PROPOSAL TITLE:
Assessment of Efficacy and Safety of a Standardized Continuous Pulse Oximetry EHR Order
PROJECT LEAD(S):
Nirav Bhakta
Aida Venado Estrada
EXECUTIVE SPONSOR(S):
Adrienne Greene
Joshua Adler
ABSTRACT - One paragraph summary of your proposed initiative – Limit 1500 characters (with spaces)
Continuous pulse oximetry (CPO) allows for real-time monitoring of oxygen saturation and heart rate 24 hours a day in hospitalized patients. For patients at risk of respiratory decompensation, CPO can lead to early intervention and prevention of medial events. However, many technical and biological factors contribute to noisy data. This noisy data creates false alarms, leads to alarm fatigue, and distracts attention from real desaturation events. In the absence of professional society guidelines, clinicians currently order and discontinue CPO at their discretion. Appropriate CPO utilization, reducing over- and under-ordering, can improve safety and reduce cost.
Our aim is to implement a CPO order in the EHR with standardized durations based on specific indications and developed by expert consensus from several specialties at UCSF. We will track efficacy and safety outcomes every month—number of CPO hours per patient per admission and number of escalation events (rapid response and ICU transfers), respectively.
Our multidisciplinary team, the CPO clinical care model group, started working on this learning healthcare system initiative in August 2022. To understand the problem, we reviewed the literature and interviewed clinicians and frontline providers. To understand the current state of efficacy and safety, we partnered with the DOM Data Core to extract data from the EHR. We will use this information to set targets for our intervention.
TEAM - Core implementation team members and titles
Aida Venado Estrada, MD, Assistant Professor of Medicine, co-physician lead of CPO clinical care model group
Nirav Bhakta, Associate Professor of Medicine, co-physician lead of CPO clinical care model group
Adriane Crouse, Assistant Unit Director | 8 South & 12 South TCU, co-lead of CPO clinical care model group
Madeline Chicas, Quality Analyst, Adult Quality Improvement
Andrea Ratti, Director, UCSF Health - Physician Services
Logan Pierce, Assistant Professor of Medicine, Hospitalist & Physician-Informaticist at UCSF Medical Center
Nader Najafi, Associate Professor of Medicine, directs the Division of Hospital Medicine’s Data Core
PROBLEM - Background of the problem. What is the cost associated with this problem? Why address this problem now? What is the current condition?
In response to delayed recognition of respiratory decompensation in patients on continuous pulse oximetry (CPO), CMO Dr. Adrienne Greene created a clinical care model group to develop guidelines for CPO use at UC Health. A goal of this guidance is to reduce both over- and under-ordering of CPO. CPO monitoring for low-risk patients increases alarm burden and the risk of missed signals in patients at risk. CPO monitoring stewardship (appropriate utilization) can contribute to prevention of patient safety events. CPO overuse also has the potential to contribute to delirium and fall risk as well as increase monetary costs to the hospital system.
Our interview of 15 clinicians at UCSF revealed that 100% use CPO to monitor for desaturation, 50% use CPO as a surrogate for a higher level of care, 64% agreed CPO is overordered, >70% felt CPO is not discontinued appropriately and left longer than needed, >50% favor developing an Apex order set to guide ordering and discontinuation of CPO, and 36% refer to nursing as guidance of when to discontinue CPO.
Our interview of 11 frontline staff at UCSF revealed that the burden of alarms is unsafe and there is alarm fatigue. Guidance for providers on when to order and discontinue CPO is needed. CPO is used by providers as a surrogate for “someone is watching the patient”. Formal training for RNs and PCAs on proper probe placement and troubleshooting would be beneficial.
The guidelines will be employed to develop APeX orders that will be required to start CPO. We are partnering with clinicians here and ZSFG, who already have a CPO orders (in part based on the currently used Parnassus telemetry orders), to create new CPO orders that we aim to be the same in both hospitals. This project will reduce variability in practice for improvement of patient outcomes and reduction of cost.
