The Impact of Integrating the Palliative Prognostic Index into Palliative Consultation on Patients with Haematologic Malignancies: A Case control study

Background The study aims to investigate the inuence of integrating the Palliative Prognostic Index (PPI) into the consultation system for patients with haematologic malignancies. Methods We retrospectively enrolled 53 patients with haematologic malignancies. The PPI was evaluated at the rst palliative consultation. Patients were divided into two groups: before the use of the PPI (23 patients) and after the use of the PPI (30 patients). Results


Abstract Background
The study aims to investigate the in uence of integrating the Palliative Prognostic Index (PPI) into the consultation system for patients with haematologic malignancies.

Methods
We retrospectively enrolled 53 patients with haematologic malignancies. The PPI was evaluated at the rst palliative consultation. Patients were divided into two groups: before the use of the PPI (23 patients) and after the use of the PPI (30 patients).

Results
We rst con rmed that the life expectancy for patients with haematologic malignancies was correlated with the PPI score ranking (p < 0.01). For patients with a PPI score > 6, agreement to attend hospice care was signi cantly higher (p = 0.01). After the use of the PPI, the mean survival time from the rst consultation to death was 131.4 ± 55.9 days, which was signi cantly longer than before the use of the PPI (p < 0.01). Meanwhile, more leukaemia patients received palliative consultation and fewer antibiotics in their end of life care. Although there was no difference in agreement for hospice care after the rst consultation, we believed that the concept of palliative care had been delivered to patients and their families.

Conclusions
The PPI score is a good prognostic index for patients with haematologic malignancies. The use of the PPI score in the rst consultation enables patients, families and haematologists to become aware of the necessity of palliative care.

Study design
Data were collected from two cohorts of patients receiving treatment at Keelung Chang Gung Memorial Hospital. The rst group of patients attended the palliative consultations at our hospital from January 2012 to August 2013. During this period, many patients expired within 24 hours, so we hoped to nd a reliable indicator for patients and their families to understand their life expectancy. Consequently, our palliative care team decided to create a PPI recording system at the initiation of the palliative consultation, and the patients involved formed the second group. These patients received consultation from January 2016 to August 2017. To gain a better understanding of the clinical impact of the PPI recording system in our practice, the period of data collection was the same as for the rst set of data. A prior study reported that the PPI index could be helpful for clinical nurse specialists to improve prognostic accuracy,(11) so we hoped to establish a useful consultation system to improve the quality of our hospice care.
In this study, the haematologic tumour board at Keelung Chang Gung Memorial Hospital con rmed the diagnoses of haematologic diseases. Studies have indicated that haematologic oncologists limit hospice referrals because they are concerned about the adequacy of the services provided. (13) Because all of the haematologists in our hospital received training in hospice care and are licensed specialists, we were able to reduce inconsistencies in providing care for these patients.
Finally, a total of 53 patients were enrolled in this study. Because this study is retrospective, the results did not interfere with the decision-making process following treatment. The PPI was determined by a clinical physician at a patient's rst consultation at the hospital. The data collected throughout the study included not only information about the patient's clinical condition and the intervention but also the reasons for discontinuing the consultations. The institutional review board of Chang Gung Memorial Hospital approved this study. (IRB No. 201507911B0D001, 202101470B0).

