top of page

FAQs
Click this icon for a list of key pages >

Frequent questions:

  • How can we send our data?

We support any standard and secure data exchange format and connection -- from real-time internet links to batch file uploads.  

  • Do we need to provide data in the East Bay Labs format?

No.  Any standard format is fine.  We simply need documentation to understand the definition of each field,
the field type and your field parsing / separation format.  

  • Can I use EMR, EHR, HRA and lab results data?

Yes.  More detailed and current data improves predictive accuracy and PreVista can process virtually any data you have -- in fact, the more data, the better.  A new EMR system with dozens of new clinical data fields?  A flood of Big Data for financial risk analysis?  No problem.         

  • Can you work with anonymized data?

Yes.  We can also anonymize your data and/or PreVista results to HIPAA / NIST standards. 

  • Can I load PreVista results into my data warehouse and reports?

Yes. You can load PreVista results directly into your data warehouse, reports, applications and third-party systems, e.g. to deliver risk alerts to providers via an EMR or report.  (Some restrictions apply, of course.)

  • How are PreVista results delivered?

We can deliver PreVista results as near real-time updates via the Internet, batch file updates via the Internet or encrypted disks -- to match your schedule and systems.

  • Do you also offer online reports?

No.  Data feeds allow you to integetrate PreVista results directly into your data warehouse, reporting system
and applications -- without more 'report design' fees etc.   

  • Do you offer custom analyses and predictions?

Yes.  PreVista can use virtually any data and predict any target.  What would you like to forecast?  

  • Can I distribute PreVista results to our physicians?

Yes.  Accurate alerts from PreVista can be very helpful to physicians and caregivers. 

  • Can PreVista forecast risk for specific populations by gender, age, condition, etc?

Yes.  PreVista forecasts can be 'tuned' for specific groups.

  • How can PreVista be 200% to 300% more accurate than the well-known XYZ risk adjustment product?

Almost all of the 'risk adjustment' products on the market today are based on simple regression algorithms and claims data 'groupers' that were pioneered by the late Barbara Starfield, M.D. M.P.H. and her team at Johns Hopkins University.  Ground breaking at the time, this approach was limited by the performance of 1990s computer technology.  Leading 'risk adjustment' products continue to use the same dated approach.    

PreVista is based on new AI technology, today's high performance CPUs, fast SSD data storage and the 'Big Data' approach.  The technical foundatation behind PreVista is radically different.  This enables a range of benefits, including (1) 200% to 300% improvements in accuracy, (2) support for healthcare and pharma claims, lab results, EMR, EHR, HRA and any other data, and (3) the ability to 'tune' forecasts for specific members, diagnoses, gender, age and geographic regions.    

  • Is improved accuracy really significant?  XYZ says their risk system is used by 100s of organizations.

Well, would you place a bet with your money, based on a prediction that is only 30% accurate?  Should you bet your budget and your organization on 30% scores?  Is it smart to manage patient care with risk scores that are so inaccurate? PreVista's accuracy can be very significant for patient care, financial planning, rate setting and acquisitions.

  • XYZ Corp. says PreVista is a 'black box'.  How can I validate and explain your results?

We chuckle every time we hear the old 'black box' canard.  If a vendor's predictive system is so simple that they can show you a formula that explains how it works, that highlights the basic problem!  To put this in perspective, how do you evaluate a new car?  Do you ask for all of the engineering design plans, or do you rely on the actual results, e.g. horsepower, mileage, etc?  If you are an automotive enthusiast you may even know the g-force your sports car will pull on a skid pad, the max torque, 0 to 60 time and stopping distance. That is exactly how you can evaluate PreVista -- based on actual results with your data. You will see how PreVista performs in the real world, the best 'proof' there is.  

  • The XYZ system can predict results for diabetes too.  Why is PreVista "more flexible"?

Every vendor's 'risk adjustment' system can create scores for specific conditions, e.g. diabetes and asthma. They use the same basic algorithms and approach for every condition, however -- with the same poor results.  PreVista's engine changes to match each condition and your available data.  This significantly improves accuracy and gives us the ability to focus on specific populations, e.g. "women with Type 2 diabetes between the age of 40 and 55." 

  • Our company covers people in New York City and rural Oklahoma.  Will their results be different?

Yes.  This is a key advantage with PreVista -- the engine flexes to match the data from each population.   Environmental, social and demographic factors in NYC are different from rural Oklahoma and the available data will be different as well.  PreVista can create a model to match each population group -- not the XYZ "one-size-fits-all" approach.

  • Will PreVista deliver forecasts every month, like the XYZ system?

Yes and no.  PreVista is not 'batch' driven.  As soon as your new data is available, we can deliver new predictive results.  If you have EMR data that is updated real-time, PreVista can provide immediate updates.  If you only have monthly batches of claims data, PreVista can provide monthly updates -- with improved accuracy.

  • Does PreVista support ICD 10?

Yes.  The added precision of ICD 10 can improve the accuracy of PreVista forecasts, and the full complexity of ICD 10 is supported by PreVista. 

  • Does PreVista use "machine learning" or "cognitive computing" technology?

Prediction Predictive modeling forecast health risk redmission EMR HIS HIE claims reports  acquisition custom accuracy 3M™ McKesson™
Ingenix™ Optum™ Milliman™ MedAI™ ISO™ DxCG™ DRG™ CMS ACG™ Verisk™ pharmacy fraud abuse actuary payer provider STARS healthcare

Yes.  A number of new names have been created for the same technical approach used in AI (artificial intelligence).  IBM™, for example, uses "machine learning", "cognitive computing" and "cognitive systems" 
to describe their research in this area.

bottom of page