IoT in Healthcare & Pharma : Successful Business Transformation

IoT in Healthcare & Pharma : Successful Business Transformation


IoT healthcare background

With the disruption in digital landscape as we see around today there are already countless applications for the internet of things ( IoT) in healthcare, but the technology is still evolving and has much more to improve and deliver upon both patients and healthcare majors. While one of the challenges of healthcare IoT is how to manage all of the data it collects, the future of IoT will depend on the ability of healthcare organizations to turn that data into meaningful insights.

Why the need for IoT at Healthcare

Before we get started it’s important to understand that the opportunity to build precision things and operate them more precisely is being made possible because of fundamental shifts in the economics of computing. Healthcare has to take advantage of the evolution in technologies and pass on the benefits to patients , while lowering the medical costs and decreasing the operational tasks while automating the predictive analysis and patient admissions, and beyond.

The move of enterprise applications (e.g. financial, healthcare, sales, marketing, purchasing, payroll, human resource management) to the cloud has been driven by an order of magnitude shift in economics. This is because the true cost of enterprise applications is not the purchase price, but instead the cost to manage the security, availability, performance and change in the application and all of the supporting hardware and software. A simple rule of thumb is that the cost to manage an enterprise application is four times the purchase price per year, which means in four years you’ll spend 16 times the purchase price to manage the application. The fundamental cost component to manage the applications is human labor. While finding lower labor rate countries has resulted in some decreases in cost, there is a floor.

Healthcare enterprise applications, cloud services are significantly lower cost because they have standardized their processes and infrastructure, and automated the management of security, availability, performance and change, thereby replacing human labor with computers. This same principle is now being applied to utilize , compute and storage infrastructure, resulting in dramatically lower costs and increased flexibility.

What is IoT framework ?

The first layer is composed of Things. We’ll use the words Thing, machine and equipment interchangeably in this article. Things focus on the machines themselves and are connected to the Internet in many different ways. Once connected, Collect refers to the technologies designed to collect the data, which are increasingly time-series data being sent every hour, minute or second. The fourth layer is Learn.Unlike in the world of IoP applications where we had to entice to type something, IoT applications will get data constantly. For the first time, we can use machines to learn from our Things at the hospital, mine or farm, for example. Finally, you should ask, what’s all this technology for? What are the business outcomes? The Do layer describes both the software application technologies and the business models affected by companies that build Things, as well as those who use them to deliver healthcare, transportation or construction services, for example.

Precision Healthcare :

U.S. News and World Report named UC Irvine Medical one of the nation’s top 50 hospitals for gynecology, cancer, digestive disorders and urology. The medical center has been home to a number of firsts — including Orange County’s first heart transplant, the West Coast’s first insulin pump implant in a patient with diabetes, and a number of research breakthroughs involving therapy for cancer and other diseases.

Under the leadership of Charles Boicey, an informatics solution architect, the medical center has piloted a new technology to frequently monitor and transmit patient vital signs. Patients in the pilot program wore a SensiumVitals patch that monitored and wirelessly transmitted heart rate, respiratory rate and temperature every minute; the objective was to have software that alerted nurses if a patient’s vital signs crossed certain risk thresholds so the patient could be immediately attended to.


SensiumVitals is a wireless system designed to monitor the vital signs of patients in all areas of the hospital setting where patients would not normally be continuously monitored. The product is a light-weight, wearable, wireless, single-use patch that takes vital signs and wirelessly communicates that data to clinicians via the hospital’s IT infrastructure.

The system takes patient measurements every two minutes, which is significantly more often than current manual practices where the majority of patients have vital signs taken only once every 4–8 hours. Infrequent monitoring can mean that deterioration in patient conditions may go unnoticed, potentially leading to longer hospital stays, more expensive treatments or even admittance to intensive care.

This patch provides continuous monitoring of three of the patient’s vital signs — heart rate, respiration rate and temperature — with a high degree of accuracy. The patches transmit data every two minutes, which is about 4,320 data points per patient, per day.

The patch’s processing power is a proprietary Sensium system-on-a-chip (SoC), which is based on the 8051 micro-controller — a complex instruction set computing (CISC) instruction set, single-chip-microcontroller series developed by Intel in 1980 for use in embedded systems. The Sensium implementation runs at 16MHz with a 32kHz sleep clock. The patch is disposable and has no operating system; they don’t need to do software updates often, although they can update the software through the electrode snaps post-production if required. Normally, they just make changes to the next batch of patches when they’re produced.


The patch is connected to the network through a proprietary ZigBee-like protocol. The ZigBee protocol is a IEEE 802.15.4-based specification for a suite of high-level communication protocols used to create personal area networks with small, low-power digital radios. The technology is intended to be simpler and less expensive than Bluetooth or WiFi. Its low-power consumption does limit transmission distances to 10–100 meters. In the hospital environment they use the hospital ISM band of 915MHz.

The devices can transmit data over long distances by passing it through a mesh network of intermediate devices to reach more distant ones. ZigBee is typically used in low-data-rate applications that require long battery life and secure networking; 128-bit symmetric encryption keys secure ZigBee networks. ZigBee has a defined rate of 250Kbit per second and is best suited for intermittent data transmissions from a sensor or input device. One of the other advantages is that you can get both 2-D position (latitude, longitude) and elevation information so you’ll know what floor the device is on. And as a separate radio frequency, there is no interference with WiFi.

