1. Introduction and Objectives
Patients requiring specialized palliative care who exhibit significant global dysfunction, face difficulties in accessing hospital-based palliative care consultations, or do not benefit from follow-up with other specialties may be referred to community or home-based palliative care teams [
1,
2,
3,
4].
Therefore, community palliative care teams monitor patients with greater motor and cognitive limitations, who are more dependent on activities of daily living and exhibit more advanced stages of the disease, often with the imminent loss of oral intake [
1,
5,
6,
7].
Percutaneous endoscopic gastrostomy (PEG) placement is a minimally invasive procedure that involves the endoscopic placement of an artificial tube between the stomach and the abdominal wall. This is a safe and efficient method that safeguards nutritional and water supply to patients without an oral route or those who are about to lose it and who have an expected survival period of more than four weeks [
8,
9,
10].
For those with the sudden loss of the oral route, and with an expected survival period of more than four weeks, it is possible to guarantee nutritional and water supply through a nasogastric tube until the PEG is placed [
8,
9,
11].
In our clinical practice, we observe a high rate of complications associated with long-term nasogastric tubes, particularly in frail patients in need of specialized palliative care.
In patients with an expected survival period of less than four weeks, the assessment and decision of a multidisciplinary team must be individualized, although the most frequent and acceptable approach is the comfort feeding method, avoiding measures of futility and therapeutic intrusion [
9,
12,
13].
There is some controversy regarding the use of PEG in patients with dementia and severe–extreme frailty [
8,
10]. The most frequent and main indications are malignant neoplasms of the upper digestive tract, as well as those of the head and neck, along with certain neurological diseases, such as amyotrophic lateral sclerosis [
8,
10]. However, for these indications, there is a gap in knowledge, and there is also a lack of publications with data on follow-up by specialized palliative care, regarding the benefits of this approach.
In this context, our study aims to evaluate the impact of PEG placement on survival and quality of life in patients followed by a community palliative care team over the last four years. As secondary objectives, we intend to evaluate the main diseases and clinical reasons that motivate the request for PEG and an assessment using the Palliative Performance Scale (PPS) before PEG placement. We also seek to determine whether the socioeconomic conditions have any influence on the PEG request.
2. Methods
Study design: This was an observational, non-interventional, retrospective cohort study. The STROBE guidelines (Strengthening the Reporting of Observational Studies in Epidemiology) were followed [
14].
Selection of participants: All 1000 patients followed by a community palliative care support team in the northern region of Portugal between March 2020 and December 2023 were considered. Of these patients, 54 underwent PEG. We also considered a control group of 20 patients that had a nasogastric tube for more than 4 weeks.
Ethical approval: This study was approved by the Local Ethics Committee (reference CE/2023/34), and the Helsinki Declaration was respected. The European General Data Protection Act was respected. All data collected were recorded and saved electronically; identifiable data were anonymized and protected by a password, to which only the researchers had access.
Measures
Data collection: Data were obtained through the electronic clinical process and the physical clinical files in March–April 2024. The information extracted included sex, age, diagnosis, referring unit, team follow-up starting date and number of days of follow-up, life expectancy at the start date of follow-up by the team (before or after to six weeks), and symptom control and quality of life assessment (with the application of the Edmonton scale [
15] and the Personal Outcomes Scale [
16], respectively). Additional information was collected regarding the presence of a caregiver and their characteristics, the level of functionality upon admission to the team (assessed using PPSv2 [
17]), the date and reason for the PEG request, the team responsible for the request (community palliative care team or previous request), and the PEG placement date. Finally, data about the place and date of death were also collected.
Social support: The existence of basic social support was measured by evaluating whether patients had a multipurpose certificate and dependency supplement, which are the main support instruments in the Portuguese state for especially vulnerable and dependent patients.
Data analysis: Statistical analysis involved descriptive statistics measures (absolute and relative frequencies, means, and respective standard deviations) and inferential statistics. In this analysis, the chi-squared test of independence, the Fisher test, Kaplan–Meier survival analysis, Cox regression, Student’s t-test for independent samples, the Mann–Whitney test, and the Kruskal–Wallis test were used. The chi-squared assumption that there should be no more than 20% of cells with expected frequencies of less than five was analyzed. When this assumption was not satisfied, the chi-squared test was utilized, using Monte Carlo simulations. Differences were analyzed with the support of standardized adjusted residuals. The normality of distribution was analyzed using the skewness and kurtosis of the graphical representation. The significance level to reject the null hypothesis was set at α ≤ 0.05. The statistical analysis was carried out using the Statistical Package for the Social Sciences (SPSS) software version 28 for Windows.
