Implementation and Evaluation of an Automated Patient Death Notification Policy at a Tertiary Pediatric Oncology Referral Center

Abstract

Background

Not knowing about a child’s death can result in poor quality of care coordination among staff and poor quality bereavement care for families. The purpose of this project was to create, implement and evaluate an automated Patient Death Notification policy and procedure (PDNPP).

Measures

Baseline and follow-up surveys of clinical staff.

Intervention

Implementation of a PDNPP that created an automated, systematic process for staff notification of patient deaths.

Outcomes

Ninety-six percent of the staff rated the PDNPP as a significant improvement; 91% reported being “very” or “somewhat” satisfied with the PDNPP, whereas only 44% of the staff was satisfied with the process at baseline.

Conclusions/Lessons Learned

Implementation of an automated PDNPP was feasible and improved staff satisfaction about how they were informed of patient deaths. Staff also reported being notified about patient deaths more quickly, performing their jobs more efficiently, being able to avoid doing something that might upset the deceased patient’s family, and being able to better provide support to bereaved families.

Case Study

JD was a 5-year-old boy with neuroblastoma. JD’s family made the difficult decision of transitioning to a primary goal of comfort and they enrolled him in home-based hospice where he died three weeks later. The next day a scheduler from the physical therapy clinic called the family to confirm JD’s appointment for later that week. This caused significant distress to JD’s family as well as to the scheduler, who was unaware of JD’s death.

Background

In 2002, members of the St. Jude Children’s Research Hospital (SJCRH) Palliative and End-of-Life Care Task Force conducted ten focus groups comprised of interdisciplinary members of the clinical staff in order to better understand the vision and goals of palliative and end-of-life care held by clinicians in the institution. One common theme that was identified was “clear communication about patient care to all members of the health care team is needed.”¹ A recommendation was made to “develop a communication mechanism to alert clinical teams of changes in patient status.” Clinicians at SJCRH, where approximately 140 SJCRH patients die each year, indicated that the most critical aspect of communication was timely formal notification of a patient’s death. In order to improve the quality of care coordination and interdisciplinary communication when a patient dies at our institution, a death notification working group (DNWG) was established to evaluate the existing death notification process and implement system changes designed to communicate information regarding the death of a patient to institutional staff in a safe, timely and effective manner.

Measures

Improvement in the quality of care coordination was evaluated by system-level performance measures before and after implementation of an automated Patient Death Notification policy and procedure (PDNPP). An instrument to collect information at baseline was developed based on the notification needs identified by clinicians in the focus groups. Two process measures and two outcome measures were utilized. Measures were critically reviewed by members of the multidisciplinary DNWG to establish consensus validity. A follow-up survey was repeated six months following implementation of the PDNPP. This survey repeated the questions from the baseline survey and also measured potential benefits and unintended effects of the intervention.

Intervention

The clinician-reported focus group themes and the baseline staff satisfaction survey revealed a high level of dissatisfaction with the process of notification about patient deaths, which led the DNWG to draft a new policy and procedure (available upon request from the authors) to address the identified concerns. The PDNPP identified a staff member responsible for immediate notification by telephone of those with an urgent need to know about patient deaths (for example, the child’s physician, fellow, chaplain, etc.). After the immediate notification process was completed, the same staff member was required to enter a “change of patient status” order into the patient’s electronic medical record. Entry of this electronic order triggered an automated e-mail to a predefined list of individuals notifying them of the patient’s death. The original list of individuals to be included on the distribution list was compiled by the DNWG after they surveyed each department to determine which staff members needed to be informed about a patient’s death. A mechanism for keeping the list current also was implemented. Every six months, individuals on the death notification list receive an e-mail mandating that they remove themselves from the list if they no longer have a need to know about patient deaths, and reminding them about the confidential and sensitive nature of the transmitted information.

Outcomes

At baseline, 333 (47% response rate) staff members completed the survey and 429 (55% response rate) responded at six-month follow-up. Only 7% (n=28) of the staff who were a part of the original distribution list asked to stop receiving the e-mail notification. Sixty-eight percent (n=19) of those who opted out did so because they did not need the information whereas 25% (n=7) opted out because the number of emails was burdensome.

Prior to implementation of the PDNPP, staff stated that knowing about a patient’s death was necessary to fulfill job responsibilities (55%), and not knowing about a patient’s death prevented them from fulfilling those responsibilities (25%) or led to an awkward or uncomfortable interaction with a parent, family member, or health care provider (54%). Most staff felt a need to learn about a patient’s death immediately after the death (20%), within a day (34%) or within a week (37%).

Ninety-six percent of respondents considered the new patient death notification process to be an improvement; 95% of clinical staff surveyed found out about a patient’s death through the electronic death notification process (Table 1). Ninety-one percent of the clinical staff were “very” or “somewhat” satisfied with the new death notification process, compared to 44% of staff who were “very” or “somewhat” satisfied prior to implementation of the new process. Eighty-nine percent of the clinical staff were “very” or “somewhat” satisfied with the timeliness of the new process, compared to 28% of staff who were “very” or “somewhat” satisfied prior to implementation of the new process. Following its implementation, clinical staff reported being notified about patient deaths more quickly (81%), being able to perform their jobs more efficiently (29%), being able to avoid doing something that might upset the deceased patient’s family (45%), and being able to better provide support to bereaved families (32%). Clinical staff also reported that the death notification process made them more aware about the number of children who die at the institution (42%). Most clinical staff (80%) did not identify any detrimental effects from participating in the new process. A small group of clinicians felt the new process created extra emails (12%), or indicated feeling upset as a result of knowing about patient deaths (7%) (Table 2).

