AI Medical Chronologies from Settlement Intelligence
Share
An Introduction to AI Medical Chronologies
Medical chronologies are an important part of handling personal injury claims. For lawyers without extensive medical knowledge, medical chronologies or medical timelines, can provide lawyers with insights about the injuries and treatments involved in their clients' claims. For lawyers with moderate to advanced levels of knowledge of traumatic injuries, a medical chronology can provide insights about the specific factors of most value in a personal injury claim.
Some lawyers use medical timelines as the content for their demand letters, although for reasons discussed on this web site, we strongly recommend you do not write demand letters that contain a medical timeline. Instead, we recommend using the data in the medical chronology in a very different way for demand letters.
Recently a wave of AI Medical Chronology companies have flooded the legal field - many out of China, and others backed by venture capital companies or litigation loan companies. These are companies that have no background with the practice of law, or understand what data is used by insurance companies to evaluate claim value. This has created a lot of bad quality AI medical chronologies for reasons we discuss in this article.
So, if you are looking for AI medical chronologies to maximize the efficiency and effectiveness of your personal injury practice, this article will provide you the facts you should consider before choosing one platform.
AI Medical Chronologies are Faster
AI medical chronologies are generated in much less time than one created by a legal nurse consultant or law firm staff member.
A properly trained AI platform will analyze your medical records quickly and generate a medical timeline within seconds to hours depending upon the size of the client file.
By contrast, hiring a legal nurse consultant or agency to review a client file almost always takes time simply to find the consultant, and then longer to review the file. While it may take a few days to complete a short file with 100-300 pages, this same file can be finished by a thorough AI system in seconds to minutes.
For larger files of 5000+ records, a legal nurse consultant may take weeks to finish the medical chronology, whereas a thorough AI medical chronology program can finish it within hours. As AI systems become better trained, the time will continue to become shorter.
The speed of an AI medical chronology will increase the efficiency of your law firm and assist you in getting out demand letters more efficiently, allowing you to settle cases more quickly.
As discussed below, research on the use of Legal AI in law firms already demonstrates a distinct advantage for those law firms using AI to the point that for firms not using AI are already at a clear competitive disadvantage.
In terms of speed, there are companies advertising that they can evaluate "thousands of documents within seconds," but even using the fastest tech stack available today, this is not possible. Test their product and watch the progress of their platform to see if that claim is true. You also need to consider the quality of what you are getting in seconds, versus minutes. Ultimately, the accuracy and usability of the AI medical chronology is the most important factor.
Direct Access vs. Indirect Access to the AI Platform
In terms of speed, you also need to consider whether you get direct access to the AI system and whether that creates limitations on when you can run the AI medical chronology.
Unlike many other companies that require you to upload your records and then they run your records through their AI system before sending it back to you, Settlement Intelligence gives you direct access to our AI program and you upload the records directly to our system. This allows you to get results any time of day, and as quickly as possible.
AI Medical Chronologies are Less Expensive
Cost is also an issue to consider.
The cost per page of an AI medical chronology is usually done per page of records evaluated, or is part of a much more expensive package of products. Using AI medical chronology platforms you can usually obtain a medical timeline for $0.50 to $1.00 per page of records analyzed. So the price for a 100 page file will likely cost $50-100. A 300 page medical records file would be $150-300.
By contrast, hiring a legal nurse consultant or agency to review a client file almost always takes time simply to find the consultant, and then much longer to review the file. Historically, the charges for a legal nurse consultant range from $100-$200 per hour in the United States, which is the only place a legal case should be analyzed due to HIPAA prohibitions on the shipment of Protected Health Information outside the United States. Despite this prohibition some lawyers still take the risk in order to get legal nurse consultants or doctors from countries like India to evaluate their cases for $50 per hour. Regardless it would be difficult to get a human to perform a high quality medical chronology at the same cost as a highly trained AI platform. A human cannot evaluate medical records as quickly or as inexpensively as a trained AI medical chronology system.
AI Medical Chronologies are Equally or More Accurate than Human Review
Accuracy is the most important factor in determining whether to use AI or human evaluation. Some AI systems (discussed below) can be relied upon for consistent accuracy in every case. Our AI system was trained by doctors, lawyers and paralegals with a high degree of training in traumatic injuries and decades of experience in highly complex medical traumatic injuries. Our AI medical chronologies have been highly trained over two years to ensure the highest accuracy before taking our product to market, unlike many other AI medical chronology companies that launched on the basis of a "minimum viable product."
The same cannot be said for the accuracy of in house legal assistants or even paralegals with minimum medical training. Legal nurse consultants are certainly better in terms of accuracy, but for reasons discussed below, they often do not report the data of most importance in personal injury cases outside the field of medical malpractice. Even against the gold standard of a doctor, they will miss some of the data of value to the insurer in a claim evaluation because they often do not know the specific claim factors the insurance company is looking for when evaluating personal injury claims. This is where our founder's 20+ years of research into insurance claim software becomes important.
Representations about the accuracy of AI Medical Chronologies
Some of the companies that provide AI medical chronologies have used wording in their marketing materials that suggests they are 99% accurate. This is not true. A claim that a company is “training for 99%” does not tell you what their actual accuracy is.
