Research

PainDA – Pain, Depression, Anxiety

Low back pain (LBP) is a global issue with a prevalence point of around 8% of the world population (~ 577.0 million people). Approximately 60% of patients with chronic pain suffer, in addition, from two or more psychological conditions like depression and anxiety. In nearly 52% of the cases, LBP patients also suffer from anxiety, depression, underdiagnosed, perpetuating the economic burden and the mental conditions.

Importantly, this synergy between pain and mental health conditions may change dynamically and not be well-understood. We are focusing on researching pain, depression and anxiety that usually change over time in daily life. We build an integrative system that will combine daily self-reported assessments, physiological and behavioral digital markers.

This Also, advanced computational approaches will be used for analysis of linguistic characteristics and behavioral markers of daily life activities over time, along with physiological measures, traced by the Fitbit smart watch device.

Better understanding of the compound mechanism that underlie chronic pain and mental health conditions will allow us to improve clinical approaches and prevent aggravation of these synergetic conditions.

The study is in collaboration with, Prof. Hagit Hel-Or (Department of Computer Science), Prof. Hadas Okon-Singer (Department of Psychology, Cognition Emotion Interaction Lab), Prof. Sigal Zilcha-Mano (Department of Psychology, Psychotherapy Research Lab) and Prof. Tzvi Reiss (Department of Statistics).

PainStory - pain detection based on facial expressions, voice and language content

Website: https://PainStory.science

The problem. Chronic pain, generally classified as pain that persists past normal healing time and lasts for longer than 3 to 6 months, affects almost 2.5 billion people, and is the main cause of the opioid crisis.1 Costs for treating and managing chronic pain conditions are greater than that for heart disease, cancer, and diabetes combined.2

While these facts are alarming, they paint an impersonal picture of the condition, detached from the actual daily suffering experienced by patients. As Kurt Tucholsky once said, “a single death is a tragedy, a million deaths are a statistic”, reflecting our protective tendency to simplify a harmful reality.

A critical challenge in pain management emerges from the fact that pain cannot be directly measured. Therefore, currently, the assessment of pain is based on a single 0-10 scale articulated by the patient with significant limitations especially for long-term usage: (1) being subjective and repetitive, it is biased and not sensitive enough, resulting in an unrealistic burden on the patient; (2) this oversimplified pain estimation does not capture the multiple aspects of chronic pain, such as its emotional and social impacts; and (3) shows low consistency with patients’ judgment about the severity of ongoing pain and its effects on daily life.3 As result, this easy-to-use approach to pain assessment means clinicians don’t have an accurate measurement tool for diagnosis or an evaluation of treatment efficacy. Therefore, objective measures of pain are needed to better inform pain management.

Our goals:

To develop an accurate objective pain assessment for clinical pain monitoring.
To reduce the stigma associated with chronic pain conditions and to increase awareness about chronic pain.

Our approach. We are developing a completely novel and intuitive digital platform, PainStory.science, that allows us to collect audio/video recordings from pain patients sharing their experiences of living with chronic pain. PainStory is a self-serving digital platform that uses advanced cybersecurity protocols to ensure participants’ privacy. The patients will be asked to describe their current symptoms, related emotions and suffering, the causes of the pain, and what makes the pain better or worse – by talking into their smartphone’s camera and microphone at home or in in their normal environment. In addition, they will rate their pain levels and complete a series of psychosocial surveys.

Using advanced machine learning approaches, the PainStory platform will “listen” and analyse patients’ narrative content, their vocal nuances and associated facial expressions to develop a personalised pain assessment. The PainStory assessment will be used for better clinical evaluation, and digital follow-up between the clinical visits. Relying on cheap and available sensing technology (microphone and camera), our solution will be easily scalable and adoptable. A select number of the pain narratives will be shared publicly (with patients’ approval) to empower others suffering from chronic pain.

PainRadar: skin-based chronic pain detection

We are developing a new biomarker for fibromyalgia with multi-modal techniques, including biochemical, brain-related, and physiological measures. Applying machine learning tools, we will attempt to classify fibromyalgia patients and their pain levels with neurophysiological and biochemical information. Further our model could be transformed into fully automated real-time detection of fibromyalgia and ongoing pain. The automated pain tracking has considerable potential to improve the efficacy of pain treatments, by providing just-in-time feedback and triggering interventions. Overall, the proposed project will contribute fundamental scientific knowledge about biochemical and neurophysiological signs of real-life pain and lay the groundwork for translational efforts to improve outcomes of pain self-management and reduce opioid use.This project is conducted in collaboration with Prof. Hossam Haick (Technion).

