Study population
Table 1 summarizes the demographic characteristics of the 2,202 study participants. Both sexes including 938 (42.6%) female and 1,264 (57.4%) male participants were equally represented. The average age of patients that did not have a palliative consultant was higher (74.9 ± 14.1) than those that did (70.5 ± 14.5). The age groups from 21–40, 41–60, 61–80, and > age 80 were 2.0%, 20.3%, 44.0% and 33.7% respectively. With regard to the 1,364 patients that had a palliative consultant, 1,353 (99.2%) of patients received either palliative care or palliative related consultation 1–3 months prior to their death. The prevalence of patients with three major causes of death and comorbidities such as lung disease, cancer, and sepsis was 61.3%, 53.5%, and 27.4%, respectively, indicating that patients may have had cancer or multiple other chronic diseases. It is also possible for patients to have more than one comorbidity at the end of their lives. The average age for patients that had not received palliative care in comparison to the group that did was 74.1 ± 4.1 years old vs 70.5 ± 14.5 years old (P < 0.001), respectively. There were significant differences in age distribution between the patients that had received palliative care and the patient group that did not (P < 0.001). Furthermore, in the subgroup of the palliative care patients, the average age of patients that received palliative care before their last hospital admission versus patients that received palliative care after their last hospital admission was 66.3 ± 13.5 versus 71.7 ± 14.6 years old, respectively. Additionally, the data exhibited a significant difference in age distributions before and after the patients’ last hospital admission (P < 0.001).
The patients that visited the Emergency Room (ER) did not exhibit a significant difference between the two groups (73.4% in the no palliative care group vs. 70.5% in the palliative care group, P = 0.148). Compared to patients without PCS, those with PCS exhibited a reduced utilization of medical resources, including the duration of their stay in ICU (n = 552 vs. n = 439, P < 0.001), endotracheal intubation (n = 68 vs. n = 151, P < 0.001), respiratory ventilator use (n = 316 vs n = 444, P < 0.001), and CPR (n = 32 vs. n = 79, P < 0.001).
Medical resource utilization
The utilization rates of aggressive care for the PCS with and without groups are presented in Table 2. The number of ER visits have no differences statistically between the two groups with or without palliative care intervention. The average hospitalization length was 19.0 ± 9.1 days versus 15.8 ± 10.3 days (P < 0.001) and the duration of the stay in ICU was 3.7 ± 6.7 days versus 5.8 ± 7.1 days (P < 0.001) in palliative care and non-palliative care groups, respectively. However, the number of blood transfusion units increased significantly from 2.6 ± 5.1 in the PCS group to 3.4 ± 5.8 in the group without PCS.
The medical resource utilization between cancer and lung disease
Table 3 compares healthcare resource utilization between patients with cancer and lung disease who received palliative care before or after hospital admission. Regarding these cancer patients, 57.8% were male and 42.2% were female, and the proportions of patients aged 21–40, 41–60, 61–80, and > 80 years were 28 (2.9%), 267 (28.1%), 484 (50.9%), and 171 (18.0%), respectively. Palliative care and consultant services before the final hospitalization or in the post-hospitalization period did not differ significantly based on sex. However, with regards to age, the age of patients receiving palliative services after hospitalization is significantly older than that of patients receiving palliative services before hospitalization (66.3 ± 11.5 years old vs 71.7 ± 14.6 years old).
In this study, there was a lower proportion of patients who were hospitalized for a period over 14 days (19.0 % vs 57.5%, P < 0.001) as well as those who were hospitalized for a period over 30 days (17.3% vs 18.3 %, P = 0.007) compared to the PCS group of patients who received care before and after hospital admission. However, in terms of the proportion of patients utilizing other healthcare resources, there was no significant difference between cancer patients before and after hospital admission (P > 0.05). Among the lung disease patients, 60.7% were male and 39.3% were female, and the proportions of patients aged 21–40, 41–60, 61–80, and >80 years old were 20 (2.7%), 124 (16.9%), 323 (44.1%), and 265 (36.2%), respectively. The proportion of individuals over age 80 was higher in the cancer group. Patients with lung disease exhibited significant differences in receiving palliative care before or after hospital admission during hospitalization over a period of 14 days: 41.0% vs 64.4% (P < 0.001), hospitalization over a period of 30 days: 13.1% vs 27.0% (P = 0.001), ICU stay: 33.6% vs 51.0% (P = 0.001), and use of respiratory ventilation: 24.6% vs 39.7% (P = 0.002).
