Healthcare Big Data Defined: Improving Care, Coordination and Coding
Lance Speck, general manager of Actian cloud and healthcare, speaks here about healthcare big data and how it can be used in healthcare to improve processes from care coordination to coding for ICD-10. In his day job, he is focused on delivering healthcare solutions to help payers and providers address an estimated $450 billion annual opportunity created through data analytics, ranging from fraud analytics to patient re-admission reduction to staff optimization to accountable care reporting and clinical auto-coding. For more than 20 years, Lance has served in a variety of management, sales and product roles in the software industry including a decade focused on SaaS, cloud and healthcare.
How can big data analytics improve patient care?
According to a recent PwC survey, 95 percent of healthcare CEOs are exploring better ways of using and managing big data; however, only 36 percent have made any headway in getting to grips with big data. All agree that big data analytics has the potential to improve the quality and cost of care, but many are still struggling with finding the right ways to infuse analytics into everyday operations. Assuming they realize that they already have access to the data, what do they do with it? What are the areas that will have the biggest impact? Where do they start?
Start with the basics. Organizations should focus in infusing big data analytics where a big impact can be recognized. They should ask themselves:
- Is there enough value in solving the problem?
- Can the problem can be predicted?
- Can the problem be prevented?
- Can the predictive action be delivered accurately, and in a timely fashion to make a difference?
Very early in the process, organizations should address how they plan to incorporate big data into the everyday workflow of clinicians, financial staff and other healthcare stakeholders for organizations to:
- Use predictive analytics against historical and external data to anticipate patient occupancy needs to adjust staffing levels to have the right care available at the right time.
- Use science to determine with accuracy health trends in specific communities and take action to prevent costly
- Determine patients’ risk of readmission before they are discharged to improve patient outcomes and reduce costs and penalties by nearly $70 billion.
- Realize that for this insight to be effective, you must put this information into the hands of the clinicians and the patients in the format that fits their daily flow.
How can healthcare providers transition to ICD-10 as simply as possible?
The deadline for ICD-10 is approaching and the cost for not getting to ICD-10 will be high. Organizations should look for ways to transform the written and verbal language of physicians and seamlessly integrate this with their existing systems.
Can the industry use algorithms to replace manual entry of ICD codes accurately?
Yes, the industry can leverage algorithms to replace manual ICD code entry. And they must. With the transition to ICD-10, the number of codes will jump by 800 percent. While the complexity of coding is increasing, so is its importance to patient wellness. Unless they’re bursting at the seams with well-trained, experienced coders, organizations must find ways to automate this critical, yet cumbersome, task.
How does that work?
Actian and Atigeo have teamed to offer applications that replace manual entry of ICD codes. The xPatterns Clinical Auto-Coding, built on the Actian Analytics Platform, combine big data analytics, data science and deep domain expertise help drive down costs while improving patient outcomes.
xPatterns Clinical Auto-Coding (C.A.C.) automatically infers clinical codes, including ICD-10, by extracting concepts from the text in clinical encounter notes such as physician notes, lab results, and admit/discharge records. xPatterns C.A.C. detects under-coding, over-coding and miscoding to deliver higher accuracy and coder productivity.
The xPatterns C.A.C. cloud-based interfaces allows coders to review, if needed, modify the generated codes. xPatterns C.A.C. has been designed with full security and HIPAA-compliance considerations, and can be seamlessly integrated with existing HIT and EMR systems via HL7. It includes workflows accommodating the roles played by administrators, coders and doctors in coding. xPatterns C.A.C. detects under-coding, over-coding and miscoding, highlighting potential opportunities where higher billing is justified.
What are other benefits of using big data analytics on medical records?
Electronic health records are the main repository of hospital data intelligence and yet the information within them is often under-detected. By analyzing EHR data against other publicly and privately available data, organizations can identify often completely unforeseen connections across these disparate information sources.
By connecting and analyzing data from a variety of sources – EHRs, claims, e-prescribing, wearables, labs and the ever-expanding “Internet of Things” – in real time into an integrated view, clinicians can begin to build health and wellness action plans that zero in on ways to help that patient right now.
How does big data reduce Medicare readmission penalties?
More than 80 percent of healthcare costs come from treating chronic patients. The cost of prevention pales in comparison, and almost 90 percent are preventable. By predicting which patients are re-admission risks before they are discharged, healthcare providers can intervene with proactive strategies, which can dramatically improve patient outcomes and reduce care costs by nearly $70 billion. Successful patient wellness plans move beyond retrospective reports, BI and rules-based analytics. They take into account environmental data, population health data, patient data, etc. to enable the patient to avoid future illnesses and health issues before they begin.