Hey, Do You Remember That Patient…? Bouncebacks in the Emergency Department

Authors: Karan Gadhok, MD (EM Resident Physician, New York Methodist Hospital) and Hilary Fairbrother, MD (EM Attending Physician, NYU School of Medicine) // Edited by: Alex Koyfman, MD (EM Attending Physician, UT Southwestern Medical Center / Parkland Memorial Hospital, @EMHighAK) and Brit Long, MD (@long_brit, EM Chief Resident at SAUSHEC, USAF)

 

Introduction

Emergency Department (ED) bouncebacks are a serious concern in patient safety. A bounceback is when a discharged patient returns to the ED for care within a short period of time. The typical Emergency Medicine (EM) physician working 1,728 hours per year for 30 years seeing 2.5 patients per hour will see about 130,000 patients in their career. An EM physician will send home about 44 patients in their career who will die within 7 days of ED discharge.1 Bouncebacks can create terrible outcomes for the patients and their family and can also emotionally devastate the physicians and nurses that took care of that patient. With the ever-increasing number of patients being seen in the ED, it is important to know how to safely discharge a patient to minimize the risk of unexpected bounceback.
Case

“Hey, remember that patient who you saw 2 days ago with chest pain?” your colleague asks you right before Monday’s morning report. You remember the shift: it was a crazy Friday night… the ED was a zoo, you saw an unbelievable number of patients, and you instantly get this horrible feeling in your gut.

“Which patient?” You ask glumly.

“You know, that 38 y/o female with chest pain and mild shortness of breath? She was diagnosed with anxiety and discharged. She came back at 5 am this morning in respiratory extremis and was diagnosed with a huge saddle pulmonary embolus on CT. She coded and died on her way up to the ICU.”

This is your nightmare. This is the ED bounceback patient who all of us want to avoid. If you see enough patients, this may happen to you.

Research

Most studies about bouncebacks are inherently limited, as they are retrospective and are better suited for identifying possible risk factors rather than showing a causal relationship. Studies in the US are also limited because of the fragmented health care system that fails to capture all patients which bounceback to another facility, health system, or out of state. There are strengths and weaknesses to these studies from a methodological standpoint, so the results should not be taken as conclusive evidence.

Mortality

One study looked at 387,334 visits at the University of New Mexico Health Sciences Center over a ten-year period and found that 132 patients died within 7 days of discharge leading to a case rate of 34.1 deaths per 100,000 discharges.1 Seventeen of these patients were expected deaths. Twenty-nine of the 35 deaths identified as medical error had abnormal vital signs on discharge. Decompensated chronic disease was also present in 21 of these 35 patients. Finally, tachycardia was present in 48 of 55 patients that had unexpected death. Evidence such as this seems compelling, resulting in physicians and administrators making “rules” that persistently tachycardic patients cannot be discharged. Tachycardia is a good example of a “sensitive” vital sign but of course it is also NOT specific (i.e. most patients discharged with tachycardia do not bounceback or die).

Another study looked at death within 8 days of discharge from the ED based on medical examiner (ME) cases.2  The number of ME cases studied was 2,665, with 42 of them meeting the criteria. Nine of these deaths were considered unexpected and directly related to the index ED visit. Three of nine of these cases were due to a ruptured thoracoabdominal aneurysm. This study is limited primarily because only medical examiner cases were included.

Are Bounceback Visits a Good Indicator of Quality?

There is some data that supports the concept that ED bouncebacks within 72 hours do not produce significantly different outcomes when compared to patients who present for their first visit.3  A retrospective study was performed with n = 218,179 comparing 72-hour revisits and all other patients seen in the ED. The study showed similar admission rates (13% vs 13%) between the groups, though at initial visit bouncebacks were less likely to be level 1 triage (17% vs 20%) and used fewer resources. This study also identified that patients with Medicare insurance and no health insurance were more likely to bounceback. They suggest that bouncebacks with subsequent admission are a better indicator of quality and safety in comparison to all revisits within 72-hours of discharge.

