Part 1: Length of Stay

The other day I heard that even with increased taxes and program cutbacks, within a few years the rising costs of medical care would overcome those efforts to keep the US financially healthy. While we might not have “signed up” to solve the country’s financial problems, you can bet that those who did sign up to solve them will be putting huge pressure on the health care industry to control costs, while maintaining care quality.

Hospitals costs make up about 31% of total healthcare costs (that’s counting dental, nursing home, drugs, clinics, etc.), by far the largest sector, growing at 8.6%/year (1). Moreover, hospital spending is about 41% of total publicspending for healthcare.

Inpatient costs break down percentage-wise as follows (2):

Nursing                 50%
OR                        10%
Lab                         8%
Radiology               4%
Supplies                 8%
Pharmacy               8%
Everything else    12%

With this in mind and from a hospital management prospective you might ask which parts of inpatient costs can be managed for lower cost at the same care quality?

Two ways to think about this:
1) Which parts are affected by length of stay? Or, if you could reduce LOS without lowering quality, what gets saved? Clearly nursing, supplies, pharmacy, and some of “everything else”. The OR, lab and radiology are more tied to “per admission”, and what is ordered for the patient.
2) Which parts are affected by productivity considerations? Clearly Nursing and OR, and to some extent Radiology and Lab.

Two very promising areas for management, then, are ways to reduce length of stay and improve nursing productivity without compromising care quality and safety.

Length of Stay (LOS)
When looked at as its effect on costs, LOS must be divided into how long a patient stays on each “type” of unit. For simplicity, let’s say there are four types of units: the ER, Intensive Care, Intermediate care, and the Floor. Cost saving are available by making sure that each patient stays no longer than necessary on each of these four, as they have significantly different costs per hour. And, for some patients, staying an extra unneeded 6 or 10 hours on the ICU doesn’t translate to either shorter overall LOS or better quality. Same for intermediate units.

So then, for LOS, the objective should be to have each patient stay on each type of unit only as long as is necessary for quality and safety, which should also decrease overall LOS (but maybe not as much as the sum of savings, depending on what part “nature” is playing in recovery).

But what’s the “correct” amount of hours on ICU for patient Jones? Or, is there some way an evidenced-based expectation for how long patient Jones “should” spend on the ICU could be established as a guideline, and as a measure against which patient Jones could be compared while he’s there on the ICU?

Here’s how our Patient Progress Manager approaches this. For a particular DRG — say the most often admitted to your hospital — a report is generated showing how long each of the last (say) 100 discharges stayed in the ICU, sorted in shortest to longest stay. Say they range from 4 hours to 72 hours in the ICU. Now a small group of clinicians who are familiar with these patients establishes a “target” LOS on the ICU for this DRG (say “16 hours”), where they agree that, after this amount of stay, it’s legitimate to ask “is this patient about ready to be transferred”.  Of course, in many cases, he won’t be ready, but if there’s a way to have the question asked the right way at the right time, some cases would indeed be able to transfer earlier than if the question were not asked.

Upon admission the Patient Progress Manager captures a working DRG for each patient, then tracks each patient’s stay on each type of unit, using links by HL-7 to the hospital’s ADT system. When a patient is approaching or has passed the “target” LOS on a unit for his DRG, and alert is automatically generated for caregivers and/or case managers. These “alerts” are a list (or screen) — always available and up to date — of all patients on the unit approaching their target, plus those past their target, to be reviewed. For those patients where it’s clear that they are going to stay past their “target”, a new target departure is set based on the current condition of the patient so that he can continue to be monitored against this updated target.

If a new working DRG is set for the patient, his targets are automatically updated  (unless his target has been manually changed after an assessment of the patient on that unit).

The same technology is applied to each patient’s stay on other units, and separately for his LOS overall.

Our users call this “managing to departure” for each patient, whether it be the next transfer or discharge.  A small gain in time here can mean big cost savings, when applied to all patients who could have left a little earlier.

See the description of the Patient Progress Manager for more detail.  

In a later blog — “Part 2″ — we’ll talk about managing costs and quality looking at productivity.

Love to hear your comments and  questions.

1) National Health Care Expenditure Data, CMS, January 2010.
2) Kane, Nancy and Siegrist, Richard, “Understanding Rising Hospital Inpatient Costs” , on www.selectqualitycare.com, 2002