Overcoming Data Quality Challenges in Healthcare
Praveen Kumar
Dec 8,
2022 |
Posted in
Data Services
Data quality is critical to healthcare as it is sensitive in nature and is also governed by regulations. However, it is not so easy to achieve considering different varieties, structures, and presentation norms. The healthcare industry in the U.S. faces many more data quality challenges like duplication, errors, and transparency issues. In this article, such challenges would be covered in depth to understand what they are, how they impact the healthcare space, and what can be done to improve data quality in healthcare.
Data Challenges in Healthcare
Data entry errors are quite common in healthcare and can have catastrophic impacts. In fact, 100,000 medication errors are reported to the FDA every year. There emerge over 12 million cases of misdiagnosis in America each year. A third of these people die due to these errors. As per the American Hospital Association, over 250,000 patients die annually because of medical errors like these. Over 9,000 of them die due to medication errors. An analysis of safety errors performed by ECRI Institute in Plymouth Meeting, PA revealed that 9% of the patient identification mistakes have caused harm or death to patients. Errors can be costly to patients in health. For healthcare organizations, this would mean a tarnished reputation and/or loss of money when they are sued by patients or governing institutes.
But why is maintaining data quality so difficult? What are the quality challenges that healthcare organizations face when collecting, processing, managing, or using data? And what does it means to have high quality data in a healthcare organization? Let us dig deeper into the idea of data quality in healthcare to answer these questions.
Data Quality in Healthcare: Concepts and Challenges
For healthcare, data quality can be seen as good if the data is able to support its processes like maintaining health records, performing diagnostics, extracting insights from disease analytics, designing medical policies, and public health surveillance. Data quality is understood from the perspective of characteristics including but not limited to:
- Availability
- Accuracy
- Validity
- Completeness
- Consistency
- Currency
- Usability
Let us try to understand the minimum requirements of quality requirements and the challenges that healthcare organizations and systems face in meeting them. Further, we will explore how maintaining data quality can support the healthcare purposes outlined above.
| Sr. | Data quality characteristics | What it means | How it works in healthcare | Challenges |
| 1 | Availability | Data must be available whenever it is needed | An electronic health record would make patient data available to doctors. This can include information on patient diseases, symptoms, medications, treatments, clinical views, risks, and lab results. | Server, network, or device failures, and security breaches are common causes that make data unavailable in healthcare. |
| 2 | Accessibility | Data must be accessible to the user who needs it | A wearable device reads patient vitals and transcribes the information to add to a medical record that doctor can access through the information system. | Visibility of data in real-time and interoperability across platforms. |
| 3 | Completeness | Data must be comprehensive | A prescription file should contain all the information including patient name, doctor name, drugs, date, time, doses, and expiry. | Three missing data problems: Missing competently at random (MCAR) (like missing lab readings), missing at random (MAR) (like missing observed data), and Missing not at random (MNAR) (like unobserved data). |
| 4 | Conformity | Data must be stored in the required formats | Universal data types and standard reporting formats are interchangeable and comparable that makes data easier to understand and use across organizations. | External data exchange between healthcare entities is difficult as organizations use different ways to capture, store, and link records. Patient matching which is done across the healthcare system to get a comprehensive view of the patient health record is one of the major healthcare challenges because of lack of conformity between sub-systems. |
| 5 | Consistency | Data must be the same irrespective of the source from which it is obtained | Diagnosis information tells the same story of a patient whether it is obtained from a lab or an EHR system. | Not every data system conforms to industry standards or is certified for quality. |
| 6 | Continuity | The data must neither be broken nor overlapping | EHR data continuity helps healthcare practitioners in classifying exposure and health outcomes for confounding judgements. | Power failures, hacks, compromises, and data loses impact continuity of data. Lack of continuity causes misclassification of treatments and their health outcomes. |
| 7 | Currency | The data must be updated in near-real-time to give the latest picture | As soon as the results of diagnostics are obtained by the lab, the patient records must be updated immediately. | Lack of real-time data updates can delay treatment decisions, thereby negatively affecting health outcomes. |
| 8 | Duplication | Data must have unique identities without duplication | Improper patient identification process often causes duplicate entries in the name of one patient. | Typically, a hospital experiences 10% duplication of records and accessing a fragmented version leads to poor decisions. |
| 9 | Privacy | Only authorized users must be able to view the patient’s data | Patient health records, genomic details and personal information are protected by healthcare organizations through privacy measures and training. | Government has defined rules and regulations for privacy that if not followed can lead to penalties. These include privacy violations that are no fault of the healthcare organization, those preventable, and resulting from willful neglect of rules. |
| 10 | Reliability | If the data is both complete and accurate, it can be trusted | Reliable data in electronic medical records serve as a useful guide that informs healthcare practitioner’s decisions about patients. | Data collected from cyber-physical devices like sensors pose questions on reliability due to varied sources that cannot always be trusted in terms of correctness of data. |
Data duplication in medical records is a common challenge that can add an estimated cost burden of up to $4.8 million. Besides the challenges in meeting quality characteristics, the most common errors in the healthcare sector are medication data errors. One in five patients receives wrong doses due to dispensing errors in the U.S. Every year, 53,000 injuries have been reported by outpatients because of medical errors. In 2016, 6.6% of patients overdosed, 7.8% received insufficient dose, and 5.4% were provided the wrong medication. In both cases, patients suffer. Some drugs that were commonly observed to suffer these errors are antidiabetic, anticoagulants, antiplatelet, and opioid drugs. These errors have caused 50% of emergency visits to hospitals. Among the most common causes of these errors is an administrative error.
Meeting the Data Quality Needs
Challenges are many but what if solution is just one? A comprehensive data quality management framework that meets all criteria and standards that make data rich.
A few measures that you must take to ensure high quality of data in healthcare are:
- Automate data workflows between data capture and electronic health records to eliminate manual errors
- Put the right data in the right place when recording any critical data value
- Ensure that data creators are adhering to the data policies
- Categorize data between providers, patients, payer, and plan data
- Implement a system that detects anomalies for correction of issues
- Automatic data drift detection can ensure relevancy, accuracy, and consistency of information
- Ensure compliance with federal and state standards like HIPAA for transparency
Overcoming Data Quality Challenges in Healthcare
Having high quality data is essential to the success of any healthcare organization. But different data varieties, structures, and presentations can cause the quality of data to weaken. Data quality issues are not only detrimental to healthcare facilities, but harmful to patients, too. Understanding data quality characteristics, how they impact the healthcare space, and the challenges they may present is the first step to creating a data quality management framework.
Apexon uses a comprehensive enterprise data strategy to establish standard methods, practices, and processes to ensure data reusability, sharing, access, accuracy, and portability in a repeatable manner. To learn more, check out Apexon’s data strategy services.