As a result of the choice to use ZigBee, Sensium has created a disposable sensor that can be used for five days for a cost of around $50. This saves nurses time, as they don’t have to clean equipment or risk contamination between patients as they would have to with a multi-use device. Its low-power consumption allows for reasonably long battery life, meaning nurses don’t have to charge or replace batteries. This is compared to some WiFi solutions that cost 50 times more and require that the product is cleaned and charged before it’s used with another patient.

Sensium has implemented a bridge to connect the patches to the hospital servers. Unlike Bluetooth, this enables the patch wearer to move around and automatically maintain the connection. The device can also be outside of network coverage for up to three hours and will store vital-sign data on the patch; the data will then upload automatically as soon as the patient comes back into contact with a base station. As a rule of thumb, a typical ward in a hospital requires 10 bridges.


All of the data is collected in a Hortonworks Hadoop environment running within a VMware instance on an on-premises Dell server. Hadoop is an open-source software framework for distributed storage and distributed processing of very large data sets on computer clusters built from commodity hardware. Hadoop is designed with a fundamental assumption that hardware failures are commonplace and thus should be automatically handled in software by the framework.

The core of Hadoop consists of a storage part (Hadoop Distributed File System, also known as HDFS) and a processing part (MapReduce). Hadoop splits files into large blocks and distributes them amongst the nodes in the cluster. To process the data, MapReduce transfers packaged code for nodes to process in parallel based on the data each node needs to process. This approach takes advantage of data locality — nodes manipulating the data that they have on hand — to allow the data to be processed faster and more efficiently than it would be in a more conventional architecture. Hadoop was used because it was a simple solution that allowed data to be processed in its native form. Some people think of Hadoop as only being used for large amounts of data, but in this case, 350 patients with a reading every minute creates less than 100Gbytes of data in three years.

At UC Irvine, the data is also collected in an electronic medical record (EMR) application provided by Sunrise, a division of AllScripts. An EMR or electronic health record (EHR) refers to an application that collects and stores patient health information in a digital format. The EMR application may include a range of data, including demographics, medical history, medication and allergies, immunization status, laboratory test results, radiology images, vital signs, personal statistics like age and weight, and billing information.

UC Irvine used NatHealth’s DeviceConX to integrate all the devices into the EMR. In many hospitals, data captured by devices is manually entered into the electronic record by clinicians. This manually entered data is often hours old by the time it is keyed in, and transcription errors are inevitable. DeviceConX is a software solution that collects and transmits device data captured from thousands of medical devices, delivering that data to an EMR, CIS or other data repository. In this case, data from the devices is stored in the EMR once every 4–8 hours and previous history is purged every 3–5 days.


By using three years of historic sensor data coupled with Code Blue and rapid response events, the UC Irvine team built a predictive model. Code Blue is used to indicate a patient requiring resuscitation or in need of immediate medical attention, most often as the result of respiratory or cardiac arrest. The model also uses other sources of data including EMR and lab data.

Three examples of the 16 types of time-series data include SpO2, which measures the amount of oxygen in the blood; more specifically, it is the percentage of oxygenated hemoglobin compared to the total amount of hemoglobin in the blood; pCO2, which measures the partial pressure of carbon dioxide in arterial blood, is a good indicator of respiratory function and reflects the amount of acid in the blood; and NBP systolic, which is measured in milligrams of mercury (mmHG) and is a non-invasive, blood pressure measurement.

A team from Tata Consulting Services built the first models using both ARIMA models and support vector machines written in R: Auto-regressive integrated moving average (ARIMA) models. These are applied in cases where data shows evidence of non-stationarity, where an initial differentiating step (corresponding to the “integrated” part of the model) can be applied to reduce the non-stationarity. Support vector machines are supervised learning models with associated learning algorithms that analyze data and recognize patterns and are used for classification and regression analysis. A support vector-machine model is a representation of the examples as points in space, mapped so the examples of the separate categories are divided by a clear gap that is as wide as possible. New examples are then mapped into that same space and predicted to belong to a category based on which side of the gap they fall on.


An ability to precisely monitor patients can have significant health benefits. The Hospital for Sick Children, a leading children’s hospital in Toronto, conducted a study of premature, low-birth-weight babies. The study showed about 20% of the children will develop an infection and of those babies, about 18% will die. Physicians theorized that by detecting signals indicating the onset of infection, also called sepsis, they could intervene earlier.

An analysis of data also surprisingly revealed that premature babies with more stable heartbeats were actually more susceptible to infections. Starting 12–24 hours before the onset of noticeable symptoms, the heart rates of infected babies became too regular and stopped varying, as they should in a healthy state.

The results of UC Irvine’s work were an ability to predict a Code Blue with reasonable accuracy within 90 seconds of the event. While that’s a good first step, the goal is to be able to make a prediction ten minutes into the future. Charles Boicey has gone on to ClearSense to continue this work.


As you all would have enjoyed the fascinating benefits and practical utilization of #IoT_in_Healthcare, I would continue my passion for Pharma, Retail, Transport and Auto Manufacturing industries too in upcoming articles sequentially. Appreciate the time spent and do suggest me if there are any more such wonderful successful Business Transformation, Case studies in healthcare which are utilizing the benefits of IoT and Technology evolution further….. Happy reading !






Have Your Say: