New data collection procedure to assist in treatment of autism

Modernization breeds the desire for effective treatment in the cheapest manner possible. One of the main challenges is the lack of integration of various data sets that are available for further research and development.

Dr. Isaac Kohane of Harvard Medical School and his research team have developed a method to send out queries to various electronic health record systems in order to form an integrated data set composed of clinical and biomedical data. The research has been able to identify sub-classes of autism spectrum disorder (ASD) successfully. The output of the research was heavy on information and more extensive in scope than any other previous studies of its kind.

Also, by sending out queries to several electronic health record systems, Dr. Kohane’s team were able to follow privacy regulations which they could have otherwise violated had they collected information for basic aggregation. Hence, this method is also a commendable alternative to data collection that is both secure and honest.

Two weeks ago at the Strata Rx 2013 conference in Boston, Dr. Isaac Kohane showed in his presentation that integration of data is plausible and effective.

Dr. Kohane’s team extracted information from physician notes because these notes reveal more reliable and more thorough information that may otherwise be unavailable when using diagnostic codes. After getting information from these notes, the team used the R statistical software. Through this method, they were able to meet the objectives of the study. Some of the ASD clusters they were able to identify were epileptic, bowel disorder, schizophrenia and autism. Each of these clusters indicate different characteristics which mean different treatment necessities.

Dr. Kohane also developed another test for autism in a different research study. He compared blood samples from both autistic and non-autistic people. By doing so, he was able to identify a certain set of genes that indicate strong possibilities for autism. This research is a step forward to finding out ways in detecting autism at the earliest stage. Early detection would consequently mean easier treatment as well, and maybe, cheaper medications.