TARGET - What is the goal? What are the expected benefits, both qualitative and quantitative?
We will assess efficacy as hours on CPO per patient per admission, and safety as number of escalation events (rapid response and ICU transfers). Our goal for FY2024 is to reduce number of hours on CPO per patient per admission by 30% compared to FY2023, keeping escalation events equal or lower than in FY2023.
As a qualitative benefit of reducing CPO hours per patient per admission we expect a reduction in alarm burden (e.g., false alarms). Another benefit expected is reduction of cost to the hospital system, which we will try to quantify.
GAPS - Why does the problem exist? Describe system issues; technological gaps; educational gaps
The ability to order CPO increased dramatically in the Moffitt-Long hospital with the centralization of telemetry and CPO through wireless technology. This change was not paired with guidance on which patients can benefit from CPO. As trainees advance and graduate, and as new physicians and APPs are hired, non-systematic efforts to improve use are not sustainable. Alarms are evaluated by telemetry techs in a centralized room on the 6th floor of the Moffitt-Long hospital. Alarms meeting certain criteria are shared with the nursing team on the floor with the patient via Voalte and through telephone calls. The primary clinical teams do not receive these alarms and are therefore do not have a direct incentive to use CPO more wisely.
Furthermore, there are no professional society or institute guidelines for the use of inpatient CPO. In contrast, guidelines from the American College Cardiology are available for telemetry. The clinical care model guidelines and APeX orders will fill these educational gaps.
(CPO remains decentralized at Mt. Zion and Mission Bay but the educational gaps filled by this project remain the same.)
INTERVENTION - Describe your proposed intervention and rationale for approach. Describe your practice setting and target population (e.g. department, unit, clinic, patient characteristics, diagnosis group, procedural group, provider characteristics, staff characteristics, etc.). Describe potential barriers to implementation. What are the possible adverse outcomes that may occur that may affect quality of care and patient safety as a result of your proposed intervention?
Our target patient population includes all hospitalized patients in UC Health.
Potential barriers to implementation:
Possible adverse outcomes: Respiratory decompensation events may occur in patients that do not meet the guidelines’ standard indications for CPO.
We will engage the Data Core to extract data for both efficacy and safety outcomes at Parnassuss. Prior work on telemetry at UCSF, and for which existing computer code and expertise is available for modification, increases the feasibility of the proposed project. We will continue work on the implementation of the new CPO orders with a potential discontinuation alert.
Given that CPO orders in Epic similar to what is being proposed have already been in use at ZSFGH for a number of years, this project has a low risk of failure. Harmonization of the orders at UC Health and ZSFGH will demonstrate scalability.
PROPOSED EHR MODIFICATIONS Note: EHR modifications are NOT required for a winning proposal
What are the clinical problems you are hoping to solve with APeX?
Currently, CPO is ordered without an indication or duration. This leads to instances of CPO being in place right up until the time of discharge, which is just one example of CPO overuse. The new APeX orders will require an indication and a duration.
What APeX tools (patient lists, reports) or workflows (orders, documentation, alerts) are you using now to achieve this goal? How would you want these modified?
What new APeX tools/workflows do you think you need to achieve the goals of your project?
COST - Estimated baseline costs to the health system and projected savings from the proposed project
SUSTAINABILITY - If successful, how will this intervention be sustained beyond the funding year? Who are the key UCSF process owners?
The work of the CPO clinical care model group is already underway and part of a larger effort to improve CPO at UC Health that is supported by the CMO (Adrienne Greene), CCO (Joshua Adler), and Dean (Talmadge King). Therefore, there is executive and clinical leadership support to keep the project going.
BUDGET - Line-item budget up to $50,000 - Briefly identify key areas of the project that will require funding, e.g., salaries, software, printing, etc
Funding for Data Core $12,000 (one time cost)
Drs. Bhakta and Venado Estrada already have salary support through the medical center for this project, which increases the feasability and reduces the requested budget.
