Over the past few weeks we have been discussing the changes that will abound in the healthcare arena for 2014. In fact, this is an unprecedented time of change for healthcare with impacts across the board clinically, financially, and organizationally. In one of our most recent posts we discussed changes that will impact the healthcare revenue cycle, and today we are going to discuss a strategy that promises to have a significant impact on an organizations bottom line…..healthcare predictive modeling. With so many things up in the air organizations are struggling to determine where they can generate new revenue, boost current revenue, and decrease waste. Savvy organizations are turning to healthcare predictive modeling with astounding results, and today we will explore why healthcare predictive modeling will be the healthcare organizations best friend in 2014. Let’s begin with a quick overview of what healthcare predictive modeling is.
Understanding Effective Healthcare Predictive Modeling
Healthcare predictive modeling is a process in which data is analyzed to create an algorithm that can assist in determining the likeliness of an event. In healthcare predictive modeling the data may be patient outcome data, historic financial data, or quality outcome data. The most effective organizations conduct healthcare predictive modeling in both the clinical and financial arenas. Healthcare predictive modeling is effective as it will provide concrete predictable scenarios which can allow an organization to know with some level of certainty where high future costs will be generated, how patients will respond to care, and what the impact on an organization will be if a specific variable changes. For instance, in healthcare finance predictive modeling scenarios can be run by an organization to determine what would happen if there was a shift in payer mix, and would illuminate, with some degree of certainty what the financial implications of an organization would be. Similarly, clinical healthcare predictive modeling can assist organizations in determining where their patient outliers are (i.e. those individuals who have significant healthcare costs with minimal health improvements). This allows an organization to effectively respond to a patients needs in providing alternative services and strategies (such as care coordination and case management/disease management) which allow the patients healthcare needs to be addressed in the most financially responsible way possible.
Whether clinical or financial, healthcare predictive modeling can provide a degree of certainty in relation to a specific healthcare metric during uncertain times
The Challenges of Healthcare Predictive Modeling
Healthcare predictive modeling requires organizational commitment and investment, and the number one obstacle which organizations face when implementing healthcare predictive modeling is the availability of robust data. Generally speaking, healthcare predictive modeling is easier to implement on the financial side of an organization, than on the clinical side, but both are worthy of consideration. On the financial side, healthcare predictive modeling is already widespread in the industry. One of the most common types of healthcare predictive modeling utilized by organizations is determination of prospective payments. In this instance patient characteristics and average cost information are utilized to estimate or anticipate the cost of providing care for that patient over a given amount of time (typically determined on an annual basis). While this is only one example of predictive modeling at work in the realm of healthcare, the potential for financial forecasting is only limited by the amount of data which is available. If your organization has reliable data on denial rates, reimbursement rates, organizational expenditures, and cost of services provided vs. reimbursement rates there is virtually no limit to the way that the information can be organized to provide answers to healthcare’s critical questions such as:
- what impact will changing x have on our denial rate
- if we do y, what will the impact be on our reimbursement rate
- what impact will a changing payer mix have on our bottom line
- what are the services that we should focus our marketing efforts on, and which are costing our organization money
Again, data can be reconfigured in a variety of ways to provider reliable answers to an organizations questions, and can be invaluable in pinpointing areas of opportunity from a financial perspective for healthcare organizations and hospitals.
But what about clinical healthcare predictive modeling?
The primary obstacle when it comes to clinical healthcare predictive modeling is lack of data over an extended period of time, however with the increased utilization of Electronic Medical Records (EMRs), Electronic Health Records (EHRs), and increased collaboration between providers across the healthcare spectrum this is changing. Clinical applications of healthcare predictive modeling require complex models (capable, for instance, of dealing with the complexity of patients and numerous variables such as age, diagnosis, admission rate, adherence rates, etc.) and the application of calculations that have enough sophistication to consider multiple factors in determining an event.
With the adoption of EHRs/EMRs clinical healthcare predictive modeling will offer great opportunities for healthcare organizations to utilize data to improve the quality of care for their patients, and the bottom lines of their practice.
What Is Predictive Modeling Already Accomplishing for Healthcare Organizations?
Reduction of Hospital Readmission Rates – this is one of the most significant areas that organizations can quickly utilize healthcare predictive modeling to positively impact revenue. Organizations are currently facing potential reduction in reimbursement and other financial penalties if they have a high 30-day readmission rate. In fact, reducing unnecessary re-admissions is on the organization to do list of nearly every hospital that we have spoken with because financial penalties for high re-admissions would have a significant negative impact on the hospitals revenue. An effective strategy to target re-admissions begins with healthcare financial modeling. Organizations, such as Parkland Health & Hospital Systems, have been able to utilize a readmission specific healthcare predictive model to track patients who are at a high risk for readmission. This allows Parkland to effectively identify at-risk patients, and more importantly deploy strategies that can impact the patients care outcomes. When a patient if first admitted they are put into a risk category, those who meet criteria which places them in a higher risk category are managed by dedicated care managers who work to ensure that they are provided with appropriate treatment and interventions to assist them. This strategy has led to a 20% reduction in re-admissions overall for the organization. Similarly, Mount Sinai Medical Center in New York began piloting a program two years ago which utilized healthcare predictive modeling to similarly target re-admissions. In this case admission history was utilized to target specific at risk patients for interventions, which includes not only medical interventions, but social interventions. The result was a 30-day readmission rate which went from 30% to 12% and a decrease in Emergency Department visits by 63%.
Management of Denials – In healthcare one of the most significant emphasis is on the optimization of organizations when it comes to reimbursement and the reduction of waste allowing for a robust bottom line. Healthcare predictive modeling is being piloted across the country by multiple organizations who are using it to obtain dynamic results when it comes to revenue improvement. One organization recently put in place healthcare predictive modeling to analyze their denial rates. The algorithm constructed looked at likeliness of denials by payer and service type, as well as the successful overturn of previously denied claims. This allowed the organization to effectively target where they put their time and energy when it came to revenue recoupment via denials, and had the offset benefit of allowing the organization to identify which areas had the highest prevalence of denials by causation so that appropriate training could be provided to the staff. In this case the organization was able to recoup more than $1.2M in denied claims over a six month period, and was able to decrease their overall denial rate by 12%.