Consumer-centered Health Care Depends on Accessible Medical Records

There is a problem with medical records—they are scattered everywhere. John Quelch discusses approaches to integrate patient data so that medical professionals and patients can make better decisions.
by Dina Gerdeman

Your patient health care data is most likely scattered throughout the medical universe, in everything from notes scribbled by various doctors to test results resting in far-flung computer systems. So when medical professionals need to pull that information together to get a complete picture of your health, the process is a bit like throwing a lasso around a tsunami.

That's why integrated medical data has now become a major goal of health care reform. Integrating patient records, such as physicians' notes, diagnoses, medical device records, surgical procedure outcomes, and treatment histories, can be used in powerful ways, ideally leading to better patient care, lower health care costs and, ultimately, healthier patients.

“Trust on the part of both the consumer and doctor is hugely important in determining who is going to be the likely winner here”

There is certainly a market demand for solutions, but which type of organization will take the lead on a patient-data integration system that enjoys widespread success: hospitals, insurance providers, medical equipment suppliers, consumer tech companies, or some other entity?

Right now the answer to that question remains uncertain, but what is clear is that the organization will have to be one that both consumers need and doctors trust, says John A. Quelch, the Charles Edward Wilson Professor of Business Administration at Harvard Business School and Professor in Health Policy and Management at the Harvard T.H. Chan School of Public Health.

Quelch says integrating patient data is crucial to propelling the health care system to become more consumer-centric. Yet with many people concerned about privacy and skeptical about personal data being shared, he believes any organization looking to take the lead in this area will not only need the resources required to pull off such a complicated task, but will have to establish great credibility with the consumer.

Not only are medical records scattered, many are still paper-based. ©

"Trust on the part of both the consumer and doctor is hugely important in determining who is going to be the likely winner here," he says. "Some organizations may have the capability to integrate patient data, but may not have enough credibility with the consumer. And some may not have the motivation to realign their businesses in ways that would allow them to deliver this extra-added value service."

Electronic Health Records Promoted

In 2014, health care spending in the United States exceeded $3 trillion and was expected to reach $4.5 trillion by 2020. To reduce health care costs and improve care, some organizations have made it a priority to invest in information analytics, attempting to get their arms around the wide array of patient data. Government incentives are helping the push. In 2011, the US Centers for Medicare and Medicaid Services established incentives for doctors' offices and hospitals to switch from paper to digital, providing up to $2 million to hospitals serving Medicare patients that adopted electronic medical records (EMRs).

Quelch and HBS research associate Margaret L. Rodriguez released a case study in April about one organization that has ramped up its patient analytics: The Carolinas Healthcare System, based in Charlotte, which in 2014 owned and managed hospitals and acute care facilities in three states.

In 2011, Carolinas launched Dickson Advanced Analytics, which incorporated complex clinical, financial, demographic, and claims data to develop analytical tools that would speak to three strategic priorities: predict the health care needs of patients, continually enhance patient outcomes, and drive transformative solutions to import community access to health care.

For example, using patient data including emergency department visits and sodium levels, the analytics group helped clinical teams identify admitted patients who were at high risk for readmission to the hospital. Patients at risk of being readmitted within 30 days of discharge have hospital staff members who work with them on managing medications and connecting them with dieticians and others who can provide follow-up care.

"If you can predict the rate of readmission on the front end, you can put extra resources against that patient to minimize the risk of readmission," Quelch says.

“The consumer's time is being eaten up unnecessarily doing bureaucratic work”

In addition, the Carolinas analytics group has organized around 2,000 data points on each of its 2.2 million patients, including clinical data, medication use, education, socioeconomic factors, and consumer spending information to segment patients into groups, such as "high risk of cancer," so these segments can receive customized preventive messages and care specific to their needs.

The data have not only provided important information to improve patient care, they have also been used to support hospital investment decisions, such as which surgical devices to purchase, since patient data includes information on the quality of outcomes.

"Carolinas Healthcare employs 120 staff working on data analytics, a big strategic investment but a major potential point of differentiation in the quest to both cut costs and improve patient outcomes" Quelch says.

Other companies have made their own strides in health care analytics. For example, Quelch and Rodriguez wrote a case in May about Philips, which started in 1891 as a light bulb manufacturer in the Netherlands and, more recently, has become a major manufacturer of medical devices and software sold primarily to hospitals.

