Appl Clin Inform 2017; 08(01): 137-148
DOI: 10.4338/ACI-2016-07-RA-0120
Research Article
Schattauer GmbH

Screening Consolidated Clinical Document Architecture (CCDA) Documents for Sensitive Data Using a Rule-Based Decision Support System

Beatriz H Rocha*
1   Wolters Kluwer Health, Waltham, United States
2   Harvard Medical School, Boston, United States
,
Deepika Pabbathi
3   Partners HealthCare System, Information Systems, Wellesley, United States
,
Molly Schaeffer
3   Partners HealthCare System, Information Systems, Wellesley, United States
,
Howard S Goldberg
2   Harvard Medical School, Boston, United States
3   Partners HealthCare System, Information Systems, Wellesley, United States
4   Brigham and Women’s Hospital, Boston, United States
› Author Affiliations
Further Information

Publication History

Received: 20 July 2016

Accepted: 04 February 2016

Publication Date:
20 December 2017 (online)

Summary

Background: The Centers for Medicare & Medicaid Services’ Stage 2 final rule requires that eligible hospitals provide a visit summary electronically at transitions of care in order to qualify for “meaningful use” incentive payments. However, Massachusetts state law and Federal law prohibit the transmission of documents containing “sensitive” data unless there is a new patient consent for each transmission.

Objectives: To describe the implementation and evaluation of a rule-based decision support system used to screen transition of care documents for sensitive data.

Methods: We implemented a rule-based document screening system to identify transition of care documents that might contain sensitive data. The transmission of detected documents is withheld until a new patient consent is obtained. The documents that were flagged as containing sensitive data were reviewed in two different time periods to verify that the decision support system was not missing documents or withholding more documents than necessary.

Results: The rule-based screening system has been in regular production use for the past 18 months. During the first evaluation period, 3% of 5,841 documents were identified as containing sensitive data (true-positive rate of 44%). After additional enhancements to the rules, the system was evaluated a second time and 4.5% of 6,935 documents were identified as containing sensitive data (true-positive rate of 98.4%).

Conclusion: The analysis of the system demonstrates that production rules can be used to automatically screen the content of transition of care documents for sensitive data. The utilization of the rule-based decision support system enabled our hospitals to achieve meaningful use and, at the same time, remain compliant with state and federal laws.

* The study was conducted while the primary author was a Principal Medical Informatician at Brigham and Women’s Hospital