Table 1.

Recurrent Qualitative Themes from Staff When Asked About PDNPP Beneficial Effects (n=110) and Recommended Changes (n=94)

Beneficial Effects	Recommended Changes
Improving staff’s ability to perform job duties (n=24)	No changes needed (n=32)
Creating a more rapid reliable process to stay informed about patient deaths (n=18)	Duplication of emails (a flaw related to being on the distribution list twice) (n=28)
Being more sensitive to the needs (including bereavement) of families (n=12)	Ensure tight regulation of list (n=7)
Making a personal impact; including gaining a better understanding of what is most important in life and helping provide closure (n=9)	Make it easier to opt out (n=6)
Improving communication within the care team (n=8)	Include more personal information about the patient (n=6)
Feeling more connected to the mission of the institution (n=8)	Include cause of death (n=5)
No impact (n=6)	Send sooner (n=4)
Makes me sad or upset (n=6)	Make the e-mail text more personal (n=3)
Makes me aware of how many children die at the institution (n=5)
Improving staff’s ability to follow up with the family (n=5)

Table 2.

Recurrent Themes and Representative Quotes from Before and After the Implementation of the PDNPP

Theme	Sample Quote from Before Death Notification Process	Sample Quote from After Death Notification Process
Personal impact on staff	It does not keep me from fulfilling my job, however I lose the opportunity to offer support and gain needed closure. I have seen nurses devastated at abrupt, delayed news. They have had to leave the work area, myself included. Sometimes, it is just too much to take when you’re really close to a family and the last to know they have died.	(The death notification process) has made me more sensitive to patients and families. Sometimes (before the process) I wouldn’t see a family for awhile and then ask about their child, only to find out he/she died and it’s embarrassing and sad. It’s easier this way. It allows me to stop what I am doing however important it may seem at the time and remember what is really important. My son died of a neurological disorder in April and I am very aware and in tune with the feelings of families or parents of a deceased child. When I get a notice I stop and say a prayer for the deceased and their family and it makes me relax for a second and refocus my nervous energy on what is important in this life.
Improving Communication Within the Care Team	The nurses on nights feel especially out of the loop because we miss out on the word of mouth that goes on during the day.	As a night shift worker it really makes me feel more involved in the process of total patient care.
Creating a more rapid and reliable process	Usually a card to the family is put on the desk in our office for our signatures, should we wish to sign. That’s usually my first knowledge. Of course, a patient whom I thought I was told by “word of mouth” had died last week showed up today for anesthesia. So, “word of mouth” ain’t very reliable.	It’s very important to know about deaths in a timely manner and unfortunately in the past, this hasn’t been effectively communicated. I have found it very beneficial.
Feeling more connected to the institution’s mission	I think that all research and clinical employees should be on the mailing list when a child dies. It does help the research nurses to know because a child might be on one of the protocols. It also helps the clinical nurses to know so that when you see the family, you can give them a big hug.	“It’s a way of putting a name with knowing the fact that we lose some children. It keeps me reminded that we are a research hospital and we don’t have all of the answers but for this reason we keep trying.” (Because of the death notification process) I never lose sight of the mission.
Being more sensitive to the needs (including bereavement) of families	But, not knowing and finding out after caring for families close to the deceased has made it more difficult meeting their needs. The child may wake up from anesthesia more afraid or upset and the family may be more anxious. Each year, when sending letters to patients - former patients -regarding the Annual Survivors Day Celebration, I will receive a call from the parent saying the child/adult had died. This places me in an embarrassing/awkward situation and apologizing profusely. This happens about two or three times each year.	It has been real beneficial for me to get (bereavement) cards to the clinic staff to send out to the families in a timely manner.
Ability to perform job duties	We make plaster hand molds as a legacy item for parent whose child has died at the hospital. There have been multiple times I have not been contacted by staff, however, the parents still request this service and I have had to go to the morgue to fulfill this responsibility and even that is sometimes not possible Not being able to do appropriate bereavement follow up and very difficult when communicating with parents that “think” you know and have not communicated. In follow-up conversations with caregivers, I have occasionally been unprepared to assist with their grief process.	It has helped me to do my job better (if a patient dies, he/she is more expediently removed from our Website). It has been an especially helpful addition because it allows us to remove any future appointments or requests for a deceased patient. Continuing to get phone calls or appointment cards from us would likely be very distressing for parents.

Conclusions/Lessons Learned

The Institute of Medicine, American Academy of Pediatrics, and the National Institutes of Health have identified end-of-life care and care coordination of children with special health care needs as priority areas for quality improvement. ^2–4 Our focus groups and baseline survey also demonstrated this need for improvement in the area of care coordination, as high levels of dissatisfaction with staff communication surrounding patient deaths and the need for improved family follow-up and bereavement care existed.

The outcomes of the quality improvement project described here indicate the feasibility of implementing a systematic policy and procedure for notifying staff of patient deaths, and the improved satisfaction among staff members about being informed of patient deaths. Improvements were quantitatively and qualitatively demonstrated including enhancing communication within the care team, improving the staff’s ability to perform job duties, being more sensitive to the needs of patients and families, feeling more connected to the mission of the institution, being more involved in the overall care process of patients and families, and helping provide closure for staff.