Implementation of Polish guidelines on futile therapy protocols in pediatric intensive care units: a multicenter retrospective review

Abstract

The futile therapy protocol (FTP) is a formal document supporting decisions to withdraw or withhold certain forms of treatment owing to their futility, with the only motivation being the good of the patient, considered individually for each patient. We conducted a multicenter retrospective review of consecutive FTP cases initiated in six Polish pediatric intensive care units (PICUs) between August 2021 and June 2025. Of 227 identified cases, 25 were excluded due to substantial missing data (> 50% missing variables) or because the protocol was ultimately not implemented. Finally, 202 patients were included in the analysis. The primary outcomes were in-hospital status (death vs. discharge alive) and continuation of mechanical ventilation after futile therapy protocol (FTP) implementation (1 = continued; 0 = not continued / palliative or spontaneous breathing). Discharge destination (hospice care or Pediatric Home Mechanical Ventilation Program) was recorded when available. The median age was 24 months, and 42.8% of patients were ≤ 1 year of age, including > 70% in the first month of life. Neurological (54.3%) and oncological diseases (16.6%) were the most frequent underlying diagnoses documented in the medical records. Median time from FTP completion to death was 6 days among 60 in-hospital deaths with complete date information. Overall, 40.6% of patients were discharged alive, either to hospice care or to the Pediatric Home Mechanical Ventilation Program. Compared with patients discharged alive, those who died in hospital were younger (p = 0.005) and more likely to have chronic diseases (p < 0.001). In exploratory multivariable logistic regression, younger age and non-neurological underlying disease were independently associated with continuation of mechanical ventilation after FTP implementation. The guidelines implemented in the participating centers provided a structured framework for decision-making regarding the discontinuation of futile therapies. FTPs were most frequently initiated in children under 1 year of age with severe underlying chronic disease. A substantial proportion of patients (40%) were discharged alive to hospice care or a home mechanical ventilation program. This highlights the role of organized palliative pathways in this setting.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-50207-9.

Introduction

The refusal to accept the failure of therapy and the death of the patient leads to the progressive process of medicalization of death. Social and cultural changes and advances in medicine have resulted in the fact that the most common place to end life in Poland is the intensive care unit. Unfortunately, this problem also, and perhaps especially, concerns pediatric patients. The death of a child is always shocking and devastating. Curing a sick child has become one of the most important challenges in medicine, and the doctor no longer works for the good of the patient but fights the disease, very often in vain. In many cases, parents are unable to come to terms with the death of their child, and some of them unknowingly make the dying child suffer by exerting various pressures on the medical team, who is convinced that they are doing it for the child’s sake¹. Futile therapy (FT) refers to treatment that maintains organ function without providing any benefits to the patient and without supporting the assumed therapeutic goals. FT which prolongs the process of dying is associated with the suffering of patients and their families and may constitute a violation of human dignity². Medical futility can be categorized into three distinct types³:

1. Physiological Futility this intervention is physiologically ineffective, as it fails to produce the desired biological response, it represents the most objective form of medical futility;

2. Quantitative Futility the treatment is deemed quantitatively futile, as the chances of a positive (result are negligible;

3. Qualitative Futility the treatment may produce a physiological effect, but the resulting quality of life is considered extremely poor; it focuses on whether the outcome is a “benefit” rather than just a “physiological effect”.

The complexity of this concept is further influenced by the patient’s personal worldview and their subjective perception of quality of life. The guidelines are specifically designed to move away from discretionary interpretations of medical futility. Their first iteration was drafted in 2011 by a group of pediatric hospice experts who highlighted the urgent need for protocols regarding the withdrawing or withholding of futile therapy in children⁴. However, this document failed to gain traction within the medical community and remained largely ignored by the public. A decade later, a dedicated Working Group of specialists in pediatric anesthesiology and intensive care, under the auspices of the Polish Society of Anesthesiology and Intensive Therapy, developed comprehensive guidelines¹. This initiative led to the creation of the Futile Therapy Protocol (FTP) - a unified national framework for withdrawing or withholding interventions deemed medically futile⁵. Before the protocol is implemented, specialist consultations related to the underlying condition as well as multidisciplinary consultations must be carried out. The decision to withdraw futile therapy should be made by at least two anesthesiology and intensive care specialists in consultation with a doctor and/or a team of other specialists involved in the therapeutic process, in consultation with the head of the ward or a person authorized and consistently followed by the entire medical team. The therapeutic procedures that will be withheld or withdrawn in PICUs include: Cardiopulmonary resuscitation, Mechanical respiratory support, Pharmacological and electrical support of the cardiovascular system, Antibiotic therapy, Parenteral nutrition, Transfusion of blood products, Surgery and other invasive procedures, Mechanical circulatory support, Extracorporeal respiratory support, e.g. ECMO, Renal replacement therapy, Extracorporeal liver support.