We are not aware of any other AI Medical Chronology company that is providing accuracy percentages to their evaluations. By contrast, our patented technology will show you the accuracy of each category of records including:
- Liability
- Causation
- ICD diagnosis codes
- CPT and treatments
- Symptoms
- Duties Under Duress
- Loss of Enjoyment of Life
- Exacerbation or aggravations
- Permanency
- Future treatment probability
- Loss of earning capacity.
After two years of training our AI system is presently varying between 85% and 100% accuracy depending upon the complexity of the health care records. Both accuracy and consistency of high accuracy will increase substantially with customer use, as our AI platform will continue to gain experience analyzing a wide variety of medical records.
At trade shows, many of the AI medical chronology companies created by computer programmers with no backgound in law, have marveled at our 97% overall accuracy. Many companies are selling services to law firms with accuracy of 30% and no idea whether the data they are extracting is of any value in the insurance claim because they have zero background in personal injury law. While launching a "minimum viable product" is commonly done in the field of technology, we know you cannot use a highly inaccurate program for use in legal products.
AI can consistently deliver highly accurate data for claim analysis if the company implements cutting edge technology and trains the system extensively. The question is, does the AI platform you use deliver highly accurate medical chronologies, or is it just unsubstantiated marketing promises?
AI Medical Chronologies - are they useful or useless?
Unlike other AI medical chronology companies, the founders of Settlement Intelligence are not new to medical chronologies or demand letters.
The benefit of deep insights into medical records and insurance claim software
Our co-founder Aaron DeShaw is a retired doctor, award winning trial lawyer, and a leading lecturer on traumatic injuries. DeShaw has written several books on the insurance claim software systems that evaluate economic damages and noneconomic damages, and understands the inside details of how settlement offers are created by insurance companies. Our co-founder and CEO, Charlette Sinclair, has been a legal consultant on demand letters and settlement practices for nearly 20 years and has personally overseen teams of doctors, nurses, paralegals and lawyers in doing claim evaluation services.
The benefit of 20 years of experience in Medical Chronologies
Starting in 2010, the co-founders of Settlement Intelligence ran a company that provided medical chronologies and demand letters for lawyers throughout the United States. We are not new to this field, and many of the companies now providing these types of services are duplicating problems we already experienced 15 years ago and have intentionally avoided at Settlement Intelligence.
One of the issues our founders previously experienced in working with professionals on the creation of medical chronologies and demand letters is that legal nurse consultants often could not find and record the facts of importance in personal injury cases outside the field of medical malpractice. Their background is in medicine, and their medical timelines almost always focused on what was done correctly or incorrectly in the health care field. Despite extensive training to locate and report known "value drivers" in personal injury cases such as auto and premises liability cases, we could not get them to change the focus of their analysis to the factors we knew to be of value in auto and premises cases.
Most of the legal nurse consultants we've worked with in the past 15 years cannot re-focus their analysis on the factors that we know are of value to the insurance claim software used to evaluate auto, premises and workers compensation claims. That also assumes they could find out all of the 20,000+ factors evaluated by insurers, which would take decades of research that we have already done.
The inherent problem is, if the in-house staff member or legal nurse consultant doesn't understand the factors of value to the insurance claim software, the data recorded in the medical timeline will not be useful to getting the best settlement offer.
In fact, much of what is recorded in a typical medical timeline is not useful at all and adds a lot of unnecessary length to the medical timeline.
Just one of many examples is that the ICD code for the injury carries an inherent noneconomic damage value for physical pain. So, extensively discussing pain in a medical timeline and demand letter is of no value to a determination of the settlement offer except in rare instances such as migraine headaches. Despite that most medical timelines and demand letters contain discussions about pain that will go on for many pages.
The same is true for lawyers who use that medical timeline for their demand letters. While medical timeline demand letters have been largely ineffective since the implementation of insurance bodily injury claims software in the mid-1990s, using a medical timeline with content that doesn't convey any value as part of the demand letter compounds the problem. It presents the adjuster with information they won't consider, and will be set aside in favor of a demand letter that gives them what they need to evaluate the claim efficiently.
Creating highly effective AI Medical Chronology Technology
Over the past two years we've trained our AI model on hundreds of thousands of medical records to create AI medical chronologies that give you the information of greatest use in your case. No more scrolling through pages of useless information. The data provided in our AI medical chronologies is usable information. Our AI medical chronologies are designed for the specific purpose of effectively feeding insurance claim software to maximize claim value.
Are the AI Medical Chronologies from Settlement Intelligence Better than ChatGPT?
Yes.
Why? Beyond the problem of ChatGPT not being confidential (and its use therefore being an ethics violation for any lawyer using a public version of an AI system such as ChatGPT, Gemini, Claude, Perplexity, etc.), large language models are not specifically trained in how to analyze medical records for personal injury cases. Even if you get a medical chronology from the entry of Protected Health Information ("PHI"), it won't be good quality information for use in a legal case. These systems certainly don't have information on how a specific insurer will evaluate your client's claim.