3. Results
The mortality rate during follow-up by the community palliative care team was 74.1% (n = 40) and the age at admission to the team was not a predictive factor of the survival time (
p = 0.397). The data analysis allowed us to verify that the survival time was significantly higher in women (
p = 0.003), as shown in
Figure 1, i.e., while males had an average survival period after PEG placement of around 9.8 months, females had an average survival period of 21.4 months.
We observed that six patients (11%) experienced minor complications after PEG placement and no patients had major complications. None of the patients who had PEG discontinued its use until the end of this study or at the end of their lives. Among the six patients with minor complications, such as mild bleeding (n = 3), soft tissue infection (n = 2), and removal of the PEG (probably related to an episode of delirium, n = 1), there were no significant differences in the survival time. In patients with mild bleeding, only local containment measures were necessary, without additional pharmacological therapy. In patients with soft tissue infections, antibiotic therapy was introduced. Four patients had their complications resolved within one to two weeks, while two patients died during this period, but their deaths were clinically unrelated to these complications (one with externalization of the PEG and one with a mild soft tissue infection).
Considering the 22 patients that had a nasogastric tube inserted before PEG placement (for 7 to 20 days), we observed a significant reduction in respiratory secretions (91%), signs of discomfort (59%), and delirium (22.7%), with a significant reduction in the dose and number of neuroleptics.
Regarding the 20 patients (10 males and 10 females) who had nasogastric tubes when they lost their oral route and who were not subjected to PEG placement, we observed a survival rate of 1.7 months (no significant statistical differences between males and females). We observed that all patients (100%) had minor complications after nasogastric tube placement during this period, such as dyspnea associated with respiratory secretions (85%), persistent cough (85%), pain and signs of discomfort (60%), and major complications such as delirium (60%) that led to the self-removal of the tube (30%) and pneumonia (25%).
No correlation was observed between the patient’s age and the reason/diagnosis for which PEG was requested (p = 0.874).
Finally, the relationship between disease and death was not statistically significant (χ
2 (3) = 6.485,
p = 0.087), as shown in
Table 1. It is worth highlighting the fact that all frail patients died during the period of follow-up by the team.
There was a significantly larger proportion of patients with dementia (χ
2 (3) = 12.647,
p = 0.007), for whom the reason for the PEG request was partial or total dysphagia. Regarding patients with oncological disease, the main reason for the PEG request was the anticipation of possible dysphagia, and the same was observed for patients with neurological diseases. For patients with frailty, the request was mainly due to established dysphagia, as seen in
Table 2.
The relationship between the pathology and place of death was not statistically significant (p = 0.479). However, the age at death was significantly higher in those with frailty syndrome (χ2 KW (3) = 10.372, p = 0.016).
In relation to other comorbidities, we considered each patient’s personal medical history and usual medication, but there was no possibility of an adequate statistical study given the very small numbers for each of these factors. The most relevant variables that we found were high blood pressure and diabetes, but, even so, they did not present any statistical interest in this study.
The age at PEG placement (χ2 KW (3) = 9.583, p = 0.022) and the age at the start of monitoring by the team (χ2 KW (3) = 9.673, p = 0.022) were significantly higher in those with frailty syndrome.
The difference in the survival time between pathologies, after PEG placement, was not statistically significant (
p = 0.422), as shown in
Figure 2.
The correlations between the results obtained by applying PPS and the age (p = 0.997), age at the beginning of follow-up by the team (p = 0.999), age at PEG placement (p = 0.996), age at death (p = 0.997), place of death (p = 0.116), and follow-up time (p = 0.092) were not significant. Moreover, the difference in the PPS values depending on the pathology was not statistically significant (p = 0.486).
The PPS values were significantly higher when PEG placement was motivated by prevention (p = 0.028) and in patients who did not die (p = 0.002). However, the PPS values did not show significant predictive value for the survival time (p = 0.061).
Survival was significantly higher in patients with adequate social support (
p = 0.028), as shown in
Figure 3. There were no other correlations with social support, such as gender, the reason for requesting PEG, the place of death, the main pathology, PPS, the follow-up time, the age at which the team started follow-up, and the age at which the PEG was placed.
The correlation between the follow-up time and survival time after PEG placement was statistically significant, positive and moderate (r = 0.457, p < 0.001). Thus, the longer the follow-up time, the longer the survival time after PEG placement. The overall survival time in patients with PEG had a median of 150 days (interquartile range [42–333]).
There was a significantly larger proportion of deaths when the PEG was requested by others compared to when it was requested by the team involved in the study (p = 0.002). Of the 23 patients who had PEG requested by others, 22 died (95.7%) during the follow-up period. In contrast, of the 31 patients who had PEG placement requested by this team, 18 died (58.1%).
Regarding the survival time after PEG placement, despite not showing statistically significant differences, patients for whom the team requested the PEG had longer follow-up times (166 days vs. 81 days, p = 0.106) and longer survival times (563 days vs. 350 days, p = 0.237).
This study found no significant differences between the various parameters analyzed and the place of death.
4. Discussion
PEG is used as the preferred route of nutritional support for patients with long-term enteral nutrition needs. For patients in palliative care with the actual or predictable loss of the oral route, and with an expected survival time of more than four weeks, it is necessary to guarantee nutritional and fluid supply, with PEG placement being a good possible option.
In this study, the data suggested that women who underwent PEG placement had a longer survival time compared to men, with males having an average survival period of 9.8 months, while females had an average survival period of 21.4 months. This situation does not appear to be related to major complications from PEG placement or a greater disease burden. It may be related to the fact that women’s average life expectancy is higher. However, no consistent data allowed us to assess this situation.
The survival rate seems to be in line with a study carried out by Akkuzu and colleagues in 2021, in which they reported an average life span after PEG insertion of 221.3 ± 330.7 days; half of the patients lived for less than six months (180 days) and about 28% of the patients lived for less than 30 days [
18]. In our study, if we also used the average as a statistical measure, we observed an overall survival time in patients with PEG of 463 ± 72 days.
Another study by Stenberg et al. in 2022 stated that mortality was around 15% at 30 days and 28% at 90 days [
19]. The authors noted that the main causes of death were malignancy and aspiration pneumonia. Additionally, they concluded that each year with PEG seemed to increase the likelihood of death by 30 days, and, differently from our study, women tended to have a higher mortality rate than men [
19]. In our study, the median survival time for women was 490 days and that for men was 85 days.
It should be noted that the above two studies were carried out in a hospital setting and without specific monitoring by a palliative care team, and our study was restricted to a community team specialized in palliative care and patients in the home context.
Stenberg et al. (2022) found some factors that may affect the survival rate, but without directly conditioning it, such as advanced age, diabetes, heart failure, C-reactive protein levels, and a very high or very low body mass index. These factors may influence the final result of the procedure, but without an impact on post-placement survival [
19,
20]. In our study, regarding the age at which the PEG was placed, there was no statistical significance regarding the notion that age directly affects survival or complications inherent to the placement of the device.
The mortality rate observed was 74.1%, but it was not possible to establish a relationship with the placement of the PEG, since the specificity of the patients monitored by the team showed that the patients’ baseline status was already compromised and all patients had a life expectancy of less than 12 months. Some studies show that the mortality rate is higher in patients with oncological pathologies compared with neurodegenerative diseases and organ failure [
18,
19,
21,
22]. Sobani et al. (2018) conducted a study with patients over 100 years and reported a mortality rate that was about 9% below what is found in the literature [
21]. In their studies, Akkuzu et al. (2021) and Stenberg et al. (2022) revealed higher mortality rates in patients with oncological diseases. However, this was not the case in the present study, since no statistical significance was found between the pathology and survival time after PEG placement or with the PPS index [
19,
21,
22,
23]. We consider that this assessment should be considered in a prospective study, with a larger sample (as the sample size may have led to the lack of statistical significance in the present study), and differentiate between the different oncological diseases, which do not have identical behaviors in terms of the increasing dysfunctionality that they present, as well as in other types of diseases, such as neurodegenerative diseases.
The main reasons for the placement of the PEG were dysphagia, as already seen in the case of patients with dementia, and predictable dysphagia in patients with an oncological pathology. These data are in congruence with the literature. Sobani et al. (2018) reported that most patients with an indication for PEG had dysphagia and a risk of aspiration [
21].
The European Society for Clinical Nutrition and Metabolism (ESPEN) does not recommend the placement of PEG in patients with a short life expectancy (30 days), regardless of the underlying pathology [
24]. The European Society of Gastrointestinal Endoscopy (ESGE) corroborates this indication [
25]. However, both recommend the individualization of the decision-making process and the suitability for each patient. They also mention that the date of placement can be a relevant factor in preventing abrupt weight loss and reducing the catabolic process common in dementia and oncological diseases [
25].
The ESGE also states that PEG placement can play an important role in some degenerative and oncological diseases that induce marked weight loss, even with adjusted oral intake [
20,
21,
23,
25,
26].
The age at which the PEG was placed was slightly higher in the elderly with frailty, just as the age at which the team started follow-up was higher in patients with frailty compared to patients with other pathologies. This fact is also corroborated by other studies that confirm the difficulty and ambivalence in the concept of frailty. It is a concept with multiple interpretations, making it difficult to assess and delaying referral to palliative care teams. Loureiro and Carvalho (2021) summarize frailty as a clinical condition related to sarcopenia, neuroendocrine dysregulation, and immune system dysfunction, with the phenotype consisting of certain characteristics, such as unintentional weight loss, self-reported fatigue and/or exhaustion, decreased muscle strength, a slow gait speed, and low level of physical activity [
27]. The concomitant presence of three or more of the aforementioned conditions defines the concept of frailty [
27].
