Abstract
Background
The Home Ventilation and Respiratory Support Service (HVRSS) at Tan Tock Seng Hospital’s (TTSH) in Singapore manages patients with chronic, progressive neurological diseases requiring home ventilation support nation-wide. We examined prevalence of advance care plans (ACPs) among a cohort of decedent patients, available ACP information, and the relationship of having completed an ACP with healthcare utilisation [acute hospital (AH) admissions, cumulative AH length of stay (LOS), emergency department (ED) and HVRSS home visits], 90 days prior-to-death. We hypothesised that some HVRSS patients had ACPs, and HVRSS patients with and without ACPs potentially have differences in end-of-life healthcare utilisation outcomes.
Methods
This was a retrospective cohort study. Patients were deceased by 2019, were discharged home with ≥1 HVRSS encounter and sufficient electronic medical documentation. We extracted healthcare utilisation outcomes, in the 90 days prior-to-death period, of: AH admissions, cumulative AH LOS, ED and HVRSS home visits. We subsequently linked out dataset to another database of ACP information, using unique identifiers, to determine whether study participants had completed an ACP during their lifetime. The factor of having completed ACPs (with and without) was incorporated into multiple regression analysis of outcomes.
Results
Among 118 patients, 75 (63.6%) were male and mostly Chinese (84.7%), who had a mean HVRSS enrolment-age was 64.2 years [standard deviation (SD) =16.1 years] and were mostly diagnosed with motor neuron disease (47.5%). The prevalence of ACPs was (44.1%), with more being doctor-facilitated discussions (67.3%). Most discussions were conducted with subjects with or without family members (82.7%). Mean ACP-to-death duration was 1.6 years (SD =1.6). Compared to participants without ACP, the cumulative hospital LOS in the 3 months prior-to-death was at least 55% lower for participants with a completed ACP [ACP: mean =5.8, SD =11.0; No-ACP: mean =14.9, SD =25.3; internal rate of return (IRR): 0.38–0.53; P<0.001]. However, the number of home care visits was up to 54% higher for those with completed ACPs (ACP: mean =5.3, SD =4.4; No-ACP: mean =4.0; SD =4.2; IRR =1.27–1.88; P<0.001). No significant between-group differences in hospitalization (P=0.79), and ED visits were detected (P=0.14).
Conclusions
This was the first known examination of ACP uprate among a nationally representative multi-ethnic group of home ventilation patients. ACP uptake appeared to be lower than similar studies conducted in other countries. The completion of ACP with higher home visits by a homecare team may have supported end-of-life care and reduced hospital LOS at end-of-life. ACPs can potentially be an important service-planning consideration for home ventilation patients.
Keywords: Home ventilation, healthcare utilisation, advance care plans (ACPs)
Highlight box.
Key findings
• Forty-four percent of decedent patients enrolled in a home ventilation service had completed an advance care plan (ACP) in their lifetime.
• Among decedent home ventilation patients who had completed an ACP, 44.2% of these plans were completed more than 1 year prior-to-death, and 78.9% were completed after enrollment with a homecare service.
• Decedent home ventilation patients with an ACP consumed <55% in cumulative length of stay in an acute hospital (AH) in the final 90 days of life, compared to those without an ACP.
What is known and what is new?
• Home ventilation patients and their families should have early and regular communication with healthcare teams that involves advance care planning.
• We found that ACPs are associated with lower cumulative AH length of stay at the end-of-life among decedent home ventilation patients.
What is the implication, and what should change now?
• Given the lower ACP prevalence among home ventilation patients, and evident acute healthcare utilisation impact at the end-of-life, efforts towards development of sustainable models of encouraging ACP update among these patients are critically needed.
Introduction
Background
The Home Ventilation and Respiratory Support Service (HVRSS), at Tan Tock Seng Hospital (TTSH), is the sole, dedicated, multi-disciplinary service provider for the specialised needs of adult ventilation-assisted individuals (VAIs) needing long-term home mechanical ventilation (HMV) in Singapore since 2009 (1). HVRSS provides home-based care to a heterogeneous group of patients [degenerative neurological conditions, spinal cord injuries (SCIs), neuromuscular disorders (NMDs), complex myopathies, severe chest-wall disorders, complex post-intensive care unit (ICU)], with additional complex home ventilation needs (Table S1) requiring close respiratory status follow up. NMD or SCI patients start with ventilatory insufficiency from breathing muscle weakness, but over time develop restrictive chest wall pathologies accompanied by pulmonary disease like bronchiectasis and fibrosis from persistent atelectasis or recurrent pneumonia, if not optimally managed. HVRSS patients usually also receive other respiratory support, like cough assist, that may be symptom relieving with rehabilitative, palliative or life-extending impacts. Even across countries of varying income levels, significant resources are required from patients, family and health systems where available to enable HMV (1), and HMV is associated with high, recurring (monthly) economic costs (2). In Singapore, a high-income South East Asian country, the survivability of HVRSS patients is comparable with other VAI cohorts in Australia and Europe, with median survival between 1.8 to 6.7 years (1).