In June 2014, Philips announced its HealthSuite Digital Platform, a cloud-based infrastructure that supports the secure collection of information related to health and lifestyle, allowing hospitals and care providers to integrate and analyze data. Most health care providers do not have a complete data set that includes all of a patient's information; Healthsuite is an open platform that provides a secure and private home for data, allowing care providers and consumers to use the information to make decisions about patient care.

Philips executives are hoping the company's early entry in the market will provide a significant advantage, since hospitals tend to seek technology partners that can grow with them over a long period, rather than bearing the switching costs of starting over with a new platform and supplier.

Ideally, a sophisticated analytics system would allow consumers to have easier access to their own health records, providing the kind of "timely, accurate, and complete information that enables you to do a better job of taking charge of your own health care—preventative or therapeutic—to speed a healthy outcome," Quelch says.

Consumer Enslavement

Aside from enjoying lower health care costs and improved care, consumers stand to benefit by saving time handling appointments, referrals and the like. Quelch says he recently had to arrange an appointment with a specialist, was told by the specialist's office he needed a referral from the primary care physician, and was given a fax number. He then had to call his doctor's office to pass along the fax number so the document could be sent.

"This is not consumer empowerment. This is consumer enslavement," he says. "The consumer's time is being eaten up unnecessarily doing bureaucratic work. When there is cost pressure—and there's a lot of criticism about health care costs—the natural tendency is to shift work to the consumer."

About 18 percent of US gross domestic product is spent on health care costs, and a certain portion of those expenses are related to operational inefficiencies that could be alleviated with big data analytics and efficient IT systems that handle a good chunk of the work, Quelch says.

Yet consumer trust when it comes to data privacy is not always a given, and consumers want to have a say. Some people fear that information about a serious disease diagnosis or a visit with a mental health provider could be shared with their employers and used against them—and the consequences could involve missing out on a promotion or even losing a job.

"No one will be able to integrate all of my data for the benefit of doctors without consumers also being involved," Quelch says.

Handicapping The Players

So which organizations have the best shot at creating an integrated health care data system that enjoys widespread adoption and acceptance? Quelch outlined the possible players:

  • Hospitals: Large organizations like Carolinas have created their own data systems. But the hospital world is fragmented, and unlike Carolinas, many hospitals don't have the resources, commitment, and management capacity to build and oversee a complex data analytics system.
  • Insurance companies: Large insurance companies, such as UnitedHealth and Humana, are increasingly focused on analytics. Some use claims and consumer profile data to target high-risk patient segments with advice about exercise, nutrition, and other programs in an effort to minimize medical visits. The problem: insurers' claims data is not as current as hospitals' clinical data. In addition, insurers have a perception hurdle to overcome: "They are not well trusted by consumers," Quelch says. "Can a health insurer counter out of the mold and break the standard prejudice against trusting insurers to develop a closer relationship with the end consumer?"
  • Device companies: Device companies like Philips have created health instruments used in hospitals and homes to monitor millions of patients. Philips generates over one-third of its global sales from health care. Other companies like GE and Samsung are also active in health care. These brands enjoy some consumer credibility, but they tend to be viewed as device companies, rather than as data companies. Partnerships could alleviate this concern. For example, Philips recently aligned itself with, a leader in customer relationship management approaches that could be applied to patients.
  • Consumer tech companies: Consumer tech companies including Apple and Google have entered the industry with health care dashboards that track and aggregate data from various health wearables, such as heart rate monitors and step counters. Apple is also partnering with several EMR companies to make medical records and lab results available to consumers on their iPhones.

    These tech giants have expertise in analytics and managing huge amounts of information, have access to demographic and location data, and enjoy broad consumer adoption of their devices. But do they have enough access to and necessary understanding of claims and clinical data? Quelch asked. And do they have enough credibility with physicians in terms of their health care expertise?
  • IBM: IBM has entered the health care data fray with IBM Watson Health, which is intended to bring together clinical, research, and social data from a range of health sources, creating a secure, cloud-based data-sharing hub that can help patient diagnosis and improve outcomes. IBM is trusted by many consumers, but because it has not specialized in health care, it may be perceived as lacking an intimate understanding of doctor-patient decision-making.

