Opioids: pulse ox on patient finger
Pulse ox on patient's finger | Wikipedia

Much of the public attention on the opioid-epidemic has been focused on the harm caused by prescription use and abuse of opioids. However, there is another facet that must be focused on: opioid-induced respiratory depression in clinical settings.

ECRI Institute has repeatedly issued warnings in its annual report on health technology hazards about undetected opioid-induced respiratory depression (OIRD). In ECRI’s 2017 report, ECRI ranked OIRD as the fourth greatest technology threat, saying:

“Patients receiving opioids… are at risk for drug-induced respiratory depression. If not detected, this condition can quickly lead to anoxic brain injury or death. …Drug-induced respiratory depression is of particular concern for patients receiving parenteral and neuraxial opioids in medical-surgical and general care areas. However, it is also of concern for hospital or ambulatory surgery/endoscopy facility patients receiving opioids during procedural sedation and while in the post-anesthesia care unit.”

Increased use of patient-controlled analgesia (PCA) and the higher acuity of patients presenting with chronic disease and comorbidities, such as obesity and sleep apnea, combined with sporadic monitoring protocols, has significantly increased the risk of adverse or fatal opioid-related events.

RELATED STORY: A Nursing Error Led to My Son’s Unexpected Death

To-date, efforts at State and Federal levels and by health system leaders to grapple with the wildfire-like spread of opioids in communities across the country are underwhelming. According to the Association for the Advancement of Medical Instrumentation (AAMI) Foundation, 50 percent of medication deaths are attributable to opioids and each year more than 20,000 patients administered opioids experience respiratory depression arrests—costing the U.S. healthcare system $2 billion each year.

 A Best Practice with Barriers

One potentially effective countermeasure is better monitoring protocols. Current practices are neither adequate nor comprehensive. For example, one of the most common methods—periodic physical spot checks by direct-care clinical staff—can leave patients unmonitored up to 96 percent of the time.

As such, continuous respiratory monitored is broadly recommended as a best practice by The Joint Commission, the Anesthesia Patient Safety Foundation (APSF), AAMI, and other healthcare advocates.

The challenges associated with reducing respiratory events are significant and many, including the financial implications of adding costly monitoring devices and full-time, direct care staff. Further, improvements to monitoring are clearly necessary: according to the Joint Commission’s Sentinel Event database, 29 percent of adverse events are related to improper patient monitoring.

There is also a concern that the implementation of new or additional devices with alarm capabilities will add to the growing problem of alarm fatigue. Many clinicians have expressed a reluctance to use monitoring devices because of the nuisance alarms that their use may entail.

However, as these devices play an essential role in capturing and utilizing real-time patient data for timely interventions, a concerted effort to should be made to mitigate this issue.

Common Monitoring Strategies

Monitoring typically falls into three broad categories, each varying in complexity and comprehensiveness of the surveillance. The first involves transmitting physiologic sensor measurement data and alarm signals sent from multi-parameter monitoring devices to a central station. Patients may not necessarily be continuously monitored in such cases.

The second is sending the preceding data and alarm signals to a telemetry room, which can be quickly and easily overwhelmed by the hundreds of alarm signals that could potentially be generated by a single patient.

The third—and arguably most sophisticated—category is the use of smart alarms—tailored to specifically identify clinically actionable notifications—that are transmitted to a device held by direct-care clinical staff. This approach provides an accurate and real-time picture of a patient’s condition, enabling direct-care patient staff and physicians to intervene before a patient begins to deteriorate. Moreover, attenuating alarm data achieves the balance between communicating contextual patient-safety specific information and minimizing false alarms or events that are not indicative of a threat to patient safety.

Finally, smart alarm strategies allow for not just the analysis of the alarm signals themselves, but also the high-fidelity physiological data associated with them, including time trends, in-depth alarm sensitivity and statistical and predictive analysis.

Alarm Reduction and Continuous Monitoring

Nine in 10 hospitals indicated they would increase their use of patient monitoring, particularly capnography and pulse oximetry if false alarms could be reduced. However, the number of alarm-enabled medical devices on the market today, narrow alarm limits, and inaccurate default settings can make alarm management a complex endeavor.

Several techniques and strategies exist for reducing alarms, including trending alarms, which expand or contract patient alarm limits on individual devices; consecutive alarms, in which patterns of a consistent alarm detected, occurring over a clinician-defined period of time; sustained alarms, which requires setting a minimum time threshold that an alarm limit must be violated prior to sounding the alarm; and combination alarms, in which multiple parameters from different devices occurring simultaneously may together indicate a degraded patient condition.

