What happens when patients call their own rapid response?

A hospital finds surprising results from safety initiative.


Clinicians have repeatedly studied their own effectiveness summoning and implementing rapid response teams, but a group of researchers recently analyzed what happens when the first step is turned over to patients and families.

The University of Pittsburgh Medical Center (UPMC) has a system called Condition Help, which allows patients and their families to summon a patient care liaison or an on-duty administrator to the bedside. The system was launched as a safety initiative in 2005 at the Children's Hospital of Pittsburgh at UPMC before expanding to other facilities.

Admitted patients are advised to call the Condition Help hotline if they have an emergency and inability to alert hospital staff, a change in clinical condition without recognition from the health care team, a breakdown in how care is given, or uncertainty over next steps.

From January 2012 through June 2015, 367 hotline calls were made by 240 patients and family members, and 43 people who made multiple calls accounted for about 46% of all calls, according to the results of a study published in the March Journal of Hospital Medicine. The most common call reason was inadequate pain control (about 48% of all calls), followed by dissatisfaction with staff (about 13%), and a change in care was made in about 41% of cases.

Elizabeth Eden MDslashPhoto courtesy of Dr Eden
Elizabeth Eden, MD/Photo courtesy of Dr. Eden.

ACP Hospitalist recently spoke with lead author Elizabeth Eden, MD, a UPMC hospital medicine fellow, about the implications of these findings.

Q: Why did you decide to study this issue?

A: Condition Help has been active in our medical system for a number of years, and we believe that it's a pretty novel intervention. After reviewing the published literature, we found that most of the data was in pediatrics and a few small pilot studies in adults. We were also curious about who is calling Condition Help, why they are calling. . . and whether we were able to pick up any safety-related issues as a result of this program.

Q: What were some of the most interesting findings?

A: One of the interesting things was just the sheer number of Condition Help calls that happened at our hospital. Most of the published studies were pilots looking at a few months at a time, and they often had single-digit numbers of calls, whereas we had about 100 calls every year for Condition Help. The other surprising thing is the number of calls made by the repeat caller group. We tried to determine. . . Is this repeat caller group calling because they're much sicker and they need a lot more attention? So we looked for a few surrogates of that, like ICU transfer, whether a traditional rapid response team was called, if there was death in the hospital. I think what was helpful for us to see was that there wasn't actually any difference in those outcomes between the people who called regularly compared to the people who just called one time.

Q: What are some potential implications of those repeat callers?

A: They definitely seemed to have a different makeup than the other patients. They were younger, they were admitted more frequently, over half of them were calling for reasons of pain control. They also called more often on their own: A much higher percentage of these calls were calls by the patient rather than the family member. We have started to look into the repeat caller group, reviewing demographics and medical comorbidities, and found that a very high percentage of them have issues with chronic pain. Over 75% are either taking opiates or have a history of opiate abuse. This is not surprising given the high frequency of calls related to pain and suggests that the approach to this group of patients should be different.

Q: Considering that only about 11% of calls were about safety issues, what kind of impact do you think this has on patient safety?

A: Our paper can't answer that question, since we're only looking at it in one facility. In our tertiary care center, Condition Help probably does not have a significant impact on patient safety, given the few safety-related events that we identified. I think that could be for a few reasons. One could be that we have a very well-developed traditional rapid response team and very clear criteria that the nurses use as to when to escalate care. So part of this could be that we're detecting deteriorating patients very early on. We also have physicians that are here 24/7. In a hospital that maybe doesn't have those resources, perhaps Condition Help could provide a little bit more of a safety benefit. Then again, maybe patient-activated rapid response teams generally do not impact patient safety. We would need additional studies to better answer that question.

Q: With these results in mind, what are the merits of systems like Condition Help?

A: Some of the merits of Condition Help are that no matter what the circumstance, a patient will always have a voice during the hospitalization, and having a system like this ensures that there's a path open to them if they feel like something's going wrong. We think of it as being kind of a last line of defense, a last line of patient advocacy if they feel like their needs just aren't being met by the current medical team. Other studies of patient-activated rapid response systems have shown that it improves patient satisfaction and that they are well received. Anecdotally, I can say I agree with that, although we didn't measure patient satisfaction in our study.

Q: Is Condition Help still in use at UPMC, and how might it be improved?

A: The system is currently still in use. We believe that it works, though it is accomplishing a different purpose: enhancing patient engagement and not really patient safety as intended. If our goal is to reduce the number of Condition Helps called and shift the focus back to targeting patient safety, then more clearly defining the criteria for calling Condition Help when patients are admitted could be helpful. This would involve re-education of our staff when they are reviewing the brochure on admission. At this point, I believe our administration is OK with the balance we have struck, and no efforts have been made to modify the system.

Q: What are your next steps?

A: Beyond modifying Condition Help, I think there's probably a lot of work that we can do to target the repeat caller group that accounts for such a high percentage of calls and demonstrates high medical utilization. Interventions that establish clear expectations about pain management and/or direct patients to other resources for patients with chronic pain would be beneficial. Condition Help does not appear to be well-suited to address those needs. We have also considered performing a qualitative study of the repeat callers to learn about their general attitudes toward hospitalization and what they feel like they're getting out of Condition Help and if there are other needs that they feel aren't being met.