Monday, July 06, 2015 'Real-Time' Data From Nursing Home EMRs Could Aid in Falls Prevention According to a new study, nursing home electronic medical records (EMRs) could contain predictive "real time" data that can shed light on resident falls risk, and make identifying those most at risk for falls more precise. The result? Fewer falls, of course, but also cost savings that could pay for the expansion of EMR use in those facilities. Researchers analyzed data from 26 nursing homes that are part of a large California-based chain that uses a centralized EMR system, 13 of which participated in a program that combined standard data collected as part of the minimum data sets (MDS) required by the Centers for Medicare and Medicaid Services (CMS) with pre-identified risk factors obtained on an ongoing basis through the residents' EMRs. Authors reviewed data from 5,129 residents (133,781 observations) between 2011 and 2014. Results were e-published ahead of print in the Journal of the American Medical Informatics Association (abstract only available for free). According to the authors, the main differences between the data collected for the MDS and EMRs have to do with frequency and the risks associated with certain medications. While more risk factors are collected in the MDS than in the EMR, EMR data is collected in real time rather than at regular (often quarterly) intervals; additionally, the EMRs used for this study collected data on resident use of opioid analgesics, anticonvulsants, antihypertensive medications, and psychotropic medications—drugs that have been associated with higher falls risk. The EMR also recorded room changes, another factor linked to higher falls risk. To get a sense of whether—and exactly how—EMRs led to better predictive results, researchers looked at how risk estimates were affected by various combinations of MDS and EMR data. Because some of the data collected for the MDS and EMRs overlap, authors of the study were able to not only look at the elements exclusive to each collection method, but also evaluate which system seemed to produce more accurate predictive results in those common areas. In the end, what they found was that the additional data available through the EMRs did not significantly improve the precision of falls risk; however, substituting the EMR data that overlapped with MDS data made a big difference in who was assessed as most at risk for falls. Specifically, researchers found that when using the MDS model alone, the top 10% of at-risk residents were shown to account for 28.6% of observed falls. When MDS variables were replaced with EMR variables, the top 10% were linked to 32.3% of falls. "Together, these results imply that the replacement of MDS risk factor measures with more frequently-updated EMR measures substantially improves identification of residents at highest risk for falls, but that the addition of risk factors available in the EMR yields little or no improvement," authors write. In terms of real-world application, authors use an "average" 100-bed nursing home that experiences an "average" rate of 150 falls per year as an example. If the facility could prevent falls among the top 10% of residents most at-risk for falls using MDS data, they write, it could expect to prevent about 43 falls. With the use of EMR data, that number climbs to about 49 prevented falls. Based on an estimated cost-per-fall of $7,307, that 6-falls change would amount to a savings of $43,842 per year. Even if a facility could only prevent 33% of the falls in the most at-risk decile, authors write, the resultant $14,000 in savings "would readily justify the additional incremental cost of incorporating EMR information into targeted clinical decision support systems." As encouraging as this may be, authors say that the real problem is that nursing homes have been slow adopters of EMR systems, and the EMRs themselves aren't always well-suited to use data in this way. "Standard EMR systems currently provide no easy way to synthesize and summarize information on the changing risk factors recorded in disparate parts of the EMR to support clinical decision-making," authors write. "Further development of such application, focused both on falls and other avoidable adverse events … should be a key priority as nursing homes expand their adoption of EMRs in the coming years." Learn how predictive analytics could shape the future of physical therapy practice: check out "The Power of Prediction" from the June issue of PT in Motion magazine. The article is open-access. Also available from APTA: a recorded webinar titled "Using Data Analytics to Work Smarter in Your Health Care Setting," presented by Michael Weinper, PT, DPT, MPH, and Nancy Rothenberg. Research-related stories featured in PT in Motion News are intended to highlight a topic of interest only and do not constitute an endorsement by APTA. For synthesized research and evidence-based practice information, visit the association's PTNow website.