Potential associated factors for receiving palliative care
In the univariate logistic regression of factors associated with healthcare services utilized in patients in our study cohort, the likelihood of palliative patients stays in regular care ward over 14 days and over 30 days were (odds ratio [OR] = 1.95, 95% confidence interval [CI]: 1.62–2.35, P < 0.001) and OR = 1.46, 95% CI: 1.15–1.86, P = 0.002). Other than the duration of hospitalization, patients who received palliative care were associated with lower odds with regard to the utilization of aggressive medical care resources, including ICU (OR = 0.47, 95% CI: 0.38–0.58, P < 0.001), endotracheal tube procedures (OR = 0.27, 95% CI: 0.19–0.37, P < 0.001), respiratory ventilator use (OR = 0.43, 95% CI: 0.35–0.53, P < 0.001), CPR (OR = 0.25, 95% CI: 0.17–1.01, P < 0.001), and blood transfusion (OR = 0.82, 95% CI: 0.67–1.01, P < 0.001). However, in this study cohort, only ICR admissions (OR = 0.47, 95% CI: 0.38–0.58, P = 0.365) did not indicate a statistical significance (Figure 2A).
In comparison, in the subgroups who received palliative care before and after the last hospitalization, the patients who had PCS either by referral or through a family consultant before their last hospitalization exhibited a decrease in healthcare resource utilization during the length of their hospital stay >14 days (OR = 0.40, 95% CI: 0.30–0.53, P < 0.001), > 30 days (OR = 0.45, 95% CI: 0.29–0.69, P < 0.001), and respiratory ventilator use (OR = 0.57, 95% CI: 0.37–0.87, P = 0.009). However, in Taiwan, most patients still have a high likelihood of visiting the ER prior to hospital admission (OR = 1.82, 95% CI: 1.32–2.53, P < 0.001) even if they have been introduced to a PCS or consultant (Figure 2B).
Utilization of aggressive medial resource among different diseases
In a comparison between the cancer and non-cancer utilization of different medical resources, we measured the effectiveness between cancer and lung disease patients. Palliative care intervention before and after the last hospital admission was utilized as a variable. Figures 2C and 2D exhibit the palliative care intervention before and after the last hospital admission, respectively, in the cancer patient cohort. There was no statistical significance for ICU stay (OR = 0.84, 95% CI: 0.55–1.27, P = 0.403), endotracheal intubation (OR = 0.8, 95% CI: 0.35–1.71, P = 0.528), respiratory ventilator use (OR = 0.69, 95% CI: 0.43–1.11, P = 0.127), CPR (OR = 1.90, 95% CI: 0.68–5.28, P = 0.221), and blood transfusion (OR = 0.98, 95% CI: 0.72-1.32, P = 0.876).
The association of palliative care intervention before the last hospital admission in lung disease patients was also associated with a shorter length of stay in the hospital >14 days, >30 days, and in the ICU (OR = 0.41, 95% CI: 0.27–0.62, P < 0.001; OR = 0.46, 95% CI: 0.26–0.83, P = 0.009; OR = 0.62, 95% CI: 0.39–0.98, P = 0.042, respectively). Additionally, respiratory ventilator use is also decreasing compared with patients receiving palliative care after hospital admission (OR = 0.56; 95% CI: 0.35–0.91, 95% CI: 0.35–0.91, P = 0.019). The utilization of an endotracheal tube, CPR, and blood transfusion did not exhibit significant differences statistically (OR = 0.54, 95% CI: 0.21–1.34, P = 0.181; OR = 1.20, 95% CI: 0.37–3.88, P = 0.756; OR = 1.14, 95% CI: 0.74–1.76, P = 0.558, respectively).