INTRODUCTION
Pediatric palliative care is becoming increasingly recognized as an important part of the health care paradigm that needs to be considered early for children with life-threatening conditions (Frager, Reference Frager1996). Insufficient clinical skills in pediatric palliative care can hamper clinicians' ability to optimize their patients' quality of life, and can lead to reduced parent satisfaction with the care received (Harper & Wisian, Reference Harper and Wisian1994). Additionally, lack of perceived competence in such cases can lead to clinicians' feelings of inadequacy and discomfort (Goldberg et al., Reference Goldberg, Guadagnoli and LaFarge1987; Graham et al., Reference Graham, Ramirez and Cull1996), which might hinder future engagement with this patient population.
Pediatric residency training may be an opportune time to acquire the skills needed to care for children with life-threatening and life-limiting conditions in ways that also focus on quality of life. Pediatricians in private practice typically care for less than three patients who die per year, yet pediatric residents care for more than 35 children who die during their three years of training (Sacks et al., Reference Sacks, Fritz and Krener1984; Hoyert et al., Reference Hoyert, Freedman and Stobino2001).
Despite the increasing recognition of the need to train pediatricians in palliative care, there is little consensus as to how to achieve this education. A preliminary study by Vazirani et al. (Reference Vazirani, Slavin and Feldman2000) at our institution found no self-assessed improvement in interns' comfort levels regarding death and dying at the end of their first year of training. A more detailed study by Kolarik et al. (Reference Kolarik, Walker and Arnold2006) concluded that, without any formal training in pediatric palliative care, residents acquired no additional comfort or self-assessed competence in dealing with these issues.
In previous studies, the notions of comfort, attitude, and competence have been used interchangeably when describing and assessing residents' experiences in learning about pediatric palliative care (Goldberg et al., Reference Goldberg, Guadagnoli and LaFarge1987; Bagatell et al., Reference Bagatell, Meyer and Herron2002; Kolarik et al., Reference Kolarik, Walker and Arnold2006). For the purpose of our current study, residents were asked to assess their comfort in carrying out tasks in the following areas: communication with parents and children, medical management, documentation of do-not-resuscitate (DNR) orders, and death certificates.
The most effective method of improving the comfort and competence of residents in pediatric palliative care remains to be established at this time. Bedside teaching, pediatric palliative care clinical services, individual lectures, and seminar series are some of the proposed and tried methods in preparing residents to care for pediatric patients with life-threatening illnesses (Berman & Villarreal, Reference Berman and Villarreal1983; Bagatell et al., Reference Bagatell, Meyer and Herron2002; Kolarik et al., Reference Kolarik, Walker and Arnold2006). However, as Bagatell et al. (Reference Bagatell, Meyer and Herron2002) suggested, the quality and content of bedside teaching may greatly vary depending on the attendings' comfort and competence in this subject, and their demanding schedules. Inconsistency in bedside teaching compounded with residents' decreased time spent in the inpatient setting dictates the need for a longitudinal palliative care curriculum. Bagatell constructed a six-part seminar series for second-year pediatric residents and found an improvement in their residents' confidence in dealing with this population. Early introduction of a pediatric palliative care curriculum at the beginning of the residency training allows interns to properly deal with these complex issues before they become supervising residents.
Therefore, to provide a more consistent palliative care training for our pediatric residents, we have introduced a longitudinal palliative care-focused curriculum into our residency training program. Through the findings of the current design, we describe the challenges and possible benefits of this curriculum on the self-reported comfort of our entering pediatric residents with respect to pediatric palliative care.
METHOD
Procedure
All categorical pediatric interns of the 2005 entering class at the University of California, Los Angeles (UCLA) School of Medicine (N = 29) were asked to complete a baseline palliative care questionnaire. The initial questionnaire was administered to the incoming interns in their first week of orientation. Upon orientation, all 29 categorical interns consented and completed the questionnaire that was subsequently collected by the medical education administrative assistant, who coded the responders' name to protect confidentiality. At the end of their first year of training, follow-up questionnaires were administered, collected, and coded in a similar fashion. Twenty-two residents (76%) completed the 1-year follow-up. This study was reviewed and approved by the UCLA Institutional Review Board.
Questionnaire
Our assessment tool was designed to measure the self-reported comfort level of the residents in several areas related to end-of-life care in children and was based on questionnaires that were previously published by Bagatell et al. (Reference Bagatell, Meyer and Herron2002) and Khaneja and Milrod (Reference Khaneja and Milrod1998). The initial questionnaire was composed of seven demographic questions along with 41 other questions related to comfort toward children with death and dying issues. Our follow-up questionnaire was composed of the same questions minus the seven demographic questions. The current study presented here examines interns' comfort in communication with parents and children, medical management, documentation of do-not-resuscitate (DNR) orders, and completion of death certificates.
Curriculum
Our pediatric palliative care formal curriculum was designed to provide a longitudinal learning opportunity for our residents to train in this subject. The curriculum was composed of mandatory noon conferences and nonmandatory Comfort Care Panels.