 Risk factors: Medicare & Medicaid insurance, psych patient, patients who leave AMA or LWBS, chronic renal & cardiac disease

Several studies have looked at identifying risk factors associated with short-term bounceback visits. One study looked at a retrospective cohort of 5,035,833 visits at 288 facilities in California with 130,526 bouncebacks (2.6% visits) to identify these associations.4  The results showed that bouncebacks were higher with those with Medicaid (OR 1.42) and Medicare (OR 1.53) insurance. In addition patients that left without being seen (LWBS) by a provider or against medical advice (AMA) had higher odds of bounceback (OR 1.9). Specific high-risk diagnoses included: chronic renal failure without end stage renal disease (ESRD), ESRD, and congestive heart failure (CHF) with odds ratios (OR) of 3.3, 2.9, and 2.5 respectively.

Another study evaluated factors associated with bouncebacks within two days, n = 17,214 patients, at an academic medical center. This study showed 17% of return visits were in patients who left AMA or LWBS. These patients also had two times the admission rate when compared to the rest of the population. Yet another study looking at all risk factors for bounceback admission within 72 hours of discharge included 104,584 patients with 493 requiring admission. This study showed patients with Medicare, elderly patients, and those with mental disorders were at highest risk for return visit.7

In a final study, n = 1,422, bouncebacks within 3 days were studied, and dehydration was identified in 25% of patients requiring admission on the second visit.5  Furthermore, the study included surveys of 50 ED health professionals who did not perceive dehydration as a major reason for a bounceback visit. Septicemia and pneumonia were the second and third most common reasons for bounceback visit.

 Access to Primary care

A significant amount of bounceback visits occur because of poor follow up with a doctor outside the ED. One study looked to see if patients who have a PCP were less likely to bounceback.6  This prospective study of collected survey data from 1084 patients. Patients who returned were more likely to have Medicaid or no insurance. Patients who bounceback were also less likely to have a PCP but the difference may not be clinically significant (86.2% vs 79.19%, p = 0.0204). In addition, the hospital where this study was conducted was in a suburban area with better access to a PCP when compared to inner city or rural hospitals. The authors conclude that access to primary care alone is insufficient to decrease bouncebacks.

 

Case: Continued…

You go back through your chart from Friday night and start to Monday morning quarterback this case. You know that this will be a Morbidity & Mortality and you want to know what you missed or what you could have done differently.

On review of the chart, the patient was noted to be on birth control pills in the nursing notes, but you never asked her and didn’t document this information. On admission to the ED, her vital signs included a heart rate of 102, but when you saw her, her heart rate was 99, and you hadn’t noted the prior heart rate. You documented a normal exam, CXR, and the patient stated that she felt marginally better before discharge. Her whole ED stay included 2 hours of waiting prior to being seen by you, followed by a quick discharge right after you saw her. You remember her pushing you to discharge her quickly because it was busy and the ED was making her anxious. Both you and the patient attributed her symptoms to her anxiety, and you told her to follow up with her primary doctor. On further perusal of the chart, her last heart rate documented prior to discharge was 108, (you noted that she got anxious talking to you about her symptoms) and she was noted to have expressed concerns about the cost of the ED visit to her nurse as she didn’t have insurance and didn’t have an established PCP for follow up.

 

Legal Consequences

Unfortunately, the literature on this topic is limited. Much of the information on this is from professional opinion by those involved in medical malpractice.

Having a “busy day” unfortunately does not work as a legal argument. Unless it is a true disaster, you are accountable for the usual standard of care. In fact, the busy ED can be used against the physician by the plaintiff’s attorney as evidence that the patient did not get the standard of care because the ED physician was in a rush to discharge the patient prematurely due to ED volume.

Don’t ever put judgmental statements in the chart. Even if you think the patient is a drug user, just describe the situation objectively.  For example, writing in your impression that the patient is clearly a drug seeker can only hurt you. If/when you are wrong, it will be damning evidence that you made assumptions about a patient that led to their poor outcome.

What Can We Do?

