1. Quality Data for a Healthy Nation by Mary H. Stanfill, RHIA, CCS, CCS-P
Note to presenter: This presentation is meant to help you train your healthcare workforce,
especially non-HIM staff, on the importance of data quality and its impact on healthcare decision making. There will be points during this presentation where you can discuss specific examples of data quality management efforts in your organization. This presentation could also be used for speaking at community events.
Welcome to the session “Quality Data for a Healthy Nation.” My name is ________ and I am ___________ (position, title). This presentation will focus on the importance of data quality, its impact on healthcare decision making, and efforts we are making to ensure our data is of the highest quality here at ___________ (name of organization).

2. What Is Healthcare Data?
Raw facts generated in the process of patient care
Can be clinical, financial, or demographic
Multiple forms, formats, and sources
Generally stored as characters, words, symbols, measurements, or statistics
Processed to provide healthcare information
To begin, I’d like to clarify what I mean when I refer to “healthcare data.”
When I use the term data, I am referring to raw facts.
The data generated in the process of patient care can be clinical, financial, or demographic.
It is frequently housed in multiple systems, involving multiple processes. And it is generated by multiple sources (both clinical and support staff).
Each data element collected during the patient care process becomes the building block for good information.

3. Data Quality can be defined as the assurance of the accuracy and timeliness of healthcare information.
Healthcare data quality can be defined as the assurance of the accuracy and timeliness of healthcare information.
Data quality is a critical component of any health information system.

4. Uses of Healthcare Data
Healthcare clinical decision-making, research, and treatment development
Public health and pandemic pattern detection
Management and policy decision-making such as actuarial premium setting, cost analysis, and service reimbursement
Business planning, accreditation, quality assurance, billing and reimbursement (revenue cycle), and compliance and risk management
Data is critical to the healthcare industry today. Sound, accurate, available, and reliable data is the foundation of many decisions at many levels. Personnel, both clinical and support staff, who perform the day-to-day operations related to patient care or administrative functions rely on information to do their jobs.
Of primary importance is the need for data to improve patient care. Providers need data in order to treat their patients and to choose among treatments. On the administrative side, payers require data to verify eligibility for treatment and determine medical necessity for care. Scientists, practitioners, and researchers need data for various initiatives such as outcomes measurements, patient care effectiveness, risk assessment and susceptibility and environmental exposure studies. Regulators and policy makers need data to make prudent and cost-effective decisions to ensure public health and to assure the availability of healthcare services.
Healthcare is an information-driven industry.

5. Quality Clinical Records = Quality Care
Quality data is the foundation for quality information that results in quality care.
The industry has always known that quality care can’t happen without quality documentation and quality clinical records, whether on paper or in electronic format
Quality data is critical for a healthy nation.
The National committee on Vital Health Statistics’ (NCVHS) workgroup on quality (in its May 2004 report) notes that:
“Efforts to improve health care in the United States can succeed only when those working toward that goal are equipped with accurate, complete and timely information...”
This is equally true in Canada and data quality is receiving national attention.

6. Characteristics of Data Quality
Accuracy: free of errors
Accessibility: easily obtainable
Consistency: recorded consistently to prevent misinterpretation or ambiguity
How do we know when we have achieved data quality?
Data quality is evaluated using a set of characteristics. These dimensions or characteristics of data quality include:
Accuracy: Is the data free of errors? Typographical errors in dictation or misspellings of names are examples of inaccurate data that could affect patient care (for example, a name misspelling could lead to filing a report in the wrong patient’s chart).
Accessibility: quality data is easily obtainable. To assess this we consider things such as whether previous records are accessible when and where they are needed, if important clinical information is available on the chart, and if computer-entry devices working properly.
Data consistency refers to the reliability of the data. Reliable data do not change no matter how many times or in how many ways they are stored, processed or displayed. (for example, an injury to the right knee should be consistently recorded as right knee.). A patient’s diagnosis of diabetes that is specified as to type I or type II (insulin dependent or not) in one system should be specified consistently as to type when transferred to another system.

