Improved Quality of Life With Pediatric Palliative Care
Hays et al. (2006) evaluated the quality of life and family satisfaction of 41 children enrolled in their palliative care community-based program. The descriptive study, using the PedsQL 4.0, compared patients' quality of life at program entry (baseline) with repeated measures at three months post-program entry (follow up). Among 21 families who completed baseline and follow-up reports, improvements with moderate effect sizes (d = 0.49 - 0.60) in quality of life over baseline were reported for the physical, emotional, and school-related domains on the PedsQL 4.0. The only statistically significant improvement was in the area of emotional well-being (p = 0.021, effect size 0.60).
Most referrals came from Children's Hospital & Regional Medical Center of Seattle, Washington. However, 38% of their referrals were from health care providers throughout the state. Although 119 children were referred into the program during the two-year study enrollment period, 41 children were enrolled who met study eligibility criteria (i.e., age, presence of a potentially life-limiting illness, Washington State residency status, eligible health plan). Nineteen children died during the course of participation in the program; thus, baseline and follow-up data were available on 22 children and their families. One family did not complete the PedsQL 4.0 at the three-month follow-up. The authors reported only the analysis of the parent proxy reports due to the small number (n = 8) of child selfreports completed. Analysis of the child's perspective would have strengthened this report; however, a sample size of eight lacks statistical strength to demonstrate improved QoL due to a palliative care program. Nevertheless, this study can be replicated to describe the effects of other pediatric palliative care programs on quality of life for children and their families. The addition of a comparison group, those who do not use a palliative care program would more completely answer the PICO question.
Wolfe et al. (2008) completed a retrospective, time series, cohort study as part of an evaluation of their palliative care hospital-based program. Survey data were used to compare perceptions of parents whose children died between 1990 and 1997 (baseline cohort n = 102) and of parents whose children died between 1997 and 2004 (follow-up cohort n = 119). Chart reviews were used to gather data about patterns of care at end of life that were considered indicators of quality of life, including discussions related to chances for cure, curative treatment versus palliative treatment, hospice/home care, resuscitation status, location of death, and pain team involvement. Findings demonstrated the timing of hospice discussion happened more often (p < 0.001, 95% Confidence Interval [CI] 73–79) and earlier (p = 0.002) in the follow-up cohort than the baseline cohort. Additionally, the authors found earlier do not resuscitate (DNR) order documentation in relation to death (p = 0.031, CI 13–25) and fewer children dying in intensive care units than on the ward for those who were hospitalized at the time of death (p = 0.024, CI 74–81) among the follow-up cohort as compared to the baseline cohort. Parents' perceptions were surveyed using a Likert-type scale of their child's pain, suffering, and anxiety. Statistical analysis compared patterns of care at the end of life between the two cohorts. Statistical adjustments were made for confounding variables, including cause of death, parent gender, and physician clustering (the children who were care for by the same physician may have had similar traits). The authors found a decrease in parent reports of their child suffering from pain (risk difference [RD] 19%, p = 0.018) and dyspnea (RD 21%, p = 0.02) at end of life in the follow-up cohort compared to the baseline cohort on a response scale of "a great deal," "a lot," "some," "a little," or "no." Results comparing parent perception of their child's suffering from anxiety among the two cohorts lacked statistical significance. Additionally, parents in the follow-up cohort reported feeling more prepared during the last months of their child's life (p < 0.001) and at the time of their child's death (p = 0.002) than the baseline cohort of parents. This was thought to be because parents in the follow-up cohort discussed hospice and end-of-life care earlier and more often with their physicians and the Pediatric Advanced Care Team than the baseline cohort parents. Strengths of the study include the large sample size. Weaknesses of the study include retrospective parent perceptions, a lack of patient opinion, and lack of variety in life-limiting illnesses because all study participants were parents of children who had cancer. The authors also listed the potential limitation of selection bias because those who declined participation may have had worse experiences than those who agreed to participate. A control group for comparison, those who received standard care without a palliative care program, would have made this a stronger study Table 1.