Although many researchers have tried to find risk factors or red flags for patients who bounceback, very few of them have evidence based recommendations on how to improve these results. One study by Martin-Gill states “by identifying high risk patients prospectively, physicians will be better able to make informed decisions when considering the depth of evaluation, timing of discharge decisions and extent of follow up care.” A good way to put this in context can be found in Bouncebacks! Emergency Department Cases: ED returns by Weinstock et al. The book discusses a two-step approach. First step is to identify “high risk” patients who are being discharged from the ED; second step is to review their ED presentation, exam, work-up, and chart before they leave to confirm that no high risk features for emergent conditions have been missed.

One journal article stated “Regular case review of short-term returns to the ED should be included in a comprehensive ED-based program of quality assurance.” From a quality improvement and education standpoint, a system can be implemented to regularly monitor bounceback visits that result in admission and evaluate them to see if there were any red flags in the presentation or anything that could have been done differently in the index visit. This can be implemented similar to morbidity and mortality (M&M). Discussion of these cases can allow medical students, residents, nurses, and attendings the ability to learn how to identify red flags in the patient presentation. Ideally this process would decrease bouncebacks and improve patient outcomes.

Further Reading

The book Bouncebacks! Emergency Department Cases: ED returns by Michael Weinstock et al. is an excellent resource and goes through specific cases both from a medical and legal perspective.

 

Case conclusion:

You realize in retrospect that this patient had some of the signs of a PE on initial presentation with a risk factor that you hadn’t recognized (exogenous estrogen). OCP’s + initial tachycardia meant that this patient could NOT be PERC’d out.

 

 Summary

What we know: Bouncebacks are bad for patient outcome and care and are a medicolegal concern in the ED. There are many studies demonstrating that we send home a small minority of patients who die within one week of discharge.

Goal: We must try to identify patients high risk for bounceback, particularly those with high morbidity and mortality.

What can we do: We can be diligent through QI programs, reviewing the charts of patients who bounceback in 2-3 days. Some studies have shown an increase in the bounceback rate in Medicare and Medicaid patients, and certainly patients who leave AMA and those who LWBS are at high risk. Patients with chronic renal and cardiac disease are also at risk.

In the end, a simple approach is to trust your gut (while thinking harder on patients known to be high risk). If you think you will worry about the patient after they are discharged, start over, review the case, consider the worst-case scenario, and document your medical reasoning thoroughly. Remember to re-assess vital signs prior to discharge, and pay special attention to patients with persistent tachycardia.

 

References/Further Reading

1. Sklar DP et al. Unanticipated death after discharge home from the emergency department. Ann Emerg Med. 2007 Jun;49(6):735-45. Epub 2007 Jan 8.

2. Kefer MP. Death after discharge from the emergency department. Ann Emerg Med. 1994 Dec;24(6):1102-7.

3. Pham JC et al. Seventy-two-hour returns may not be a good indicator of safety in the emergency department: a national study. Acad Emerg Med. 2011 Apr;18(4):390-7.

4. Gelareh Z. Gabayan et al. Factors Associated with Short-Term Bounce-back Admissions Following Emergency Department Discharge. Ann Emerg Med. 2013 Aug; 62(2): 136–144.e1.

5. Gordon JA et al. Initial emergency department diagnosis and return visits: risk versus perception. Ann Emerg Med. 1998 Nov;32(5):569-73.

6. Moskovitz JB et al. Emergency Department Bouncebacks: Is Lack of Primary Care Access the Primary Cause? J Emerg Med. 2015 Jul;49(1):70-77.e4.

7. Martin-Gill C and Reiser RC. Risk factors for 72-hour admission to the ED. Am J Emerg Med. 2004 Oct;22(6):448-53.

8. Weinstock, Michael B., Ryan Longstreth, and Gregory L. Henry. Bouncebacks!: Emergency Department Cases: ED Returns. Columbus, OH: Anadem, 2006. Print.

9. Pierce JM1, Kellerman AL, Oster C. “Bounces”: an analysis of short-term return visits to a public hospital emergency department. Ann Emerg Med. 1990 Jul;19(7):752-7.

10. http://www.ncbi.nlm.nih.gov/pubmed/24036997

11. http://www.ncbi.nlm.nih.gov/pubmed/20604996

Leave a Reply

Your email address will not be published. Required fields are marked *