Mismatched patient data is the third leading cause of preventable death in the United States, according to healthIT.gov, and a 2016 survey by the Poneman Institute revealed that 86 percent of all healthcare practitioners know of an error caused by incorrect patient data. Patient misidentification is also responsible for 35 percent of denied insurance claims, costing hospitals up to $1.2 million annually.

Melanie Mecca, Director of Data Management Products & Services for CMMI Institute calls this situation “A classic Master Data and Data Quality problem.” A multitude of different vendors is one of the causes, she said, but “there’s really no standard at all for this data.”

The Health and Human Services Office of the National Coordinator (HHS-ONC) wants to make it safer for patients needing health care by improving those numbers.

“They’re trying to lower the number of duplicates and overlays in the patient identification data – the demographic data – so that they can have fewer instances of record confusion and ensure that records can be matched with patients as close as possible to a hundred percent,” she said.

In the article Improving Patient Data Quality, Part 1: Introduction to the PDDQ Framework Mecca remarked that, “duplicate patient records are a symptom of a deeper and more pervasive issue – the lack of industry-wide adoption of fundamental Data Management practices.” Sources for this case study also include a presentation by Mecca and Jim Halcomb, Strategy Consultant at CMMI, as well as the Patient Demographic Data Quality (PDDQ) Framework, v.7.

The Challenge

Government sources (and CMMI) estimate that the average hospital has 8-12 percent of duplicate records, and as many as 10 percent of incoming patients are misidentified. Sharing of patient data from disparate providers increases the likelihood of duplicates during health information exchanges due to defects in Master Patient Indexes.

Preventable medical errors may include:

• Misdiagnosis and incorrect treatment procedures
• Incorrect or wrong dose of medication
• Incorrect blood type
• Allergies not known
• Repeated diagnostic tests

In addition, inaccurate and duplicate records can increase the risk of lawsuits, and can cause claims to be rejected. The cost to correct a duplicate patient record is estimated at $1000.

Previous attempts have been made to address these issues using algorithms that search for data fragments, but due to a lack of standardized practices:

“Algorithms alone have failed to provide a sustainable solution. Patient record-matching algorithms are necessary, but they are reactive, and do not address the root cause, which is the lack of industry-wide standards for capturing, storing, maintaining, and transferring patient data,” she said.

Finding a Solution

According to the HHS-ONC website, the Office of the National Coordinator for Health Information Technology (ONC) is located within the U.S. Department of Health and Human Services (HHS). The ONC:

“Serves as a resource to the entire health system, promoting nationwide health information exchange to improve health care. ONC is the principal federal entity charged with coordination of nationwide efforts to implement and use the most advanced health information technology and the electronic exchange of health information.”

In line with this mission, HHS-ONC decided to craft a solution to the patient data problem with built-in participation and support across all areas of health care. To this end, ONC assembled a community of practice that included 25 organizations ranging from health Data Management associations such as AHIMA (American Health Information Management Association), to government offices like OSHA (Occupational Safety and Health Administration), and large health care organizations such as Kaiser Permanente, as well as other health care providers.

This community was charged with finding a set of standards and practices that could be used to evaluate existing patient Data Management processes, and a comprehensive tool for bringing organizations into compliance with those standards. “What they were looking for was a Data Management Framework that was complimentary to what they were trying to accomplish,” said Mecca.

They chose CMMI’s Data Management Maturity (DMM) ^SMModel as the best approximation of what they were looking to accomplish. “The DMM’s fact-based approach, enterprise focus, and built-in path for capability growth aligned exactly with the healthcare industry’s need for a comprehensive standard,” she said.

Developing a Tool

CMMI, as a sub-contractor with health information technology company Audacious Inquiry, then used the Data Management Maturity model to determine which practices were essential “specifically for patient demographic Data Quality,” Mecca said. Out of that process came the Patient Demographic Data Quality Framework (PDDQ). The PDDQ offered the HHS-ONC a health care-focused, “sustainable solution for building proactive, defined processes that lead to improved and sustained Data Quality.”

The Patient Demographic Data Quality (PDDQ) Framework: The Solution

The PDDQ Framework:

“Allows organizations to evaluate themselves against key questions designed to foster collaborative discussion and consensus among all involved stakeholders. Its content reflects the typical path that most organizations follow when building proactive, defined processes to influence positive behavioral changes in the management of patient demographic data.”