Review and Critical Appraisal of the Evidence
Hays et al. (2006) evaluated the quality of life and family satisfaction of 41 children enrolled in their palliative care community-based program. The descriptive study, using the PedsQL 4.0, compared patients' quality of life at program entry (baseline) with repeated measures at three months post-program entry (follow up). Among 21 families who completed baseline and follow-up reports, improvements with moderate effect sizes (d = 0.49 - 0.60) in quality of life over baseline were reported for the physical, emotional, and school-related domains on the PedsQL 4.0. The only statistically significant improvement was in the area of emotional well-being (p = 0.021, effect size 0.60).
Most referrals came from Children's Hospital & Regional Medical Center of Seattle, Washington. However, 38% of their referrals were from health care providers throughout the state. Although 119 children were referred into the program during the two-year study enrollment period, 41 children were enrolled who met study eligibility criteria (i.e., age, presence of a potentially life-limiting illness, Washington State residency status, eligible health plan). Nineteen children died during the course of participation in the program; thus, baseline and follow-up data were available on 22 children and their families. One family did not complete the PedsQL 4.0 at the three-month follow-up. The authors reported only the analysis of the parent proxy reports due to the small number (n = 8) of child selfreports completed. Analysis of the child's perspective would have strengthened this report; however, a sample size of eight lacks statistical strength to demonstrate improved QoL due to a palliative care program. Nevertheless, this study can be replicated to describe the effects of other pediatric palliative care programs on quality of life for children and their families. The addition of a comparison group, those who do not use a palliative care program would more completely answer the PICO question.
Wolfe et al. (2008) completed a retrospective, time series, cohort study as part of an evaluation of their palliative care hospital-based program. Survey data were used to compare perceptions of parents whose children died between 1990 and 1997 (baseline cohort n = 102) and of parents whose children died between 1997 and 2004 (follow-up cohort n = 119). Chart reviews were used to gather data about patterns of care at end of life that were considered indicators of quality of life, including discussions related to chances for cure, curative treatment versus palliative treatment, hospice/home care, resuscitation status, location of death, and pain team involvement. Findings demonstrated the timing of hospice discussion happened more often (p < 0.001, 95% Confidence Interval [CI] 73–79) and earlier (p = 0.002) in the follow-up cohort than the baseline cohort. Additionally, the authors found earlier do not resuscitate (DNR) order documentation in relation to death (p = 0.031, CI 13–25) and fewer children dying in intensive care units than on the ward for those who were hospitalized at the time of death (p = 0.024, CI 74–81) among the follow-up cohort as compared to the baseline cohort. Parents' perceptions were surveyed using a Likert-type scale of their child's pain, suffering, and anxiety. Statistical analysis compared patterns of care at the end of life between the two cohorts. Statistical adjustments were made for confounding variables, including cause of death, parent gender, and physician clustering (the children who were care for by the same physician may have had similar traits). The authors found a decrease in parent reports of their child suffering from pain (risk difference [RD] 19%, p = 0.018) and dyspnea (RD 21%, p = 0.02) at end of life in the follow-up cohort compared to the baseline cohort on a response scale of "a great deal," "a lot," "some," "a little," or "no." Results comparing parent perception of their child's suffering from anxiety among the two cohorts lacked statistical significance. Additionally, parents in the follow-up cohort reported feeling more prepared during the last months of their child's life (p < 0.001) and at the time of their child's death (p = 0.002) than the baseline cohort of parents. This was thought to be because parents in the follow-up cohort discussed hospice and end-of-life care earlier and more often with their physicians and the Pediatric Advanced Care Team than the baseline cohort parents. Strengths of the study include the large sample size. Weaknesses of the study include retrospective parent perceptions, a lack of patient opinion, and lack of variety in life-limiting illnesses because all study participants were parents of children who had cancer. The authors also listed the potential limitation of selection bias because those who declined participation may have had worse experiences than those who agreed to participate. A control group for comparison, those who received standard care without a palliative care program, would have made this a stronger study Table 1.