2 Interoperability Computers need to interoperate so they can store and act on data from other computers in a reliable and beneficial way without human intervention and translation. [ Marshall Ruffin, M.D., M.P.H., M.B.A., F.A.C.P.E., Accenture ] W hat does interoperability mean? Imagine you have traveled to London from the United States and arrive in the morning after an overnight flight; you fall asleep in your hotel to overcome jet lag and awake in the midafternoon somewhat revived and eager to clean up and get something to eat. You shower and prepare to use your corded electric razor only to discover that the electrical outlet in England does not receive the electrical plug of your American razor. You suffer from a failure of mechanical interoperability — the plug doesn’t match the wall jack. So you ring up the front desk and you ask for an adapter. A member of the hotel staff brings the adapter quickly and you plug it into the wall socket. Now your razor can interoperate with the electrical system in England, right? Not necessarily. When you turn on the razor nothing happens. The electrical standards in England, the voltage and amperage, are different than they are in electrical systems in the U.S. So while you have mechanical interoperability with the converter given to you, you do not have electrical interoperability. For your electric razor designed for the U.S. to work in England, you need both an adapter for mechanical interoperability (plug-to-jack interoperability) and a transformer for electrical interoperability, which changes the voltage and amperage to what your razor requires. At this point either you ask the hotel staff if they have a transformer, or you ask the hotel staff to sell you a manual razor and shaving cream. The electrical appliance you brought from the United States does not interoperate with the electrical system in the U.K. on two levels because mechanical standards and electrical standards do not match. Similarly, imagine that after you shave with the manual razor you dress and leave your hotel hoping to find an appealing restaurant. Looking about you do not see one, so you approach the near- est person to ask directions. That person happens to be a visitor from Germany who speaks little English and does not know the city at all, having just arrived the day before from Berlin. Here again you face a failure of interoperability. That person does not speak your language well enough to converse with you. You turn to another person walking by dressed in a tweed jacket and kilts and ask him about restaurants in the neighborhood and he replies in exemplary English that he is a visitor to London from Inverness, Scotland, and cannot recommend any restaurant. You have interoperated in language but not in content knowledge. The third person walking by is Asian and smiles agreeably so you ask your question with little hope of interoperating. To your surprise and delight she answers in a British accent with a “cheers” and a question about what type of restaurant you would like. She is second-generation English and her parents are on the faculty of University College London. With her you can interoperate with language and content. Now think of interoperability as it relates to computers. The computers must connect together in some way, over some network, in order to interoperate — meaning they must physically connect to the network in the same way with the same sort of plugs, such as ethernet plugs. They also need to share operating systems and network protocols in order to connect one to another, so one computer recognizes the other computer on the network. Anyone who has tried to add a computer to a network knows the bewildering complexity of protocols, switches, IP addresses, authentication and identification, and other facts needed to achieve interconnectivity between PC and network. Interconnecting a PC to a network is the first step, but not the only step, to interoperability. Once the PC, whether wired or wireless, connects to the network it will not give its user access Marshall Ruffin, M.D., M.P.H., M.B.A., F.A.C.P.E., is a partner of Strategy and Business Architecture for the Health & Life Sciences practice of Accenture. He recently served as the initial clinical director of the National Care Records Service for the National Health Service in the United Kingdom. Dr. Ruffin currently is the executive director of the Interoperability Consortium (Accenture, Cisco, CSC, HP, IBM, Intel, Microsoft, Oracle) that is developing the technical reference architecture for the regional health information organizations in California. 064|Healthcare Technology, Volume 3 Achieving Interoperability & Collaboration to an application, such as an electronic medical record (EMR), without proper authentication by ID, password, retinal scan, voice print or other means of reliably identifying the user and determining that he or she may access the contents of the EMR application. Once the PC has connected to the network and the user has been given access to the EMR application, the user may be able to access the content of patients’ records. A human can interpret the language on the screen and make sense of it, provided the coded numbers for diagnoses and laboratory results and pharmaceuticals and procedures are accompanied by their translations in English. A computer cannot do these things. A computer needs codes it expects, otherwise it will not understand the content of a message it receives from another computer. A human who understands English can see an ICD-9CM code and, if the English translation accompanies the code, understand the meaning of the code. A human can do the same with a SNOMED-CT code; that is, understand it, as long as the English translation of the code accompanies the code. A computer cannot understand, so to speak, an ICD9CM code if it is expecting a SNOMED-CT code, and vice versa. Interoperability requires standardized language (terminologies, data, information models and document structures) and messages so the integrity of the content of the information exchanged is preserved. Without interoperability, there can be no useful exchange of information. Without interoperability of clinical content between computers, one must print paper records from one computer, a human must interpret the information content on the paper and then that human must type into the second computer the information in a format the second computer will store reliably. This happens in the hospital pharmacy when a fancy physician order entry system used by doctors on the inpatient care units is not integrated with the pharmacy information system, so the fancy CPOE system, unbeknownst to all those hospital employees and doctors proud of the new CPOE system, prints its orders in the pharmacy on paper and a pharmacist types the order into the pharmacy information system. In most