I published a few papers during and right after my first M.S degree all in Artificial Intelligence (AI) and its use in healthcare. For years after that, while working in the corporate world I did not publish. So its very exciting and gratifying to see that so many papers cite my research. It tells me that the use of AI is increasing and being utilized more widely in Healthcare.
My next paper will be published this year.
A wise Corporate Strategy professor at MIT said “Strategy without numbers is poetry and organizations are not in the business of writing poems”
Most executives and managers in different industries use dashboards to monitor their top areas of interest a well as keep track of industry news and trends. As Peter Drucker stated “If You Can’t Measure It, You Can’t Improve It. ”
In healthcare, patient monitors have been around for years. Patients are monitored for their vitals signs and the data is also available at the nurses station. However dashboards that have hospital data or individualized data based on function are now coming of age.
A C-Suite executive at a hospital should have a customized dashboard to monitor Quality, Financial metrics, Re-admission rates and Patients Satisfaction for the hospital on a daily basis rather than wait for the monthly or annual report from the particular department. The dashboard would have not only descriptive analytics, but also predictive and prescriptive ones providing insights and plans.
Why should the doctor not have a customizable dashboard to monitor critical patients vital signs and trend as well as lab results, clinical plans and pharmacy approvals? So also the Director of Quality should be able to regularly monitor Quality goals and how the current measures compare to it.
In the era of Medical Interoperability, Electronic Health records and Internet of Things, customizable healthcare dashboards will be more readily available so turn around time on critical issues facing healthcare improves.
All newborns born in the USA are federally mandated to undergo a few different tests within the first 24-48 hours of birth. These include a blood test taken from the baby’s heel called dried blood spot (heel) test because the blood spots are collected on a special filter paper and then dried before sending the paper to the state appointed lab. This test checks for developmental, metabolic and genetic conditions. The other tests are a hearing test and one for congenital cardiac heart disease.
Since these tests are federally mandated, each state has to keep records of the tests conducted. If a baby is found to have a metabolic or genetic condition (about 10% of babies need further testing), follow up care is provided by the state to a certain extent.
Only a few states, however, have systems to keep track of all the tests for all the babies born in the state. In most states, each test is handled by a different system. Because there is no one system for all the information, these states find it very difficult to get data on percentage of babies born with a particular condition. If the baby gets adopted or their name changes, the state resources are stretched even further trying to find the baby data.
States need one system (One baby, One record) to store and retrieve all newborn screening records for each baby. This system needs to be connected to the birth hospital electronically so demographic data can be messaged over, thereby reducing costs and errors of re-entering the information. And the system needs to be connected to birth certificate records so that each state can be certain that every baby had the necessary tests and are being followed up as needed.
Medical Interoperability is being able to collect all the healthcare data pertaining to a patient in one location so a clinician can make value based decisions based on comprehensive patient information. For true medical interoperability, there needs to be a patient oriented focus where data is collected across the different healthcare disciplines software and medical devices.
Most outpatients are connected to at least 3 medical devices and interact with at least one healthcare software during their visit and inpatients are connected to many more. These devices and software are often not sending data to the same repository.
Medical devices and software have an internal collection and storage of patient data for viewing and creating trends. For this data to be made available in a patient centric location, it would need to be retrieved from the source and then converted to a format that is understandable. The conversion of data from all the disparate sources into a common understandable format is called semantic interoperability. Semantic interoperability is needed for making sense of the data and analyze it for meaningful interpretation.