One of President Obama’s strong points during his campaign was the use of technology to reach millions of people and deliver a message. Part of that message was increasing the use of modern technology in government, and advocating modern technology in certain private sector industry that require government oversight. One such sector is energy, with the Obama Administration calling for the energy industry to green up the grid using 21st century technology. Another industry is the medical industry.
Electronic medical records has been a hot debate for years now. With advances in software, hardware, and networking, our technology is now able to provide the information doctors need about patients during a diagnosis, treatment, or surgery. About ten percent of American hospitals have some sort of electronic medical record system and about sixteen percent of private practices have implemented health IT.
The problem with the electronic medical records software that’s been used so far is interconnectivity. As more and more hospitals and private practices adopt software to manage patient records, the portability of those records will be in jeopardy. There are no standards for what data is stored in a medical record and in what format that data is stored. Senator John D. Rockefeller has proposed the government work with private sector companies such as software providers, medical hardware providers, and health care and insurance providers to implement open-source software and open standards for medical records. By having open standards, all stakeholders in universal medical records will be able to read and manipulated the data. Patients will be able to keep their records portable across any system any provider offers. Customers changing insurance, primary care physicians, or relocating, will be assured their records can come with them. Transcription errors can be reduced significantly.
By creating open-source software to manage these records, small hospitals and clinics will be able to implement a solution without having to fork over a lot of money. Larger hospitals, insurance companies, and conglomerates like Kaiser Permanente will be able to purchase closed-source yet open-standard software from software providers that provide technical support. Studies show that after implementation, usually two to three years, improved efficiency from such a system is shown to become profitable, generally paying for itself in about 5 years. Decreasing the cost of the software, and added services such as online access, will increase profits faster for smaller providers. Doing so will eventually lead to lower prices and better quality of service.
Defining Specifications
The biggest foundation needed for universal, highly detailed and accurate, and efficient medical record keeping is a standard specification of the storing of those documents linked to a record. By providing a standard developed by all interested parties, adoption rates will increase. If every party can agree on the specification, the industry as a whole will be far better off. But what should the specifications include? The following is a (somewhat incomplete) list of what should be included inside of a modern electronic medical record:
- Properly formatted patient data, such as name, social security number, drivers license, and common medical information such as blood type, known allergies, etc.
- Proper vital signs data with attached history. All recorded blood pressures, heart rates, O2 absorption rates, etc.
- Detailed medical history of the patient, cross-referenced with any blood-relatives in the system on an opt-in approach. Relatives would need to opt-in their histories to be linked.
- History of prescribed drugs and medical reason for each drug. Current dosage information should also be included. Interactions between drugs should be shown in the event a new drug needs to be prescribed to the patient.
- Laboratory Tests should be saved for current and future referencing. Since no test can be completely irrelevant, all should be saved for future review.
- Images of scans such as X-Rays, MRIs, CTs, and photos. Being able to look through a series of dated scans can show doctors problems they might not be able to see if the data isn’t available to them.
Data Mining
While most people will be opposed to anonymously sharing their medical records for 3rd party companies, some patients might decide to opt-in on allowing companies to mine the data in their records. Having universities and private companies doing research using medical data from thousands or hundreds of thousands of participants could add valuable insight into drug trials, or disease trending. If DNA profiles are eventually added to medical records, that data can also be used to form correlations between genes and disease at a much higher rate than simply doing unrelated studies. Privacy of a patients records is of the utmost priority, and if the data can be handled in a complete anonymous manner, it can be a useful research tool as long as patients know their personal identities will not be used in any studies.
Building Infrastructure
One of the biggest drawbacks of moving toward electronic medical records is a storage and retrieval mechanism that works for everyone. With paper records, each provider has alot of duplicated data in many different places. No longer will doctors offices and hospitals need rooms and rooms of filing cabinets. But medical records in electronic form can swell to gigabytes of images, test results, notes, and other information per patient. Since internet access is becoming ubiquitous, the best solution is a central storage location for medical records. The most probable location would be a patient’s insurance company. Since HIPAA regulations allow for patients to change insurance providers, being able to migrate their electronic medical records during their change of coverage would be simple to implement.
