Abstract
Abstract
Objectives:
To determine: (1) residents' knowledge base in key concepts in hospice and palliative care, (2) residents' comfort level in discussing end-of-life (EOL) care and managing EOL symptoms, and (3) if a brief educational intervention improves residents' knowledge of and comfort in dealing with EOL issues both immediately following and six months after the intervention.
Methods:
This prospective cohort study was performed at a level-1 community trauma center. A survey was developed regarding knowledge of palliative care and its emergency department (ED) utilization and administered to emergency medicine (EM) residents. Subsequently, residents underwent four hours of palliative care training. An identical post-education survey was administered immediately and then six months later.
Results:
Education improved knowledge of hospice qualifying diagnoses which was maintained at six months. Improvement in residents' ability to convert between oral and intravenous formulations of opioids was not retained at six months. Education was not shown to improve EM residents' comfort in managing pain or dyspnea but it was shown to improve their comfort level in discussing EOL care with patients and families Inpatient hospice admissions from the ED increased 88% after the education. Identified barriers to providing palliative care in the ED decreased.
Conclusions:
There are gaps in EM residents' hospice and palliative care knowledge. Education is shown to improve and maintain knowledge of hospice qualifying diagnoses, comfort level in discussing EOL care with patients and families, and to increase inpatient hospice admissions from the ED.
Introduction
While some specialties have developed curricula in palliative care for residents, the authors are unaware of any studies examining palliative care education in the ED.4–6 Given increasing numbers of patients seeking ED care at EOL, it is important to understand the gaps in palliative care education in EM. Although an Education in Palliative and End-of-Life Care for Emergency Medicine (EPEC-EM) course was developed, there are no studies examining its efficacy in teaching palliative care to EM residents. 7
This study seeks to determine: (1) residents' knowledge in key concepts in hospice and palliative care, (2) residents' comfort in discussing EOL care and managing EOL symptoms, and (3) if a brief educational intervention improves residents' knowledge of and comfort in dealing with EOL issues both immediately following and six months after intervention.
Methods
Study design and population
This was a prospective cohort study of EM residents in a combined allopathic and osteopathic residency program at a level-1 trauma center. The study protocol was approved by the institutional review board (IRB).
Residents were self-selected for participation based on availability to attend any of the educational interventions and to complete at least one survey.
Survey content and administration
The primary intervention was four hours of lecture during protected mandatory resident education time. The lectures were developed utilizing the curriculum from EPEC-EM and reviewed by both emergency physicians and hospice and palliative care physicians. The topics covered during the educational intervention are listed in Table 1.
A survey was developed using both quantitative and qualitative design. Multiple choice and standard 5-point Likert scale questions were used. Questions were designed to assess factual knowledge as well as participants' comfort level in caring for the dying patient. Residents were surveyed regarding prior formal training in managing patients at EOL. A single open-ended question queried perceived barriers to ED utilization of hospice and palliative care services. (The survey is available online at www.liebertonline.com/jpm).
One week prior to intervention, residents completed the survey. The educational intervention occurred over the course of two weeks, after which the survey was readministered. A third survey was administered six months later to assess long-term retention of knowledge.
Data analysis
Categorical data were analyzed using Fisher Exact. Numerical data were analyzed using descriptive statistics. As data were not normally distributed, comparisons between items were analyzed using nonparametric methods. Open-ended survey questions were explored using grounded theory, and responses were analyzed using Fisher Exact and descriptive statistics. All analyses were performed using VassarStats (http://faculty.vassar.edu/lowry/VassarStats.html).
Results
Eighty-three percent (33/40) of residents completed the survey prior to intervention, 31 (78%) completed the survey immediately after intervention, and 25 (63%) residents completed the survey six months after intervention. Twenty-one (53%) residents completed all three surveys. Due to the statistical methods utilized, residents were not required to complete all three surveys, as groups of residents were being evaluated, not any one individual. Twelve residents reported no prior training in EOL care. Thirteen residents had one hour of prior training, five residents had two hours of prior training, and three residents had three hours or more of prior training.
Hospice qualifying diagnoses
Residents were able to identify acquired immune deficiency syndrome (AIDS), cirrhosis, amyotrophic lateral sclerosis (ALS), malignancies, and cerebrovascular events as hospice qualifying diagnoses both prior to and following intervention (p>0.05, with >90% identifying these correctly). After intervention, residents were better able to identify cardiomyopathy, aortic stenosis, debility, failure to thrive, congestive heart failure, multiple sclerosis, Alzheimer's disease, and chronic obstructive pulmonary disease (COPD) as hospice qualifying diagnoses (p<0.05 for all diagnoses), and this was maintained at six months (Table 2).
AIDS, Acquired Immune Deficiency Syndrome; ALS, Amyotrophic Lateral Sclerosis; CHF, Congestive Heart Failure; COPD, Chronic Obstructive Pulmonary Disease; MS, Multiple Sclerosis.
Opioid conversion
Prior to intervention, 30% of residents knew the correct intravenous conversion ratio of oral morphine, and 36% knew the correct ratio for hydromorphone. Immediately after intervention, 90% of residents could identify the correct conversion for morphine (p<0.0001) and 81% could identify the correct conversion for hydromorphone (p=0.0005). This knowledge was not retained at six months post-intervention.
Comfort managing EOL symptoms
Prior to intervention, residents felt comfortable caring for dying patients, with a median of 4 on a 5-point Likert scale. They felt comfortable managing pain, with a median of 4 (IQR 3–4). These values did not change significantly either immediately after intervention or six months later (p>0.05).
Residents felt less comfortable managing dyspnea in the dying patient, with a median of 3 on a 5-point scale. This improved immediately following the lecture series, with a median of 4 (IQR 3–4, p=0.02). Although it did not reach statistical significance, some improvement was maintained at six months, with a median of 4 (IQR 3–4, p=0.06).
Comfort discussing EOL issues
Residents rated their comfort discussing EOL issues as a 3 on a 5-point scale prior to intervention. Immediately following, this improved to a median of 4 (IQR 3–4, p=0.03), which was maintained at six months (median of 4, IQR 3–4, p=0.05).
Reported referral patterns
Prior to intervention, 61% of residents reported having never referred a patient to hospice or palliative care from the ED. Immediately following, 61% reported having never referred a patient to hospice (p=1). At six months, 80% of residents reported referring someone to hospice or palliative care in the intervening six months (p=0.002).
Retrospective review of the inpatient hospice admission logs indicates that during the six-month period prior to intervention, nine patients were admitted directly from the ED. In the six months following the study intervention, 17 patients were transferred to inpatient hospice directly from the ED, representing an 88% increase.
Identification of barriers to palliative care utilization in the ED
Upon exploration, the open-ended question yielded seven categories of perceived barriers (Table 3). Prior to intervention, 94% of residents reported at least one barrier to the initiation of palliative care in the ED (53 total barriers). Immediately following intervention, 65% of residents reported at least one barrier to palliative care utilization, and at six months, 76% of residents reported at least one barrier (p=0.02). Time was consistently identified as the greatest barrier.
Limitations
This study utilizes survey data, which are subject to recall bias. To minimize this in regards to hospice referrals from the ED, inpatient hospice admission logs were queried in a retrospective fashion. Unfortunately, due to inadequacies in the electronic medical record, it is impossible to track how often home hospice or outpatient hospice was arranged. However, the actual referral patterns seem to corroborate the self-reported numbers; therefore recall bias seems minimal in this study.
This was a single center study, and its results cannot be generalized to EM residencies nationwide.
Discussion
Our study demonstrates a previously unidentified gap in EM residents' knowledge base regarding hospice qualifying diagnoses and symptom management at EOL. This gap seems correctable with minimal education, and is maintained for at least six months.
We also identified a deficit in knowledge regarding pain management, specifically in conversion between formulations of opioid drugs. This deficit is not unique to emergency medicine; in one study of 1455 medical students, 30% reported not receiving instruction in basic pain management, and internal medicine residents rate nonpharmacologic interventions for pain control as their lowest EOL skill.3,8–9 Unfortunately, the knowledge gained during the intervention was not retained long term. Further studies should address optimal ED pain management in the dying patient.
The increase in transfers from the ED to the inpatient hospice unit is likely multifactorial. First, the EM residents' comfort with EOL discussions increased after intervention, which may have led to an increased likelihood of EM residents discussing EOL care with their patients. Second, intervention increased EM residents' knowledge of hospice qualifying diagnoses, which may have assisted them in identifying and referring more appropriate ED patients for hospice services. Additionally, the increased referrals to inpatient hospice can be explained by the decrease in perceived barriers to providing palliative care. Finally, the increase in utilization of hospice resources may be simply explained by residents being supplied with the knowledge of available hospice resources.
Conclusion
A brief educational intervention is sufficient to teach hospice qualifying diagnoses to EM residents with retention at six months. The intervention was inadequate for EM residents to retain knowledge regarding formulation conversion of opioids. However, residents expressed increased comfort in discussing EOL care. In the six months following intervention there was an increase in self-reported number of hospice referrals and recorded number of hospice admissions from the ED.