Quelch is planning to write additional cases, including one on an insurance company as well as one about IBM Watson Health, to explore advances by other players in the integrated patient data market.

Note to readers: John Quelch invites Working Knowledge readers to comment on which type of organization they think offers the most convincing combination of capability and credibility to pull off widespread health care data integration?

About the Author

Dina Gerdeman is a senior writer for Harvard Business School Working Knowledge

Post A Comment

In order to be published, comments must be on-topic and civil in tone, with no name calling or personal attacks. Your comment may be edited for clarity and length.
    • Charles Hewitt
    There is another player which should be included in creating the integrated health care data system: health information exchanges (HIEs). An example is CurrentCare, Rhode Island's statewide HIE. With your consent, CurrentCare collects clinical data from your health care providers, aggregates the data into a single, longitudinal record, and makes it available, with your permission, when and where needed to treat you. While the concept is simple, the execution is extremely complex for reasons financial, technical, legal, operational, political and social. (The writer recently retired as the director of the program that implemented CurrentCare.)
    • Dave Bath
    • Private individual
    The discovery and authorization to access various types of records needs a brokerage system with well-defined standard interfaces, and open specification with support from major systems manufacturers, and a trusted supplier of the specifications.

    This would permit easy assembly of components that fit together, developed by different suppliers with low barriers to market entry, and thus competitive forces would apply to quality and cost.

    Such a series of specifications, interfaces already defined in a number of languages, was defined by the respected Object Management Group ( back in 2001, and can happily deal with health related data formats such as HL7 and DICOM.

    The components include Clinical Observations Access Service (COAS), Retrieve Locate and Update Service (RLUS), Resource Access Decision (RAD), Terminolody Query Service (TQS, aka Lexicon Query Service LQS), and a host of others that allow direct access to laboratory equipment ... so that the broker can identify who is requesting what information, check authorization to view the information (not all records are necessary and appropriate for all health professionals), map the concepts where terminologies can differ (consider MeSH and WHO ICD lexicon mapping), find the most recent relevant authorized information and present it where it is needed. and integrate it with the Clinical Decision Support System (CDSS).

    The ability to locate the relevant information combined with the support for granular access controls allows data wranglers to get the maximum relevant data while navigating through tricky privacy requirements.

    Using these components, already well-defined, and the data format specifications, has to come at the requirements definition stages - or you end up with attempts to create monolithic systems, usually from a single supplier running things, creating messes such as those of the Australian NEHTA (National e-Health Transition Authority) which has led in well over a decade to no useful uptake by hospitals, let alone integration with smaller clinics, and with government auditors-general slamming all the efforts.
    • Sonny Abel-opurum
    • Consultant Lecturer, North London College
    At the heart of this topic is communication management as a vital constituent of Quality in healthcare. Centralized information that is accessible to all pertinent stakeholders is therefore crucial in delivering Consumer centered health care. A well informed and organised professional is better placed to deliver responsive, reassuring, sensitive and less costly care, that is in economics and legalistic sense. In other words it helps to eliminate wastage, reduce untidiness, perceived unprofessionalism and duplication. Moreover as the consumer is made pivotal in the care development and delivery system, through empowerment to articulate and express their views and preferences of the care service, to the participating team in real time, the prospects of enhanced care experience improves as information is captured and made accessible. Whereas the opposite would likely be the case, if information about the consumer is scattered and unaccessib
    le. Indeed a recipe for disaster.
    • Ira H. Kirschenbaum, MD
    • Chairman, Department of Orthopaedic Surgery, Bronx-Lebanon Hospital Center
    I believe that independent data aggregators/curators will eventually win in this space. Too many of the parties "handicapped" above are actually handicapped by their own biases, current customer bases, and long-term business strategies. Hospitals are too regional. Insurance companies only know risk management not healthcare management. Device companies need to sell metal. Apple and Google connect people only on their own corporate terms but not on the consumer's because they have to sell a piece of media or advertising to sell product. IBM is well situated to develop some innovation but deployment to the masses is not their strength. Once the fundamental dataset for the average patient is actually determined (no small task) then a social network aggregator/curator, that has no other interest except connectivity, where users opt-in and in essence opt-out of a certain level of privacy, with add-on third party apps, will give
    every user the range of information that each need.
    • Paul Jones MBBS FAMA
    • Family Physician/Director, Primary Health Care Ltd
    In Australia we wrestle with the same issues and have had government throw huge amounts at this problem over many years. We have a national Patient Controlled Electronic Health Record which is incomplete and which physicians have do NOT trust (inter alia as it has serious capacity to be incomplete). Many consumers are also wary of the potential privacy issues as well as having ongoing concerns around third parties (Government, Insurers) controlling or interfering with there care, a concern also shared by many doctors. Moreover, there is ZERO incentive for doctors to participate. The end result is a very expensive incomplete system with very low uptake, to the extent that Government appears set to roll out an "opt-out" model. I guess politically in the USA government role is unlikely so if I were to be a cheeky "outside observer" my answer in terms of trust might well be the consumer tech companies which clearly we all a
    s consumers knowingly or unknowingly already trust with huge amounts of data, often sensitive.
    • Kapil Kumar Sopory
    • Company Secretary, SMEC(India) Private Limited
    The problem of not having a patient related complete health record is the outcome of treatments taken from various sources many of which do not maintain such records. If everyone is on digital platform and a distinct identification number is allotted to each and every citizen, entire data would be readily available. This will facilitate proper understanding of case history and appropriate medication based on that.
    India is very backward in this respect and it may take ages to evolve such a system across the nation.
    • Merle Bushkin
    • Founder & CEO, Health Record Corporation
    Read this with great interest.

