tag:blogger.com,1999:blog-6010387589635528818.post5859782280036569721..comments2024-03-29T05:12:25.517-04:00Comments on The KHIT Blog: Are structured data the enemy of health care quality?BobbyGhttp://www.blogger.com/profile/03807934795994985233noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-6010387589635528818.post-61703395070279789362015-12-15T14:29:41.965-05:002015-12-15T14:29:41.965-05:00Thank you, Dr. Carter. Great observations.
"...Thank you, Dr. Carter. Great observations.<br /><br />"it is not clear to me how useful routinely collected data, structured or otherwise, is for research or other analytical purposes."<br /><br />Well, my example here of MU "Smoking Status" acutely begs that very question. As clunky a metric set as it is, you'd think we'd have seen some CMS aggregate tabulations by now. Maybe it's just as well. The snazziest MU Core 9 crosstabs in the world will not tell us anything we don't really already know.<br /><br />When I first started in health care analytics in 1993 for one of the Medicare "PROs" we were able to extract a good bit of value from simple old "HCFA" UB-82 claims data sets. You would think that actual structured CLINICAL data emanating from EHRs would be even more worthy. Back then we had to send out teams of nurse abstractors lugging laptops to go to indicated facilities and cull clinical data from the charts.<br /><br />That aside, the beef as set forth by Margalit and HIT critics more broadly is that such data don't add any value to the patient visit and only serve to interfere with an already too busy interval of time in the exam room. Again, that's a mix of UX suboptimality and "productivity treadmill" imperatives.BobbyGhttps://www.blogger.com/profile/03807934795994985233noreply@blogger.comtag:blogger.com,1999:blog-6010387589635528818.post-76241457638692518182015-12-14T14:34:42.559-05:002015-12-14T14:34:42.559-05:00Hi Bobby, you raise an interesting point. Struc...Hi Bobby, you raise an interesting point. Structured data entry can be very time-consuming, so it is not unreasonable to see it as a problem in some instances. However, like all things, results vary. I have been involved in EHRs since the 1990s when physicians bought systems without incentives. Many of them wanted templates, especially to improve E&M coding reimbursement. Fact is, not all templates are the same. Some are poorly designed, others not. <br /><br />The effects of MU on interface/template design cannot be overlooked. In trying to keep up with cert requirements, vendors kludged new data requirements onto interfaces. Further, many vendors do not offer screens that vary by specialty. <br /><br />EHR design has moved from where a narrative was the norm 20 years ago, with templates used for specific patients (say for diabetes tracking), to a number of quality improvement efforts today with greater data requirements. Some EHR designers handle this better than others. The ultimate question is: What role poor design plays in data entry misery? I would like to see more experimentation in ways to capture structured data. The extremes now are filling out on-screen forms on one hand, and the quest for natural language processing/machine learning on the other. There must be a middle ground. Perhaps a clinical markup language, or limiting structured data capture to specific conditions or patients instead of every patient. <br /><br />In addition, it is not clear to me how useful routinely collected data, structured or otherwise, is for research or other analytical purposes. Sure, one can collect data, but what reality do those data actually represent? Data quality in EHR systems varies greatly, even worse there are no formal metrics for defining data quality. <br /><br />Weiskopf NG, Weng C. Methods and dimensions of electronic health record data quality assessment: enabling reuse for clinical research. J Am Med Inform Assoc. 2013 Jan 1;20(1):144-51.<br /><br />Köpcke F, Trinczek B, Majeed RW, Schreiweis B, Wenk J, Leusch T, Ganslandt T, Ohmann C, Bergh B, Röhrig R, Dugas M, Prokosch HU. Evaluation of data completeness in the electronic health record for the purpose of patient recruitment into clinical trials: a retrospective analysis of element presence. BMC Med Inform Decis Mak. 2013 Mar 21;13:37.<br /><br /><br />At my last visit, my doctor said he would avoid using the EHR because it got in the way. So, how much of what clinicians hate is bad design, versus a bad idea (.i.e., this should be computerized at all)? I have no idea, but I’m working on it…<br /><br />Jerome Carter, MDJerome Carter, MDhttp://www.ehrscience.comnoreply@blogger.com