Insurance companies would work with providers they support to allow access to the records. The software specification should include some kind of compression format and incremental download solution since large files would take a long time to download. One method would be to download only the information needed immediately, and optionally selecting data that might be useful to be background downloaded. Obviously the cost of bandwidth is a concern, especially to facilities that handle hundreds of patients a day and require more complete records such as hospitals. Intelligent caching of records could speed up access and retrieval. If insurance companies have facilities in multiple states with datacenters at each facility, medical records could be distributed geographically to best serve all interested parties. Patients could also register with hospitals and primary care facilities and the software could dynamically update local copies of patient records with whatever changes have happened since the last update. Ideally, something like git or svn could be built around patient records.
Building the centralized storage solution for medical records will also have some cost. However, using server software such as Linux and open source database and security software will cut costs. Essentially, the cost of building the server portion of the solution is simply the cost of the hardware. Hiring technicians to install and maintain the system will be the same no matter what software is used. Its been proven that using open-source software can dramatically reduce upfront cost, and security is next to none.
Integrating Devices
Many different manufacturers make many different medical devices that adhere to that manufacturer’s specifications. In order to increase efficiencies, manufacturers would need to provide the specifications for their file formats so software engineers can write conversion functions. Eventually manufacturers can use the records specs to program their devices for a more open format. As an example, if a better MRI image format is introduced with the open medical record specs, GE could update the software running on their MRI machines to produce the updated format. Other devices such as blood-glucose readers and heartrate monitors or any other remote wearable device would still function appropriately, and with plugins written by manufacturers or open-source projects, the uploaded data could be properly translated into the open format.
The most important thing with an open format is keeping the data normalized. All the data stored for individual patients would be stored in the same units for easy comparison to normality, or to other records during research.
As Part of the Stimulus
Healthcare in the United States is definitely on life support right now. With the new administration wanting to change the landscape of how healthcare is provided in this country, moving to a more technologically advanced means of keeping records is an important step to cutting costs, easing malpractice suits, and generally providing a better quality of healthcare. Doctors will be able to better provide for the patients under their care and be able to do it cheaper. Nurses will be able to see visually in real time how hospital patients are reacting to treatments, and can better plan how to provide care for someone who’s just had major surgery. Primary care physicians will be able to better diagnose patients coming in with aches and pains if a proper medical history and crowd sourced data is available for particular patients. Getting there is the biggest obstacle. Luckily funding for such an upgrade, or at least the first basic steps, has been provided as part of the stimulus. Health IT will create many good paying jobs all across the country for software developers, IT professionals, teachers, and with new efficiencies the money saved by healthcare providers can be used to hire more nurses at better salaries.
Moving forward, the groundwork for an Open specification needs to happen very soon. All interested parties should sit down and come up with standards everyone can agree with and publish them online for review. From there, existing software firms can update their software, open source software can make use of the new specifications, and new companies can be formed in order to roll out and support these new systems. Server farms will need to be built to handle the massive amount of data pouring into the system, and many jobs would be created in this whole process.
On a personal note, a friend of mine has a sister-in-law who is now without both of her legs because of a clerical error in a hospital. Its an error that could have been avoided if proper testing and records were available at the time. I’ve been told that particular hospital has now upgraded some of the technology it uses, but it didn’t happen in time to save her legs. Recently, the court found in favor of the doctors who presided over her, but considering the jury couldn’t really understand what happened and how it happened, the judge has granted a retrial. If all of this hadn’t happened, insurance companies wouldn’t have to be paying out so much in malpractice, premiums on malpractice insurance would go down, premiums on health insurance would go down, and quality of care could be increased dramatically. I’m sure that not only would I, nor my friend, nor anyone else reading this could not benefit from having their medical history available to any doctor currently treating them.
Referenced links
Open letter to Obama: Uncle Sam should go open source
Testing Drug Efficacy Using Electronic Medical Records
Electronic medical record
Medical record
Studies Show Electronic Medical Records Make Financial Sense
As always, if you find the information in this article interesting or informative, please subscribe to my RSS Feed, or leave a comment and tell me what you think.
