Domain

Clinical Informatics

Type

Protocol

Theme

quality; operations

Start Date

7-6-2014 1:15 PM

End Date

7-6-2014 2:45 PM

Structured Abstract

Introduction

The proliferation of Electronic Health Records (EHR) greatly advances the opportunity to capture electronically captured clinical data for QI, CER, and PCOR-focused analyses. The purpose of this study is to determine the contribution of data validation reports (DVR) in improving the completeness and correctness of hospital-based EHR data. Insights gained from this study inform a methodology that measures and improves EHR data quality and integrity.

Background

The Perioperative Outcomes Initiative (POI, www.poi-cqi.org) is a collaborative of 26 Michigan hospitals with clinical OR data from EHRs for QI purposes. Participating hospitals feature mature EHRs that extract data elements from local data warehouses. The sample included 17 POI hospitals that submit general surgery cases and perioperative data elements on a quarterly basis. A total of 57,000 cases were included between 2010 and July 2013.

Data Collection: The EHR data are extracted according to the POI Data Dictionary, de-identified, and uploaded to a secure POI server. The DVRs are released daily through a secure website for review by hospital IT and OR staff. Further telephone correspondence is provided to address any questions to address data resubmission.

Analyses: Review of changes in DVR results from 2010 – 2013 (Quarters 1 and 2) aimed to determine the number of missing cases (completeness) and the correctness of values (logical consistency) of the data presented in descriptive reports.

Findings

The DVR approach was well received by participating hospitals and led to improvements in the data extract. The DVR approach allowed multiple resubmissions till the data was correct. The results show that DVR, with telephone and/or a site visit contributed to a statistically significant decrease in the missing rate of patient cases from 37.1% in 2010 to 13.7% in 2013 (p-value < 0.05), which represents a 63% improvement in the missing case rate. Hospital staff confirmed more complete data and reduced inconsistencies based on EHR reports. Since site visits and telephone calls occurred within the same timeframe, it is not possible to establish the independent effect of each of these interventions. However, the combined effect confirmed increased completeness and correctness of the data.

Discussions

Developing a complete data validation program (DVP) for EHR data is a top priority to further the understanding of EHR data. The DVR approach proved effective in improving the completeness and correctness of perioperative data elements. Despite documentation of all interactions and reports, independent contributions of the data validation approaches could not be established in this study. Comments from hospital IT and OR staff confirmed little internal insight in local EHR data and limited use of EHR data for hospital-based QI, CER, or PCOR initiatives. Submission of standardized data to the CDR provides valuable cross-institutional benchmarking and data validation opportunities to improve the quality of the EHR data.

Conclusions

A preliminary data validation program to establish the completeness and correctness of EHR data provides a framework to learn about and improve EHR data quality. Further development and evaluation of the data validation program will stimulate collective learning about the quality of EHR data.

Acknowledgements

We would like to acknowledge BlueCross BlueShield of Michigan for their generous funding of the Perioperative Outcomes Initiative (POI, www.poi-cqi.org)

Creative Commons License

Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License.

Share

COinS
 
Jun 7th, 1:15 PM Jun 7th, 2:45 PM

Evaluation of a Data Validation Protocol to Use EHR Data for Improvement Purposes

Introduction

The proliferation of Electronic Health Records (EHR) greatly advances the opportunity to capture electronically captured clinical data for QI, CER, and PCOR-focused analyses. The purpose of this study is to determine the contribution of data validation reports (DVR) in improving the completeness and correctness of hospital-based EHR data. Insights gained from this study inform a methodology that measures and improves EHR data quality and integrity.

Background

The Perioperative Outcomes Initiative (POI, www.poi-cqi.org) is a collaborative of 26 Michigan hospitals with clinical OR data from EHRs for QI purposes. Participating hospitals feature mature EHRs that extract data elements from local data warehouses. The sample included 17 POI hospitals that submit general surgery cases and perioperative data elements on a quarterly basis. A total of 57,000 cases were included between 2010 and July 2013.

Data Collection: The EHR data are extracted according to the POI Data Dictionary, de-identified, and uploaded to a secure POI server. The DVRs are released daily through a secure website for review by hospital IT and OR staff. Further telephone correspondence is provided to address any questions to address data resubmission.

Analyses: Review of changes in DVR results from 2010 – 2013 (Quarters 1 and 2) aimed to determine the number of missing cases (completeness) and the correctness of values (logical consistency) of the data presented in descriptive reports.

Findings

The DVR approach was well received by participating hospitals and led to improvements in the data extract. The DVR approach allowed multiple resubmissions till the data was correct. The results show that DVR, with telephone and/or a site visit contributed to a statistically significant decrease in the missing rate of patient cases from 37.1% in 2010 to 13.7% in 2013 (p-value < 0.05), which represents a 63% improvement in the missing case rate. Hospital staff confirmed more complete data and reduced inconsistencies based on EHR reports. Since site visits and telephone calls occurred within the same timeframe, it is not possible to establish the independent effect of each of these interventions. However, the combined effect confirmed increased completeness and correctness of the data.

Discussions

Developing a complete data validation program (DVP) for EHR data is a top priority to further the understanding of EHR data. The DVR approach proved effective in improving the completeness and correctness of perioperative data elements. Despite documentation of all interactions and reports, independent contributions of the data validation approaches could not be established in this study. Comments from hospital IT and OR staff confirmed little internal insight in local EHR data and limited use of EHR data for hospital-based QI, CER, or PCOR initiatives. Submission of standardized data to the CDR provides valuable cross-institutional benchmarking and data validation opportunities to improve the quality of the EHR data.

Conclusions

A preliminary data validation program to establish the completeness and correctness of EHR data provides a framework to learn about and improve EHR data quality. Further development and evaluation of the data validation program will stimulate collective learning about the quality of EHR data.