More particularly, how alarms may be configured may significantly impact the number of nuisance alarms received and whether alarms may be missed. Recently, ECRI Institute issued its Top 10 Health Technology Hazards for 2018. One of ECRI’s 2018 concerns is how missed alarms may result from inappropriately configured secondary notifications devices and systems.

In a recent clinical education podcast, “Improving Patient Safety and Reducing Alarm Fatigue,” Marc Schlessinger, RRT, MBA, FACHE (Senior Associate, ECRI Institute’s Applied Solutions Group) discussed ECRI’s concern, saying:

“The true value comes when the devices are used properly. When we do an integration for middleware and secondary alerting devices at a hospital, the most common request is that pretty much every alarm goes to the device initially, which is totally the wrong approach. So, what we try to do is convince the nurses and respiratory therapists that the only alarms they really want to see and hear on their personal devices are critical alarms and actionable alarms. However, again, you don’t want to just take the ventilator or the physiologic monitor, and make every alarm go to your device, because, at that point, you’re doubling your alarm fatigue, because not only is the device alarming and the central station alarming, you then have the alarm on the person.”

Appropriately configured secondary notifications devices and systems rest on algorithms. As Maria Cvach, DNP, RN, FAAN (Director of Policy Management and Integration, Johns Hopkins Health System) explained about the development of algorithms that has enabled Johns Hopkins to better manage alarms:

“It is important when you’re using middleware to develop algorithms and those algorithms need to be developed at the level of the alarm. So, for instance, you can have an algorithm for a ventilator alarm in which it goes to the respiratory therapist first, then it goes to the nurse, then it goes to the secondary nurse, and then it goes to the charge nurse. But, it’s really important that you have those algorithms built at the level of the alarms, because, like Mark said, a lead fails, you may not want that to go immediately, because you know somebody is moving around or they’re just readjusting something, so you don’t want that to go immediately, but certainly you don’t want to wait 30 minutes or have your algorithm be 30 minutes apart, you have to be rational on how you develop your algorithm. And, that’s the beauty of the middleware, is you can write all these rules within the middleware. I think that we have at least 10 to 20 algorithms for our devices and how they work. And, one of the things that we have done is in case someone has failed to put their escalation system in – and, of course, that’s subject to change as your staff change. So, we have it set that if somebody fails to put in the primary or the secondary person, it’s still going to go automatically to the charge nurse. The charge nurse is automatically going to get it. So, if the charge nurse starts getting all these alarms, they know that somebody hasn’t been put in as the primary or the secondary, and they know to go and figure that out.”

Data Delivery, Communication, and Integrity

 Certainly, physiologic devices are critical components in continuous patient monitoring and capture a more complete and real-time picture of a patient’s condition. Capnography, along with continuous pulse oximetry monitoring, could provide a sensitive and early predictor of opioid-induced respiratory depression. Capnography is used to measure exhaled end-tidal carbon dioxide (EtCO2) and inhaled carbon dioxide (FiCO2) to determine a patient’s respiratory rate and generate waveforms (i.e., capnograms) of exhaled carbon dioxide over time. (See sidebar).

Middleware Requirements

Just as critical is implementing a device-agnostic middleware platform for interfacing with bedside devices. Middleware can be leveraged to pull data from medical devices and combine it with other data in the patient record to create a more holistic and complete picture of the current patient state. Combining analysis with real-time data at the point of collection creates a powerful tool for prediction and decision support. The ability to track patients throughout the hospital, continuously add new devices, and distribute real-time patient monitoring to centralized dashboards and mobile devices should be a major consideration in technology selection.

Minimally, middleware needs to be able to retrieve episodic data from a medical device and translate it into a standard format. Additionally, middleware should be able to retrieve data at variable speeds to meet the requirements of various clinical operational settings (e.g.: operating rooms vs ICUs versus medical-surgical units). Because data will be used for real-time intervention, any delay in their delivery to the correct individuals can have deleterious effects. As such, it is vitally important to understand the implications of requirements on data delivery latency, response, and integrity.

These include FDA requirements, such as those required for Class II clearance. For a middleware vendor to claim clearance for active patient monitoring, they must have all the checks and balances in place to ensure the receipt and delivery of all active patient data for intervention purposes from end to end—from collection point (medical device) to the delivery point (the clinician). Again, the ability to deliver on the timing and receipt of data necessary for interventions and active patient monitoring is an important distinction.

Collaborating on Continuous Monitoring

Many of the doomsday scenarios associated with failed technology adoption and implementation can be mitigated with adequate planning, training, and collaboration. By listening to, engaging with, and educating direct-care staff, hospitals can dramatically increase their chances of success with continuous monitoring.