Settlement Intelligence uses significant technologies and expert training with doctors, lawyers, and paralegals with decades of experience in traumatic injuries, insurance claim software, and demand letters, in order to train our AI. This is done by using a combination of Agentic AI and Retrieval Augmented Generation or "RAG" to accomplish highly effective AI medical chronologies.
Leading AI chip maker, NVIDIA, has a nice legally-related explanation of the concept of RAG using a courtroom analogy:
"To understand the latest advancements in generative AI, imagine a courtroom.
Judges hear and decide cases based on their general understanding of the law. Sometimes a case — like a malpractice suit or a labor dispute — requires special expertise, so judges send court clerks to a law library, looking for precedents and specific cases they can cite.
Like a good judge, large language models (LLMs) can respond to a wide variety of human queries. But to deliver authoritative answers — grounded in specific court proceedings or similar ones — the model needs to be provided that information.
The court clerk of AI is a process called retrieval-augmented generation, or RAG for short."
Retrieval Augmented Generation helps provide very specific answers for specific goals that a Large Language model like ChatGPT cannot.
Settlement Intelligence's RAG is a specially trained model for auto and premises liability cases. Unlike a court clerk accessing a library or system such as Westlaw or Lexis that has the ability to be queried with legal questions, our system has already been trained on 100,000s of medical documents specifically from auto and premises liability cases for over two years. It was then trained for two years to look for the specific facts of a case that we know are evaluated by insurance claim software.
Since we know that auto and premises liability cases are analyzed almost exclusively by insurance claim software, and we know the factors of value in creating the settlement values in those systems, our AI platform has specifically been trained to find, extract and report the factors that we know will be provided value in the evaluation by an insurance company.
We have trained our system not to generate massive medical timelines filled with data that is of no value to a claim. Instead, our Retrieval Augmented Generation, focuses on what matters for a specific type of legal case. This is why using our AI medical chronology is far superior to working with a legal nurse consultant who often looks for the factors of value in a medical malpractice case instead of focusing on the factors of value in an auto case. (We are separately creating a different RAG for medical malpractice cases that will be released in the future.)
How Retrieval-Augmented Generation Works
RAG differs from large language models (LLMs) through training on specific documents and retrieving specialized knowledge sources and industry specific training. Which is why entrusting your cases to a company that understand the personal injury legal field, insurance software, and medical records is critical, rather entrusting your clients cases to a technology company or venture capital company that has no background with law or medicine.
When a user submits documents, the system queries the RAG, and retrieves relevant data from the database to return the most accurate and context-specific responses.
Why did Settlement Intelligence take longer to release its AI Medical Chronologies?
Some of the flood of new AI medical chronology companies are just dressing up public LLMs like ChatGPT with graphics that make it look like a legally related product. Some are from China and use LLMs that will be trained on your confidential client information. Others claim to be using AI, but the work is actually done by former insurance adjusters.
Many of the AI medical timeline companies have launched "minimum viable product" AI systems that can do the bare minimum with poor accuracy, simply in order to get to market early.
Shockingly, some of the AI medical chronology companies that claim to also retrieve records know so little about the field of law that they believe they will get all of the records by getting access to a major hospital system - missing all records from small clinics that don't use large EMR systems such as Epic. (Bizarre) They claim that most of their clients are fine only getting the records they can find (much less than half the records in most personal injury cases.) This demonstrates how disconnected these companies are from the practice of law, and how little they understand about the purpose and use of a timeline.
By contrast, Settlement Intelligence created a specially trained AI model for analyzing a specific types of records for personal injury cases – analyzing 100,000s of documents and trained over two years.
New uses for AI medical chronologies - Case Selection
The ability to use fast and inexpensive AI medical chronologies provides lawyers the opportunity to use medical chronologies in new ways:
- Case selection. Given the speed, you can run a medical chronology before accepting the case so that you can choose to accept or reject a case based an analysis of the case facts. This helps you decrease the likelihood of accepting a case with significant records of pre-existing conditions that will refute a potential client's injury claims. If you choose to accept the case, you understand the challenges and can inform the potential client of the problems with their case from the beginning.
- Early Demand Letter. Sometimes a case has sufficient proof to result in a clear policy limits settlement before they are medically stationary. Running a quick AI medical chronology may help you understand when there is sufficient proof to send an early demand letter for policy limits.
Firms Using Legal AI have a Clear Competitive Advantage
Studies by Harvard Law School on the use of AI in law firms demonstrates that law firms that use AI, have a clear competitive advantage over law firms that are not using AI.
Highly Accurate Medical Chronologies, more quickly and at a fraction of the cost.
Considering every factor of value in a medical chronology, Settlement Intelligence offers a highly accurate medical timeline that has been trained to find and report the factors of value in an insurance claim evaluation, faster and at a far lower cost.
Settlement Intelligence AI Medical Timeline are available to Settlement Intelligence licensees first as the company starts its beta testing.
License Settlement Intelligence to get access to highly accurate AI Medical Timelines.
Author
This blog was written by lawyer Aaron DeShaw, author of the only legal treatise on insurance claim software, nationally renowned expert on personal injury demand letters, and founder of Trial Guides.