The aging population, and the growing need for long-term care among younger people affected by COVID, is putting great demands on skilled nursing professionals. Rising resident rates make for a more complicated risk landscape too. The key is to be proactive rather than reactive.
Yet long-term care facilities face many challenges. While demand increases, staffing shortages plague the industry; medical and prescription errors, falls, resident elopement and advanced directive confusion can all cause issues and the inherent risk has not decreased despite the current pandemic landscape.
It’s a lot to monitor and manage. At the same time, clinicians are deluged with data in the Electronic Health Record (EHR). Each nursing home building creates an average of 34,400 EHR data points in a month. Tracking the 19,800 vitals, 7,500 notes, 4,300 lab results and so much more is too much for one human to handle alone. Now more than ever staffing efficiency is so important to managing the ever-changing landscape.
Resident Changes may be subtle. Or documentation may be incomplete. The essential information may be recorded in a different place by various members of the clinical team. It is impossible for a clinician to manually review every aspect of the medical record every day.
Fortunately, clinicians can benefit from technology to accurately identify risk and anticipate clinical adversity today. With artificial intelligence scrubbing the copious amounts of data in the EHR, Avante Group is able to identify different, high-risk areas allowing for timely intervention to avoid adverse outcomes.
Anticipating risk in long-term care and taking action
Avante, like many skilled nursing facilities, monitors its return to hospitalization rates and looks for ways to reduce these events. Rehospitalization disrupts residents’ lives, increases the risk of healthcare-associated infections and cognitive loss in vulnerable residents, and is costly. To make prevention a priority, Avante implemented SAIVA’s machine learning platform, which uses medical information already entered into EHR, to identify the top fifteen of our residents most likely to be readmitted within the next 72 hours.
Yet this is not the only point of focus across Avante’s 11 facilities. Our facility leadership needs to stay ahead of potential concerns that can hurt patient safety and satisfaction, risk regulatory noncompliance, and erode the bottom line. That’s why we also have the AI technology identify other clinical medical areas where we need to put interventions in place proactively rather than reactively.
Take elopement, for example. By directing the machine learning to search nursing notes and other patient documentation for keywords such as “wandering,” “exit seeking,” “searching for the exit doors” and “attempting to leave room/facility” our interdisciplinary team can learn, in advance, of a potential elopement risk. A plan can be put in place to resolve the situation and the impact can be regularly monitored over the next month for quality assurance (QA).
Individualizing interventions for patient-centered care
Another big issue right now is behavioral management, in particular resident-to-resident behaviors. By setting the platform to pick up trigger words such as “hit” or “smack” or inappropriate language (be it sexually inappropriate or verbally abusive) Avante can identify high risk areas in which there is the potential for abuse. We have the ability to develop a resident centered plan of care to reduce the potential for adverse outcomes.
The AI review of the medical record also allows us to ensure documentation is done correctly (e.g. meds are recorded in the right place) and catch trends. A floor nurse might think they are innocently documenting something that is not an issue for them but for higher level clinicians and risk managers the text triggers concern of a potential problem. It allows for the facility leadership to identify areas of clinical opportunity and a need for further education and training.
I don’t have time to look through volumes of charts. It’s impossible for me to look at the patients that we have every day. This new technology helps identify risk and lets us put interventions in place to specifically meet residents’ needs. Being proactive versus reactive allows for the facilities to deliver a resident centered plan of care that allows for the safety and well being for those we care for.
Preserving the future of the long-term care facility
Elopement or resident-to-resident abuse can result in high scope and severe regulatory consequences. A citation is bad in and of itself but there’s also the potential for financial penalties, loss of licensure, and huge reductions in star rating.
Having the AI offering predictions and recommending actions on Avante’s daily risk report allows the organization to act, bring interventions forward, and ensure appropriate documentation. We can keep a higher level view of what is going on from a risk perspective and act before it becomes an issue.
With the platform identifying documentation issues as well, Avante’s leadership gains peace of mind that we can reduce the cost of any future litigation. The more frequently and aggressively we document reduces how that litigation will impact the organization. Gone are the days of settling quickly because the facility doesn’t have the documentation needed to launch a defense.
The best part of the platform’s support is the ability for the facility to utilize their QA committee and process to develop plans for continued compliance. We aren’t waiting for compliance bodies to find a problem. Instead, we can self identify and self correct, which often avoids a citation because at the time of survey the facility is in compliance.
The biggest thing for me on the executive level, when you have to monitor 11 facilities during the times of COVID, time is precious and essentially time is money. I have to be very focused in all areas of operation of the facilities and SAIVA really guides the focus and allows us to see areas of opportunities for improvement. The AI platform saves us a lot of fines, saves us a lot of citations, and it saves us a lot of headaches on the back end.