Statistical analysis
We used descriptive statistics to describe the participants' demographic characteristics. Basic demographic data were summarised as n (%) for categorical variables and median with the interquartile range (Q1-Q3) for continuous variables, respectively. We used Pearson χ 2 or Fisher's exact test to examine the statistical signi cance between the variances. An independent-sample t test was performed to compare the mean PPI score before and after the PPI was used in the consultation system. Overall survival was calculated using the Kaplan-Meier method. Eight potential prognostic factors were included in the univariate and multivariate analysis. All factors used in the univariate analysis were examined in the multivariate analysis, but only those factors with statistical signi cance were displayed. All factors that were at least marginally associated with overall survival (p ≤ 0.2) were entered into the multivariate analysis. To understand the impact of independent factors on overall survival, we used multivariate Cox proportional hazard model using forward logistic regression analysis. All analyses described above were performed using the Statistical Package for the Social Sciences for Windows, version 21.0, and results were considered signi cant when p < 0.05. Table 1 shows the demographic characteristics of the 53 patients with haematologic malignancies (men, 40; women, 13; mean age, 76.7 years; age range 69-83 years) who received hospice care. Lymphoma was predominant among these patients, followed by leukaemia, multiple myeloma and myeloproliferative disorder/myelodysplastic syndrome. Of the 53 patients, 32 (60.4%) died within 24 hours after their rst palliative consultation. After 2 years of follow-up, we assessed whether the life expectancy of patients with haematologic malignancies after the rst palliative consultation was associated with the PPI score. Figure 1a shows that the median overall survival in patients with a PPI score > 6 and ≤6 was 23.1 ± 7.6 days and 269.3 ± 118.2 days, respectively. The clinical impact of the PPI intervention on survival is shown in Fig. 1b. Before the PPI score was used, the median overall survival was only 12.0 ± 12.3 days. After we started our programme, the median survival time was 131.4 ± 55.9 days, which was signi cantly longer (p < 0.01). Table 3 shows the impact of the PPI score on patient care. Although more leukaemia patients were included after the PPI intervention, the distribution of PPI score (> 6 or ≤ 6) also showed no difference. Patients with a PPI score greater than 6 required more support, especially in terms of blood transfusion (54.72%, p = 0.05), antibiotic use (58.49%, p = 0.02), and oxygen supplementation (73.58%, p < 0.01). The number of deaths within 24 hours after the rst consultation and the patient or family's agreement to hospice care were also signi cantly associated with a PPI score > 6 (p < 0.01 and 0.02, respectively).

Results
Next, we enrolled disease type, age > 65 years, gender, PPI score > 6 and aggressive interventions, including blood transfusion, antibiotic use, oxygen supplementation and pain control, as prognostic factors in the univariate and multivariate analyses. Both antibiotic use and PPI score > 6 were independent factors of overall survival. The group with a PPI score > 6 showed an increased risk of death

Discussion
According to previous studies, the PPI is considered a predictor of life expectancy. (12,14) In our study, we demonstrated the clinical impact after the integration of PPI into the consultation system for haematologic malignancy, which has not been previously reported. At rst, the baseline data indicated that the total number of cancer patients receiving hospice care increased. Chiang et al also demonstrated that the quality of end of life improved in Taiwan from 2002 to 2011.(15) Unlike cases of solid tumours, the number of patients receiving hospice care for haematologic malignancy in our hospital did not increase dramatically. However, the type of disease obviously changed, and more patients with acute leukaemia could receive palliative consultation. At the same time, after we started our consultation earlier, the number of patients who expired within 24 hours declined signi cantly.
Second, the mean PPI score at the rst consultation was > 6, which suggested that the life expectancy for those patients was less than 3 weeks. (14) In cases of haematologic malignancies, it is crucial that the concept of hospice care is promoted at an earlier stage. (7) Based on the overall survival of patients after the rst consultation, our programme reached the goal of starting hospice care earlier.
Third, this study investigated the outcomes before and after the integration of the PPI score. Our results showed that patients with a PPI score > 6 required more interventions to alleviate their symptoms. Although all caregivers involved in this study were well trained in hospice care, the frequency of aggressive intervention was still high. Previous studies showed that one-third of patients undergo blood transfusions, and 90% receive antibiotic treatment during the last week of their lives.(16) Aggressive interventions such as blood transfusions, antibiotic use and oxygen supplementation were therefore still commonly used during the patients' last weeks or days in our hospital. It has, for example, been reported that blood transfusions could relieve symptoms with minimal harm. (17)  This study has at least two limitations. First, only 53 patients were enrolled. Thus, some bias will be present when using the statistical methods. The second limitation is the lack of other palliative care services in our hospital, such as home hospice care. Focusing only on symptomatic relief is not so attractive for patients with terminal-stage haematologic malignancy. It is necessary to expand our work in palliative care to provide a better quality of end of life.
We discovered that patients with a PPI score > 6 are good candidates for initiation of hospice care.
However, more effort is needed to improve the quality of end of life for patients with haematologic malignancies.

Page 7/13
The implementation of the PPI score for predicting the life expectancy of terminally patients with haematologic malignancies could help promote hospice care among physicians, patients and their families. Through the early initiation of palliative consultation, appropriate care can be provided, including reducing ineffective treatment.

Declarations
Ethics approval and consent to participate The study was approved by institutional review board approval (IRB number: 201507911B0D001, 202101470B0)

Consent for publication
The authors declare no con icts of interest.

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests    Although the total number receiving hospice care increased, the total case number for haematologic malignancies did not change signi cantly. By using the PPI score evaluated, more leukemia patients and family agreed to receive palliative care. Meanwhile, physicians also reduced the use of antibiotics obviously to avoid unnecessary treatment.