Monday, 21 May 2012 09:42


Written by  Scott R. Davidson

The beauty of computers, well actually the functional benefit of them, is the ability to rapidly analyze data and present it to the user.   This can be an effective tool in managing patient care.  The problem arises when the data presentation is so frequent that it becomes routine or overwhelming.

Alerts and alarms in an acute care facility are a different matter from those in an ambulatory setting.  Medical devices are designed to make care providers aware of critical changes in the monitored subject so that prompt action can be taken.  There are reports that “alarm fatigue” has become the top health IT problem in the last few years.  Efforts are being made to identify ways of prioritizing alarms amid the normal duties of the medical staff to address these problems.

The typical medical practice doesn’t experience a level of device alarms over and above the fire alarm, telephones, or a few non-critical laboratory devices.  But a provider using EMR software on a tablet or desktop computer is faced with a barrage of notices, messages, and warnings.  The hardware and operating environment generate (hopefully infrequent) alerts and alarms about resources, battery life, network connectivity, etc. These issues are generally addressed according to some predefined protocol or over time, and typically (with the exception of true outages) don’t affect the patient encounter.  Messages generated by the EMR are a different story.

During a patient visit, a physician will see messages about drug interactions, allergies, health management alerts, practice level messages, duplicate patients or patients with the same name, required data for meaningful use, refill requests, abnormal lab results, patient requests, and more.  Sometimes the software will confirm an action that you have taken, e.g., saving data, that you already knew.  Physicians are inclined to be “click efficient”; meaning that creating the progress note with the minimum number of clicks is important.  Having to constantly, and sometimes repetitively, respond to these events is not efficient.

One problem is that all events have the same apparent priority.  Most systems support priority levels, typically normal, urgent, or emergent, but they are often not used or misused (everything has top priority).  Creating guidelines for your practice on using these priorities can help to ensure that problems and tasks are addressed in the necessary order.

The second problem is somewhat more difficult to address.   The same drug interaction warnings have a tendency to pop up every time a certain medication is prescribed.  Or lab results for a patient may be outside of the defined normal range, but they are normal for that particular patient.  The problem arises with the tendency to “click through” the warnings. The messages are seen so frequently that they are not even read.  The danger is that something truly important might eventually be missed because the message appears routine, just like the one seen many times before.   Some adjustments make it possible to reduce the frequency, or to change the priority levels, but you can’t turn them off.  The liability is too high and besides, you might be disabling one of your meaningful use measures.

The solution is more intelligent software that is certainly on the horizon.  More reliable systems and improved clinical decision support capabilities will go a long way to remedy some of the problem.  In the meantime, practice policies and workflow solutions are the only viable answers.