Software

The Benefits of Observability for Healthcare Organizations

Healthcare IT teams can increase reliability and uptime with observability, which uses AIOps to predict issues.

Tom Stafford is the chief technology officer for CDW Healthcare. He helps to provide guidance to customers around CDW capabilities and service offerings, the creation of new solutions and services customized for healthcare providers, and the creation of new partnerships and rebranded solutions from other verticals. Tom was the vice president and chief information officer at Halifax Health for eight years.

When someone gets in a car accident at midnight or a child has a concerning stomachache early in the morning, hospitals are always there to provide care. That’s why it’s so important that health systems mitigate downtime.

Many healthcare organizations operate 24/7, 365 days a year. They can’t afford to be down due to the potential for financial loss and, more important, the impact on patient care. To support clinical care, healthcare organizations should be able to provide “five nines”: 99.999% availability, or no more than 5 minutes and 15 seconds of downtime a year. This can be challenging to pull off in the healthcare industry, where organizations have large networks and a long list of systems to manage, including the electronic health record (EHR).

Most software companies today are creating reliable products, but healthcare organizations are still left with ensuring the reliability of their own networks and infrastructure to ensure that those software systems don’t go down. That’s where observability comes in.

EXPLORE: Maintain the health and performance of complex applications with observability.

The Importance of Observability Maturity for Healthcare

Monitoring involves assessing all of a healthcare organization’s systems and receiving alerts when something goes wrong. However, monitoring doesn’t necessarily explain what is wrong, which can result in several calls to the service desk and time spent investigating the issue. The other limitation with monitoring is that it informs IT teams only when something has happened, meaning it’s too late to prevent the issue from occurring.

Observability involves using AIOps, or artificial intelligence (AI) for IT operations, to analyze network or system data to predict a failure. This is the highest level of maturity and a key capability when it comes to mitigating downtime in healthcare. According to Amazon Web Services, the observability maturity model includes four steps:

Stage 1 - Foundational Monitoring: Collecting Telemetry Data
Stage 2 - Intermediate Monitoring: Telemetry Analysis and Insights
Stage 3 - Advanced Observability: Correlation and Anomaly Detection
Stage 4 - Proactive Observability: Automatic and Proactive Root Cause Identification

Ultimately, healthcare organizations want their networks and infrastructure to be self-healing so that systems never go down. Observability paired with redundancy can go a long way toward achieving that goal. Think of Amazon: It never goes down and consumers are always able to buy something. They are at Stage 4 of the observability maturity model. So, while they may have issues, they are able to predict them. And in case something does go wrong, they have redundancies in place so they can fail a system over, meaning that consumers never feel the impact of that issue. That’s where we need to get to in healthcare.

Observability will help healthcare organizations get there much quicker than simply relying on redundancy and hoping for five nines.

Click the banner below for expert guidance to help optimize your IT operations.

ht-itoperations-animated-2024-uncover-desktop

ht-itoperations-animated-2024-uncover-mobile

Solving Healthcare Service Challenges with Observability

CDW is currently working with a hospital to implement an observability platform from ServiceNow called Metric Intelligence that uses AIOps to predict issues and automate resolution. We’re partnering with this hospital to help it build out more basic use cases. While it’s important to crawl, walk and then run, these first use cases are incredibly important and needed.

For most hospitals, each unit uses workstations on wheels, or carts with mobile devices that allow clinicians to access the EHR wherever they are caring for patients. However, the problem is that when there’s an issue, clinicians typically don’t call the service desk. They’re more likely to grab a sheet of paper, write “broken” on it, tape it to the cart and push it into the corner because they don’t have time to call the service desk.

Through observability, healthcare IT teams have a configuration management database that knows which devices are attached to each WoW. It can identify if one of those devices has been dormant for a long period of time, such as a 12-hour nursing shift. That lets the IT team know that there’s a good chance it’s in need of attention. The observability platform will automatically create a ticket and assign it to a field tech without anybody having to call in the problem.

The easier we can make it for our customers and their clinicians to do their jobs, the easier it is for them to create a better care environment for their patients.”

This use case creates delight for our customers because if a WoW goes down, a tech will appear and fix it without being called. It feels like magic. This level of insight also prevents emergency situations such as a nurse waiting to call the service desk until several WoWs are unavailable, leading to an impact on clinical workflows and patient care.

Another use case we’re working on involves battery life. With observability, IT teams can tap into charging stations to find out when a battery is decaying and in need of replacement. The hospital can set the system to automatically create a ticket when the battery is down to only 20% of its remaining life. You can also go a step further and have it order a new battery, from CDW for example, through ServiceNow.

Those are basic use cases, but they’re very helpful. The easier we can make it for our customers and their clinicians to do their jobs, the easier it is for them to create a better care environment for their patients.

The Role of Artificial Intelligence in Observability

Pairing AI with observability enables healthcare IT teams to scour a network and see how data is flowing. If it starts bottlenecking somewhere, the team can predict the failure before it happens and either fix the issue or fail over to a redundant system until the issue is resolved.

AI has allowed observability to move to the forefront because we now have software systems, AI engines and machine learning predicting failure. AIOps does the causation and correlation across the board.

This insight makes health systems’ networks and infrastructure more reliable in case of a hardware failure, cyberattack or natural disaster. Observability helps the IT environment self-heal and prevents clinicians and patients from experiencing the impact of a system failure. While that failure may still happen, the IT team is able to at least keep the end users out of the situation by using observability. Clinicians can do their jobs without being hindered by failing IT.

Why Organizations Should Lean on a Partner for Observability

The best way to approach observability in healthcare is by looking to other industries, because it’s not yet happening in hospitals today. Observability has skyrocketed in other industries due to interest in AI.

At CDW, we’re starting to introduce observability into healthcare. We’re working with one hospital to build out a program and solution set that we can then use to help our other customers.

We’ll start with simple use cases, verify that they work and then get into more network self-healing to solve a majority of the hospital’s issues. This will become a future solution that we can provide to our customers through our ServiceNow solutions practice. Observability is a process over time, but once healthcare organizations reach a high maturity level, they’ll be able to take advantage of self-healing and prevent downtime.

If we’re able to help several hospital systems increase their reliability when it comes to uptime, that’s a big win for everybody, and for healthcare too.

This article is part of HealthTech’s MonITor blog series.