimes more usefulFor example, among cell phonelicensee, and event (e.g., outbreak),health tools. These support patientthan is that obtained face-to-faceusers today, African Americansthus allowing users to more easilyengagement in health care but canand filtered by health care pro­ and Hispanics are more likely toexplore and understand informa­ also facilitate information ex­ viders.39 look up health information usingtion in context. An early examplechange for public health surveil­ Internet polls and surveys aremobile devices than are Whiteis the “child health record” linkinglance, health promotion, research,becoming more important asnon-Hispanics.43 Uneven diffu­ information from multiple sources,and other purposes. Patient ac­ landline and cell phone surveyssion of technology in the commu­ such as birth records, newbornceptance of sharing PHR informa­ lose representativeness.40 Constit­ nity also affects health departmentscreening, lead screening, and im­ tion for public health purposesuents and policymakers will alsocosts. For example, persistence of munization records.46 (with appropriate privacy protec­ likely expect health departments,parallel paper systems may frus­ Similar efforts are needed fortions) is high and increasing.33 like other successful businesses, totrate anticipated savings fromsystems such as those to manageSocial media (eg., http://www.use Internet feedback to improveelectronic reporting. foodbome outbreaks in whichpatientslikeme.com) offer sharingservices and products.41 Method­ complex information on patients,and networking about health is­ ological issues abound to ensureINFORMATION ANDlaboratory specimens, food prod­ sues beyond one’s medical team.that the noise of high volumes ofKNOWLEDGE OVERLOAD ucts, and food-handling licenseesThe capability and inclination oflay information can be appropri­ must be interpreted in concertindividuals to send or broadcastately filtered and structured toThese developments, togetherwith speed and efficiency. Infor­ information over the Internetreveal meaningful signals forwith increased use of genomic andmation can continue to be stored(sometimes called Web 2.0, con­ health departments, but as in otherphenomic data and the network­ separately (for security or othertributed data, or crowdsourcing) isbusinesses, rising quantities ofing of sensors (Internet of things)considerations) but must be ac­ growing rapidly and becomescontributed and social media in­ in home, work, and the environ­ cessible to integration applicationsmore potent with ubiquitous smartformation are coming whetherment, will rapidly increase thethat can assemble it meaningfullyphones and mobile tablets, whichdepartments are prepared or not. volume of information that healthfor use in different ways (e.g.,can add photos, video, and geo­ The public also expects un­ departments manage.44 Healthcaring for an individual, under­ graphic position.34 Standards fa­ precedented information access indepartments will need to sort thestanding an outbreak, protectingcilitating patient downloading andreturn. “Data liberation” is a fed­ data flood into actionable infor­ vulnerable populations insharing of their EHR informa­ eral policy to make informationmation for various users’ needsa disaster). tion,33 combined with the ubiq­ that public agencies hold accessi­ through 3 critical capabilities. Finally, this automation and in­ uity of smart mobile devices, mayble to both individuals and appli­ The first is to leverage interop­ tegration must be designed toenable powerful platforms forcation developers.42 Health de­ erability for automation: to use thesupport the specific workflows ofpublic health surveillance, tailoredpartment data stewards shouldincreasingly standardized formatsdifferent types of public healthalerting (eg., notifying asthmaticsexpect to wrestle with increasedand vocabularies of high-volumeworkers: to help them performof air quality problems), and per­ public data sharing while manag­ data streams to automate tasks oftasks efficiently, effectively, andsona] health decision support (eg,ing privacy hazards (eg., mosaicreceipt, validation, sorting, distri­ safely (i.e., a user-centered design).when to seek care).36 effect reidentification hazards bution, storage, filtering, and For example, EHR systems September 2014, Vol 104, No. 9 | American Journal of Public Health Foldy et al. | Peer Reviewed | Government, Law, and Public Health Practice | 1595GOVERNMENT, LAW, AND PUBLIC HEALTH PRACTICE Assuming ongoing progress in departments have convertedprinciples gained from accumu­ designed to support episodicserver closets to other functions. lated information and inference, cloud reliability, speed, security,health care delivery may frustrateand cost, cloud technology willlongitudinal case management or e.g., evidence-based practices, ex­ TOMORROW’S CAPABLElikely allow health departments tooutbreak investigation, and sys­ perience) is growing alongside theHEALTH DEPARTMENT lease technology online less ex­ explosion in data and information.tems focused on collecting data forpensively than by locally purchas­ This is driven by 3 factors: anstate or national use may neglectHealth department roles willaging public health workforce ing and maintaining servers andthe workflow needs of local casechange with health care financewhose turnover requires transfer software. Such doud-based ar­ investigation. Analyzing and im­ reform, but their need to use andrangements can also facilitate se­ proving business processes and of experiential knowledge; grow­ protect personal health informa­ cure information sharing betweentheir subsidiary workflows is ing volumes of public health andtion will still increase. Higher vol­ organizations, programs, and ju­ prevention research; and the ev­ a critical first step in user-basedumes of more timely informationrisdictions when appropriate. Fordesign. Because many of the most ermore interdisciplinary naturewill need to be integrated andexample, BioSense 2.0 leveragesof public health work.48 Theseurgent and complex informationcloud capabilities for syndromic used more rapidly to improverequire access to and sharing oftasks related to case and outbreakoutcomes. Capabilities such assurveillance, facilitating data sub­ knowledge and expertise, includ­ management occur at the localmission from EHRs nationwide those in Table 1 will be neededing policies, procedures, and prac­ level, redesign should build fromand permitting the voluntary shar­ to manage information to thetices; bibliographic and traininglocal workflows upward to ensureing of data, software, and analysis greatest advantage. resources; and subject matterand improve the information sup­ products horizontally across juris­ Stand-alone local solutions willexperts who often as not workply chain (even when statewide orbecome obstacles to necessary in­ outside a particular health depart­ dictions and vertically from localnational information systems sup­ teroperability and informationment. Technologies to share through federal levels.50 port these functions)47 sharing (with health care pro­ knowledge synchronously (live, The power and cost-efficacy ofAll 3 capabilities (automation,viders, the public, and betweenincluding webcasts and telecon­ such shared platforms may proveinformation linkage, and user-local, state, and federal levels).suffidently enticing to overcomecentered design) are required to sulting) or asynchronously (storinghealth department commitments Cloud-based solutions offer econ­ useful knowledge for searchingmanage overload and convert in­ to separate jurisdictional and pro­ omies of scale and simplified in­ and retrieving on demand) areformation into better outcomes.formation sharing, but privatedeveloping rapidly. Using such grammatic data and hardwareThey require an enterprise-widesolutions will not emerge sponta­ technologies effectively is an silos. Programs and departmentsinformation architecture (the ca­ neously. Public health is a nicheemerging core capability for will have to accede to standardspability to access and use infor­ market complicated by conflictingfor defining and coding data ele­ mation across silo systems to meet future health departments.49 jurisdictional and program re­ ments and greater uniformity ofbusiness needs). This is increas­ quirements. Successful platformsCLOUD COMPUTING workflows before they can enjoyingly considered a core capabilitywill require agreement on infor­ convenient, scalable cloud solu­ of modem organizations both formation governance, data stan­ Improvements in Internet ac­ tions. For example, information in­ routine business processes (e.g.,dardization, and, critically, healthputs and outputs to manage a caseservices) and for performance im­ cess, speed, and distributed com­ department requirements andputing now enable practical access of tuberculosis must become moreprovement. National standards arecapabilities such as those pro­ to massive computing power, ap­ uniform before cloud-based solu­ necessary but not sufficient. Pri­ posed in Table 1. plications, and data sets “in the tions become practical nationwide. oritization, planning, and execu­ These capabilities (whethercloud” (i.e., on the Internet) instead Health departments will needtion of information architecturegreater focus on information managed locally or supported byof on local servers. This allows the(inside and between health de­ state or national cloud services)purchase of infrastructure as a ser­ management (the competenciespartments and programs, sup­ have important, near-universalvice (i.e., online computing power), associated with public health in­ ported by national standards andimplications for health departmentsoftware as a service, or entire formatics) than on technologyaligned program funding) are stillsustainability, planning, budget­ platform as a service (i.e., an online management (server and networkrequired to ensure that timelyadministration) during and after ing, workforce, and technology,environment combining access toactionable information reachesthis transition. Fortunately, these making them appropriate for dis­ computing, software, and datathose who need it. are the same competencies that will cussion in the context of accredi­ sets) from an expanding collectionThe challenge of managingremain in demand long after health tation. We hope there will be of public and private providers. public health knowledge (truth or American Journal of Public Health | September 2014, Vol 104, No. 9 1596 | Government, Law, and Public Health Practice | Peer Reviewed | FoMy et at"