The U.S. spends more per capita on healthcare than any other country. Yet, a 1999 report from the Institute of Medicine called To Err is Human found medical errors to be responsible for approximately 98,000 deaths in the U.S. every year. To put this in perspective, the number of deaths attributed to medical errors is about the equivalent to crashing a fully loaded 747 every day for a year.
Those two statistics illustrate the main challenges facing medicine today: runaway costs, and an unacceptable number of poor outcomes. To be sure, the Institute of Medicine’s eye-opening report is a decade old. But if the number of deaths due to medical error has changed, the change hasn’t been dramatic enough.
As we all add more gadgets and devices to streamline our daily lives, the next question is, what can the tools and advancements of modern technology do to fix healthcare?
On the national level, with the current administration’s focus on passing healthcare legislation, there has been a push to try to fix healthcare’s woes from a policy perspective. Much of the legislative effort has been aimed at making healthcare affordable and accessible to all. There is also a push, however, to try to change the healthcare system’s approach to medicine.
“The goal is to move to outcome-based medicine rather than the traditional reactive, treatment-based medicine. To the degree that you can keep people out of hospital beds by getting them to do pre-emptive and preventative things that is the degree to which you are focusing your healthcare dollar on better outcomes,” says Jeff Hawley, Senior Manager in Accenture’s Health and Public Service practice.
There are several steps that modern technology can take to begin to improve the efficiency and the outcomes of our healthcare system, and they begin with data. Retail and online merchants have perfected data management— the concept of capturing and storing the purchases we make, the items we view online, and all other interactions and putting this information to use to make recommendations and targeted offers. Those data and analytics practices hold strong potential for the healthcare industry as well.
“In healthcare we have notions of what kinds of people “walk the aisles” because we know their diseases, but do we really know them? Do we know their genetic profiles? Is that a possibility in the future? The answer is yes,” says Hawley. “Using proteomics and genomics we could look at genetic profiles and identify candidates who might be at higher risk for cancer. If you are cancer free today, but there is a high probability of cancer based on your profile and family history, your doctor could recommend a surveillance program for early detection and treatment. . We could design treatments based on genetic profiles. That is one of the future outcomes of medicine: using data analytics to effect much more enriching outcomes.”
All of this begins with capturing the data. You can’t analyze what you don’t know, and the task of capturing the data will involve encouraging increased use of electronic health records and an attempt to transition the healthcare industry to a more paperless environment. The American Relief and Recovery Act of 2009 (ARRA) set aside more than $30 billion in health IT, much of it to encourage hospitals and doctors to adopt electronic health records. Hospitals are incented to show “meaningful use” of electronic health records, and the bill similarly incents doctors to implement electronic health records through two different programs – one through Medicaid and the other through Medicare. Electronic health record adoption is currently around 18 percent, says Hawley, which is actually up dramatically from eight percent two years ago.
The data, of course, must be captured and stored in standardized nomenclatures so it can be transportable and usable. One such format is Logical Observation Identifiers Names and Coding (LOINC). There are others. The office of the National Coordinator for Health Information Technology created the Health Information Technology Standards Panel (HITSP) and HITSP’s Health Level Seven (HL7) protocol is a messaging standard for electronic data exchange in healthcare. The World Health Organization’s International Statistical Classification of Diseases and Related Health Problems (most commonly abbreviated as ICD) is a format for diagnoses.
“In the U.S. we use ICD9 and we are woefully behind many of the developed countries,” says Hawley. “ICD9 has about 45,000 codes for diagnoses. In Europe they have been using ICD10 for a number of years, which is much more descriptive and has 155,000 codes.”
The transport and sharing of data is an area where Health Information Exchanges (HIE) are emerging. These organizations, often regional, manage contractual relationships with the providers, maintain and further standards, and arrange means of electronic exchange of health data. Hawley suggests this might be one of the areas where concepts like cloud computing, open source and virtualization might aid the industry.
“You don’t have to host the infrastructure inside the walls of the facility, just tap into it and exchange information using the cloud,” he says. “Trends like open source, cloud computing and virtualization are creating an environment that might lessen the load and the need for integration and interface engines to transport data back and forth while reducing cost to serve.”
To be sure, there are privacy and security issues surrounding the capture, transport and management of healthcare data. But Hawley says there are data techniques currently in place aimed at addressing them.
“Data privacy is paramount in the minds of the consumer as well as the practitioner. Systems must have robust security protocols and schema in place to protect private healthcare information. At the same time, we need to meet the needs of the healthcare community in a way that protects patient confidentiality yet serves the public need. Medical data for secondary purposes must be protected. The concept of data pseudonymization and anonymization are key techniques in meeting these needs, he says.vacy and security for HIPAA, but the ability for data to be synonymized and anonymized,” he says.
“Anonymizing” means permanently deleting any patient-specific identifiers from medical data, so that data on diseases can be collected without infringing on anyone’s privacy. Population medicine would be a typical use for such data. “Pseudonymizing ” data is similar, except with pseudonymized data, identities are scrambled but the scrambling can be undone with appropriate security. This way, for example, if someone looking at psudonymized data spots a problem that needs to be brought to a patient’s attention, the patient identity can be retrieved so that patient can be notified.
All this data collection, transport, and management is paving the way for data analytics. The move towards evidence-based medicine is facilitated by analyzing data from disparate sources and combined with data analytics, provides a data-rich environment for effective decision making. Keeping in mind the 98,000 medical errors noted earlier, effective use of data and information coupled with technology is providing the platform for reducing these errors., Hawley says. When the data from disparate sources is analyzed and clearer pictures emerge as to which method of treatment has been most effective based on a larger sample, treatment processes are optimized. As data drives decisions in retail, where suggestions are made as to what we might want to buy, eat, or wear, rules based on empirical data and alerts can suggest methods of treatment to doctors, or suggest tests that should be performed before treatment proceeds.
Summing it all up, reduction of medical errors, appropriate and secure access to medical information and the ability to manage data information that lead to better outcomes is the goal that these varied technologies bring. Data collection, by means of electronic records, would ensure that health data would not be lost. Standardized aata transport and management would allow a patient’s medical record to be accessible by varied parties – from primary care physicians, to surgeons, anesthesiologists, radiologists, lab technicians and all others who might have a hand in his or her healthcare. Data analytics has the potential to inform and alert the caregivers to allergies, medical histories and underlying conditions and recommend courses of treatment, thus potentially avoiding an adverse outcomes and unnecessary cost.. The combination of these technologies has a potentially larger role as well. Enabling the transport and sharing of data gives the industry a larger study group of patients and diseases to examine. Layering sophisticated analytics on top of that data means medical researchers could potentially spot more quickly patterns and trends that it might have taken an individual doctor a lifetime of experience to see. The investment in technology in conjunction with a paradigm shift in how we think of the art of medicine—driving towards empirical data supported by those essential “soft skills” promise a more effective healthcare delivery process.