    The truth is that if you want all your medical records from all your providers available at the point of care anytime, anywhere even without Internet access, there is one--and only one--product available that meets your need. It's called MedKaz(r). Our company, Health Record Corporation, created it. It is the only product that provides total interoperability today and for the foreseeable future. We are the type of "independent data aggregator" Dr. Kirschenbaum referred to in his comment.

    To give you an idea of MedKaz scope and power, I have all my records for the past 30 years on my MedKaz. They include records of more than 320 encounters with more than 40 care providers in six cities and three states, in all formats, both paper and electronic (from 14 different EMR systems). And with two or three clicks, both I and my providers can bring up specific records. They are displayed as searchable PDFs in a browser. It is updated for me after every encounter.

    MedKaz is unique, as is its business model which pays doctors and hospitals to update their patients' MedKaz. There's nothing like it and it's available today!

    Our website,, tells all about it. The Video, FAQs and News & Announcements sections are especially informative. Also, the HBS Alumni Bulletin profiled it online at

    Merle Bushkin, AB 1956, MBA 1960
    • Sherpa
    The question posed implies that in order for a project to be highly successful it must contain elements of elaborate competition. In my opinion, this type of activity merely postpones the result that is needed sooner than later. An uber-collaboration, in my opinion, would bring this process to bear quickly and with a positive and sustainable product outcome. This product being a costs savings process rather than a revenue generating one.
    • Marlanges Simar
    • Customer Experience Architect, Prime Therapeutics
    I believe insurance companies are well positioned in terms of having the capabilities to pull off widespread healthcare data integration, primarily via medical claims. Medical claims include the structured data (labs, diagnoses, medicines, ambulantory care, long-term care, etc.) needed to better serve most participants of the healthcare industry (consumers, doctors, hospitals, medical devices, etc.) and possibly consist of the most complete picture that exists today (minus device-based trackers, health apps, over-the-counter remedies, etc.).

    Lack of trust by consumers in insurance companies however, as you stated, is a hurdle that must be overcome. Consumers tend to look to their doctors for health related recommendations, often as the single source of credibility. And, although insurance companies sometimes make health related recommendations to prevent disease progression, I don't believe consumers necessarily accept or adopt such advise as willingly as they do with advise from their doctors. The integrated healthcare data from insurance companies pushed to doctors via the EMR in my mind is well-positioned to deliver and enable a higher level of consumer-centered healthcare.
    • Anonymous
    The computerisation of patients medical records Pharma industry specifically clinical trials companies will benefit for easy harvesting of test results. Their stocks are human beings/patients.

    People are intentionally made ill so that they know the onset and the developement of diseases. I was intentionally made ill then tricked into signing the New Prescription Service. It was not mentioned that I will loose the Data Protection Act, that my record will be access the whole of NHS.

    I become a property of NHS like a slave, repeatedly subjected to clinical trials. Trying to find justice is like facing a brick wall.