exchanges of healthcare information between providers, between providers and payers, between doctors and pharmacies and durable medical equipment A scanned document, unless formatted carefully and passed through an optical character reader software application, is an image in the recipient information system, available for viewing but not for extraction of data for entry into a database. Computers are not as facile with English and data as we humans. To store, communicate and act on information they need data carefully formatted, with codes they expect, properly formatted and in messages in the right order. To interoperate, computers need a carefully defined structure of messages, using standards such as Health Level 7 (HL7); and information, using a reference information model (RIM) such as HL7 RIM, and documents; using common document formats such as HL7 clinical document architecture (CDA). Without these standards, computers cannot reliably exchange the information content of laboratory results, operative reports, radiology interpretations, diagnosis codes, discharge summaries, prescriptions and lists of allergies without human intervention and translation. Without these standards, the computers cannot interoperate — though they may be on the same network. So why is interoperability important? It permits the automatic exchange of clinical data from one computer to another while maintaining the integrity of the information content, so diagnosis codes sent from one computer are interpreted and stored as diagnosis codes by the recipient computer, and allergy data sent by one computer are received by another computer as allergy data. The network is essential but not sufficient for interoperability. vendors and between physicians and hospitals, the translation function takes place by people interpreting paper records printed from one computer before entering by hand the content into the second computer. Think of all the claims that pass from doctors’ offices to insurers on paper, having been created in the doctors’ office computers and printed to paper for mailing to the insurers, and the legions of clerks who take the paper records and type their contents into the insurers’ claims processing systems. Think of all the paper prescriptions hand written or computer generated by physicians and typed into pharmacy information systems by pharmacists or clerks. Think of all the paper reports from consulting physicians and reference laboratories and diagnostic radiology groups received by doctors’ offices and either scanned or typed into those physicians’ office practice electronic medical records. A scanned document, unless formatted carefully and passed through an optical character reader software application, is an image in the recipient information system, available for viewing but not for extraction of data for entry into a database. Scanned documents generally do not contribute content to databases. Interoperability of content with standardized terms and messages, www.HCTProject.com|065 2 Interoperability and data and information models permits computers to exchange coded data and to interact on the data exchanged, so that an electronic medical record can alert a physician when it receives an abnormal laboratory result from a laboratory system, such as from a blood gas measurement in an intensive care unit that can trigger changes in oxygen concentration, or ventilator rate per minute or tidal volume of the ventilator oxygenating a patient. Computers need to interoperate so they can store and act on data from other computers in a reliable and beneficial way without human intervention and translation. The velocity of actions speeds up and treatment decisions can occur more quickly than they can with human intervention for translation. The economic benefits of interoperability are enormous. Imagine you had to speak with an operator every time you wanted to make a cellular telephone connection because the various cellular networks did not interoperate. You use Sprint. Your friend uses AT&T. You would need to call a Sprint operator, who would call an AT&T operator, who would dial your friend’s telephone number before establishing a circuit for you and your friend to speak before both operators can hang up. For this model to work, almost the entire population of the U.S. would have to become operators in order to facilitate the number of telephone calls we Americans take for granted every day. All the telephone companies in the U.S. have adopted interoperable standards for dialing and connecting telephone calls so that computers can switch calls and set up connections without expensive and slow intervention by human operators. The cellular and landline telephone networks interoperate, and so our society enjoys enormous productivity benefits. So too with the Internet and the World Wide Web — the standards permit interoperability insofar as Web pages on a server attached to the Internet anywhere can be viewed by computers with Web browsers anywhere. Through a Web browser you can schedule a hotel room or a rental car or reserve an airline ticket, all because your Web browser formats data in the way that the computer systems of the hotel and rental car company and airline expect the data to be formatted. With this same technology for interoperability, you can check the weather, the value of your stocks or your medical record without another person having to translate your request into the format of another computer. The economic and political benefits of interoperability derive from the automatic exchange of information between computers. 066|Healthcare Technology, Volume 3 Interoperability reduces labor costs and increases the velocity of information exchange; it expedites and lowers the costs of commerce. Interoperability allows computers to exchange information without human interference, assuring doctors that laboratory their patients’ results will be available in their patients’ electronic medical records as soon as the computerized laboratory equipment have calculated the results; it means a doctors’ staff can instantly send claims for a day’s work to the patients’ various payers, and the doctors’ practice management system will receive from the payers’ computers, just as quickly, results of claims adjudication and remittance advice for those claims found acceptable by the payer. Delays with paper processing that may be on the order of days and weeks in the interoperable digital world shrink to seconds and minutes. Cash flow improves and clerical staff work declines — provided the physician’s computer system has been set up to interoperate with the computer systems of the payers used by his patients. There are costs to interoperability in hardware, software and network access, and in standardization of data and information in clinical and financial transaction systems that exchange