A great example of a hospital monitoring its patients while managing alarms is Virtua Memorial Hospital. Leah Baron, MD, who is Chief of The Department of Anesthesiology at Virtua Memorial Hospital, in a clinical education podcast spoke about the experience of Virtua Memorial Hospital in improving patient safety and reducing alarm fatigue:

“ … when we first introduced capnography to monitor patients for respiratory depression related to their opioid therapy, we very quickly found out that the number of alarms that were bombarding our healthcare workers was unmanageable. And, the reality was that they could not respond to all of them, but a lot of them were just pure noise. And, that’s why we realized that if we want to use this effectively, we needed to figure out how to identify these actionable alarms and filter the noise, and that’s why, subsequently, we decided that we’re going to do another study and see if we can achieve better results with that.”              

By connecting capnography to middleware, Virtua Memorial Hospital has been able to distinguish between actionable and non-actionable alarms and help them to escalate actionable alarms when they occurred. As Dr. Baron describes:

“So, we selected patients with what we thought have a higher chance of having this respiratory depression – patients with significant serious sleep apnea undergoing major surgery. And, we created these algorithms that we wanted to test on our med-surg floors and see if they meet our expectations. So, that was our goal in this study. And, we actually were able to significantly reduce our alarms, without having a single patient event that went unnoticed.”

How will this new technology impact how nurses deliver patient care? What adjustments in workflow and practice need to be made—at go-live and beyond? Starting with these questions can foster buy-in from the staff that will be utilizing this equipment. If end-users are not involved in the selection, adoption, and implementation of a technology, then the likelihood that they will be an enthusiastic user of that product is significantly lower.

For example, nursing staff are charged with the proper setting of the alarms and the prompt response when any of the devices send an alert. As the presence of alarming equipment continues to grow, nurses find that their workflow and ability to engage with patients is disrupted as they chase down hundreds of (often non-actionable) alarms. Without proper education and implementation of alarming devices, it’s all too easy to imagine clinical staff arbitrarily adjusting alarm settings—or even turning them off entirely.

Thus, an expert project team should be formed, ideally comprised of leaders from myriad stakeholders; from IT networking and facilities, to informatics nurses and direct-care clinical staff. This team will be responsible for every phase of deployment, including goals identification, vendor evaluations, business and clinical requirements, and progress assessments.

In describing the experience of John Hopkins Hospital, Ms. Cvach emphasized that putting together an interdisciplinary team is a critical first step to better alarm management:

First of all, if you’re going to get started at your hospital, you need to convene an interdisciplinary alarm management committee. I emphasize the word “interdisciplinary,” because a lot of times when I’m asked to consult, I’ve noticed that they’re asking one department to fix the problem, like nursing or engineering. It really is an interdisciplinary approach and you need to include people from various disciplines – respiratory therapy, nursing, physicians and include the bedside user, because they’re the ones who can really give you what the problems are and tell you whether or not you’re possible solutions are going to work. So, convening an interdisciplinary committee is step one. 

Designating an executive stakeholder, nursing champion—or empowered super-user—at the outset is also recommended.

Finally, avoid minimizing clinical workflow, as this largely defines how data is collected, how it is displayed, and what is displayed. Hospitals should incorporate clinical workflow as quickly and as early as possible in the process.


Both patient-managed and staff-administered pain medication are necessary for the patient’s well-being in the hospital. However, their use presents the real risk of overdosing and death, especially for patients with complex chronic conditions and comorbidities. Continuous monitoring can help improve patient safety while keeping the patient comfortable, but the careful implementation is necessary to avoid a negative impact on the staff and environment the patient depends on for care.

In addition, improved inpatient care contributes to better post-discharge outcomes and can lead to fewer readmissions.

Given the rising incidences of in-hospital opioid-related adverse events and deaths, hospitals should commit, soon, to a strategy for ending them.

Michael Wong, JD
Michael Wong, JD is the founder and Executive Director of the Physician-Patient Alliance for Health and Safety (PPAHS). He has been at the forefront in driving practical solutions that reduce healthcare costs, decrease medical errors, and improve patient health outcomes. He has been particularly active in these areas that most affect patient safety: • Improving patient adherence (i.e. helping patients to take their medications as prescribed by their physicians) • Enhancing patient access to healthcare • Reducing medical errors (PPAHS), is an advocacy group of physicians, patient advocates, and healthcare organizations. Supporters of and commenters for PPAHS include highly respected physicians and healthcare organizations, including the The Joint Commission, Anesthesia Patient Safety Foundation, Anesthesia Quality Institute, Johns Hopkins School of Medicine, Harvard Medical School, Stanford University School of Medicine, and the Cleveland Clinic.



Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.