J P Systems, Inc.

data quality collage with doctor entering prescription data

Clinical Data Quality Improvement Services


According to a recent US government study, the vast majority of EHR systems contain unacceptable data quality. Often data in clinical documents is put in the wrong place. Or dozens of clinical documents arrive and either lack key data fields or contain superfluous data. We have made great progress in the last 5 years in being able to move data between EHR systems. But when doctors have to spend precious appointment time looking at data they don't need and trying to find data they do need, it adversely affects the quality of patient encounters. According to NIH: 'Quality assessment of healthcare data used in clinical research is a developing area of inquiry. The methods used to assess healthcare data quality in practice are varied, and evidence - based or consensus “best practices” have yet to emerge.' Granularity problems arise quickly and data can degrade as clinical data is sent from provider to provider.

 

Once clinical data leaves its original EHR "home", local context can be lost. Without the foundation of reliable standardized codes in all data fields, the conclusions of AI systems and researchers have reliability problems. These problems can be alleviated with mapping to standard terminologies like SNOMED. J P Systems is working hard to help providers improve the validity of clinical data exported to or imported from clinical documents known as CDAs. Meaningful use has driven us to adopt data exchange standards and take the first step to true interoperability - move the data! Now we need to focus on populating these documents with clean and useful data.

 

Data Quality improvement efforts are a combination of many processes including preventing the entry of invalid data, standardizing clinical value sets internationally, and improving the quality and location of existing data. Over many year,s database field contents change. Some data fields are added while others are re-purposed. Thus data in a single field may be inconsistently coded. It is critical for a DBA to know the history of the data file structures. Without a detailed knowledge of the data's history and evolution, valid queries can not be composed. The older the data is, the harder it is to know how to properly compose a SQL query as data files develop quirks as the files age. Add to this the problems of transporting data to another hospitals EHR system, and you have data that is in 'culture shock', if you look at it from the data's perspective. Local codes can not be understood unless they are reliably mapped to international standards like SNOMED and LOINC.

 

 Our Data Quality Improvement Brochure


J P Systems, Inc. specializes in the improvement of clinical data quality. Clinical data is complex and its granularity, or the degree of detail, is always related to the mission of the organization. A neonatal facility is going to collect different data about patients than an Allergist or an ER. Thus, healthcare domain specific knowledge is as important to data quality improvement as is the technical knowledge of IT data architecture. Proper standardization of the clinical data is a key first step.

 

According to American Health Information Management Association (AHIMA), "Data quality and consistency are critical to ensuring patient safety, communicating delivery of health services, coordinating care, and healthcare reporting. Assessing the quality and consistency of data requires data standards."


Data Quality is improved by:

  • Standardized terminologies which limit the values which can be entered. These are referred to as Reference Terminologies and are often shown to a user on a data entry screen in a drop down box.
  • Analyzing an EHRs data to find misspelled words and redundant synonyms
  • Using the same semantics as a basis for logical and physical database model generation, software component and service generation, rule development (e.g., in production rule-based systems), etc.
  • Labeling data received from outside sources so its reliability can be assessed

According the AHIMA Information Governance Principles for Healthcare (IGPHC)™ the foundation of data and information governance employs eight key principles:

Accountability: Designation or identification of a senior member of leadership responsible for the development and oversight of the IG program.

Transparency: Documentation of processes and activities related to IG are visible and readily available for review by stakeholders.

Integrity: Systems evidence trustworthiness in the authentication, timeliness, accuracy, and completion of information.

Protection: Program protects private and confidential information from loss, breach, and corruption.

Compliance: Program ensures compliance with local, state, and federal regulations, accrediting agencies’ standards and healthcare organizations’ policies and procedures and ethical practices.

Availability: Structure and accessibility of data allows for timely and efficient retrieval by authorized personnel.

Retention: Lifespan of information is defined and regulated by a schedule in compliance with legal requirements and ethical considerations. Disposition: Process ensures the legal and ethical disposition of information including, but not limited to, record destruction and transfer.