<img height="1" width="1" style="display:none" src="https://www.facebook.com/tr?id=1196109850479190&amp;ev=PageView&amp;noscript=1">

QUALITY TALK

A podcast dedicated to advancing better understanding of health care.

Episode 38 - The 7 Stages of the Health Data Life Cycle

Posted by Jodie Jackson, Jr. on Jul 5, 2018 6:58:00 AM

 

Quality Talk is a podcast dedicated to advancing a better understanding of healthcare, and considering the massive changes unfolding in the ways providers are paid for their services, few topics are as timely and crucial as the subject of Episode 38. We dive into the deep ocean of healthcare data with Primaris COO Michael Levinger.

What is the health data life cycle, why is it important in our new, value-based care environment, and why is it imperative that each step of the data lifecycle lead to the ultimate goal: better patient care? Episode 38, The 7 Stages of the Health Data Life Cycle, answers those questions.

The data life cycle Levinger introduces are the stages recognized for Primaris’s data abstraction services.

  • Find the data.
  • Capture the data.
  • Normalize the data.
  • Aggregate the data.
  • Report the data.
  • Understand the data.
  • Act upon the data.

Find the data.
To achieve success in the new realm of quality improvement, patient satisfaction, and better financial results, data must be accurate, reliable, timely, and ultimately, actionable. Data will come from multiple sources, including:

  • Clinical.
  • Quality.
  • Claims.
  • Financial.

There is a veritable flood of data, and it is coming from sources that are often siloed or segregated. This data needs to be integrated to allow effective management of risk and reward

Risks of using poor data:

  • Inaccuracy clinical quality measures.
  • Lost reimbursement as patients aren't placed into the proper risk stratification.
  • Inaccurate real-time care alerts leading to patient safety risks.
  • Inefficient use of care management and care coordination systems.
  • Safety issues around drug interactions, drug doses, allergies and medical history.
  • Inability to perform population health analytics.
  • Security risk.

Primaris CEO Authors Becker's Post on the Health Data Life Cycle


Here’s a prime example of the importance of this first step – find the data. One ACO Primaris worked with was a group of many ambulatory practices. The group did not have a standardized EHR and had over 50 EHRs. Knowing which patient’s information was on each one was very difficult. Finding the data required access to each EHR including security and access rights as well as knowing the configuration of that EHR.

Capture the data.
Once you know where the data is, the next step is getting it from the source system into a common systems for analytics. Much of this can be done using vended population health systems that have the ability to gather data from many sources. But none of these systems connects to everything.

Quality Today blog – Registry Abstraction: Tracking Down Missing Data

For example, even if you can connect to the multiple clinical systems being used, what about an older custom developed system? Or the clinical information on paper? And you must address how to capture structured vs. unstructured data. Structured is easier but still not easy. Unstructured can be very difficult.

Normalize the data.
Once the data is in one place, you need to ensure it can be consolidated. This requires making sure the data structured so it can be consolidated and analyzed – called data normalization. Examples of normalization: the same format and the same units.

Normalization can range from straightforward to very difficult:

  • Units: Making sure dollar items are not in thousands for some data and millions for other data.
  • Syntactic structure of the data – for example last name first or first name first?
  • Semantic structure – How do different sources of data encode structured information – for example are symptoms all coded using SNOMED? Are they using the same version?
  • Is there clear mapping between different standards like SNOMED, ICD 10 and LOINC?Collage of a group of people portrait smiling, indicating a population of patients.

Aggregate the data.
This stage consolidates the data into groups or pools of patients – often referred to as cohorts. It’s a crucial step – especially for managed care – since data aggregation is essential to managing groups of patients which is at the core of managed care.

Report the data.
Healthcare is a highly regulated industry and there are many required reports – around quality, financial performance, safety and more.  And increasingly reporting is tied to financial results – by CMS, private payers and other organizations:

Understand the data.
Data alone doesn’t get the job done. You must analyze and understand the data. What has been effective - Clinically? Financially? Other criteria?

The movement to managed care necessitates understanding your data. What is the clinical point of view versus a dollars/cost point of view? How are these two points of view reconciled to get the “right” results? When Drug B is half the price but equally as effective as Drug A, that is an example of evidence-based medicine, which was the result of the data life cycle.

Act on the data.
Once you understand your data and use it to decide on improvement initiatives, you need to act effectively. Remember: If you don’t have the right data and accurate data, you won’t have the right results. For example: Having the right units – Is $24 savings per member per month (PMPM)? Per member per year? Or something else?

Summing it up.

  • Can you rely on the data to make critical decisions about patient care and financial performance?
  • Will the data be consistent every time? Is it error free?
  • For example, can clinical and claims data be consolidated the same way every month to ensure that trends can be analyzed?

Data must be timely:

  • Different types of data are available at different times. Too old of data leads to the wrong actions.
  • For example, claims data from CMS can often take months to be available. So decisions on readmissions management (to reduce readmissions penalties) or care coordination (to optimize costs and improve outcomes) could be done based on old information.

Data must be actionable:

  • Management needs data that will support action. You need things like the right granularity and the right selections.
  • For example – To help physicians in a clinically integrated network make optimal transitions of care decisions requires having information pertinent to particular patients and particular facilities. So have to be able to select that data.

Thanks for listening!
If this episode leads you to more questions about how to best find and use your data to achieve quality care, feel free to contact us today to learn how Primaris can help. You can contact us through our website, www.primaris.org, where you’ll also find our blog, “Quality Today,” and other resources to help you learn more about and comply with the requirements of MACRA, MIPS and the Quality Payment Program.

Thanks again for listening. And remember: Primaris is your partner in healthcare quality.

Topics: Quality Talk podcast

SUBSCRIBE TO THE QUALITY TALK PODCAST

Recent Posts

Interested in being a podcast guest? Email us


Please view our Online Comment Policy here: http://primaris.org/online-comment-policy/