Our study aimed to describe the implementation of FTP over a four-year period using documentation from six major clinical centers. Specifically, we sought to characterize the age and disease groups in which FTPs were initiated, summarize which treatments were discontinued, report in-hospital outcomes, and explore factors associated with discharge alive (vs. in-hospital death). This study is a continuation of an earlier pilot project involving three centers⁶.

Methods

Study design and setting

This retrospective, multicenter observational study analyzed pediatric patients in whom a futile therapy protocol (FTP) was initiated in tertiary pediatric intensive care units (PICUs) in Poland. Data were collected at six participating centers located in Krakow, Lublin, Katowice, Poznan, Wroclaw, and Warsaw, during the period from August 2021 to June 2025. The participating PICUs were:

1. University Children’s Hospital of Krakow, Department of Anaesthesiology and Intensive Therapy (Krakow);

2. Medical University of Lublin, Department of Anesthesiology and Pediatric Intensive Care (Lublin);

3. Upper Silesian Child Health Centre, Medical University of Silesia, Department of Anesthesiology and Intensive Care (Katowice);

4. Poznan University of Medical Sciences, Department of Pediatric Anesthesiology and Intensive Care (Poznan);

5. Wroclaw Medical University, Department of Pediatric Anesthesiology and Intensive Care (Wroclaw);

6. Medical University of Warsaw, Department of Paediatric Anaesthesiology and Intensive Therapy (Warsaw).

Study population (inclusion and exclusion criteria)

We identified all consecutive cases in which a futile therapy protocol (FTP) was initiated in the participating PICUs during the study period (n = 227). A case was considered ‘futile’ when initiation of the futile therapy protocol (FTP) was explicitly documented in the medical record according to the guideline; the decision was made by the treating team, not by the investigators. No cases were selected based on age, diagnosis, or outcome. We excluded 25 cases due to insufficient data (> 50% missing variables) or because the FTP was ultimately not implemented (the child died before a final decision was made or the process was withdrawn due to strong parental opposition). The final analysis set included 202 patients. For variables with occasional missingness in the included cohort, analyses were performed using available data for the respective comparison (complete-case analysis per variable). A flow diagram illustrating patient identification and exclusions is provided in Fig. 1.

Fig. 1.

Flow diagram of cohort identification and selection across six participating pediatric intensive care units (PICUs).

Data collection

Data were extracted retrospectively from medical records at each participating center. Collected data included demographic variables (age, sex), clinical setting (PICU/department completing the FTP), primary diagnosis, comorbid conditions. Pre-hospitalization chronic illness was coded as a binary variable based on the medical history available in the records: “chronic illness absent” was defined as no documented chronic condition prior to admission, whereas “chronic illness present” required explicit documentation of a chronic condition. Management following the FTP was recorded, including whether life-sustaining treatment was continued (e.g., ongoing invasive mechanical ventilation) or whether a palliative approach was implemented. Data were entered into a predefined Excel spreadsheet and subsequently exported for statistical analysis.

Outcomes

The primary outcomes were in-hospital status (death vs. discharge alive) and continuation of mechanical ventilation after futile therapy protocol (FTP) implementation (1 = continued; 0 = not continued / palliative or spontaneous breathing). Among patients who died during hospitalization, we additionally described the time from FTP to death (days), which was calculated only when both dates were available in the medical record. Discharge destination (hospice care or Pediatric Home Mechanical Ventilation Program) was recorded when available. No post-discharge (out-of-hospital) outcomes were collected.

Statistical analysis

We used descriptive statistics to summarize the study population. Categorical variables are presented as counts and percentages (n, %), and continuous variables are presented as medians with interquartile ranges (IQRs). Fisher’s exact test was used to compare categorical variables. Monte Carlo simulation was applied for contingency tables larger than 2 × 2 (B = 10,000). For continuous variables, we used the Wilcoxon rank-sum test (for two groups) or the Kruskal–Wallis test (for more than two groups). We also performed subgroup analyses limited to in-hospital deaths occurring more than four days after FTP completion. Time from FTP completion to death was summarized descriptively and compared across groups only among in-hospital deaths with complete date information. In addition to descriptive analyses, we performed an exploratory multivariable logistic regression to identify factors associated with continuation of mechanical ventilation after futile therapy protocol (FTP) implementation (1 = continued; 0 = not continued / palliative or spontaneous breathing). Age was included as a continuous predictor (years). Underlying disease was modeled in the primary analysis using three clinically motivated categories (neurological, oncological, other) to ensure stable estimates, while center was included as a fixed effect to account for inter-center practice differences. Results are reported as odds ratios (OR) with 95% confidence intervals (CI). As a sensitivity analysis, we refitted the model using the original five disease categories; estimates for less frequent categories were imprecise (wide CIs), therefore this specification was not used as the primary model. Diagnostic checks supported the logistic regression specification: there was no evidence of non-linearity for age (LRT p = 0.285), no multicollinearity (GVIF_adj ≤ 1.08), and calibration was acceptable (Hosmer–Lemeshow p = 0.46; supportive). Screening for influential observations did not reveal major concerns. Missing data were handled using complete-case analysis per variable; no imputation was performed. All statistical tests were two-sided, and a p-value < 0.05 was considered statistically significant. Analyses were conducted using R, version 4.3.3 (R Foundation for Statistical Computing, Vienna, Austria).

Results

This study included 202 patients treated at 6 centers in Poland (Table 1). The median age was 24 months (IQR: 1–120); 42.8% were children up to 1 year of age, including over 70% in the first month of life (Fig. 2); and male patients predominated (58.9%). As many as 77.7% of patients had documented chronic illness from birth, which coincided with the observed impaired functioning in life. Neurological (54.3%) and oncological conditions (16.6%) were the most frequent underlying (primary) diagnoses documented in the medical records. Almost half of the patients (42.7%) had multiple organ failure. The suspicion of futile therapy was most often raised by anesthesiologists/intensivists (58.9%).

Table 1.

Characteristics of the study group.

Variables	N	Total N = 2021	Sex		p-value2
			Males N = 1191	Females N = 831
Hospital location (city)	202				0.8
Krakow		103 (51.0%)	59 (49.6%)	44 (53.0%)
Lublin		16 (7.9%)	11 (9.2%)	5 (6.0%)
Katowice		15 (7.4%)	8 (6.7%)	7 (8.4%)
Poznan		10 (5.0%)	6 (5.0%)	4 (4.8%)
Wroclaw		15 (7.4%)	11 (9.2%)	4 (4.8%)
Warsaw		43 (21.3%)	24 (20.2%)	19 (22.9%)
Age (months), median (IQR)	201	24 (1–120)	24 (1–132)	24 (1–120)	0.4
Missing data		1	1	0
Primary disease	199				0.9
Neurological disease		108 (54.3%)	66 (56.9%)	42 (50.6%)
Oncological disease		33 (16.6%)	17 (14.7%)	16 (19.3%)
Prematurity		25 (12.6%)	14 (12.1%)	11 (13.3%)
Severe heart defect		23 (11.6%)	13 (11.2%)	10 (12.0%)
Trauma/cardiac arrest		10 (5.0%)	6 (5.2%)	4 (4.8%)
Missing data		3	3	0
Comorbid condition	185				0.3
Severe neurological damage		69 (37.3%)	40 (36.0%)	29 (39.2%)
Multiorgan failure		79 (42.7%)	44 (39.6%)	35 (47.3%)
Chronic respiratory failure		20 (10.8%)	14 (12.6%)	6 (8.1%)
Heart failure		17 (9.2%)	13 (11.7%)	4 (5.4%)
Missing data		17	8	9
Duration of the primary disease	202				0.6
Since birth		157 (77.7%)	96 (80.7%)	61 (73.5%)
One year		15 (7.4%)	9 (7.6%)	6 (7.2%)
6 months		7 (3.5%)	3 (2.5%)	4 (4.8%)
More than 5 years		10 (5.0%)	4 (3.4%)	6 (7.2%)
Sudden onset		13 (6.4%)	7 (5.9%)	6 (7.2%)
Did the child previously have a life without disease (normal life)?	202				0.7
No		156 (77.2%)	93 (78.2%)	63 (75.9%)
Yes		46 (22.8%)	26 (21.8%)	20 (24.1%)
Actions taken after the FT protocol	201				0.9
Continued mechanical ventilation		99 (49.3%)	58 (48.7%)	41 (50.0%)
Continued spontaneous breathing		102 (50.7%)	61 (51.3%)	41 (50.0%)
Missing data		1	0	1
Patient outcome	202				0.5
Discharged to hospice		82 (40.6%)	51 (42.9%)	31 (37.3%)
Died in hospital		120 (59.4%)	68 (57.1%)	52 (62.7%)
Days from FT protocol to death, median (IQR)	60	6 (3–15)	7 (2–17)	6 (3–11)	0.6
Missing data		142	89	53
Who first suspected FT (grouped)	202				> 0.9
Anesthesiologist		119 (58.9%)	71 (59.7%)	48 (57.8%)
Pediatrician		26 (12.9%)	16 (13.4%)	10 (12.0%)
Other specialist		16 (7.9%)	10 (8.4%)	6 (7.2%)
Neonatologist		41 (20.3%)	22 (18.5%)	19 (22.9%)

IQR interquartile range, FTP futile therapy protocol, FT futile therapy.

¹Percentages may not sum to 100% due to rounding.

²Wilcoxon rank sum test; Fisher’s exact test. Monte Carlo simulation was applied for contingency tables larger than 2 × 2 (B = 10,000).

Fig. 2.

Age distribution of the included patients (n = 202). The main panel shows the number of patients by age in completed years, and the inset provides a detailed breakdown for infants aged 0–11 months. Bars are labeled with counts and percentages.

After discontinuation of futile therapy, mechanical ventilation was continued in almost half of the patients, whereas the rest breathed spontaneously; this differed significantly by underlying diagnosis (Table 2; p < 0.001). Children with neurological conditions constituted the majority of those managed with continued spontaneous breathing (74.7%), whereas the group with continued mechanical ventilation included relatively more patients with prematurity (20.2%), oncological disease (20.2%), severe heart defects (16.2%), and trauma/cardiac arrest (9.1). In contrast, recorded comorbidities were not significantly associated with respiratory management (p = 0.147).

Table 2.

Respiratory management after FTP according to primary diagnosis and comorbidities.

Primary disease/Outcome	Total N = 198	Continued mechanical ventilation N = 99	Continued spontaneous breathing N = 99
Neurological	108 (54.5%)	34 (34.3%)	74 (74.7%)
Oncological	32 (16.2%)	20 (20.2%)	12 (12.1%)
Prematurity	25 (12.6%)	20 (20.2%)	5 (5.1%)
Severe heart defect	23 (11.6%)	16 (16.2%)	7 (7.1%)
Trauma/Cardiac arrest	10 (5.1%)	9 (9.1%)	1 (1.0%)
p < 0.001 (Fisher’s exact test with Monte Carlo simulation) Comorbidity/Outcome	Total N = 184	Continued mechanical ventilation N = 92	Continued spontaneous breathing N = 92
Severe neurological damage	69 (37.5%)	27 (29.3%)	42 (45.7%)
Multiorgan failure	78 (42.4%)	44 (47.8%)	34 (37.0%)
Chronic respiratory failure	20 (10.9%)	11 (12.0%)	9 (9.8%)
Heart failure	17 (9.2%)	10 (10.9%)	7 (7.6%)

p = 0.147 (Fisher’s exact test with Monte Carlo simulation).

Values are n (% within each respiratory management column). Denominators differ due to missing data (primary diagnosis: N = 198; comorbidity: N = 184). P-values from Fisher’s exact test with Monte Carlo simulation.

In the multivariable logistic regression model adjusted for center and underlying disease group, older age was associated with lower odds of continued mechanical ventilation after FTP implementation (OR per 1-year increase 0.93, 95% CI 0.87–0.99; p = 0.031). Compared with neurological conditions (reference category), the “other” underlying disease group was strongly associated with continuation of mechanical ventilation (OR 9.18, 95% CI 3.79–22.24; p < 0.001). The association for the oncological group was not statistically significant and the estimate was imprecise (OR 2.51, 95% CI 0.90–7.04; p = 0.079). Substantial inter-center differences were observed (see Table S1).

The median time from FTP completion to death was 6 days, calculated among 60 in-hospital deaths with complete date information available in the medical records. As many as 40.6% of patients were discharged either to hospice care or Pediatric Home Mechanical Ventilation Program. Patients who died in intensive care units were statistically younger (p = 0.005) and more likely to have chronic diseases (p < 0.001) than those discharged to hospice care (Tables 3 and 4). Among in-hospital deaths with complete date information, the time from FTP completion to death did not differ significantly across underlying diagnostic categories (Table 5).

Table 3.

Patient outcomes vs. primary disease.

Outcome/Primary disease	Total N = 199	Neurological N = 108	Oncological N = 33	Prematurity N = 25	Severe heart defect N = 23	Trauma/Cardiac arrest N = 10
Discharged alive (hospice or Pediatric Home Mechanical Ventilation Program)	80 (40.2%)	61 (56.5%)	8 (24.2%)	5 (20.0%)	6 (26.1%)	0 (0.0%)
Died in hospital	119 (59.8%)	47 (43.5%)	25 (75.8%)	20 (80.0%)	17 (73.9%)	10 (100.0%)

p < 0.001 (Fisher’s exact test with Monte Carlo simulation).

Post-discharge outcomes were not available; mortality after discharge (including at home) could not be assessed.

Table 4.

Patient outcomes vs. disease duration.

Outcome/Duration	Total N = 202	Since birth N = 157	For a year N = 15	Half a year N = 7	More than 5 years N = 10	Sudden onset N = 13
Discharged alive (hospice or Pediatric Home Mechanical Ventilation Program)	82 (40.6%)	73 (46.5%)	5 (33.3%)	2 (28.6%)	1 (10.0%)	1 (7.7%)
Died in hospital	120 (59.4%)	84 (53.5%)	10 (66.7%)	5 (71.4%)	9 (90.0%)	12 (92.3%)

p = 0.011 (Fisher’s exact test with Monte Carlo simulation).

Post-discharge outcomes were not available; mortality after discharge (including at home) could not be assessed.

Table 5.

Time from FTP to death by primary disease (> 4 days).

Time from FTP/Primary disease	Total N = 34	Neurological N = 15	Oncological N = 4	Prematurity N = 4	Severe heart defect N = 7	Trauma/Cardiac arrest N = 4
Time from FTP to death (days)	11 (7.25–20.5)	15 (10–21)	13 (7.75–21)	6 (6–6.25)	11 (9–16)	11.5 (5–24.75)

p = 0.119 (Kruskal‒Wallis test).

Discussion

Over the following years, progress and access to advanced life-saving procedures in intensive care units significantly intensified the problem of decision-making and rationalized the use of all possible measures for each patient. Withdrawal of life-sustaining treatment (WLST) for some seriously ill children was first described in the 1970s⁷ and has manifested in different ways related to multiple factors, including physician, patient, culture, geography, economics, laws, religion, religiosity, and ethics education⁸. In Poland, discussions on this topic began several years ago, and there is still a need for clear policies and guidance to support transparent decision-making. The recommendations for the discontinuation of persistent life-sustaining treatment (LST) were published by specialist scientific societies^1,9–11. On the basis of the position of the appointed team of experts, the Patient Ombudsman issued a publication on the implementation of the patient’s right to a dignified death, containing standards for end-of-life medical therapies¹². In this context, the Ombudsman acts as a bridge between bioethical principles and clinical practice. In the issued statement, the Patient Ombudsman highlighted the issue of futile therapy and the need for legal regulation to support doctors in shifting the focus from prolonging life at all costs to palliative and hospice care. Our study revealed that most of the analyzed protocols were related to neonates. Decisions made during this early clinical stage are critical for planning ongoing care (pediatric advance care plans - pACP). Integrating pACP is crucial in this process, guiding the withdrawal of LST to prioritize holistic, symptom-focused care for the patient and family. Perinatal palliative care (PPC) is an increasingly important option for pregnant women carrying fetuses with life-limiting conditions who choose to continue their pregnancies¹³. As a multidisciplinary and holistic approach to care, PPC aims to minimize suffering and enhance quality of life, respecting the family’s values and preferences¹⁴. An increasing number of these deaths follow a decision to withdraw or limit LST, highlighting the importance of a clear, coordinated approach to palliative care in neonates.

Half of our patients were not intubated, or it was decided that mechanical ventilation would be a futile therapy, or non-invasive ventilation was used, some patients maintained spontaneous respiration. The withdrawal of mechanical ventilation in our country is an exceptionally difficult decision, as it is commonly viewed in Poland as tantamount to euthanasia. The only exception is the declaration of brain death which is strictly regulated by law. Under the current Polish legal framework and socio-cultural lens, clinicians often consider forgoing the initiation of mechanical ventilation a safer course of action than performing palliative extubation, fearing severe legal repercussions, including potential imprisonment. The use of mechanical ventilation is usually associated with multiorgan failure and the inability to transfer the patient to hospice care. The decision not to intubate the patient appears crucial in the Polish clinical context in relation to the determination of futile therapy. This decision is often tantamount to refusing admission to the intensive care unit and allowing the patient to die naturally while providing palliative care. Withdrawing catecholamines, blood transfusions, renal replacement therapy, or withholding intubation and mechanical ventilation may help to avoid prolonging death. Importantly, respiratory management after FTP differed by the underlying diagnosis in a between-group comparison. Neurological conditions predominated among patients managed with continued spontaneous breathing, whereas continued mechanical ventilation was relatively more frequent in patients with prematurity, oncological disease, severe heart defects, and trauma/cardiac arrest. This likely reflects differences in baseline respiratory dependence and the feasibility of maintaining spontaneous breathing in specific diagnostic groups in routine clinical practice. Most of the analyzed FT discontinuation protocols involved patients who had underlying chronic diseases, which is consistent with data from other countries^8,15,16. Several studies have shown that children who die after the decision to withdraw or limit treatment are more likely to have chronic diseases^17,18. Children with chronic critical illness (CCI) are high-risk patients with persistent multiple-organ dysfunction and functional morbidities requiring recurrent or prolonged critical care. The current definitions of the pediatric CCI, while variable and often subjective, form the basis for a more precise definition of the patient population based on the examination of factors related to complexity and chronicity (including prolonged PICU length of stay and hospital readmission)¹⁹. The time from FTP completion to death after discontinuation/withholding of futile therapies (based on available in-hospital records) is broadly consistent with that reported in most published studies²⁰ but shorter than that reported in our pilot study⁶, which may be due to differences in the profile and internal policy of the hospital ward regarding patient qualification, as demonstrated in multivariable logistic regression model adjusted for center.

Our findings regarding discharge to hospice care are consistent with the literature^8,15,21. Avoiding invasive, burdensome procedures may improve quality of life and, in some cases, prolong survival. Optimal end-of-life care is based on the principles of palliative medicine. The World Health Organization defines palliative care as measures aimed at improving the quality of life of patients and their relatives who face a terminal disease. Its goal is to provide patients with benefits without physical, mental, or spiritual suffering²². Patient well-being has become the main goal of therapy. Establishing effective communication and mutual trust with parents is imperative to ensure that clinical decisions remain aligned with the patient’s best interests. Formal guidelines do not mandate parental consent for FTP implementation; however, our dataset inherently excludes cases where parents dissented from the medical recommendation. In the majority of instances, clinical teams prioritized providing families with psychological support, sufficient time for reflection, and clinical reassurance to validate the necessity of the intervention. Nevertheless, it must be underscored that the decision to initiate an FTP is fundamentally a medical judgment and does not strictly require parental authorization. This distinction can precipitate ethical and legal conflicts, as evidenced by recent high-profile cases in Poland²³.

Parental involvement in end-of-life (EoL) decisions varies significantly across countries. For instance, a study from Slovenia revealed that only 26% of intensivists believe a child’s legal guardian should participate in such decisions²⁴. In contrast, research from Belgium reported that physicians consulted parents in 84% of cases involving neonatal and infant deaths²⁵.

Although Polish guidelines do not strictly require parental consent for a Futile Treatment Protocol (FTP), clinical practice often dictates that if both parents oppose the procedure, the protocol is not initiated. Consequently, parental opinions were not analyzed in the present study. However, it is argued that judicial intervention should be sought in cases where parental refusal to discontinue non-beneficial treatment flagrantly violates the best interests of the child. Ultimately, EoL decisions regarding futile care must be made by qualified physicians focused on the patient’s well-being, while maintaining essential communication with the family to clearly explain the medical reasoning²⁶.

Limitations

This retrospective, multicenter study is subject to potential bias, including selection bias and information (misclassification) bias due to record-based ascertainment. Missing data were handled using complete-case analysis without imputation, and cases with > 50% missing variables were excluded. Our study captured only in-hospital outcomes. Pre-hospitalization chronic illness status was derived from available documentation and may have been misclassified when medical histories were incomplete, potentially leading to under-ascertainment of chronic conditions. Patient- and family-reported outcomes (quality of life, symptom burden, and family satisfaction) were not available in this retrospective dataset. No post-discharge follow-up was available for patients discharged to hospice or home care; therefore, survival and other longer-term outcomes could not be assessed in this group. Moreover, incomplete documentation after FTP completion limited estimation of life expectancy for some patients. In addition, the cause of death among patients who died in hospital was not uniformly recorded, limiting cause-specific analyses. Future studies require prospective observation with participation of additional PICUs in Poland.

Conclusions

In this multicenter retrospective cohort from six tertiary PICUs in Poland, futile therapy protocols (FTPs) were most commonly initiated in neonates and infants, typically in children with pre-existing chronic illness and neurological conditions. After FTP implementation, approximately half of patients continued invasive mechanical ventilation, while 40.6% were discharged alive to hospice care or a pediatric home mechanical ventilation program. These data provide real-world evidence that the national FTP framework is being applied in routine practice and that post-FTP management often follows two distinct trajectories (continued invasive support vs. transition to hospice/home care). The structured protocol may support transparent, team-based decision-making and documentation; however, our findings also indicate that clearer guidance on post-FTP pathways and prospective assessment of longer-term and family-centered outcomes would strengthen future evaluations.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

M.D., B.R., A.Ś-B., M.Z. designed the study and drafted the manuscript. M.D., B.R.,A.Ś-B., E.B-M, M.M-S. and A.K-C. collected the data. M.G. conducted the data analysis. M.D. and B.R.and Ł.W.prepared the final version of the manuscript. I.P-K, I.D. and A.A. critically reviewedand edited the manuscript. All the authors read and agreed to the publishedversion of the manuscript.

Data availability

https://data.mendeley.com/datasets/5r8gr3rrmv/1.

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval

The study protocol (a multicenter retrospective review of medical records) received a single central ethical approval from the Ethics Committee of the Medical University of Lublin (KE/651/07/2024), applicable to all participating centers. Local investigators at each participating site were authorized to extract the required variables and transfer only de-identified data to the central study database. Owing to the retrospective, non-interventional design and the use of de-identified data, the requirement for informed consent was waived by the approving ethics committee. All methods were performed in accordance with relevant guidelines and regulations.