Rationale and knowledge gap
Despite HVRSS patients’ predictably limited survivorship, little is known about their end-of-life preferences and circumstances. In Singapore, advance care planning (ACP) involves discussion of end-of-life preferences, potentially leading to formal documentation. Key recommendations of palliative care guidelines for patients with chronic, progressive neurological diseases include early integration of palliative care access and ACP initiation, especially when deterioration is expected to impair decision-making and communication abilities (3). A study of patients with amyotrophic lateral sclerosis (ALS) found they believed ACP is essential (4). However, ACP may not be prevalent due to communication challenges, although communication aids can support this (5), and preferences require periodic review (6). Acute healthcare utilisation generally increases at the end-of-life (7-10). End-of-life preferences like no cardiopulmonary resuscitation (CPR), comfort or limited care during end-of-life deterioration can be accommodated in at home, without escalation to an acute healthcare facility, lowering acute healthcare utilisation (11). ACP prevalence varies widely across patient populations and healthcare settings, ranging from <5% among Australian chronic disease patients in hospital and community settings and patients presenting with an acute medical emergency in the UK, to around 50% for older Australians (65 years and older) accessing health and aged care services, and high levels (77–97%) for Swedish nursing home patients (12-15). Crucially, ACP shifts the care focus to patients’ quality-of-life and improves families’ bereavement experiences (11). It was found in a systematic review that ACP with specialist palliative care can lower hospital admissions among advanced heart failure patients (16). However, the prevalence of ACP and relationship with end-of-life healthcare utilisation among home ventilation patients is unstudied. Such information would help other home ventilation care providers, especially in Asia, public health policy decision making for end-of-life programmes or future impact evaluation of ACP initiatives among VAIs.
Objectives
We aimed to examine the prevalence of ACP among HVRSS patients. Additionally, we investigated if ACP was associated with end-of-life healthcare utilisation in the last-90-days-of-life [acute hospital (AH) admissions, AH cumulative length of stay (LOS), emergency department (ED), and HVRSS home visits]. We hypothesised that some HVRSS had completed an ACP and ACP is associated with lower end-of-life acute healthcare utilisation. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2024-2057/rc).
Methods
Study design and population
The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Domain Specific Review Board (DSRB) of National Healthcare Group (No. 2019/00854) and individual consent for this retrospective analysis was waived. Study subjects consisted of HVRSS patients who: were enrolled into the service between 2009–2019, were deceased by 2019, were discharged home with ≥1 HVRSS home or outpatient clinic visit after programme enrollment; reflected in their medical documentation.
Main study outcomes
Acute healthcare utilisation outcomes during the last 90-days-of-life were measured. Firstly, we counted how many times each subject was admitted to TTSH, their cumulative LOS at TTSH, visited TTSH ED and home visits by the HVRSS team during the last 90-days-of-life. Particularly for LOS, we did not include in the tabulation any part of the admission that went across the 90-day reference date. Thereafter, this dataset was linked to a database maintained by the ACP department in TTSH, using unique patient identifiers. This ACP database is routinely updated with information captured when any patient in TTSH completes ACP using a hardcopy form with the assistance of any TTSH employed healthcare professional. Data recorded on these hardcopy forms are entered into this ACP database by an ACP department administrative staff, who will conduct further checks and verifications to ensure accuracy and consistency where necessary. We determined whether study subjects had completed ACP during their lifetime by cross-checking the study subject list with the custodians of the TTSH ACP database, and study subjects were categorised into those who had completed ACP and those without ACP during their lifetime of receiving care at TTSH. The study team were then provided by the TTSH ACP database custodians, information in the ACP database that includes formally documented end-of-life preferences [(I) cardiopulmonary resuscitation (to proceed or not during cardiac arrest); (II) medical intervention (comfort measures, limited intervention or full treatment during deterioration of condition); (III) place of terminal care; (IV) place of death (17)], ACP documented date, circumstances surrounding the discussion (number of iterations, setting, language, and people involved), designation of ACP facilitator, and the identity of the participants’ Nominated Healthcare Spokesperson(s) (NHS).