The PDDQ is composed of 76 questions, organized into five categories with three to five process areas in each, representing the broad topics that need interrogation by the health care organization to understand current practices and determine what activities need to be established, enhanced, and followed.

The questions are supported by contextual information specific to health care providers.

“Data Governance is highly accented, as is the Business Glossary – the business terms used in registration, and terms that providers, and claims, and billing have to agree on, like patient status,” Mecca said.

Examples include illustrative scenarios, such as how a patient name should be entered, and what to enter if the patient has three middle names, for example. The questions and supporting context are intended to serve as an “encouraging and helpful mechanism for discovery.” While the framework encourages good Data Management,

“It does not prescribe how an organization should achieve these capabilities. It should be used by organizations both to assess their current state of capabilities, and as input to a customized roadmap for data management implementation.”

One of the features of the PDDQ is its flexibility to address Data Quality in a variety of environments. It is designed for any organization creating, managing or aggregating patient data, such as hospitals, health systems, Health Information Exchange (HIE) vendors, Master Data Management (MDM) solution vendors, and Electronic Health Record (EHR) vendors. “An organization can implement any combination of categories or process areas, and obtain baseline conclusions about their capabilities,” she said.

The organization can focus on a single process area, a set of process areas, a category, a set of categories, or any combination up to and including the entire PDDQ Framework. This allows flexible application to meet specific organizational needs and address for resource and time constraints.

Using the PDDQ, organizations can quickly assess their current state of Data Management practices, discover gaps, and formulate actionable plans and initiatives to improve management of the organization’s data assets across functional, departmental, and geographic boundaries.

“The PDDQ Framework is designed to serve as both a proven yardstick against which progress can be measured as well as an accelerator for an organization-wide approach to improving Data Quality. Its key questions stimulate knowledge sharing, surface issues, and provide an outline of what the organization should be doing next to more effectively manage this critical data.”

Results

The PDDQ assessment can deliver actionable results within three weeks, leading directly to the implementation phase. For HHS-ONC, Kaiser did pilots (in Oregon) where they went on site and did data profiling and cleansing of the patient records. During this effort, they used several of the process areas that we wrote for the PDDQ Framework, and they applied them to the organizations.”

A month later when Kaiser checked with the pilot sites, all had made improvements in the way they were managing that data, she said. “And it showed because the matching algorithms had lower incidence of duplicates.”

According to the presentation by Mecca and Halcomb, use of the PDDQ leads to decreased operational risk through improvements to the quality of patient demographic data. Specifically, patient safety is protected and quality in the delivery of patient care improves due to:

• Increased operational efficiency, requiring less manual effort to fix data issues, fewer duplicate test orders for patients, and adoption of standard data representations.
• Improved interoperability and data integration through adopting data standards and data management practices that are followed by staff across the patient lifecycle.
• Improved staff productivity by expending fewer hours on detecting and remediating data defects to perform their tasks.
• Increased staff awareness for contributing to and following processes that improve patient identity integrity.

Patient data is a common thread throughout the health care system, she said. Capturing or modifying patient data differently magnifies the potential for duplication. The PDDQ helps uncover unexamined processes around patient data and means that health care organizations don’t have to guess about how they’re managing patient data. It clearly identifies gaps, creates awareness about individual responsibility for quality of patient data, engenders cooperation and participation, and sets a baseline for monitoring progress.

“Once gaps and strengths have been identified, organizations can quickly establish timelines for new capabilities and objectives,” she said.

Steps Moving Forward

According to Mecca, managing data is “first and foremost a people problem, not a system problem.” No one individual knows everything about the patient data. Adoption of consistent data standards industry-wide would increase interoperability and minimize duplicates. The PDDQ provides organizational guidance and “an embedded path for successive improvements along with a concentrated education for everyone dealing with patient demographic data,” she said.

“Health care Data Management consultants can employ the PDDQ for their client organizations as a powerful tool to quickly identify gaps, leverage accomplishments, focus priorities, and develop an improvement roadmap with the confidence that all factors have been examined and that consensus has been reached.”

Access the PDDQ

• The PDDQ and evaluation scoring tool are available at the following location: https://www.healthit.gov/playbook/pddq-framework/
• A condensed version of the PDDQ, the Ambulatory Guide, contains a core set of questions aimed at very small health care practices, to help them get started in improving Data Quality.  It is available at: https://www.healthit.gov/playbook/ambulatory-guide/

 

Photo Credit: Micolas/Shutterstock.com