AWS for Industries

Improving Patient Engagement in clinical trials using voice and chat with AWS

Life sciences companies are rethinking patient engagement and legacy workflow processes in clinical trials due to low enrollment numbers and concerns around data quality. Voice and Chatbot solutions like Alexa and Amazon Lex, a fully managed conversational artificial intelligence (AI) service, can improve patient experience and increase patient engagement.

An estimated 48% of clinical trials do not meet their patient enrollment targets and 80% of trials are commonly delayed due to low patient recruitment and increasing time and costs. In addition, trial quality is also a concern as an estimated 30% of patients drop-out of trials after sign-on making the creation of positive patient experiences key to the success of a clinical trial.

As per clinicaltrials.gov, there are ~130,000 studies registered in the US alone, and ~20,000 of those studies are recruiting studies (involving trial patients) in the US alone. Each of these ~20,000 studies will recruit anywhere between several hundred to about 3,000 participants. This data provides a view into the large extent of impact that a solution around patient experience can make.

Using AWS preferred cloud technology solutions including Amazon Lex and Amazon SNS, our solution goals were to provide patient remote access, send proactive reminders, and capture responses in real time to daily questionnaires sent to respondents.

Summary of Challenges

Patient experience in completing questionnaires – Participants can drop out of a clinical trial if it is not convenient for them to answer questionnaires, or if there is a lack of guidance on the questions.

Questionnaire data quality and timeliness – Participants may not provide accurate questionnaire responses as scheduled, due to a lack of reminders. Delayed responses can create data quality issues.

Study results – Participants may not remember to take their medications as scheduled due to a lack of reminders. This could impede the usability of the results of the study.

Data analysis – Clinical trial administrators and coordinators lack an automated method to gather and analyze trial data.

Solution Objectives

Leveraging the latest voice and chat technology, we aimed to:

Provide remote access to patients via Amazon Alexa (Alexa) and SMS text to:

  • FAQs,
  • site information
  • patient medication diaries
  • questionnaires
  • information on transportation to clinical trial sites

Administer proactive reminders to patients to:

  • take medication
  • complete their patient diary

Capture near real-time patient responses to:

  • daily questionnaires via Alexa and text
  • report out through a dashboard for immediate intervention and action

Solution Overview

Assessment
Clinical trial questionnaires have been used for decades to assess the efficacy of a drug and the quality of life of the patient where objective measurements are not available. These questionnaires contain various types of questions, such as multiple choice or number-based questions. Questionnaires are also administered on various dates in order to accurately capture data and milestones.

A web interface enables creation of questionnaires for the clinical trial. From the web interface, a study administrator can create a set of questionnaires with multiple questions in each. They can also apply a schedule to the questionnaires, and have those questionnaires automatically administered on the respective days the trial patient is scheduled to answer them.

The notification process checks to see what activities/questionnaires are scheduled for the day. Based on that an Amazon Pinpoint segment is generated, and the respective campaign is updated with the segment. Trial patients then receive SMS notifications to complete their questionnaires on the scheduled date.

The response to a questionnaire via chat is enabled by Amazon Lex. Amazon SNS enables message management. The questions are fetched from an Amazon Simple Storage Service (Amazon S3). Amazon DynamoDB (DynamoDB) acts as the interface database and the responses are recorded into  Amazon S3 through Amazon Kinesis Data Streams and Amazon Kinesis Firehose.

When a patient invokes Alexa, an Alexa skill checks to see if the patient has any scheduled questionnaires for the day. Accordingly, the skill begins to ask the related questions. As patients provide answers, the skill logs their responses via the response application programming interface (API).

From there, the data is streamed into an Amazon S3. DynamoDB acts as the interface database to exchange real time data leveraging Amazon Kinesis Data Streams and Amazon Kinesis Firehose.

Architecture diagram depicting enabling text and voice interaction for patients in clinical trials, by deploying a design pattern using various Amazon services.

Reporting

The responses captured in Amazon S3 is queried by Amazon Athena to provide a dashboard in Amazon QuickSight. Amazon Glue provides the required data integration services. Clinical trial sponsors can run these reports to understand the response rate for questionnaires. They can further drill down to patient level data.

Lack of engagement by patients reflected by them not answering questionnaires or their responses indicating that they are not taking study medication, indicates possibility of potential drop out. Based on these near real time reports, intervention can happen to help such patients continue with the trial and prevent dropouts.

Here is a sample dashboard, which helps clinicians understand how the trial participants are responding, and aid with real time intervention:

A circle graph indicating patients’ responses about their medications and a side stacked bar graph indicating patients’ medications responses by date.

A circle graph indicating patients’ responses about their medications and a side stacked bar graph indicating patients’ medications responses by date.

In addition to making it easier for trial patients to respond to trial related questions leveraging voice technology, having real-time data on how the patients are responding also enables real-time intervention to prevent patients from dropping out of the trial.

FAQs

As part of patient engagement, FAQs play an important role to answer patient’s questions about the trial. These FAQs can be tailored for each individual trial and for specific trial sites. The FAQs can range in various topics including the study drug, duration of study, different phases of the study, screening, purpose of the study and many others.

The web interface provides the ability to configure the FAQs. The administrator can create multiple questions and provide answers, then link them to individual trials.

The Alexa skill interacts with the FAQ search API. This triggers an AWS Lambda function to Amazon OpenSearch Service to fetch the right responses to the questions asked by the patients through Alexa.

Reminders

One of the key outcomes of this solution is to improve patient adherence and enable patients to answer questionnaires in a timely manner. It improves the quality of data collected during the trial. This solution can also provide reminders via chat for medication, in addition to reminding the patients to respond to daily questionnaires.

Conclusion

Using voice and chat technology in patient engagement during trials has the potential to improve patient experience and reduce clinical trial dropouts. The cost of replacing patients during trials is three times the cost of recruiting a trial patient at the beginning. This also impacts the quality of data captured. Hence improving patient experience creates benefits through reduced dropouts, timely capture of data, and improving the overall quality of data captured. This solution can be used by pharmaceutical companies conducting trial or by their contract research organizations. Given the cost, quality and timeline impact due to patient dropouts in clinical trials, pharmaceutical companies should enable such digital transformation to enhance patient interactions during a clinical trial.

Contact your AWS Account Representative for a demo of this solution.

Kannan Raman

Kannan Raman

Kannan is with AWS ProServe Health care and Lifesciences practice at AWS. He has over 22 years of life sciences experience and provides thought leadership in digital transformation. He works with C-level client executives to help them with their digital transformation agenda.

Alex Emilcar

Alex Emilcar

Alex is a Conversational AI Architect, helping customers build digital experiences with AWS AI technologies. Alex has over 10 years of IT experience working in different capacities from developer, infrastructure engineer, and Solutions Architecture. In his spare time, Alex likes to spend time reading and doing yard work.

Mebz Qazi

Mebz Qazi

Mebz Qazi is a Senior Consultant working on global projects for AWS. He very much enjoys working on technological innovation in natural language and AI/ML in the fields of healthcare and education.

Brittany O'Sullivan

Brittany O'Sullivan

Brittany O’Sullivan is a Global Account Manager within the Healthcare and Life Sciences vertical at AWS. She has worked in the life sciences industry for over 15 years with a special interest in molecular diagnostics, precision medicine, and innovation. She is motivated by people and projects that aim to make the world a healthier, more sustainable, and patient friendly place.