data. However, once the standards have been implemented and the computers can interoperate, the velocity of commerce increases, the friction of interaction between businesses and their customers declines and the costs of commerce decline. In the short run, adoption of digital communications and standardization of information exchange hurts and costs money. It hurts because it requires change in the ways business happens. It costs money because hardware and software and network access cost money, and because the language of work must change from free text to coded information interpretable and storable by computers. In the long run, the standardization reduces the costs of commerce and reduces the costs of both producers and consumers, especially for activities that require searching many sources of information, or updating sources of information remotely. Think of the friction of finding sources of information in paper form on obscure topics, the publications of a little-known author or the medications used to treat an unusual tropical disease. Visits to libraries and searches through card files and stacks of journals often take days or weeks. Now with a Web browser and the English language, you can perform those searches in seconds with PubMed or Google. How can a large multispecialty group practice using paper medical records alert its patients taking Vioxx that the Achieving Interoperability & Collaboration medication has been recalled by the manufacturer and they should contact their physicians to discuss other treatments? It will have to alert all of its patients by mail, since it has no simple way to scan paper charts for those patients who have received prescriptions for Vioxx, and no way of knowing if those patients who responded to the letter represented all of the patients taking Vioxx. On the other hand, with an EMR and software for e-prescribing, the group practice can identify and contact only those patients who have received prescriptions for Vioxx and make certain it reaches each and every one of them directly, increasing the velocity and accuracy of care and reducing the friction of information management. What steps can a healthcare community of doctors, hospitals, pharmacies, nursing homes, home health agencies, durable medical How do interoperable EMRs form in communities? In some countries with centralized healthcare systems, such as in England, the government defines the interoperability standards. The National Health Service (NHS) has defined the data and messaging standards for a summary EMR for every person in England, and calls that summary record the “National Spine.” Right now the computer systems used by all the providers of healthcare in England, including dentists, home health nurses, nursing homes, diagnostic centers, pharmacies, general practitioners, hospitals and consultants are receiving new EMR software that will communicate with the National Spine. England has one source of organization and funding for this gargantuan effort, the NHS, which is planning to spend more than $20 billion over 10 years on the effort. England has one source of organization and funding for this gargantuan effort, the NHS, which is planning to spend more than $20 billion over 10 years on the effort. equipment vendors and dentists (to name a few of the providers of care in a community) take to establish interoperable digital health records which they can share? For providers, payers and patients to share the myriad of standards for messaging and content of medical records, they must share a complete technical reference architecture for defining the documents and content to be exchanged between participants on the network. A community will have as much a political as a technical challenge to establish such a reference architecture. When any two people or organizations want to share data, they need only agree on standards among themselves. When a company wants to exchange data in standard format with all its customers or all its suppliers it faces a one-tomany challenge, and needs to establish the standards for communication and convey them to all those with which it wishes to communicate electronically. However, in healthcare, a patient may have many providers of care and a provider will have many patients, with many forms of insurance and many pharmacies of choice. A community sharing medical records electronically produces a many-to-many challenge that requires some third party to define the standards for the community — either the government or the community itself in the form of a shared organization. The community needs a trusted broker to establish the shared standards for all participants in the exchange of EMRs. In the paper world, the analog are the carriers of paper — the U.S. Postal Service, FedEx and UPS. In the digital world, for interoperability, every computer needs to connect to a shared network and share messaging and format and content standards in order to exchange medical information reliably. No aggregation of individual point-to-point exchanges will serve the purpose, at least not economically. In the U.S., with our pluralistic and fragmented industry and dislike of big national government programs, we are moving toward a model for interoperable EMRs that promotes standards for interoperability and gives providers of care and payers for care economic incentives to modify their information systems to send and receive patients’ data according to those standards. The standards will be defined by consortia of private sector organizations such as the Interoperability Consortium (started by Accenture, Cisco, Computer Sciences Corporation, HewlettPackard, IBM, Intel, Microsoft and Oracle) and endorsed by the federal government through the Office of the National Coordinator for Health Information Technology. HHS and other payers now are expressing their strong interest in the widespread adoption of interoperable EHRs and are formulating the government grants and contracts to initiate development of regional health information organizations (RHIOs) that will establish communication-based interoperability, data security and privacy standards for all providers and payers and others who would exchange healthcare data in digital form. Because the costs of interoperable electronic health and medical records will be lower than those costs for maintaining and communicating paper records, the federal government believes that after catalyzing the conversion from paper to EMRs in some communities — with grants — the business case for EMRs will become obvious to most medical communities, and regional health information organizations (RHIOs) will appear spontaneously, adopting the interoperability standards that the first governmentfunded RHIOs will adopt, based largely on the Federal Health Architecture that already exists. ■ www.HCTProject.com|067
© Copyright 2026 Paperzz