    Centralised medical records will ease Pharma industry to pick and chose who to target.
    • Charles Hewitt
    Allow me to elaborate on my earlier comment which suggested that health information exchanges (not to be confused with health insurance exchanges) have a role to play. Health information exchanges (HIEs) facilitate the collection and disclosure of patient-level clinical information. With proper implementation and governance, I believe HIEs satisfy the trust requirements which are raised in the article.

    In the case of Rhode Island, if you choose to participate, the statewide HIE (CurrentCare) collects clinical data from your providers, aggregates and normalizes it in a single, longitudinal medical record, and with your permission, makes it available to your health care providers when and where they are treating you.

    The trust factor is huge. In Rhode Island,
    ? You must give your consent for CurrentCare to collect your data and to disclose it only to the providers that you designate;
    ? If you are being treated at a substance abuse program, you must give a separate consent to the program provider to enable them to send your data to CurrentCare;
    ? Providers must sign a data sharing agreement to enable them to send data to CurrentCare and sign a separate data use agreement in order to access CurrentCare records.
    As you might guess, most consumers are reluctant to sign up. Normally, it requires the intervention of one's primary care physician to convince a person to do so. It has taken Rhode Island about five years to get about half the state (500,000 people) enrolled. The whole process is expensive and results in a very slow ROI.

    Getting providers to use the CurrentCare services is also not easy. Here the main issues are insufficient panel penetration and workflow. Docs are understandably reluctant to use CurrentCare if it requires a separate logon and only a fraction of their patients are enrolled. Busy practices resist changing their workflow unless the advantages are obvious and compelling.

    On the other hand, providers who operate under value-based contracts view CurrentCare as the trustworthy entity which can enable them to track their high-risk patients as they access care outside the providers' walled gardens. However, they need a solution which tracks all their high-risk patients, whether or not they're enrolled in CurrentCare.

    Today, CurrentCare collects data from over 150 sources - hospitals, labs, diagnostic imaging centers, community mental health organizations, pharmacies, community health centers, urgent care clinics and ambulatory practices. The data types include encounters, lab results, imaging reports, medications and care summaries. Except for medications, data is automatically pushed to CurrentCare - no human intervention is necessary. Medication data is pulled to CurrentCare automatically when a provider accesses the patient's record.

    CurrentCare provides two basic services to providers: access to patient records and notifications to providers when clinically significant events occur to their patients. Access is either through a visual display (portal) or via a feed directly into the provider's electronic health record (EHR). Notifications happen when, for example, the patient is discharged from a hospital emergency department.

    The data in CurrentCare is incomplete today and will be for the foreseeable future. It lacks connection to most hospitals and other providers outside of Rhode Island. It lacks access to diagnostic images. It lacks access to ambulatory practices which still rely on paper charts or which use an EHR which does not share data with CurrentCare. Although CurrentCare is continually expanding the number sources and data types, it is unlikely ever to have a person's entire medical history.

    Nevertheless, most docs find value in CurrentCare today, especially if they can pull data from CurrentCare directly into their EHR. Although the information may be incomplete, the two in combination are better than what is available from any stand-alone EHR.

    CurrentCare is owned and operated by the Rhode Island Quality Institute (RIQI), a private non-profit entity. It operates as Rhode Island's HIE under a contract with the Rhode Island Department of Health and according to policies set forth in state law, the Rhode Island HIE Act of 2008. The RIQI board of directors consists of CEO-level leaders of various stakeholders in Rhode Island's health care delivery community.

    To date private voluntary donations are funding CurrentCare. However, at some point in the fairly near future, the funding will have to come primarily from fee-for-service revenue. There are interesting challenges ahead.
    • shane busby
    • executive director, strategy, Alberta Health Services
    Great article. Completely agree with the premise, the structure, the function, the purpose and ultimately the innovation and strategy behind Electronic Health Records. Now, if we can only convince users that we are using their health information, for "good" and not "evil" and that we will and can protect their privacy, then we will get (someday) to where we want to go and need to be.
    • Dr (Major) Shishir Basarkar
    • chief operating officer, CARE hospitals , India
    What the Dina has said I agree with all the points raised. The MRD is the core of any and every healthcare institution and it's sanctity should be maintained with utmost care.