Alert Fatigue in Hospitals

Errors made while prescribing medication constitute the most common type of medication errors in hospitalized patients (7-15%) and result in significant morbidity, mortality and costs. However, many errors are preventable. Electronic prescriptions with computerized physician order entry systems (CPOE) and integrated computerized clinical decision support systems (providing online alerts) could resolve Drug Related Problems (DRP) at the point of care and reduce prescription errors by approximately 50%. CDSS includes drug-allergy, dosing and renal adjustment, duplicate therapy and drug–drug interaction checking.

Clinicians have to deal on a daily basis with a flood of data and information that they need to make decisions upon; mistakes are more likely to happen. CDSS helps clinicians to easily identify areas that need attention. Examples like drug-drug interactions, drug-allergy interactions, high doses and duplicate therapy, or labs that need to be checked before prescribing medication. Possibilities are endless, and here lies the problem.

Due to the fact that there is so much information to present to the clinicians, and while everything is important, nothing is important anymore! Here is where the alert fatigue starts to take place.

The introduction of CDSS is often met by opposition due to the flood of alerts, and most prescribers eventually ignore even crucial alerts (“alert fatigue”).

How many is too many?While in healthcare systems, alert fatigue has been observed for clinicians presented with as little as 12 alerts per day in a busy clinic seeing almost 40 patients a day. That means that 1 alert every 3 patients, was enough to kill the concept of CDSS, cause alert fatigue, and drive the override rate all the way >90% for all types of alerts. Why?

Relevance is the key!

The more we look into alert fatigue cases, the more we realize; it’s mainly around alerts relevance rather than numbers. There are many factors that can lead clinicians to override alerts, but relevance is the most important one.

Once we present clinicians with multiple irrelevant alerts that they have to override without a positive impact on their current plan of care, their perception of the CDSS will shift down. We (as humans) are programmed to filter unimportant information.

Ignoring crucial alerts (“alert fatigue”), reducing the CDSS’s effectiveness. Therefore, optimizing and customization of the CDSS to a hospital department’s specific needs is necessary, reducing the alerts to “actionable alerts” that are likely to result in a change of prescription or follow-up instructions.

Not all alerts are stopping alert!

To fight this phenomenon called “alert fatigue”, we must categorize our alerts according to number of factors. For example; before you accept an alert in the system, ask yourself:

Does the clinician need to take an action NOW?
Should the alert stop the clinician’s current workflow or notifying the clinicians unobtrusively achieve the same goal?
Are there any valid reasons to override the alert? What’s the acceptable override rate?

Going through such process for every type of alert you have will help you achieve your objective with as little disruption as possible to the clinicians, which is a main factor in compliance with your CDSS.

Improving CDSS

CDSS systems provide reports on the number of fired alerts, clinicians’ response to the (Override rate, override reasons, changes), triggers, trends, etc. These reports should drive us to continuously change our design and review our alerts system, and provide feedback to clinicians on their behavior. It’s critical to regularly evaluate these data and look closely for trends like:

Rise or fall of alerts fired (Especially the top 10)
Rise or fall of override rate (%)
Significant changes in the type of fired alerts by customization tools
Clinical outcomes related to these alerts

Clinical Decision Support systems should support your decisions in their design, as much as they support clinicians in providing better and safer care for the patients.

The main purpose our study is to examine the impact of multilevel factors affecting drug prescribing errors in inpatients, and how implementation and customization of CDSS affect the safety of medication treatment, with the purpose of reducing the number of alerts to minimize alert fatigue. Moreover, we aim to examine the mediating role of alert fatigue rate in the effect of CDSS implementation on drug prescribing errors.

TrainPain - a new science based digital product designed to improve mental and emotional well-being for people living with neuropathic and fibromyalgia pain

Trainpain is a unique project aimed at examining the feasibility and effectiveness of a gamified sensory perceptual training programme for patients with fibromyalgia. Using a remote app-based somatosensorial training programme the project will examine changes in pain perception. The role of catastrophising, depression and anxiety will also be examined in relation to pain.

There is no known cure for fibromyalgia and pharmacological pain relief is only reported in a minority of patients. Perceptual training has been shown to reduce the impact of fibromyalgia along with other chronic pain conditions. Specifically, in fibromyalgia temporal discrimination may be altered; however, no study has examined training in this domain as a moderator for changes in pain.

There is a need for tools that fibromyalgia patients can use independently at home to manage their chronic pain and self-care. Thus, Trainpain offers a novel digital platform to enable patients to perform gamified perceptual training at home.