The late recognition of this condition or the devaluation of the reports of patients and family members leads to late referrals, which partly explains the age of referral to the team and the request for PEG placement. However, monitoring by differentiated teams and increased awareness of patients’ frailty seems to enhance the survival time, and social and family support are strong contributors to prolonged survival.
There was moderate and positive statistical significance between the time of PEG placement, the time of follow-up by the team, and the survival time. It was observed that the longer the follow-up time, the longer the survival time. These data corroborate the empirical perception that monitoring by palliative care teams tends to increase the survival time. Some studies carried out in patients with specific pathologies, such as amyotrophic lateral sclerosis, show that follow-up and post-PEG placement monitoring increases patient survival [
18,
22,
26,
27]. Gaspar et al. (2021), Vieira et al. (2017), and Stenberg et al. (2022) show, although without specifying the types of patients, that patients with higher levels of follow-up have higher survival rates [
19,
22,
28].
The present study found that the proportion of deaths was larger in patients who were referred to the team with the PEG already placed than those in whom the team requested PEG. This factor can be explained by late referrals and the patients’ greater symptomatic and disease burdens at the time of entry into the team.
The mortality rate for patients with PEG prior to referral to the team was 95.7%, while patients with the PEG placed at the request of the team and with follow-up by the team exhibited a rate of 58.1%. This difference can be explained by the previous evaluation of the patient and family and by the window of opportunity at the time of PEG placement. Close and consistent team monitoring can facilitate the early detection of complications and mitigate their evolution [
20,
21,
26].
Concerning the place of death, there does not seem to be any valuable relationship between the fact that the patient has a PEG and the place of death.
Considering the control groups (patients with a nasogastric tube until death and patients with a nasogastric tube until the PEG was placed), we observed a significant increase in symptom control and quality of life with PEG placement, and, inversely, we observed earlier death with a nasogastric tube and worse outcomes related to the patient’s quality of life, with more symptoms and associated suffering.
We can conclude that timely referrals and robust family and social support increase the survival of patients with PEG. The timing of PEG placement is important, but there are other factors as well. Monitoring by differentiated and specialized teams seems to contribute to an increase in the survival rate [
1,
3,
27]. The primary reason for the placement of PEG is dysphagia or the risk of it in oncological diseases [
12,
13,
18,
19,
21]. The present study shows that most PEG placements were in patients with dementia, followed by patients with oncological pathologies.
We can conclude that the proximity of care after PEG placement seems particularly relevant to the survival time after the procedure. Thus, the timing of referral to differentiated teams and consistent and close patient assessments can be predictors of adjusted times for PEG placement.
5. Limitations and Future Perspectives
First, this study’s observational nature may have introduced confounding factors that could not be fully controlled, potentially impacting the results. Potential selection bias could have occurred, as the study relied on participants already receiving care in specific settings. Second, the sample size was relatively small, limiting the generalization of the findings to broader populations. We need more and multicentric studies, with larger samples. Additionally, data collection was restricted to a single center, which may not fully represent the diversity of the patient population.
In conclusion, this study highlights the importance of PEG in specialized palliative care. The findings suggest a need for further exploration in this area. Future research should focus on larger, multicenter trials to confirm these results and explore the possible outcomes. Additionally, integrating methodologies like prospective registries can offer more precise predictions and treatment pathways. Expanding the research to include more diverse populations and longer follow-up periods will also be crucial in fully understanding the implications of PEG on clinical practice and patient outcomes.
6. Conclusions
The expected survival time for these patients would be less than four to six weeks if they had not undergone PEG (as presented by the control group who had a nasogastric tube until death), while, with its placement, we observed an average survival period of 9.8 months in males and 21.4 months in females.
However, these patients could have had a nasogastric tube inserted, but the evidence reinforces what we observe in our daily practice: in the long term, the nasogastric tube presents clear disadvantages compared to PEG, due to the more frequent complications that affect patients’ survival and quality of life. In this study, we only observed 11% of patients with mild complications, which were resolved within one to two weeks. On the other hand, all patients with a nasogastric tube had minor complications and several patients (25% to 60%) had major complications.
The main diseases and clinical reasons that motivated PEG placement were dementia and oncological diseases, particularly in those patients with minor or severe dysphagia, and the PPS were significantly higher in patients who received PEG for prevention. Although we did not identify a cutoff for PPS, all patients who received PEG in anticipation of potential dysphagia for fluids had PPS greater than 50%, while the remainder with PEG had PPS of less than 40%.
We found no correlations between our variables and the place of death, but we identified that basic social support (which we considered to be a multipurpose certificate and the dependency complement, which are additional supports granted by the Portuguese state) is correlated with a greater survival time.