Mass Deduplication

Overview
Netsmart’s Mass Deduplication solution provides powerful, sophisticated matching algorithms to locate duplicate records in existing data and helps eliminate new duplicates.

Business Benefits

• Lowers costs
• Improves patient care
• Reduces medical errors
• Removes complexity from the deduplication process

Features
Mass Deduplication simplifies the deduplication process through an easy-to-use graphical user interface (GUI). The software can be installed and start deduplicating data in less than an hour. Even though the Mass Deduplication software is simple to use, it uses a sophisticated proprietary process to sharply divide the real duplicates from similar records (e.g. twins). Key features include:

• Processes large databases.
• Sophisticated typographical error recognition.
• Based on a proprietary pattern probabilistic scoring technology, so accurate that most matching decisions can be made without human intervention.
• No “tuning” needed – is pre-calibrated for health data.
• Flexible input data sources.
• Can run on a stand-alone workstation, disconnected from the database. Doesn’t tie up valuable server and network resources.
• Reports less-confident matches for human resolution.
• Includes National Death Index and census name frequency data when deduplicating small databases.
• Can print multiple reports at different quality and confidence settings quickly, without re-running the duplication.
• Single license fee entitles customers to future upgrades with no ongoing maintenance costs.

Netsmart offers options for agencies dealing with data from multiple sources – ranging from stand-alone databases to Regional Health Information Organizations (RHIOs).

Option 1
Netsmart Mass Deduplication
A powerful, stand-alone deduplication and analysis tool for patient records or health data. Uses pattern-matching algorithms that go beyond the current generation of probabilistic matching.

Option 2
Insight-Match
An automated Web Service that supports real-time data deduplication and record matching. Utilizes the same level of sophistication as in Netsmart Mass Deduplication.

Option 3
Deduplication Services
Custom solutions for closely integrated data in existing systems, including real-time record matching.

Option 4
Insight-Master Patient Index (MPI)
Used for data coordination and deduplication in multiple systems. Builds on the deduplication technology of Mass Deduplication and the real-time Web capabilities of Insight-Match, adding record linkage among multiple systems with multiple identifiers.

Best-of-breed Solution

• Complete: Initial query and scoring is fully configurable to user requirements for accuracy versus automation rate. Selecting a higher level of accuracy reduces the percentage of true matches that are linked without human intervention. Reports can be run quickly, with different sort orders to evaluate deduplication accuracy.
• Fast: Designed to achieve >99.9 percent accuracy, measuring false positives on all automated decisions. Initial scoring speed depends on the processor count and speed, available memory, bulk-storage system speed and system activity. A machine dedicated to deduplication out-performs a general-purpose server.
• Accurate: Finds easily-recognizable duplicate records, as well as less obvious typographical and other errors in names, addresses and dates. The software also provides an audit trail of linked records, so that mistakes can be corrected. Twin-recognition is built-in from the start rather than being an add-on feature.
• Feeds Automatic or Manual Link/Merge Tasks: Provides data files to automatically link or merge client records that are determined to represent the same person. Users can also opt for human review of the reports prior to automatic link or merge.
• Queues Questionable Records for Human Evaluation: Records with insufficient data for automatic link or merge are queued for human evaluation. Reports and interactive decision tools are available to deal with these problem cases. These records are maintained in the database as separate records until the issue is resolved. Questionable matches that are researched by staff and that do not represent the same person are flagged to prevent re-evaluation by automatic computer processes or resubmitting the records to humans for re-evaluation.
• New Queries Examine Only New and Changed Data: Only new and updated data are evaluated to find new matches that may not have been previously apparent. These are always evaluated against the entire database. Typically, this occurs quickly, since new and updated data are only a small fraction of a registry database.

Contact us today for additional product information.