Home Commentary Utilization matters

Utilization matters

Practices that add days to stays

June 2019
Getty Images: Sorbetto

HOSPITAL LENGTH OF STAY (LOS) is a common if imperfect barometer of throughput efficiency, cost and quality. Discharge patients too quickly and you run the risk of incomplete care, poor outcomes and readmissions—and patients who feel they were given a bum’s rush out the door.

But chances are your hospital leadership is a lot more focused on high LOS. Inappropriately long LOS reduces bed capacity and erodes profit margins, while slowing patients’ access to services. Plus, excess days across a health system translate into millions of dollars of additional nursing costs required to care for patients lingering on the wards.

After many years as a hospitalist and physician advisor, I’ve witnessed various care practices that add days to stays. In the spirit of the Choosing Wisely initiative and the “Things We Do for No Reason” column in the Journal of Hospital Medicine, here’s my list of the top five ways LOS may be unnecessarily prolonged.

Old habits die hard.

The dubious GI workup: Often a dubious work-up is associated with needless continuous IV fluids causing hemodilution, coupled with daily phlebotomy that results in mild acute blood loss anemia. Initial unrecognized hemoconcentration can contribute as well. A modest drop in hemoglobin/hematocrit (H/H) despite a lack of bleeding stigmata leads to occult blood testing, serial labs (making the anemia worse), a restricted diet, IV Protonix and a GI consult— which may lead to the patient being kept in-house for endoscopy. Try instead to establish what the patient’s euvolemic baseline H/H is, manage volume status closely and don’t succumb to this knee-jerk bundle.

• High-dose IV steroids for COPD exacerbations: Most physicians are aware of the murky evidence surrounding IV steroids for acute COPD exacerbations. But old habits die hard. I continue to see both veteran and new hospitalists and pulmonologists order high doses of steroids for prolonged durations. What’s being lost in this practice is the recognition that associated delirium and frank psychosis may be triggered in elderly and/or cognitively impaired patients. It’s sad to see a previously compensated elder in restraints, on a 1:1 sit or being considered for placement because of iatrogenic harm.

Similarly, severe hyperglycemia in diabetic patients from large doses of Medrol can lead to shortsighted diabetic medication adjustments that impose cost, inconvenience and potential harm. Connect the dots regarding cause and effect, then solve the problem by disconnecting them.

• CHF/AKI on CKD ping-pong: This phenomenon is almost absurdly comical, yet it’s too common to laugh at. Many patients with chronic heart failure have comorbid chronic kidney disease. When a heart failure patient presents with an exacerbation and receives IV diuretics, a bump in creatinine typically follows.

The prudent hospitalist, cardiologist and nephrologist note this and monitor appropriately. But pair a fearful hospitalist with a conservative cardiologist and an aggressive nephrologist, and watch out! A minimal worsening of labs in a patient who can now breath and ambulate comfortably suddenly turns into complicated medication holding and adjustment, a useless renal ultrasound, alarmist discussions about hemodialysis, IV fluids, and occasionally sending the patient back into volume overload.

A prominent cardiologist who is now the CEO of an insurance company once told me, “Nobody ever died of a creatinine of three.” He wasn’t being cavalier. When I recently shared this with a grounded nephrologist, he had to agree.

• Low/no-yield blood cultures: Does this sound familiar? You inherit a patient admitted yesterday with “bilateral lower extremity cellulitis, failed outpatient management.” After review and exam, you determine that the patient has been afebrile, with a normal white blood count, lactic acid, CRP, and vitals, is morbidly obese with chronic lymphedema and stasis dermatitis, and was sent in after three doses of Keflex didn’t improve the chronic but slightly worsened erythema.

As you consider discharge, the lab calls to report that the blood cultures drawn in the ED are showing one out of two bottles positive for gram-positive organisms. But further sensitivity and identification will take another two days because, after a recent merger, your micro-lab operations are no longer in-house and are instead sent across town. You realize the culture is likely a contaminant and shouldn’t have been drawn in the first place. But reflexively you repeat cultures, start vancomycin (if the patient wasn’t already on it), order an echocardiogram and consult ID.

Even if you don’t go this far, many providers will hold onto this patient despite any convincing evidence that the cultures are legitimate. Work with your ED, infectious diseases, primary care and admitting providers to reduce low-value blood cultures. Instead, develop mechanisms to safely monitor clinical status and follow up final culture reports outside of the hospital.

• Inadequate analgesia: I’m not speaking of the chronic pain or narcotic-dependent patient. Many simple and common scenarios beg for improvement.

An elderly demented osteoporotic who’s suffered a mechanical fall with resultant pubic rami or rib fractures, contusions or hematomas, and is now unable to ambulate comes to mind. She lived independently before, but now can’t do her daily activities of living. So what is ordered for pain? PRN IV narcotic alone or perhaps coupled with a PRN oral narcotic.

Realize that discharge will not be on an IV narcotic and that without constant analgesia, effective participation in therapy services is unlikely. Standing acetaminophen with well-timed narcotics (such as just prior to mobility) can minimize suffering and maximize function, potentially averting a rehab stay and expediting transition out of the hospital.

Don’t forget that judicious use of NSAIDs in selected patients can be very effective. Topical medications and scheduled ice packs are other safe, underutilized options that may provide enough additional relief.

I’ve seen all these practices (and others) frequently over my 15-plus years in hospital medicine. If you can avoid these pitfalls and be thoughtful stewards of patient evaluation and management, your patients will receive more effective, efficient care. Plus, LOS will stay in line, earning you the appreciation of your hospital CEO and CFO.

John Krisa, MD, 
is a former regional medical director for a national hospitalist group and currently serves as a physician advisor/hospitalist for St. Peter’s Health Partners, a large integrated health system in Albany, N.Y. You can contact him at johnkrisa@hotmail.com.

Published in the June 2019 issue of Today’s Hospitalist
Notify of
Inline Feedbacks
View all comments