More Than A Monolith – The Various Ways Attorneys Can Leverage Relativity Beyond Traditional Document Review

Taft Stettinius & Hollister


Established in 1885, Taft is a nationally recognized law firm serving individuals and businesses worldwide, in both mature and emerging industries.
As big data is here to stay, attorneys must find ways to adapt and work with seemingly impossible amounts of information from their clients.
United States Food, Drugs, Healthcare, Life Sciences
To print this article, all you need is to be registered or login on

As big data is here to stay, attorneys must find ways to adapt and work with seemingly impossible amounts of information from their clients. Attorneys must understand how to leverage tools, like Relativity, to effectively and efficiently manage large datasets and provide exceptional client service. Relativity is a tool that can be used across various disciplines – beyond the typical document review functions for which the software is commonly known. The power behind a tool like Relativity is its organizational and collaboration capabilities. Whether a white-collar attorney uses the tool for investigations, a civil litigator uses it for discovery, or a data privacy lawyer uses the tool for security incident response, Relativity transcends practices.

Below is a high-level overview for getting started, regardless of the project type.

Traditional Document Review With a Twist

Relativity can accommodate traditional document reviews but can also be leveraged in ways unique to each practice area.

1. Determine How Relativity Can Benefit the Project

Relativity should be considered when a project involves a considerable number of documents. The core functions include:

  • Document Categorization/Organization – Data of all volumes can be ingested through the platform. Using Relativity to organize your project is a major benefit. A Relativity workspace can be customized for a project's specific organizational needs using various issue designation fields to categorize and sort documents.
    • Analytical Data. Within this categorization/organization function, statistical information can be obtained about a data set. Relativity also can export coding data in Excel format for a client-ready deliverable.
  • Document Production/Sharing – Relativity is often the go-to tool when producing documents in response to a subpoena or discovery request. The platform can also be used when working on large-scale projects with multiple legal teams, and sharing documents between firms is a necessity. Relativity provides a great way to manage which documents are shared and produced between non-adverse parties.
  • Soliciting Attorney Feedback – An often-overlooked capability of Relativity is the attorney feedback function. The platform allows multiple attorneys to review and comment on the same document simultaneously. This feature may be used for contract reviews and risk assessments.

2. Identify the Data Set

At the start of any document-related project, an early step is to determine:

(i) The types of documents within the data set (e.g., internal policies, medical records, manufacturing plant data); and

(ii) The information you want to learn from the data set. Are you looking for documents from a particular time, or privileged communications or documents with sensitive personally identifiable information (e.g., SSNs)? Because a data set can contain hundreds of thousands of documents, using search terms or other filters to cull the data pool to be reviewed can save clients valuable time and money. For example, if a civil investigative demand requests information for a specific time frame, documents outside that window should be removed from the review pool. Search terms are another way to limit a data set. In some instances, the review pool must include the entire data set. In those situations, attorneys should consider using Active Learning. Active Learning is a technology-assisted review tool that helps organize data and predict which documents are most likely to be relevant to reviewers. Active learning works by using a technology called support vector machine that continuously learns from the coding decisions of reviewers. The coding specific for Active Learning is binary: AL Relevant or AL Not Relevant. These coding decisions are ingested by the active learning model where machine learning takes place. As reviewers code more documents, the model becomes better at discerning what makes a document relevant or not relevant and begins to serve the best documents to reviewers.

3. Develop a Document Review Protocol

Having a document review protocol that outlines what the review is about, how the documents should be coded, and who questions should be directed to is the best way to ensure a document review is productive. Without a protocol, document reviewers are left trying to read minds. The basics of a document review protocol should include an explanation of the data set and the project's goal. For example, is the client a health care provider responding to a grand jury subpoena for various patient records? The protocol should also include any specific subject area information necessary for a productive review. Is the health care provider mainly providing medication-assisted treatment for patients with substance abuse diagnoses, and is there a concern that the records will reveal a provider was prescribing a particular drug outside the normal course of medical practice? The review protocol should also inform reviewers on when and how to code documents. For example, if a document is non-responsive, should the coder continue to code for privilege or stop coding as soon as a document is marked non-responsive? In some projects, it may also help to have a cast of characters that reviewers will likely see during the project. Finally, designated points of contact should be listed for reviewers to contact with substantive and/or technical related issues.

4. Get Started and Leverage the Results

Combining steps one through three, the final step is to begin the project and examine the results. To illustrate, let's walk through a HIPAA risk assessment example. In this scenario, the task is to assess certain policies based on their effectiveness in guarding against HIPAA security violations using a sliding risk scale. As for step one, thinking about how Relativity will benefit the project will require thought about the final deliverable – "What will the assessment entail, and what will it identify for the client?" In this example, the answer is twofold: the assessment will require a review of existing client policies and procedures and determine the risk/vulnerability posed to the client based on those policies and procedures. The Relativity function most beneficial for this project is likely the document categorization/organization feature, to categorize the client's policies and procedures based on risk level.

Moving to step two, identifying the data set, the client's external and internal-facing policies, procedures, and agreements relating to HIPAA and the collection and use of protected health information (PHI) would form the data set. The information that would be obtained from those documents is the threat each policy and procedure poses to the client. If a previous risk assessment was completed in 2013, narrowing the data set to exclude documents and policies predating 2013 for efficiency may be beneficial.

Next, a document review protocol would:

(i) Describe the client's internal and external HIPAA-related policies, procedures, and contracts as the data set;

(ii) Detail the document categorization task (in this case, the document reviewer would review relevant documents and code them "low," "moderate," or "high" risk); and

(iii) Outline any secondary document categorizations (e.g., additional coding to determine whether the HIPAA-related policies or procedures need revisions using "yes/no" buttons). This review protocol would then be given to everyone working on the project.

Finally, the project team would conduct the review consistent with the protocol and review the results. Relativity can then generate a report outlining the results of the document categorization, such as the number of policies that fall within each risk category, the percentage breakdown of policies by risk category, and the number of HIPAA-related documents that require revisions and export documents by risk category.

The illustration above is just one potential function of Relativity. Relativity can add significant value to various projects where voluminous data is involved, especially for non-ligation projects where Relativity is not typically considered as a solution. Attorneys are encouraged to get creative and start thinking about how this tool can make their practice more efficient.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.

See More Popular Content From

Mondaq uses cookies on this website. By using our website you agree to our use of cookies as set out in our Privacy Policy.

Learn More