Archive for the tag 'Business Intelligence'

“You know my methods, Watson”: IBM’s Watson to enter the Healthcare world

Technology in healthcare is taking a huge step forward. Wellpoint, Inc has announced that they will be using a commercial version of IBM’s Watson supercomputer.

Not too long ago, a room full of computer hardware once computed at a power less than what our cell phones currently do. Now, a room full of computer hardware will equate to a computing entity with the intelligence to assist physicians with medical decisions.

You may know Watson best for its performance on the Jeopardy game show. Watson demonstrated swift decision making after indexing over 200 million pages of data. Watson would only answer if the system crossed a certain confidence threshold.  The confidence threshold was a predefined percentage set inside the system. When Watson referenced the data, it determined the percentage to which it was sure the top three answers were correct. If the percentage of the top answer crossed the confidence threshold, Watson would signal for the answer. The IBM machine proved itself successful against two humans competing in the game show by winning both rounds.

Certainly physicians and members have much to gain from the assistance of a machine that can reference millions of pages of data to ascertain a diagnosis or treatment.  While physicians may always hold the upper hand to interpret the context of the situation for a presenting patient, Watson’s assistance can certainly supplement any decision using vast amounts of data in a quicker time frame.

In an article posted by EMR and HIPAA, it noted that “One of the keys in the AP article above and was also mentioned by Dr. Nick from Nuance was that the Watson technology in healthcare would be applied differently than it was on Jeopardy.  In healthcare it wouldn’t try and make the decision and provide the correct answer for you. Instead, the Watson technology would be about providing you a number of possible answers and the likelihood of that answer possibly being the issue.” The article later went on to state:  “Saying that perhaps 25 percent of all healthcare errors are errors of diagnosis, Kohn [IBM Chief Medical Scientist Dr. Marty Kohn] noted how getting the diagnosis right can prevent all kinds of unnecessary complications and spending. “Of course, if you’ve made the wrong diagnosis, picking the right course of treatment becomes a challenge,” Kohn said.

So how might this affect the EHR world? The electronic EHR would be used as a reference for the Watson system. Previous prescriptions, orders, lab results, presented problems, among others, would all contribute to Watson ascertaining a confidence threshold.  Once a confidence threshold is reached or passed, the system would suggest a route of possible treatment, or determine a possible diagnosis.

With the advances in accuracy, these decisions can come back to the EEHR and certainly provide more efficiency and cost savings for the practice. The technology undoubtedly proves to be a win-win situation for all players in the healthcare industry.

 What do our readers think?

Allscripts Analytics CrossTab

Analytics Analysis, also known as crosstabs, is a tool for sorting data. It allows users to drill down into the data using fields available from the database and is very flexible. It is often used for viewing data at multiple levels, comparing data across sites or providers, and providing detailed worksheets about the data.

For more Galen webcasts visit: http://www.galenhealthcare.com/calendar/

Interface Transaction Processing Analysis

Issue:

A recent issue came up with one of our clients in that interfaced patient appointments from their Practice Management system were not making it in a timely manner to the EHR. The client witnessed that appointment messages built up in the interface queue and there was a delay in processing the messages. The client desired a resolution that would assist in speed up of the processing of the messages such that appointments booked in PM would render in the EHR quickly without a disruption to workflow.

Investigation:

Enter the ConnectR Toolbelt “Transaction Processing Time” report:

This report extracts transaction count, minimum, average, and maximum ConnectR processing time per hour. Using the report, the following analysis was conducted.

Findings:

Based on the aforementioned analysis, it was determined that in the clients Live Reg/Sched system target, blocked messages were being logged. Having blocked messages logged can be invaluable when first designing and developing interfaces. However, as evidenced in the analysis, it can lead to performance degradation as the system requires much less processing time when messages are not logged.

Outcome:

Logging of blocked messages in the Live Reg/Sched target was disabled on 6/30/2010 and as witnessed in the analysis spreadsheet the number of transactions decreased by roughly 70% and peak transaction processing time decreased by roughly 90%.