Re-Indexing Historical Medical Records

Document Separation, Classification and Indexing



Solution

Project Scope:

The project had an implementation deadline of four weeks to achieve the following:

  • process six million pages.
  • separate batches into distinct document types.
  • classify 100 document types.
  • index patient PII to match record in EMR system.
  • index date and description of each sub-document.

A leading healthcare provider approached our partner with a significant challenge they were facing in re-indexing historical medical records into their Hyland Onbase system.

Their primary challenge was the time-consuming nature of manual document classification and indexing. With a massive backlog of 6 million pages, the process was not only slow but also error-prone. Different types of documents required separation, around 100 distinct document types needed to be accurately classified, and sensitive patient Personally Identifiable Information (PII) had to be indexed to match records in the Electronic Medical Records (EMR) system. Additionally, each sub-document needed to be indexed with its corresponding date and description.

The project involved developing a robust workflow within the Hyland Onbase system that could automate much of the indexing process. Our partner's team worked closely with the customer's internal stakeholders to design and implement the solution.

The solution involved the following key steps:

  • Document Separation and Classification: Aluma created a systematic process to separate distinct document types within the batches. Advanced machine learning algorithms were employed to classify the documents into approximately 100 categories based on their content and structure.
  • PII Indexing and EMR Integration: To ensure compliance and accuracy, patient PII such as names, dates of birth, and medical record numbers were extracted from the documents and cross-referenced with the existing EMR system. This integration helped match the historical records with the correct patients.
  • Sub-document Indexing: Each sub-document within the medical records was meticulously indexed with its corresponding date and description, allowing for easy retrieval and reference.

Impact

Streamlined Workflow Automation

The project successfully automated the indexing process, addressing the challenge of time-consuming manual indexing.

Efficient Document Management

Advanced machine learning algorithms enabled the systematic separation and classification of approximately 100 document types, significantly enhancing the efficiency of document handling and retrieval.

Enhanced Data Accuracy and Compliance

Through PII extraction and EMR integration, the solution ensured high data accuracy, enabling the matching of historical records with the correct patients and resolving compliance issues. Sub-document indexing improved record organization and retrieval.

Printable PDF Version


Medical records dispalyed on a tablet and a stethoscope

16 Oct 2023


Next case study

Insurance documents with car key and digital lock

Ready to get started?

Get in touch or request a free trial.
X

Product enquiry

If you want to contact us about one of our products then use the form below.