7. Data Characteristics (continued)
Currency and Timeliness: data should be up to date and recorded at or near the time of the event or observation
Comprehensiveness: all the required data elements are captured
Definition: Users of the data must understand what the data mean and represent
Dimensions or characteristics of data quality also include:
Currency and timeliness refer to the requirement that healthcare data should be up-to-date and recorded at or near the time of the event or observation. Timeliness of the documentation or data entry has a direct effect on the quality of the data.
Comprehensiveness means that the record is complete. In both paper-based and computer-based systems, having a complete health record is critical to the organization’s ability to provide excellent patient care and to meet all regulatory, legal, and reimbursement requirements.
Definition refers to the meaning of the data and information in the health record. Users of the data must understand what the data mean and represent. Every data element should have a clear definition and, where applicable, a range of acceptable values.

8. Data Characteristics (continued)
Relevancy: relevant to the purpose for which it is collected
Granularity: Collected at the appropriate level of specificity
Precision: measurements are close to the actual size, weight, etc.
Finally, dimensions or characteristics of data quality include:
Relevancy—which refers to the usefulness of the data. The reason for collecting the data element must be clear to ensure the relevancy of the data collected. Accurate data may be collected about a patient’s color preferences, but is that relevant to the patient’s care?
Granularity is a characteristic that needs to be considered when establishing data definitions. It requires that the attributes and values of data be defined at the correct level of detail. For example, numerical values for laboratory results should be recorded to the appropriate decimal place needed for the meaningful interpretation of test results.
Precision is the term used to describe expected data values. As part of data definitions, the accepted values or value ranges for each data element must be defined. For example, a precise data definition related to gender would include three values: male, female, and unknown. Precise data definitions yields accurate data collection.

9. Threats to Data Quality
Design flaws
Methods for data collection
Technical errors
Interpretation differences
Interfaces, transferring data from one system to another
It is important to assess data quality because there are many threats to data quality. For instance:
Design flaws in paper forms or computer systems can result in poor data. (for example, a poorly designed form may be missing specific prompts for relevant information or has ambiguous prompts; systems design flaws include inappropriate use of mandatory fields or an automatic default entry that requires action to change the data)
Methods for data collection affect data quality. It is important to consider at what point it is best to capture certain data and who is the best staff person to do that. (for example, there is a greater potential for error if support staff are entering vital signs taken by the nurse)
Technical errors can occur as well if the data owner is not responsible for accuracy of the data. For example, formulas used for calculating medication dosages should be validated by someone with the clinical expertise to recognize an error.
Interpretation differences often lead to data quality issues. Use of nonstandard vocabulary or abbreviations can cause others to interpret the data differently than the source intended.
Interfaces from one application to another are often problematic. Though both vendors may claim they are HL7 compliant, they may be on different version of HL7 and thus cause problems in interfacing the data.

10. Barriers to Data Quality
Poor documentation practices
Outdated coding classification system in the US
Lack of data sets and data standards
Inconsistencies in reporting requirements
In addition to these threats within our systems and processes, there are also barriers to data quality within the healthcare industry.
Some of these barriers include:
Poor documentation practices (for example, documenting hours or days after the fact) may result in incomplete data or insufficient levels of detail.
The ICD-9-CM classification system is approximately 30 years old and no longer reflects current medical practice. Outdated diagnosis codes often do not reflect a sufficient level of detail needed for healthcare data reporting today.
Standardization of data sets and data definitions is a key to quality data that is comparable between healthcare entities. There is a lack of data standards across healthcare settings in the healthcare industry.
And we struggle with reporting requirements that differ significantly. For example, the uniform bill (UB92) is not used in a “uniform” manner. Only about 15 data elements are truly common across payers. For the other elements, payers have varying definitions for the field values.
Canada has moved forward to adopt the newest International Classification of Disease - 10th Version (Canadian adaptation) to build consistency at the global level.

11. Common Mechanisms to Ensure Data Quality
Audit and monitoring activities
Database, data warehouse design
Organizational data dictionary
System design including testing and initial evaluation
Maintenance and ongoing evaluation
This is why specific mechanisms are put in place to ensure data quality. These mechanisms include auditing and monitoring activities (especially of paper-based data).
(Note to presenter: Give examples of your organization’s auditing or monitoring activities.)
Mechanisms employed in an electronic record include database and data warehouse design, developing an organizational data dictionary, and system design including testing and initial evaluation.
(Note to presenter: If appropriate, give examples of mechanisms built into your database that prevent duplicate entries or ambiguous data entries.)

