A few years ago, my then flatmate Gabriel came home frustrated and angry after being asked to manually analyze more than 150 patients’ medical records. He was doing clinical rounds as a medical student and spent eight days reading records and manually extracting information, which he then had to copy into an Excel file before he could perform a simple data analysis. One evening he said, ‘It’s unbelievable that we don’t have any tools capable of doing this’. Thinking more about this dilemma, we came up with the idea for IOMED Medical Solutions. It took us three years to understand that we were dealing with something that could actually become a company, but we eventually quit our jobs in order to work full-time on IOMED.
The root of the problem is that all the information collected about a patient is simply written down as text in a medical record and saved to a server. That means there are millions of inaccessible medical records saved in hospital servers all around the world. In Catalunya, there are more than 300 million medical records, and that’s only in the public sector. Now imagine what kind of information that is: insights on diagnoses, symptoms, treatments, findings, outcomes. Information that we have but can’t touch—that’s a huge problem. Studies have shown that structuring this data can help hospitals save up to eight percent of their total annual healthcare expenses.
IOMED transforms the information found in written medical records into structured data that can be more easily analyzed. The technology we have developed scans the raw material of the medical records, finds keywords, extracts them, codifies them and saves them onto a database. That way we can find patterns in the data—information that the hospitals can then plug into their business intelligence software to help doctors make more informed decisions.
With the structured data, hospitals can also identify patterns and predictions, which allow them to tackle inefficiencies within the healthcare system. A clear example of this is readmissions (when a patient leaves the hospital with a certain diagnosis and treatment plan, but comes back with the same symptoms in less than 30 days). By analyzing millions of similar cases, we can generate a percentage of how likely a patient is to be readmitted. Our predictive algorithm can also pinpoint patients who are at risk of developing sepsis in a hospital, rejecting a prosthesis and so on.
Very few hospitals actually generate and maintain their own systems, instead paying large IT companies to do the job for them. These IT companies are our direct clients. They implement our solution and make sure the right indicators appear in the business intelligence tool of the hospital.
Learn more about IOMED at iomed.es.