Other information
We collected additional information that could potentially be used as covariates for risk adjustments in regression analysis. These include subject socio-demographic factors (age, gender, ethnicity, marital status, housing type, availability of caregiver and identity of caregiver familial or otherwise). If there was a familial caregiver identified, we collected their socio-demographic information (age, gender, ethnicity, relationship to subject, employment, whether they lived with the participant). As Singapore is geographically divided into three areas covered by three regional healthcare networks [Singapore Health Service (SHS), National Healthcare Group (NHG) and National University Hospital Service (NUHS)] and HVRSS being an NHG-led country-wide service that attends to home ventilation patients who may not reside in its regional catchment, we extracted the regional location of the subjects’ residence. Subject clinical information was collected, consisting of main diagnosis leading to HVRSS enrollment, duration of service enrollment, pre-existing comorbid conditions (e.g., diabetes, hypertension, hyperlipidemia, ischaemic heart disease, heart failure, atrial fibrillation, chronic obstructive pulmonary disease, asthma, stroke or seizures), tracheostomy status, type of ventilation support (at enrollment and death). In the disease trajectory of home ventilation patients, it is not uncommon to have changes in their ventilatory requirements, for example going from non-invasive ventilation (NIV; i.e., without a tracheostomy) assisted (ventilation support <16 hours/day), to NIV and dependent, needing ventilation support ≥16 hours/day. Thus, we additionally captured if subjects had transitions in ventilatory requirements from enrollment to death.
Statistical analysis
Univariate analysis
We determined the percentage of study subjects who had completed ACP. Then, we carried out between-group (with or without ACP) comparisons of outcomes collected, using chi-squared tests for categorical variables, and independent t-tests for continuous variables. Following which, we computed descriptive statistics of ACP-specific information for study subjects who had completed ACP during their lifetime.
Multiple regression
We conducted multiple Poisson regressions for specific healthcare utilisation outcomes [(I) AH admissions at TTSH; (II) AH LOS at TTSH; (III) TTSH ED; and (IV) HVRSS home visits] by looking back for 90 days from the date of death. An independent variable categorising study subjects into those with and without ACP was included in the regression models. In our modelling approach, covariates were incorporated in an incremental manner in our regression models of specified healthcare utilisation outcomes. First, we ran models with the ACP-grouping variable only. Second, socio-demographic characteristics of gender, race, marital status were included. Third, we added the regional catchment area of the study subjects’ residence. Last, we incorporated clinical information (diagnosis, comorbidities, ventilation transition status [type of change in ventilation requirements from enrollment to death], and HVRSS programme enrollment duration). This statistical adjustment approach was undertaken in each of the four end-of-life healthcare utilisation outcomes of hospital admissions, cumulative LOS, ED and home visits by the HVRSS team. As the AH utilisation outcomes were collected from one AH (TTSH), we additionally conducted sensitivity analysis using all AHs’ utilisation data across Singapore.
Results
Study participants
We had 118 HVRSS decedent subjects included in this study. The process of determining the inclusion of HVRSS patients into the study is outlined in Figure 1. We found that 44.1% (n=52/118) of study subjects completed ACP. Socio-demographic and clinical characteristics of study participants, by ACP groups, are shown in Table 1. Among 118 subjects, 75 (63.6%) were male and mostly Chinese (84.7%). Mean HVRSS enrolment-age was 64.2 [standard deviation (SD) =16.1] years for all subjects. Those with ACPs were younger at HVRSS enrolment (P=0.005) and death (P=0.001). More non-ACP subjects were married (87.9% versus 61.5%, P=0.004). subjects with ACP were enrolled in HVRSS longer than those without ACP [ACP: 2.3 years (SD =2.0 years); Non-ACP: 1.4 years (SD =1.5 years), P=0.01]. Subjects’ clinical information, overall and by ACP groups, can be found in Table 2. The most common diagnosis for participants was motor neuron disease (47.5%), and there was no between-group difference in diagnosis distribution. Study subjects mostly maintained ventilation status from HVRSS enrollment to death, and were not different in ventilatory states at enrollment, death or ventilatory transitions by ACP-group. Study subjects with ACP had fewer comorbidities, with more subjects without ACP having hypertension and cardiovascular conditions.
Figure 1.
Diagram of study subjects’ inclusion into study. HVRSS, Home Ventilation and Respiratory Support Service; TTSH, Tan Tock Seng Hospital.
Table 1. Socio-demographic and clinical characteristics of study participants by ACP groups.
| Variables | ACP | No-ACP |
|---|---|---|
| Socio-demographic information | ||
| Gender, n (%) | ||
| Female | 19 (36.5) | 24 (36.4) |
| Male | 33 (63.5) | 42 (63.6) |
| Ethnicity, n (%) | ||
| Chinese | 43 (82.7) | 57 (86.4) |
| Malay | 6 (11.5) | 3 (4.5) |
| Indian | 1 (1.9) | 5 (7.6) |
| Others | 2 (3.8) | 1 (1.5) |
| Marital status, n (%) | ||
| Married | 32 (61.5) | 58 (87.9) |
| Single | 11 (21.2) | 4 (6.1) |
| Widowed | 9 (17.3) | 4 (6.1) |
| Age at enrolment (years), mean ± SD | 59.5±17.8 | 67.9±13.7 |
| Age at death (years), mean ± SD | 61.7±17.9 | 69.3±13.6 |
| Housing type, n (%) | ||
| Housing Development Board (3-room) apartment | 13 (25.0) | 6 (9.1) |
| Housing Development Board (4-room) apartment | 15 (28.8) | 16 (24.2) |
| Housing Development Board (5-room) apartment | 11 (21.2) | 11 (16.7) |
| Private condominium apartment | 5 (9.6) | 10 (15.2) |
| Landed property | 4 (7.7) | 12 (18.2) |
| Others/missing | 4 (7.7) | 11 (16.7) |
| Family member(s) living together, mean ± SD | 2.4±1.5 | 2.6±1.5 |
| Foreign domestic work(s) employed by family, mean ± SD | 0.9±0.6 | 1±0.6 |
| Familial caregiver [1] living together, n (%) | 42 (80.8) | 57 (86.4) |
| Familial caregiver [1] employed, n (%) | 23 (44.2) | 28 (42.4) |
| Familial caregiver [2] living together, n (%) | 16 (30.8) | 25 (37.9) |
| Familial caregiver [1] employed, n (%) | 15 (28.8) | 25 (37.9) |
| Catchment, n (%) | ||
| National Healthcare Group (within catchment) | 32 (61.5) | 36 (54.6) |
| National University Hospital Service (outside catchment) | 12 (23.1) | 19 (28.8) |
| Singapore Health Service (outside catchment) | 8 (15.4) | 11 (16.7) |
| Clinical information | ||
| Diagnosis, n (%) | ||
| Motor neuron disease | 27 (51.9) | 29 (43.9) |
| Neuromuscular disease | 7 (13.5) | 8 (12.1) |
| Spinal cord injury | 5 (9.6) | 11 (16.7) |
| All others† | 13 (25.0) | 18 (27.3) |
| Home Ventilation and Respiratory Support Service enrolment duration in years, mean ± SD | 2.3±2 | 1.4±1.5 |
| Ventilation status at death, n (%) | ||
| IV dependent | 18 (34.6) | 29 (43.9) |
| IV assisted | 3 (5.8) | 2 (3) |
| NIV dependent | 14 (26.9) | 12 (18.2) |
| NIV assisted | 13 (25.0) | 12 (18.2) |
| NIV non-compliant | 0 (0) | 2 (3.0) |
| Not on vent | 4 (7.7) | 9 (13.6) |
| Ventilation transition: enrolment-to-death, n (%) | ||
| Improved | 2 (3.8) | 7 (10.6) |
| Maintained | 40 (76.9) | 44 (66.7) |
| Worsened | 10 (19.2) | 15 (22.7) |
Familial caregiver [1], primary; familial caregiver [2], secondary. †, Central Pathology, Obstructive Pathology, Restrictive Pathology and other complex or rare diagnosis. ACP, advance care planning; IV, invasive ventilation; NIV, non-invasive ventilation; SD, standard deviation.
Table 2. Breakdown of those involved in ACP discussions of study subjects with ACP.
| People involved in ACP discussions | Study population with ACP (n=52), n (%) |
|---|---|
| Designation of ACP facilitator | |
| Doctor | 35 (67.3) |
| Nurse clinician | 7 (13.5) |
| Medical social worker | 6 (11.5) |
| Medical social worker & nurse | 1 (1.9) |
| Nurse | 2 (3.9) |
| Senior coordinator | 1 (1.9) |
| Nominated Healthcare Spokesperson [1] | |
| Spouse | 19 (36.5) |
| Children | 19 (36.5) |
| Parent | 5 (9.6) |
| Sibling | 6 (11.5) |
| NA | 3 (5.8) |
| Nominated Healthcare Spokesperson [2] | |
| Children | 10 (19.2) |
| Parent | 1 (1.9) |
| Sibling | 1 (1.9) |
| Grandchildren | 1 (1.9) |
| Children-in-law | 1 (1.9) |
| NA | 38 (73.1) |
| In attendance during ACP discussion | |
| Patient | 19 (36.5) |
| Patient & 1st Nominated Healthcare Spokesperson | 19 (36.5) |
| Patient, 1st & 2nd Nominated Healthcare Spokesperson | 5 (9.6) |
| 1st Nominated Healthcare Spokesperson | 5 (9.6) |
| 1st & 2nd Nominated Healthcare Spokesperson | 4 (7.7) |
Nominated Healthcare Spokesperson [1]: relationship of the nominated healthcare spokesperson to patient; Nominated Healthcare Spokesperson [2]: relationship of the 2nd nominated healthcare spokesperson to patient. This is when there’s more than 1 nominated healthcare spokesperson involved in the ACP discussion. ACP, advance care planning; NA, not applicable.
ACP overview information
Figure 2 shows ACP completion rates across enrollment years. The ACP completion rates were notably higher in certain years (2012 and 2013; P value =0.002). Temporal information about HVRSS patients’ ACP that we were able to gather include: Majority (86.5%, n=45/52) of the ACP were completed on the same day when started; Slightly less than half (44.2%, n=23/52) of ACP was completed before the last year of life; ACP was mainly (78.8%, n=41/52) performed after HVRSS enrollment; mean ACP-to-death duration was 1.6 years (SD =1.6 years). We found that 19.2% (n=10/52) of subjects who had ACP revisited their care preferences, demonstrated by having more than 1 iteration of completed documentation of ACP. Nearly half (48.1%, n=25/52) of these conversations were not fully conducted in English, and 57.7% (n=30/52) were done in the home setting.
Figure 2.
Rates of ACP completion, by year of enrollment into HVRSS programme. ACP, advance care planning; HVRSS, Home Ventilation and Respiratory Support Service.
Table 2 contains ACP-related information collected regarding persons involved in the ACP of study subjects. ACP was more commonly physician-facilitated discussions (67.3%, n=35/52), with majority done by HVRSS-specific physicians. NHS were family members like spouse or children, unsurprisingly. Most discussions were conducted with subjects and/or family members (82.7%, n=43/52).
End-of-life care preferences captured in ACP
Table 3 reports end-of-life care preferences documented in an ACP document; 80.8% (n=42/52) did not want to be resuscitated in the event of a cardiopulmonary arrest, and 73.1% (n=38/52) only wanted comfort or limited medical interventions. 59.6% (n=31/52) would like a trial of home treatment before considering a hospital transfer, and 21.2% (n=11/52) wanted to remain at home for treatment. A 73.1% (n=38/52) majority indicated that they wanted to pass on at home.
Table 3. Breakdown of end-of-life care preferences documented in ACP of study subjects.
| ACP preferences | Study population with ACP (n=52), n (%) |
|---|---|
| Cardiopulmonary resuscitation | |
| Do not resuscitate | 42 (80.8) |
| Resuscitate | 10 (19.2) |
| Medical treatment | |
| Comfort or limited interventions | 38 (73.1) |
| Full treatment | 14 (26.9) |
| Preferred place of care | |
| Trial of treatment in own home before considering transfer to hospital | 31 (59.6) |
| Remain in my own home | 11 (21.2) |
| Remain in hospital | 4 (7.7) |
| Transfer to hospital | 4 (7.7) |
| Trial of treatment in nursing home or hospice before considering transfer to hospital | 1 (1.9) |
| No preference | 1 (1.9) |
| Preferred place of death | |
| Own home | 38 (73.1) |
| Healthcare institutions | 8 (15.4) |
| Not discussed or no preference | 3 (5.8) |
| Own home or specific healthcare institution/s | 2 (3.8) |
| Not own home | 1 (1.9) |
ACP, advance care planning.
End-of-life healthcare utilisation outcomes
The unadjusted results of healthcare utilisation in the last 90-days-of-life, by ACP groups, are presented in Figure 3. We can observe some unadjusted between-group differences for cumulative LOS and HVRSS home visits, but these appear minimal for hospital admissions and ED visits in the last 90-days-of-life. The graphs in Figure 4 depict multivariable modelling outputs of: (I) hospital admissions; (II) cumulative LOS; (III) ED; and (IV) HVRSS home visits, in the last 90-days-of-life (reference: No-ACP group), accounting for socio-demographic characteristics, regional catchment area and clinical information. Of note, subjects with ACP had at least 55% lower cumulative LOS, compared to those without ACP [95% confidence interval (CI): 0.38–0.53, P<0.001]. Additionally, subjects with ACP had up to 54% more HVRSS home visits in the last 90-days-of-life (95% CI: 1.27–1.88, P<0.001), compared to those without ACP. The hospital LOS findings from analysis of acute healthcare utilisation data from all AHs across Singapore, yielded similar results. We did not find multivariable between-group differences in hospital admissions (P=0.79) and ED visits (P=0.14), unsurprising given our unadjusted findings.
Figure 3.
Unadjusted results of healthcare utilisation outcomes in the last 90-days-of-life, by ACP groups. ACP, advance care planning; ED, emergency department; LOS, length of stay.
Figure 4.
Results of multivariate analyses (incidence rate ratios & 95% confidence intervals) of hospital admissions, cumulative length-of-stay, emergency department and HVRSS home visits in the last 90-days-of-life. Reference = No-ACP group; adjustments = gender, race, death age, marital status, catchment area, diagnosis, comorbidities, ventilation transition status, duration of HVRSS programme enrolment. ACP, advance care planning; HVRSS, Home Ventilation and Respiratory Support Service; LOS, length of stay.
Discussion
Key findings
We confirmed our primary hypothesis and determined the prevalence of ACP completion among decedent study subjects was 44%. Among those with ACP, 44.2% were completed more than a year before death, and 78.9% after HVRSS enrollment. We found evidence of different profiles of subjects who had varying prevalence rates of ACP. We uncovered that the current manpower required for conducting ACP among this patient population in Singapore was heavily reliant on bilingual physicians, and many subjects had their ACP performed more than a year before death. A vast majority of study subjects indicated they preferred limited medical intervention, and home-based care and death in their ACP, despite their complex medical condition requiring medical equipment to keep them alive and comfortable at home. Specifically, we found that having ACP was associated with healthcare utilisation findings of lower AH cumulative LOS and more home visits by the HVRSS team in the last 90-days-of-life.
Strengths and limitations
To our knowledge, this is the first examination of ACP uptake & associated end-of-life healthcare utilisation outcomes of the only adult home ventilation service in Singapore. The patient population studied is a nationally representative sample, multi-ethnic group who received home ventilation support, capturing the local cultural context of this patient group. Due to the Chinese ethnicity majority in the study sample, the end-of-life preferences revealed in this study may be applicable to other home ventilation patients of similar cultural backgrounds. A key limitation of this study is some incomplete data capture, like national-level ACP information and associated costs of healthcare utilisation, and concordance of preferences. However, we anticipate that numbers of ACP among our study subjects, recorded outside of TTSH to be small. This was because the care of these home ventilation patients was strongly anchored with their home ventilation care team, employed by TTSH, and we think that subjects would more likely want for their end-of-life preferences to be known by staff of TTSH. We were unable to measure healthcare utilisation cost outcomes, but given that healthcare utilisation, particularly LOS, and associated costs are highly related, we think we are able to surmise that our findings on LOS would give a good indication on associated healthcare utilisation cost outcomes. We were not able to verify if end-of-life preferences were concordant with actual end-of-life circumstances, and thus study the potential mediation of concordance on healthcare utilisation in our analysis.
Comparison with literature
The prevalence of ACP among decedent HVRSS patients appears substantially lower, when compared to another group of similar patients in the US where ACP completion rate was 76% (3), and similarly lesser in comparison to older health and aged care services users and nursing home patients (14,15). Evidence of different profiles of patients who had varying prevalence rates of ACP has been highlighted in population-level ACP studies in other countries, like Australia (18). The end-of-life preferences of our study subjects were quite dissimilar to the general Singapore population, especially relating to preferred place of death, where only about 40% of the general Singapore population wanted to pass on at home (17).
Explanations of findings
The lower rate of ACP prevalence and awareness in our study is likely due in part to the known understanding of cultural taboos around such topics among Asian populations, leading to low awareness, knowledge and engagement in topics regarding ACP (19,20). In the Asian context, family involvement in ACP is crucial, which in turn makes ACP conversations more complex and challenging to initiate for Asian healthcare professionals (21,22). Another factor for low ACP rates, as mentioned earlier, communication barriers faced by such patient populations (5). The reliance on HVRSS physicians as the main facilitator for ACPs and provider of care as patients’ disease progresses, is likely related to the close partnership and rapport with patients, and HVRSS physicians’ holistic understanding of patients’ medical conditions. HVRSS physicians, with their high levels of experience and expertise, are keenly aware of disease progressions and are able to give patients’ an accurate depiction of this and issues these patients may encounter as they are dying. However, there are few HVRSS physicians, meaning the number of ACP sessions is therefore limited in numbers. Also, anecdotal knowledge from the HVRSS team is that home ventilation patients can have changes in care preferences due to evolving medical or social circumstances, and as such there may be lower numbers of formal ACP documented due to the reluctance in formalising preferences that could be transient in nature, despite regular engagements and end-of-life discussions.
The HVRSS team endeavors to manage patients’ care needs in the home as much as possible, which is demonstrated by our healthcare utilisation results of higher home visits in the last 90-days-of-life. Through their clinical efforts, HVRSS team maximises the patients’ time at home, before transiting to AH care only when home care measures according to patient preferences are exhausted and if medically necessary, thus potentially shortening their AH LOS. The difference in end-of-life preferences of our study participants, compared to the general Singapore population, is quite understandable given the nature of the disease trajectory of home ventilation patients, whereby the disease burden is gradually progressive over years, and individuals may come to terms with lowering medical care intensity at the end-of-life. Also, due to aforementioned communication difficulties, many home ventilation patients experience, they might fear hospitalisation as their disease advances, as they may feel their needs may not be understood by hospital staff who are not familiar with their non-verbal communication methods. So, we think that the patient management of the HVRSS team and end-of-life choices of our study participants in their advance care plans are closely aligned with our main healthcare utilisation findings of lower hospital cumulative LOS and more home visits by the HVRSS team in the last 90-days-of-life. The reduced desires of hospital-centric medical management together with heightened wishes of home-based care, when facilitated by the availability of a home-based service to support the home-based care needs of these patients at end-of-life is perhaps the mechanistic pathway of the healthcare utilisation results in this study. This muting of healthcare resource use was not picked up in the other healthcare utilisation outcomes of hospital admissions and ED visits, likely due to the low event rates in the first place, as this group was receiving regular support from a home-based team.
Implications and actions needed
Our lower ACP prevalence is of concern for this specific clinical sub-group, due to the high healthcare utilisation impact from having ACP done. Nevertheless, it was encouraging to note there were higher 2012 and 2013 ACP prevalence among decedent study subjects, which was likely connected with the introduction of a national ACP programme in 2011, to encourage conversations about end-of-life care preferences (17). This demonstrates that such policy on a country-wide level may have some trinkle down effects to individual clinical programmes, which then peters out without sustained impact. However, it does imply that a policy-based mechanism can be influential of ACP rates among the home ventilation population, perhaps due to the population needs being met by a single-institution mode of care delivery. We think that end-of-life discussions may need tailoring for specific profiles to improve ACP uptake for those receiving home ventilation support, targeting those older, married, with more comorbidities, coinciding with enrollment into HVRSS. If ACP were to be encouraged among this clinical patient population among multi-ethnic population cohorts, healthcare systems need to be cognisant that this may heavily burden scarce healthcare manpower, like bilingual physicians, who were the majority primary healthcare profession involved in facilitation of ACP for individuals receiving home ventilation support in our study. This means that moving forward, we need more sustainable models of encouraging ACP update, for example more frequent ACP triggers during inpatient stays, or enrollment, and having multiple sources of ACP input to improve completion rates, as physician manpower is limited. Importantly, these suggested models for home ventilation users require evaluation of feasibility, effectiveness and sustainability. It would be encouraging for healthcare providers considering such models of ACP uptake implementation for home ventilation patients to note, that there should be ample lead-time for the majority of ACP to be completed before end-of-life, with many ACP of study subjects performed more than a year before death, and benefits in reducing cumulative LOS in the acute setting were nevertheless latently expressed at the end-of-life. However, as communication challenges faced by home ventilation patients (limited verbal and more non-verbal, i.e., eye-blinking or gesturing) are high, especially further along in the disease trajectory, it would be our research team’s recommendation for early ACP. This will allow care goals and end-of-life preferences to be precisely captured when home ventilation patients have better communication ability, and they may have more adequate time to reflect on their circumstances. This can be done with an initial understanding between the patients and care teams that these wishes can be revisited and updated later to ensure that information captured in the ACP documentation are up-to-date and relevant for the patients through regular check-ins. Some other system-level implementation facilitators that have been highlighted by researchers in this area, include approaches that emphasise effective communication and care integration between healthcare teams, transparent processes and upskilling healthcare professionals involved with such patients in the proficiency of conducting of end-of-life care preference discussions (23).
In summary, pertinent actions needed to increase ACP uptake in similar patient populations may potentially encompass national-level public health policy campaigns, targeting of specific profiles prone to not being receptive to ACP considerations, and the development and testing of sustainable models of early ACP uptake and regular review.
Conclusions
In conclusion, 44.1% of decedent HVRSS patients in our study completed ACP, and among them: 44.2% were completed more than a year before death, and 78.9% were completed after a home ventilation service enrollment. Importantly, HVRSS patients with ACP had 50% lower cumulative usage of AH LOS that was complemented with 53% more HVRSS home visits during end-of-life.
Supplementary
The article’s supplementary files as
Acknowledgments
None.
Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Domain Specific Review Board (DSRB) of National Healthcare Group (No. 2019/00854) and individual consent for this retrospective analysis was waived.
Footnotes
Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2024-2057/rc
Funding: This work was supported by Palliative Care Centre for Excellence in Research and Education (PalC) Research Grant 2020 (No. PalC-RG-20/P004).
Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2024-2057/coif). The authors have no conflicts of interest to declare.
Data Sharing Statement
Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2024-2057/dss
References
- 1.Tan GP, Soon LHY, Ni B, et al. The pattern of use and survival outcomes of a dedicated adult Home Ventilation and Respiratory Support Service in Singapore: a 7-year retrospective observational cohort study. J Thorac Dis 2019;11:795-804. 10.21037/jtd.2019.02.18 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Nonoyama ML, McKim DA, Road J, et al. Healthcare utilisation and costs of home mechanical ventilation. Thorax 2018;thoraxjnl-2017-211138. [DOI] [PubMed]
- 3.Oliver DJ, Borasio GD, Caraceni A, et al. A consensus review on the development of palliative care for patients with chronic and progressive neurological disease. Eur J Neurol 2016;23:30-8. 10.1111/ene.12889 [DOI] [PubMed] [Google Scholar]
- 4.Moss AH, Oppenheimer EA, Casey P, et al. Patients with amyotrophic lateral sclerosis receiving long-term mechanical ventilation. Advance care planning and outcomes. Chest 1996;110:249-55. 10.1378/chest.110.1.249 [DOI] [PubMed] [Google Scholar]
- 5.Karlsen MW, Holm A, Kvande ME, et al. Communication with mechanically ventilated patients in intensive care units: A concept analysis. J Adv Nurs 2023;79:563-80. 10.1111/jan.15501 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Silverstein MD, Stocking CB, Antel JP, et al. Amyotrophic lateral sclerosis and life-sustaining therapy: patients' desires for information, participation in decision making, and life-sustaining therapy. Mayo Clin Proc 1991;66:906-13. 10.1016/s0025-6196(12)61577-8 [DOI] [PubMed] [Google Scholar]
- 7.Kaur P, Wu HY, Hum A, et al. Medical cost of advanced illnesses in the last-year of life-retrospective database study. Age Ageing 2022;51:afab212. 10.1093/ageing/afab212 [DOI] [PubMed] [Google Scholar]
- 8.Ní Chróinín D, Goldsbury DE, Beveridge A, et al. Health-services utilisation amongst older persons during the last year of life: a population-based study. BMC Geriatr 2018;18:317. 10.1186/s12877-018-1006-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Hamblin R, Minko N, Shuker C, et al. What happens at the end of life? Using linked administrative health data to understand healthcare usage in the last year of life in New Zealand. Health Policy 2018;122:783-90. 10.1016/j.healthpol.2018.05.011 [DOI] [PubMed] [Google Scholar]
- 10.von Wyl V, Telser H, Weber A, et al. Cost trajectories from the final life year reveal intensity of end-of-life care and can help to guide palliative care interventions. BMJ Support Palliat Care 2018;8:325-34. 10.1136/bmjspcare-2014-000784 [DOI] [PubMed] [Google Scholar]
- 11.Kononovas K, McGee A. The benefits and barriers of ensuring patients have advance care planning. Nursing Times 2017;113:41-4. [Google Scholar]
- 12.Jeong S, Barrett T, Ohr SO, et al. Prevalence of advance care planning practices among people with chronic diseases in hospital and community settings: a retrospective medical record audit. BMC Health Serv Res 2021;21:303. 10.1186/s12913-021-06265-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Knight T, Malyon A, Fritz Z, et al. Advance care planning in patients referred to hospital for acute medical care: Results of a national day of care survey. EClinicalMedicine 2020;19:100235. 10.1016/j.eclinm.2019.12.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Detering KM, Buck K, Ruseckaite R, et al. Prevalence and correlates of advance care directives among older Australians accessing health and residential aged care services: multicentre audit study. BMJ Open 2019;9:e025255. 10.1136/bmjopen-2018-025255 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kastbom L, Falk M, Karlsson M, et al. Prevalence, content and significance of advance care planning in nursing home patients. Scand J Prim Health Care 2022;40:29-38. 10.1080/02813432.2022.2036429 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kernick LA, Hogg KJ, Millerick Y, et al. Does advance care planning in addition to usual care reduce hospitalisation for patients with advanced heart failure: A systematic review and narrative synthesis. Palliat Med 2018;32:1539-51. 10.1177/0269216318801162 [DOI] [PubMed] [Google Scholar]
- 17.Tan WS, Bajpai R, Low CK, et al. Using Routinely Collected Data to Ascertain Concordance With Advance Care Planning Preferences. J Pain Symptom Manage 2018;56:659-666.e2. 10.1016/j.jpainsymman.2018.07.017 [DOI] [PubMed] [Google Scholar]
- 18.Buck K, Nolte L, Sellars M, et al. Advance care directive prevalence among older Australians and associations with person-level predictors and quality indicators. Health Expect 2021;24:1312-25. 10.1111/hex.13264 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Shen JJ, Dingley C, Yoo JW, et al. Sociocultural Factors Associated with Awareness of Palliative Care and Advanced Care Planning among Asian Populations. Ethn Dis 2020;30:459-68. 10.18865/ed.30.3.459 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Vashisht A, Gutman G, Kaur T. ACP Conversations with Chinese and South Asian Patients: Physicians' Perspectives of Barriers and Facilitating Factors. Can Geriatr J 2023;26:486-92. 10.5770/cgj.26.691 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Martina D, Lin CP, Kristanti MS, et al. Advance Care Planning in Asia: A Systematic Narrative Review of Healthcare Professionals' Knowledge, Attitude, and Experience. J Am Med Dir Assoc 2021;22:349.e1-349.e28. 10.1016/j.jamda.2020.12.018 [DOI] [PubMed] [Google Scholar]
- 22.Mori M, Chan HYL, Lin CP, et al. Definition and recommendations of advance care planning: A Delphi study in five Asian sectors. Palliat Med 2025;39:99-112. 10.1177/02692163241284088 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Kavanagh E, Rowley G, Simkiss L, et al. Advance care planning for patients with chronic obstructive pulmonary disease on home non-invasive ventilation: A qualitative study exploring barriers, facilitators and patients' and healthcare professionals' recommendations. Palliat Med 2023;37:1413-23. 10.1177/02692163231192130 [DOI] [PubMed] [Google Scholar]