12. Data Quality Is No Accident
Ask not what your data can do for you, but what you can do for your data.
We need more than data—we need data we can trust.
Confidence in data is due in large part to its quality. This can become especially challenging to achieve when data is submitted from multiple facilities and different organizations.
Data quality is no accident.
We must be very methodical and intentional about how we manage our data if we want it to remain trustworthy.

13. Data Quality Management
Identify and resolve data quality issues
Routinely monitor and assess quality
Provide preventive maintenance
Support data users
Facilitate good data management
A data quality management program provides a rigorous means of routinely monitoring and improving the trustworthiness of the information that informs your decisions.
Data quality management typically involves:
Identifying and resolving data quality issues
Routine monitoring and assessment of quality
Preventative maintenance efforts
Support of data users (education, training)
Other activities that facilitate good data management
(Note to presenter: You may want to note the individuals involved in data quality management in your organization.)

14. Examples of DQM Efforts
Clinical documentation improvement programs
Assessment of clinical coding accuracy
Master Patient Index integrity
Examples of data quality management efforts include:
Clinical documentation improvement programs
Auditing for coding accuracy
Master patient index integrity
(Note to presenter: Give specific examples of your organization’s efforts.)

15. Data Quality has an impact both internally and externally
Internal Impact:
Quality data has a direct effect on quality patient care and patient safety.
[Note to presenter: challenge the audience to think of instances where data has affected patient care, positively or negatively. Some examples are availability of a specific report needed for patient care, or timely information that lowered a patient’s risk of complications]
External Impact:
The quality of data within your organization serves you outside your organization as well. We are heading into a time where your data will be used on a national level much more than it has been in the past. Examples in the US are: pay for performance (where care givers are reimbursed based on how they manage their information) and in Canada, includes performance accountability to government funders. In both countries, quality report cards are on the rise (on which our organization may be judged by the consumer).
While quality data it is important for delivering care at the patient level, it also defines the quality of care that you are delivering from an external perspective.

16. Quality Data  Accurate, Timely Information  Knowledge for a Healthy Nation
Quality data leads to accurate, timely information, which becomes the knowledge that clinicians need to improve healthcare.
Quality data is the foundation of quality information, which is critical to healthcare decision making for a healthy nation.

17. This is Health Information and Technology Week
This week we recognize the importance of health information and technology working together.
We recognize the vision of improved healthcare will be largely met by leveraging health information and technology to produce quality data.
November 6-12, 2005

18. HIM Vision
HIM is the body of knowledge and practice that ensures the availability of health information to facilitate real-time healthcare delivery and critical health-related decision making for multiple purposes across diverse organizations, settings, and disciplines.
Technology is certainly transforming the way healthcare is delivered, managed, and assessed, with a continued shift from records management to data management.
Electronic health records enhance, but don’t ensure, data quality. Data quality will remain a key focus and central HIM competency in an interconnected health system. We can expect the requirements to be far greater because the information in electronic records will be more accessible and more widely used.
HIM has a crucial role to play in helping the industry address its current and future health information needs to sustain and improve quality healthcare.

19. Information Management Is Critical to achieve Data quality
The need for more and better data requires a concentrated movement toward processes that place value on how data is defined, understood, analyzed, and interpreted.
A core fundamental function of HIM is to ensure quality health information for all healthcare settings and organizations that create and use it.
Data quality is everyone’s job, from registration to discharge. Each of us needs to focus on the data we work with and that passes in front of us to ensure that it is of the highest quality. This takes everyone working together, in a concerted effort.
If you have questions about how you can facilitate data quality, look to the HIM professionals in your organization as a resource on credible, accessible, and meaningful health data. [Note to presenter: modify this to direct the audience to a specific individual perhaps, as appropriate in your organization]

20. Questions?
Thanks for helping us celebrate Health Information and Technology week 2005!