Contract research organizations (CROs) should implement risk-based monitoring (RBM) as a top priority for medical device clinical studies. Use of modern data technologies for remote risk-based monitoring can help reduce non-value added rework, and  dramatically improve patient compliance in medical device clinical trials and help speed up time to statistical report.

The goal of a medical device clinical trial is to test the efficacy and safety for the intended use of the device. In many cases, medical devices are used at home by patients and caregivers for chronic diseases such as diabetes, migraine, heart failure, chronic constipation.   Medical devices are connected wearables for treating chronic diseases (like the Theranica Therapeutics migraine wearable for neuromodulation therapy) or implanted connected devices (like  the Vectorious minimally-invasive implanted device that is packaged with an external home-unit; the in order to increase CHF patient quality of life and reduces readmission rates).

CROs are charged with the task of conducting medical device clinical trials that are valuable to the healthcare industry and produce near-real-life results. RBM helps assure that data collected is reliable, and is highly applicable to help develop innovative home-use medical devices by assuring high levels of patient compliance to the study protocol.

In this post, we are going to go over how remote risk-based monitoring can help your current study and also provide value for your medical device development.

What is risk-based monitoring?

Risk-based monitoring entails the mitigation of risks during a medical device clinical trial that a CRO is conducting through the process of identifying, assessing, and then monitoring the risks associated with patient safety throughout the course of the study.

In August 2013 the FDA issued draft guidance for “A Risk-based approach to monitoring”

“The overarching goal of this guidance is to enhance human subject protection and the quality of clinical trial data by focusing sponsor oversight on the most important aspects of study conduct and reporting”

The draft guidance includes three steps in a risk-based approach to monitoring:
1. Identifying critical data and processes. To accurately monitor the quality of a study and the safety of its patients, sponsors must know which elements are vital for each particular study, including informed consent to eligibility screening and tracking of adverse events.

2. Performing an assessment of risk. Risk assessments requires determining specific causes of risk and the effect of study errors pertaining to risk.

3. Developing a comprehensive monitoring plan. According to the FDA’s guidance regulations, RBM plans should “describe the monitoring methods, responsibilities, and requirements of the trial.” Planning is responsible for communicating risks and monitoring procedures to each party involved in trial RBM.

With Case Report Forms (CRFs) now being completed with cloud electronic data capture (EDC) software systems, (or entered directly using eSD (electronic source documents) or ePro (electronic patient reported outcomes),  it is possible to to collect CRF data into a centralized database, accessible by all parties, along with a full register of operational, quality controls, efficacy and clinical safety data from all sites, subjects and devices across multiple studies.

Remote risk-based monitoring algorithms are designed for scanning the data for specific and calculate specific metrics that trigger alerts to the clinical data management team – for example excessive query rates or the cycle time from patient visit complete to eCRF data entered both of which are indicative of data quality issues at sites

Remote risk-based monitoring metrics may be considered in two main categories:  patient safety risk (for example trends of vital signs) and data quality risk (for example long data entry cycle times).

When human study monitors visit sites once every 4-6 weeks, it is hard to catch slow-moving, high risk events.  Human brains are not good  at identifying and classifying  small numbers of slow-moving events, although they can easily identify a face moving quickly across a basketball court. 

Computer algorithms are outstanding at identifying anomalies in time-based clinical data found in clinical trials – and this is where RBM will have its biggest payback for you in your study.

What comes next?

What comes next is to make an accurate assessment of the data model. A single number, or a graph displaying data points, data in any form, has no meaning until the data is coherent and understood by a clinician with how the data is understood, and interpreted with how it can produce increased or mitigated risk to the patient, and the study as a whole.

Making RBM your study’s mantra

Consider this example: if a medical device study has a greater than expected negative-result reporting rate, are we talking about the great attention to detail of the study staff, or the negative performance of the interventional device? How do we determine this?

The answer is that you need to dive in deeper, and have monitors in place that are not only looking at the data, but observing whether the site is performing at the maximum study level, or whether the data is plainly reflecting that the device is effective for the study subjects.

An essential component in performing RBM is that it is required to guarantee that it can ensure that your data collected either indicates that risk is assessed, and that analysts can repeat, from site-to-site (if multiple sites are used) consistent performance that will protect study subjects, and reduce the time and effort it takes to identify risks.

The next crucial course of action during the study is to use the findings to take a direct course of action. With a clear understanding of a site’s specific risk level, CROs should be designated to visit the site(s) and take appropriate measures to alleviate site risks.

With a coherent and thorough understanding of study risks, CROs will take the necessary steps to not only reduce patient risks, but reduce future patient risks, helping your studies run more efficiently and save future costs.

What to look out for, and how to approach challenges

In our experience it has been proven that combined analysis, EDC system automation, visualisation tools, along with the monitors’ innate abilities, established processes for corrective actions have resulted in an efficient and effective uncovering of risks. This can, and will, put data and patients, and thus the entire study in jeopardy. What CRO monitors should look out for include:

– Fabricated, false or manipulated data (remember Theranos – where the financials were as cooked as the results?

– Missing data omissions

– Deviations from protocol, poorly trained clinician data submissions

Take for example if there was data that showed an alarmingly  high rate of a rise in blood pressure for study subjects. Monitors should not only be looking into the increase in blood pressure, but also examining whether the data being submitted is accurate. 

Monitors should consider whether the submitted data is wrong, or if the data is sound and the device is at fault or patients were taking conmeds that contraindicated the treatment.

RBM as your guiding light

RBM is not just second guessing every step of your clinical trial. It is a culture of study operations. The goal is to ensure that a CRO is approaching studies with not only patient care in mind, but also study efficiency and lower study costs. This is based upon past assessment of data monitoring and monitoring abilities. RBM is a CRO’s past, present, and future for data assurance and patient safety.