Healthcare and IT

First it was the office, then banking/finance and now the healthcare space is the next domain to be revolutionized by information technology. Almost 15% of US GDP is spent on healthcare expenditure, so the market opportunity is enormous for the solutions that would save the costs in this field. Information Technolgy application in healthcare has been in the form of cutting edge computer graphics/visualization solutions for medical imaging. There are specialized software for effective visualization of the data gathered e.g. 3D reconstruction of organs from Xray/CT/MRI scans. But data management in healthcare has been a forgotten child. Also with the advent of Web 2.0, mobile and ubiquitous computing paradigms, we need to think of innovative approaches to improve healthcare IT. Below are some of the applications I have thought of :

-RFID for hospital asset management and location tracking. RFID is the new wave in asset tracking. They are being used right from theft prevention in Barnes and Nobles to inventory management in Walmart. RFIDs in a hospital can be a powerful technology. They can be used to track the equipment right from the time it arrives in the hospital. They can also be used for sensing the location of machines etc in real time. Imagine a controller's console which gives an interactive map of every floor/room of the hospital and lists the items that are in there. Another application is to account for all the Biohazardous waste or tracking the objects in an OR. So the next time there is an intelligent system to point out that there is a gauze piece missing. This system again can be linked to the hospital's real time order management system. So for example all the equipment and medicines/food needed can be ordered just days in advance for the surgeries scheduled and hospital wastage can be reduced.
Patients tagged with RFID bracelets can be used to track them through various treatment workflows and standard workflow improvement techniques like parallelization/pipelining can be used to increase efficiency for reducing patient wait time and increasing personnel/equipment utilization.

-Hospital planning : Right now in most of the ER rooms there are whiteboards detailing the doctors on call. All this can be easily automated by provided large screen interactive information screens (like Microsoft Surface) which update the status of the doctors in real time based on location sensing , lets say of the doctors' PDAs.

-Information Integration : As has already been done in the Azyxxi system all the patient information can be integrated for a unified view of the patient data including MRI scans , Xrays etc captured directly at source and associated with the paient ID sensed automatically lets say the RFID bracelet . Thus the information on a patient can be made available in real time. Visualization of this integrated information is a possibility , an example being all the prescriptions given to the patient depicted on a timeline or a slideshow to gauge a fracture's healing progress.

Integration of the data at the hospital or the hospital system level is another possibility. Innovative dashboards coupled with BI data processing , can be constructed to give a high level view of the different activities in the hospital. Consider an app through which the planners can see in real time all the financial, operational parameters like ER admissions and perform better capacity planning. They can drill down in different dimensions of the data and see what kind of specialists are more in demand, what procedures are more frequent and what equipment is most used.

-Patient Data availability : The patient data can be stored in a central repository which is accessed from a web portal like the MS healthVault. In health data privacy has been a major concern, so the data can actually be encrypted and stored locally with the patient, in lets say USB drives or the patients can opt in for the data to be stored within the portal. The idea of a persistent storage with the patient is compulsive because the storage itself can have the encryption keys in it to access the public data. This way the patients would not have to fill in the information forms everytime they go to a new provider. The thumb drive can plug in to any windows computer running "doctor's version of Windows" and the data accessed/updated replicating it with the encrypted public vault too. Imagine wearing a thumb drive which can act as a life saviour by providing all the vital medical data (drug allergies etc.) in an ER.

-Web 2.0 : Web 2.0 has enabled easy collaboration betweeen people using intutive interfaces. Also it has demonstrated the power of crowdsourcing. Web 2.0 apps again have not targeted the health domain barring a couple of companies like zocdoc, InSTEDD and a handful of other health oriented social networks. There is a great need for a portal with different dimensions of health mapped in the mashups. For example track my IP location and show me all the doctors in my area for psychological disorders. Then enable me to set-up an appointment and remind me automatically on my smartphone on the specified date. Also the doctors can share the research and can be effective in detecting pandemic based on the sysmptoms observed by doctors in areas where the outbreak has actually happened. Consider a doctor in China who has dealt with patients suffering from H5N1 and he can share his findings with the doctors in US through this social network.
Also the plain old social networks of patients are invaluable in detection of community problems. Imagine a social network of patients with elaborate dashboards. You can input any symptom and it gives you details of what can be the causes. So if someone enters the symptom "tingling sensation in fingers" the portal shows the dashboard stats based on the other users who have reported the same symptom. So it shows 90% of the people have tunnel carpal who reported this sysmptom, 5 % have MS and so on. You can input demographics like age and the system spits out more precise/focussed information. This system would calm down the hypochondriacs and give early warning signs to at-risk individuals.


-Crowd sourcing : can be used for early diagnosis of diseases. As the healthcare costs rise, a collaborative framework can be built which uses the doctors (available at lower than US rates) in countries like India and China for early detection of diseases. They get paid for each positive identification they refer to a provider (and maybe penalized for a false positive). This network can also be used for time sensitive tasks like detection of tumors in xrays/CT scans. The doctors can opt in to for consultation or medical students can even be trained on such a collaborative set-up. Imagine an ebay for medicine. It might sound unethical but it is just discussed as a possibility. After all we hoped to find Jim Gray by such a collaborative approach.

Also digitization of legacy health records has been a big problem. Again crowdsourcing can be used for this purpose by using workers in low cost countries for digitizing the data. Benefits of this digitization can be demonstrated to the insurance providers to foot the bill for such an effort.

A rather benign scenario is having your own personal trainer sitting in a remote part of the world. You come in and sync your workout data (collected from a device like nike pod) over the web and he gives you food and exercise advisories.


It is estimated that post-hospitalization neglect costs around $10B in healthcare costs. These stem from missed medication to poor monitoring. The monitoring aspect can be solved by having remote patient monitoring personnel (akin to a healthcare call center) who monitor the vitals over web enabled, bluetoothed devices and can also give reminder calls for medications, doctor visits, perscription fills etc. Such a network coupled with intelligent devices can be effective in geriatric care too.

-Convergence : all the information sources need to be unified and given the technologies available today it can be easily done. e.g. a prescription can be pushed into a smartphone in the form of SMS alerts for drug time reminders. This data can also be relayed to a MSN-Direct enabled watch worn by the patients. Also bluetooth interfaces can be built into BP/Glucose monitors which automatically communicate the vital data to the Portal and onto the doctor if anomalies are detected. Inside the hospitals the different machines can all feed the data onto a single visual monitor like Microsoft Surface, which will contain an intutive interface for the doctors and nurses on rounds to check the vitals, history etc.