In this special guest feature, Guy Meger, CTO, VP R&D of EarlySense, (a leader in contact-free, continuous monitoring solutions for the healthcare continuum) discusses how with automated compliance monitoring in place, healthcare staff can focus their expertise and time on healing, while patients can rest and recuperate without any anxiety-induced disturbances. For these reasons, many global hospitals and medical facilities are turning to automated compliance technology to support big data systems that operate at their full potential and reach their clinical and economic goals. Guy received his B.SC in Computer Engineering from the Technion Israel Institute of Technology and his MBA from Tel Aviv University. He is also the author of several issued and pending patents.
The medical sector, specifically general care wards which make up 90% of the available beds in a given hospital, has encountered an unexpected challenge. In an era where nearly all of our digital messages are carefully recorded and analyzed, the medical sector is experiencing a drought of continuous patient data documentation. We are all aware of the value of data, and how companies like Facebook and Google use their data to further their enterprise goals. While this is crucial for business growth and strategy, reliable medical data answers to a higher calling, as it serves as the foundation of today’s and tomorrow’s powerful medical technologies, helping doctors and nurses save lives.
It’s surprising that in 2018, at the peak of technological advancement, the vast majority of global health teams continue to use antiquated “spot checking” methods to monitor general care patients. This approach is highly outdated and problematic for three reasons. First, hospital general ward staff only spot check their patients once every 6-8 hours, looking at key vitals and patient-specific issues. This is dangerous because it creates a “data desert”, where clinicians and health staff are not being fed the latest patient data, impacting their decision-making process. This is significant as patient conditions can oscillate in-between those manual checks. Second, nurses spontaneously entering hospital rooms to check on patients contributes to “White Coat Syndrome,” a disorder in which patient readings can go awry due to psychological parameters changing when a patient is aware that they are being measured. This is manifested, for example, when checking blood pressure levels, as patients exhibit levels above the normal range, as well as a higher heart rate during these spot checks. This also apparent with incongruous respiratory rate measurements. Third, infrequent spot checking leads to a lack of reliable data points. Without consistent and dependable data points, true big data generation is unattainable, preventing any form of data analytics.
Big Data Analytics Requires Quality Input
Big data analytics, in all fields, is characterized by three factors: volume, velocity and variety. Volume meaning that the amount of gathered data is too large to be easily managed, velocity being that the gathering of the data is speedy, and variety implying that the range and type of data are various. These criteria are highly dependent on the data that they are compiled from. When quality data is fed into an analytics system, algorithms run through the data and produce results that are reliable, accurate and ready to be acted upon. These results can then be used to further organization goals or signal key staff of noteworthy trends in their respective field. However, if the original data is poor or infrequent, then the resulting analytics will reflect that data and not be worthy of use.
Complications with Patient Data
Within the medical field, and specifically within hospitals that spot check their patients, patient data is only registered by health staff once the data is plugged into the EHR. This gathering of input is meant to be objective, but the reality is that spot checking varies across the board depending on measurement tools/equipment and measurement methods. Being that spot checking is not a very consistent process, the results of each check are highly dependent on staff mood, training, skills, and approach towards patient conditions. For example, one nurse may track the respiratory rate of a specific patient at 19 breaths per minute while a different nurse may track the same patient at 32 breaths per minute, due to different measurement approach (rate estimation/patient awareness to the measurement etc.) Spot checking can create “noisy input”, meaning that the entered data is influenced by multiple factors, making it “impure”. These influences include the aforementioned factors, as well as infrequent data entry, entry that is impacted by staff mood and frame of mind (estimating patient numbers based on patient health rather than performing time consuming actual patient evaluations), staff who are not tech savvy or staff who do not fully embrace the input system, and breaking data entry routine. Noisy input leads to unreliable results that true systems cannot generate reliable analytics from.
Even hospitals that consistently spot check their patients and enter data correctly are not able to generate enough patient data for analytics systems. A patient who is admitted for three days will on average only generate 12-14 data points until his discharge at an average spot checking hospital. This is nowhere near the thousands of data points required to generate the volume necessary for true big data accumulation.
A Solution in Automated Compliance Data
Within healthcare, automated compliance technology, commonly referred to as non-compliance technology, is essential for creating quality data for patient analytics. Automatically gathering data without input or action from patient or staff, automated compliance tracking creates a steady stream of patient readings. The key to generating successful data collected for analytics also depends on the algorithms that gather these patient readings and filter out the ambiguous data. Only then can the reliable and accurate patient readings be digested by big data systems. In this way, each patient generates vast amounts of quality labeled data (the measurements are taken at the same times every day in optimal conditions), and the big data that develops from the many varieties of patients, dependent on their diagnosis, and users is very rich data for algorithm systems to operate on. It is this kind of data that leads to breakthrough analytics.
Additionally, the automated compliance approach helps eradicate “White Coat Syndrome” amongst patients, reducing fluctuations, errors and bias, and generating true data. By tracking key vitals like heart rate, respiratory rate and movement, automated reliable data generation can provide more than 100 data readings per minute, giving health staff a truly complete picture of patient health. With this data, trends in patient status as well as baseline changes can be easily tracked, analyzed and detected, helping to create new opportunities for effective intervention.
There are several health focused companies that have already begun implementing automated data generation monitoring technologies in an effort to create an oasis in the aforementioned medical data desert. Their efforts have proven fruitful, as they have succeeded in sparking early intervention for patients in need, reducing code blue events by 86% [1] (strongly correlated to cardiac arrest reduction), pressure ulcers by 64% [2], ICU days2 and patient length of stay2.
With automated compliance monitoring in place, staff can focus their expertise and time on healing, while patients can rest and recuperate without any anxiety-induced disturbances. For these reasons, many global hospitals and medical facilities are turning to automated compliance technology to support big data systems that operate at their full potential and reach their clinical and economic goals.
[1] Zimlichman, E et al, ATS. 2012
[2] Slight, SP et al, Society of General Internal Medicine. April 2013
Sign up for the free insideAI News newsletter.
That’s an insightful read! Automated Compliance Technology can actually save healthcare institutions from data disparity due to manual checks and provide reliable analytics in healing patients.