Noon conferences included eight sessions (Table 1), held every 6 to 8 weeks throughout internship. All interns and residents on-site are mandated to attend noon conferences as a part of their core training requirements. All sessions were formatted uniformly and included a lecture in the first 20 minutes, followed by a sample case and related questions. These questions were intended to encourage residents to interact and apply the knowledge gained from lecture in a semistructured environment. Each session ended with a 10-minute open discussion.
Table 1. Pediatric palliative care lectures
We also introduced the concept of Comfort Care Panels (CCP). These were designed to examine the palliative care treatment options for patients who suffered from complex diseases with significant morbidity or end-of-life issues. These CCPs met bimonthly and were open to all faculty, staff, and residents. Resident attendance was not specified as mandatory and all classes were notified of the time and the place by e-mail, posted memos and alphanumeric pages. The panel of experts included attending physicians from different subspecialties, ethicists, nurses, psychologists, social workers, child-life workers, and others who were critically involved in the care and decision-making process. Members of the audience had the opportunity to pose questions or contribute to the discussion. Recommendations generated by these CCPs were subsequently discussed with the patients and their families. To promote “practice-based learning” among our house staff and faculty, we dedicated some of these sessions to the follow-up of previously discussed cases.
Statistical Analysis
We compared baseline and follow-up survey responses for each of 22 questions about comfort and competence related to end-of-life care. Two of the questions had 5-point Likert scale responses ranging from “strongly disagree” to “strongly agree,” and 20 questions had 5-point Likert scale responses ranging from “least comfortable” to “most comfortable.” Comparisons were conducted using the Wilcoxon signed-rank test for nonparametrically distributed categorical variables. Because the 22 outcomes were highly correlated, Bonferroni-type corrections were not used when assessing statistical significance. Correlation studies were conducted using the Pearson correlation method. We used the Statistical Package for Social Sciences (SPSS v. 14, Chicago, IL) software to calculate our statistics.
RESULTS
The 29 interns represented 24 different accredited U.S. medical schools, with no foreign graduates. Sixty-two percent of the interns were female with a mean age of 26 years (ranging from 24 to 33 years). Most interns were single (76%), with 7% “living as married” and 17% married. Only one of the residents had a child. The members of the intern class identified themselves as subscribing to an array of faiths and religious beliefs, including Buddhist (3%), Christian (31%), Hindu (14%), Muslim (3%), Jewish (7%), and agnostic (7%). The remainder (35%) did not define a religious preference. Residents described their racial and ethnic background as Caucasian (24%), Asian (17%), Indian (14%), Iranian (10%), and “other” (35%). A comparison of the responders versus nonresponders to the follow-up questionnaire across demographic variables yielded a significant difference between genders, such that female residents responded more than males (p = .006).
More than half of the interns (55%) indicated some experience with taking care of a dying pediatric patient during their medical school training, and 79% indicated that they had taken care of a dying adult. All indicated some form of experience dealing with death and dying involving their family and friends. Fifty-two percent identified some formal training in grief and bereavement during their medical school training, whereas 35% had no training and 14% were unsure. A smaller portion of our interns (38%) indicated formal training in death and dying in different cultures, whereas 52% had no training and 10% were unsure. Only 7% of the interns felt adequately prepared to deal with death and dying.
However, all respondents indicated interest in further training in pediatric palliative care. During their first year of training 55.2% of the interns attended at least one of the eight palliative care lectures. The mean number of lectures attended by this group was 1.27. However, our CCP attendance records indicate that, despite a considerable attendance by the second and third year residents, none of the interns attended any of these sessions.
We compared the overall competence and comfort levels of 22 residents who had responded to the questionnaires before and after the first year of training (Table 2). Overall, the responses indicated greater self-reported comfort after 1 year. We also asked them to rate their comfort level in carrying out 20 different related tasks before and after the first year of training (Table 3). After 1 year, the survey results demonstrated a significant improvement in 7 of the 20 measured areas. The areas in which residents reported greater comfort included managing bowel and bladder problems, fluids and nutritional needs, pain, seizures, anxiety, and respiratory needs, and pronouncing a child's death.
Table 2. Comfort in end-of-life care before and after internship
1 = strongly disagree, 5 = strongly agree.
*p < .05.
Table 3. Self-assessed competence in dealing with pediatric end-of-life care
1 = least comfortable, 5 = most comfortable.
*p < .05, **p < .01, ***p ≤ .001.
There appears to be no significant statistical correlation between the number of lectures attended or resident demographics and improved comfort levels mentioned above.
DISCUSSION
Our initial survey indicated that about half of the interns, representing 24 U.S. medical schools, had had some form of medical school training or experience with patients' deaths. However, a mean comfort level of less than 3 (1 indicating the lowest comfort level and 5 indicating the highest comfort level) in 17 of the 20 palliative care areas on the initial questionnaire (Table 3) suggests that our interns were not adequately prepared to feel comfortable in medical school on this subject. They have additionally indicated that they do not feel adequately prepared to deal comfortably with death/dying among their patients (Table 2). This supports previous studies indicating a fragmented palliative care education that mostly takes place in a classroom setting during the first 2 years of medical school, without any longitudinal clinical exposure (Mermann et al., Reference Mermann, Gunn and Dickinson1991; Rappaport & Witzke, Reference Rappaport and Witzke1993; Holleman et al., Reference Holleman, Holleman and Gershenhorn1994; Hill, Reference Hill1995).
