Randomized Trial of a Best-Practice Alert in the Electronic Medical Record to Reduce Unnecessary Telemetry Monitoring
Randomized Trial of a Best-Practice Alert in the Electronic Medical Record to Reduce Unnecessary Telemetry Monitoring
This is a randomized trial of a best-practice alert delivered via the electronic medical record (Apex) to physicians on the Medicine service at UCSF Medical Center over a six month period. The alert notifies physicians that their order for continuous cardiac monitoring (telemetry) for a given patient has exceeded the duration recommended by national guidelines and offers them the opportunity to discontinue monitoring if they feel it is clinically appropriate. Physicians will be randomized to either receive the BPA or not, with the anticipation that physicians in the intervention arm will discontinue unnecessary telemetry.
Physicians often order continuous cardiac monitoring (telemetry) for their patients in the hospital. Telemetry provides real-time heart rate and rhythm monitoring and can be useful for detecting arrhythmias and evaluating the response to cardiac therapies. Leads on the patient's chest detect the electrical signature of the heartbeat, relay the information to a box in the patient's shirt pocket, and this box relays the signal in real-time to a control room where technicians monitor for abnormalities. If a technician notes an abnormality he or she will contact the appropriate provider. Currently, physicians can order telemetry at UCSF Medical Center with or without providing an indication for its use. In addition, telemetry remains active until the provider discontinues it or the patient leaves the hospital. Studies at UCSF and hospitals nationwide demonstrate that telemetry is overused. Physicians either order it without a valid indication or they do not remember to discontinue it when the indication is no longer valid. Guidelines published by the American Heart Association provide clarity regarding what clinical conditions warrant telemetry and for how long. However, numerous studies demonstrate that physicians do not faithfully follow these guidelines. This led the American Board of Internal Medicine Foundation to urge physicians in Hospital Medicine to develop protocols governing the use of telemetry outside of the intensive care unit in a 2013 campaign called Choosing Wisely.
The harms of telemetry overuse are myriad: 1) Frequent false or unimportant alarms lead to alarm fatigue and studies have documented patient harm when subsequent real alarms were ignored. False alarms also often necessitate a visit to the patient by the nurse and this distracts the nurse from other patients, 2) Monitoring patients without an active cardiac condition often reveals clinically unimportant abnormalities that obligate physicians to work them up, just by virtue of having seen them on monitor. The work-up then results in unnecessary cost and anxiety, 3) Leads on the patient's chest often fall off with movement and have to be replaced by the nurse. This discourages patients from getting more exercise while they are hospitalized, which is a risk factor for muscle atrophy, and getting enough sleep, which is a risk factor for delirium, 4) Telemetry is not available for all hospital beds at UCSF and even units where it is available have limitations for the types of arrhythmias they can monitor. Thus, telemetry monitoring for a patient who does not need it can prevent another patient from having timely access to a monitor-capable bed.
In order to reduce unnecessary telemetry use, we propose to use a best-practice alert (BPA) delivered to physicians by the electronic medical record (Apex). We propose to study this BPA in a randomized trial. Physicians on the Medicine service would be randomized to either receive the BPA on their patients or not during a six month study period. We would examine both groups for the following outcomes: physician response to the alert, total hours of telemetry used, number of rapid response activations for arrhythmic events, and number of code blue events.
Inclusion Criteria:
Exclusion Criteria: