Managing Surges: Inpatient Triggers

AdobeStock_30549195.jpeg

Surge plans are created so hospitals and health systems can mobilize to meet the challenge of census surges with responses that are early, organized and complete. In my last post, I discussed how overcrowding in the emergency department is often used as a trigger for surge responses. The ED is “the canary in the coal mine” indicating when the hospital census is reaching critical levels. However, it is also possible to use direct measures of inpatient capacity to trigger surge responses.

In choosing accurate triggers, the aim is to find the right mix of “sensitivity” and “specificity.” Triggers that are high on sensitivity catch most search situations, but also produce false alarms. Triggers that are high on specificity minimize false alarms, at the cost of sometimes not triggering with true surge events.

Traditional census measures, such as midnight and noon census, are blunt instruments. Midnight and noon are not the most critical times for census. More importantly, many hospitals have specialty beds that can only be used for a narrow range of patients, like obstetrics, pediatrics, and epilepsy monitoring. It makes sense to exclude the specialty beds, and have a measure of non-specialty bed utilization, preferably updated hourly.

The combination of non-specialty utilization, known incoming admissions, and anticipated discharges, can be a very effective, accurate, and early trigger for surge action. While utilization percentages in the high 80s and 90s often signal the need for action, it is important to “road test” any new measure to get a feel for which level of utilization translates to log jammed systems, and compromised clinical care.

It is also worth going beyond the overall non-specialty utilization rate, to look at specific levels of care and determine which will become bottlenecked worst and first. Targeting that level of care, for instance, telemetry beds, allows focused efforts to be made by case managers, nursing, hospitalists, and others towards prioritizing discharges and judicious use of beds for that level of care.

While it is possible to do some of this work without dedicated IT systems, the data collected and stored by Electronic Health Records and electronic bed management platforms is invaluable. An advanced operations management platform, combining and modeling this data from across all systems, together with forecasts of incoming admission volume, can produce accurate surge warnings days in advance which can not only maintain or even improve the level of patient care during a surge but also minimize or reduce unnecessary costs and prevent staff dissatisfaction.

Rather than wait for the ED to be bursting at the seams, it makes sense to trigger surge responses directly based on inpatient bed utilization. The key is getting the measure right.