The lack of formal and consistent medical school training in this field underscores the importance of palliative care education during residency training. Additionally, because most of the clinical training of these future pediatricians occurs during their residency, graduate medical educators have a greater opportunity and responsibility to provide meaningful training with measurable results. Kolarik's study has demonstrated that without any pediatric palliative care training, most residents gain no significant knowledge or comfort in managing end-of-life issues in children (Kolarik et al., Reference Kolarik, Walker and Arnold2006). Vazirani et al.'s (2000) earlier study of UCLA interns found no significant improvement in their self-assessed comfort after the first year of training. However, to the best of our knowledge, there are no published reports on longitudinally implemented pediatric palliative care training during all 3 years of training.
Our study aims at exploring the changes in the self-assessed comfort of interns before and after the first year of training with respect to their exposure to the new curriculum. Our assessment of the interns' comfort in this subject showed significant improvement in several areas after the completion of internship. However, these improvements cannot be directly attributed to the educational intervention.
Our study underlines the challenges involved in providing an effective, practical palliative care curriculum. Although repeated efforts were made to facilitate the attendance of the house staff at the CCPs, none of the interns attended these sessions. Despite the presence of the second and third year residents, the lack of intern attendance can be explained by their demanding schedule. This may have limited their availability to attend nonmandatory educational sessions. Mandatory conferences along with modification of the work hours may facilitate and improve the interns' attendance in palliative care related training sessions.
Furthermore, the improved areas appear to be independent of the number of noon conferences attended. Each of these palliative care noon conferences was composed of didactic lectures and interactive discussions, intended to provide resident participation. This “mixed” approach (didactics plus interactive discussions) is supported based on Davies et al.'s (Reference Davies, O'Brien and Freemantle1999) meta-analysis findings, such that mixed conferences seem to have a significantly positive impact on improving the practice behavior of the physicians. Although Davies et al.'s meta-analysis on educational curriculum suggests that that mixed conferences improve physician practice behavior, this educational approach may not be sufficient for improving levels of physician comfort in dealing with such difficult topics as pediatric palliative care.
Lack of a relationship between attendance and areas of improvement may be due to the small size of our sample, limited duration of our study, and/or lack of adequate intern participation in palliative care noon conference discussions. Any improvement may simply reflect increased comfort with repeated exposure and practice.
On the other hand, lack of any measurable direct relationship between the introduced curriculum and improved comfort parameters does not refute the usefulness of such educational interventions. Increased exposure through the palliative care educational curriculum may have created a culture of awareness regarding this topic. Thus, introduction of the palliative care formal curriculum may have created an informal curriculum that facilitated learning outside a classroom setting, where faculty and trainees alike engage in discussions and management of otherwise ignored palliative care issues.
Interestingly, six out of the seven areas of improvement included end-of-life medical management of children whereas all areas where no improvement occurred involved communication with the patients and their families, communication with a faculty role model, or seeking help in the community. This deficit is of particular concern because communication with parents of dying children is a task often incumbent upon the pediatrician. Harper and Wisian's (1994) study indicated that parents were generally dissatisfied with their pediatrician's communication, sharing of grief, and provision of support during their child's end-of-life care. This supports the need for further training in this area. Role modeling by the more experienced faculty, involvement of the residents in family conferences, parent advisory panels, and communication workshops are some of the venues for improving the communication skills of the training residents.
The areas where no improvement occurred deserve further study and assessment in the next 2 years. Upcoming introduction of a clinical Palliative Care Team may facilitate better role modeling, bedside teaching and practical didactic sessions. Additionally, introduction of role-playing workshops and inclusion of the residents in the bedside discussions led by the faculty may provide added experience and improve communication skills of the residents.
Limitations
We recognize that our study has multiple limitations that include lack of a control group and poor attendance at the CCPs. Additionally, we did not track the number of dying patients cared for by the interns during their first year of training. Due to the multicampus nature of our residency program, we decided to introduce the lectures only on the main campus, and, therefore, some of our interns missed some of the offered didactic and clinical teachings while on away rotations.
Conclusion
Preliminary findings of our study suggest that interns' comfort improved in managing certain areas of palliative care after the first year of their training. The role of the introduced longitudinal curriculum in improving the interns' comfort level is unclear at this time. However, we intend to further study the 3-year effect of the existing curriculum on our interns at the time of their graduation and distinguish the components of their formal and informal training that they deem as most valuable. Lack of improved comfort in communication with the parents and faculty involving a dying child suggests a need for further study and educational interventions. Further studies are needed to examine the direct and indirect effects of a longitudinally implemented palliative care curriculum in pediatric residency programs.
ACKNOWLEDGMENT
This research has been made possible by a grant from the Unihealth Foundation. We would also like to thank all our residents for contributing to this study.
